UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bone-conduction auditory steady-state responses Small, Susan Anne 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-267950.pdf [ 15.52MB ]
Metadata
JSON: 831-1.0100422.json
JSON-LD: 831-1.0100422-ld.json
RDF/XML (Pretty): 831-1.0100422-rdf.xml
RDF/JSON: 831-1.0100422-rdf.json
Turtle: 831-1.0100422-turtle.txt
N-Triples: 831-1.0100422-rdf-ntriples.txt
Original Record: 831-1.0100422-source.json
Full Text
831-1.0100422-fulltext.txt
Citation
831-1.0100422.ris

Full Text

BONE-CONDUCTION AUDITORY STEADY-STATE RESPONSES by SUSAN ANNE S M A L L B.Sc, The University of British Columbia, 1986 M.Sc, Simon Fraser University, 1988 M.S., The University of Wisconsin-Madison, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS OF THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Audiology and Speech Sciences) T H E UNIVERSITY OF BRITISH COLUMBIA March, 2007 © Susan Anne Small, 2007 Abstract The purpose of this dissertation was to determine whether multiple auditory-state responses (ASSRs) to bone-conduction stimuli can be used to investigate bone-conduction hearing, an essential part of any audiometric assessment. Infant bone-conduction testing methods, maturation of bone-conduction hearing, and ipsilateral/contralateral asymmetries in ASSRs were also assessed. The results show that bone-conduction ASSRs can be used to estimate thresholds in infants and adults with normal hearing. It was also found that choice of electroencephalogram (EEG) conditioning and processing can avoid spurious ASSRs due to aliasing. Non-auditory ASSRs (probably vestibular and indistinguishable from an auditory response) were also identified for high-intensity air-conduction stimuli (problematic when diagnosing residual hearing). Investigation of infant testing methods on bone-conduction threshold shows that: (i) bone-oscillator coupling method (elastic-band vs. hand-held) has no effect on threshold, (ii) use of different oscillator locations on the temporal bone does not affect threshold but a forehead placement results in elevated thresholds, and (iii) infants do not appear to have an occlusion effect (thus one can can assess with or without earphones). Young infants have much better low-frequency bone-conduction hearing compared to adults, which increases with maturation beyond 24 months of age. Infant bone-conduction hearing is slightly poorer in the high frequencies, improving significantly with age only at 2000 Hz. Within all infant groups, low-frequency thresholds are better than high-frequency thresholds; for adults, 500-Hz thresholds are poorer than high frequencies and there is no difference among thresholds above 500 Hz. Bone-conducted signals are much more effective for infants across frequency, especially at low frequencies. Normal levels for bone-conduction hearing in young and older infants are proposed. Ipsilateral/contralateral asymmetries in air- and bone-conduction ASSRs are clearly present more ii often and are larger in infants compared to adults, and suggest that most infants have 10-30 dB of interaural attenuation. These asymmetries have potential as a clinical tool for isolating the cochlea that is contributing to the response in infants. The results of these studies indicate that infants can now be screened for normal bone-conduction hearing with ASSRs; however, infants with hearing loss must be tested before elevated bone-conduction ASSRs thresholds can be interpreted. in Table of Contents Abstract ii Table of Contents iv List of Tables xii List of Figures xvi List of Abbreviations xxi Preface xxiii Acknowledgements xxiv Co-authorship statement xxvi Chapter 1: Introduction to bone-conduction auditory steady-state responses 1 Assessment of Bone-Conduction Hearing in Infants 2 What Are Auditory Steady-state Responses? 5 Stimulus and recording parameters for multiple ASSRs 9 Analysis of multiple ASSRs 13 Bone-Conduction Mechanisms 17 Bone-Conduction Threshold Estimation in Infants 25 Maturation of Bone-Conduction Hearing 30 Potential clinical applications of bone-conduction ASSRs 36 References 48 Chapter 2: Artifactual responses when recording auditory steady-state responses 62 Introduction 63 General Materials and Methods 68 Participants 68 iv Stimuli 69 Calibration .' 70 Recording 70 Procedure 72 Data Analyses 72 Results 72 Experiment 1: Artifactual responses to bone-conduction stimuli 72 Experiment 2: Artifactual responses to high-intensity air-conduction stimuli 76 Discussion 78 References 93 Chapter 3: Multiple auditory steady-state responses to bone-conduction stimuli in adults with normal hearing 98 Introduction 99 Materials and Methods 102 Participants 102 Stimuli 103 Calibration 104 ASSR recording 105 Procedure . 106 Data Analyses 107 Results 108 Normal bone-conduction ASSR thresholds 108 v Monaural air-conduction stimuli (single- and alternated-stimulus polarities) 108 Bone-conducted stimuli (single- and alternated-stimulus polarities) 109 Bone-conduction versus binaural air-conduction ASSRs 110 Discussion 112 Summary 115 Conclusion and Clinical implications 116 References 128 Chapter 4: Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants 131 Introduction 132 General Materials and Methods 137 Stimuli : 138 Calibration 138 ASSR recording 139 ASSR data analyses 140 Statistical analyses 140 Experiment 1: Bone-oscillator coupling method 141 1 A. Effect of coupling method on variability in force level applied to oscillator 141 Materials and methods 141 Participants and procedures 141 Statistical analyses 142 vi Results 143 IB. Effect of coupling method on adult behavioural thresholds 144 Materials and methods 144 Participants and procedures 144 Statistical analyses 144 Results 145 IC. Effect of coupling method on infant ASSR thresholds 145 Materials and methods 145 Participants and procedures 145 Statistical analyses 146 Results 147 Experiment 2: Effect of bone-oscillator placement location on infant ASSRs 148 Materials and methods 148 Participants and procedures 148 Statistical analyses 149 Results 150 Experiment 3: Effect of occluding the ear canal on infant ASSRs 152 Materials and methods 152 Participants and procedures 152 Statistical analyses 153 Results 153 Discussion 154 Bone-oscillator coupling method 154 vii Bone-oscillator placement 157 Unoccluded vs. occluded ears '. 160 Conclusions 162 References 177 Chapter 5: Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing 181 Introduction 182 Materials and Methods 184 Participants 184 Stimuli - 184 Calibration 185 ASSR recordings 186 Procedure 187 Data Analyses ? 188 Results 189 Comparison of bone-conduction ASSRs in infants and adults 189 Threshold.... 190 Amplitude and phase 192 Discussion : 194 Conclusions 199 References 207 Chapter 6: Maturation of bone-conduction auditory steady-state responses 211 Introduction 212 viii Materials and Methods • 216 Participants 216 Stimuli 217 Calibration 217 ASSR recordings 218 Procedure 219 Data Analyses 220 Results 221 Detectability ,. 221 Threshold 222 Amplitude 224 Phase delay 225 Discussion 227 Clinical implications 238 References 248 Chapter 7: Normal ipsilateral/contralateral asymmetries in infant bone-conduction multiple auditory steady-state responses 253 Introduction 254 Materials and Methods 257 Participants 257 Stimuli ..; 257 Calibration 258 ASSR recordings 259 ix Procedure 260 Data Analyses 261 Results 263 Occurrence of ipsilateral/contralateral aysymmetries 264 ASSR threshold 266 ASSR amplitude 268 ASSR phase delay 270 Discussion 272 Clinical implications 278 References 291 Chapter 8: Conclusions and clinical implications 296 Chapter 2: 297 Clinical implications 300 Chapter 3: 300 Clinical implications 301 Chapter 4: 302 Clinical implications 302 Chapter 5: 303 Clinical implications 304 Chapter 6: 305 Clinical implications 306 Chapter 7: 306 Clinical implications 308 x Future Research 308 References 311 Appendix A: UBC Research Ethics Review Board Certificates of Approval 313 Appendix B: Single-subject data for Chapter 2 318 Appendix C: Single-subject data for Chapter 3 326 Appendix D: Single-subject data for Chapter 4 335 Appendix E: Single-subject data for Chapter 5 345 Appendix F: Single-subject data for Chapter 6 349 Appendix G: Single-subject data for Chapter 7 358 xi List o f Tables Chapter 1 Table 1.1 Aliasing and stimulus artifact in the E E G 40 Table 1.2 Comparison across studies of brief-tone ABR thresholds to bone-conducted stimuli in infants 41 Chapter 2 Table 2.1 Hearing characteristics of participants with severe-to-profound hearing loss 86 Chapter 3 Table 3.1 Mean pure-tone behavioural air- and bone-conduction thresholds for adults with normal hearing 117 Table 3.2 Air-conduction mean amplitude and phase delay values for multiple ASSRs in adults with normal hearing to single and alternated monaural stimuli 118 Table 3.3 Air-conduction: 2-way repeated measures A N O V A comparing amplitude and phase delay values for multiple ASSRs in adults with normal hearing to single and alternated monaural stimuli 119 Table 3.4 Mean slope for amplitude and phase delay for bone- and binaural air-conduction stimuli for single and alternated stimuli 120 Table 3.5 Bone-conduction: 2-way repeated measures A N O V A comparing amplitude and phase delay values for multiple ASSRs in adults with normal hearing to single and alternated stimuli 121 xii Table 3.6 Bone- and binaural air-conduction: 2-way mixed A N O V A comparing slope of intensity-amplitude functions and intensity-phase-delay functions 122 Table 3.7 Across-study comparison of ASSR thresholds in dB HL 123 Chapter 4 Table 4.1 Force levels produced by four assistants trained to use the elastic-band and hand-held bone-oscillator coupling methods 163 Table 4.2 Mean behavioural thresholds for bone-conduction pure tones with normal hearing 164 Table 4.3 Mean ASSR thresholds using elastic-headband and hand-held coupling method in infants with normal hearing 165 Table 4.4 ASSR mean amplitudes elicited by bone-conduction stimuli using elastic-headband and hand-held coupling method in infants with normal hearing 166 Table 4.5 Bone-conduction ASSR mean thresholds for three bone-oscillator placements in pre-term infants with normal hearing 167 Table 4.6 Bone-conduction ASSR mean amplitudes at 40 dB HL for three bone-oscillator placements in pre-term infants with normal hearing 168 Table 4.7 Mean bone-conduction ASSR thresholds for unoccluded and occluded ears in infants with normal hearing 169 xiii post-term infants and adults with normal hearing 201 Table 5.2 Mean phase delay results for multiple ASSRs to 40 dB HL bone-conduction stimuli for pre-and post-term infants and adults with normal hearing 202 Chapter 6 Table 6.1 Bone-conduction ASSR amplitude and phase delay: Three-way mixed ANOVAs showing comparisons between intensities (30 and 40 dB HL), across age groups (young, older infants and adults) and carrier frequencies (500, 1000, 2000 and 4000 Hz) 240 Table 6.2 A preliminary estimate of "normal" ASSR levels in dB HL for 500-, 1000-2000- and 4000-Hz bone-conduction stimuli for young and older infants, and adults with normal hearing 241 Chapter 7 Table 7.1 ASSR mean threshold elicited by bone-conduction stimuli for E E G channels ipsilateral and contralateral to the stimulus at 500, 1000, 2000 and 4000 Hz in infants and adults 280 Table 7.2a Differences in infant ASSR bone-conduction thresholds between E E G channels at 500, 1000, 2000 and 4000 Hz 281 Table 7.2b Differences in adult ASSR bone-conduction thresholds between EEG channels at 500, 1000, 2000 and 4000 Hz 282 Table 7.3 Air-conduction ASSR amplitude: Two-way repeated measures ANOVAs showing comparisons within and between age groups at 60 dB HL for 500, 1000, 2000 and 4000 Hz 282 Table 7.4 Bone-conduction ASSR amplitude: Three-way repeated measures (within age groups) and mixed (between age groups) ANOVAs showing comparisons for four intensities at 500, 1000, 2000 and 4000 Hz 284 Table 7.5 Bone-conduction ASSR phase delay: Two-way repeated measures (within age groups) and mixed (between age groups) ANOVAs showing comparisons at 40 dB HL at 500, 1000, 2000 and 4000 Hz 285 xv List of Figures Chapter 1 Figure 1.1 Examples of transient-evoked potentials and how they are related to steady-state potentials 42 Figure 1.2 Stimuli used to obtain multiple ASSRs 43 Figure 1.3 A typical ASSR for an individual subject to a 1000-Hz carrier tone modulated at a rate of 91 Hz (100% amplitude modulated) presented at 60 dB SPL: time domain waveform, polar plot and amplitude spectra resulting from FFT analysis 44 Figure 1.4 Auditory steady-state responses to four simultaneous stimuli with different carrier: polar plots and amplitude spectra resulting from FFT analyses from a typical subject 45 Figure 1.5 Large stimulus artifact in the unfiltered and filtered E E G when 500-Hz bone-conduction stimulus is presented at 50 dB H L 46 Figure 1.6 Two-channel recordings of the brain stem response elicited by bone-conducted stimuli recorded in a 3-month-old infant with bilateral atresia 47 Chapter 2 Figure 2.1 Polar plot for an individual participant for bone-conduction ASSR recorded at 1000 Hz A/D rate 87 Figure 2.2 Polar plots indicating artifactual bone-conduction ASSRs for different A/D rates and stimulus polarities 88 xvi Figure 2.3 Percent occurrence of responses elicited by single-polarity bone-conduction stimuli for different A/D rates.. 89 Figure 2.4 Percent occurrence of responses elicited by alternated bone-conduction stimuli for different A/D rates 90 Figure 2.5 Polar plots indicating artifactual air-conduction ASSRs for different A/D rates and stimulus polarities 91 Figure 2.6 Percent occurrence of responses elicited by single and alternated air-conduction stimuli for different A/D rates 92 Chapter 3 Figure 3.1 Bone- and binaural air-conduction stimuli (single polarity): Mean ASSR amplitudes for adults with normal hearing 124 Figure 3.2 Bone- and binaural air-conduction stimuli (alternated polarity): Mean ASSR amplitudes for adults with normal hearing... 125 Figure 3.3 Bone- and binaural air-conduction stimuli (single polarity): Mean ASSR phase delays for adults with normal hearing 126 Figure 3.4 Bone- and binaural air-conduction stimuli (alternated polarity): Mean ASSR phase delays for adults with normal hearing 127 Chapter 4 Figure 4.1 Cumulative percent of trials that resulted in errors produced from the target force for each individual assistant and across assistants.... 170 Figure 4.2 Representative bone-conduction ASSRs for an individual post-term infant for the elastic-band and hand-held coupling methods 171 xvii Figure 4.3 Representative bone-conduction ASSRs for an individual pre-term infant for three bone oscillator placements 172 Figure 4.4 Cumulative percent occurrence of pre-term subjects with significant responses across frequency for temporal, mastoid and forehead placements 173 Figure 4.5 Representative bone-conduction ASSRs for an individual post-term infant for test ear unoccluded and occluded 174 Figure 4.6: Mean bone-conduction ASSR amplitudes across frequency for ears unoccluded and occluded for post-term infants 175 Figure 4.7: Mean bone-conduction ASSR phase delays across frequency for ears unoccluded and occluded for post-term infants 176 Chapter 5 Figure 5.1 Representative bone-conduction ASSRs for an individual pre-term infant, post-term infant and adult with normal hearing with normal hearing 203 Figure 5.2 Cumulative percent occurrence of participants with significant responses for pre-and post-term infants and adults with normal hearing 204 Figure 5.3 Mean bone-conduction ASSR amplitudes for pre-and post-term infants and adults with normal hearing 205 Figure 5.4 Mean bone-conduction ASSR phase delays for pre-and post-term infants and adults with normal hearing 206 xviii Chapter 6 Figure 6.1 Representative bone-conduction ASSRs for an individual young infant, older infant and adult. Shown are amplitude spectra resulting from FFT analyses (75-105 Hz) of the ASSRs 242 Figure 6.2 Cumulative percent occurrence of subjects with significant responses for young infants, older infants and adults across frequency 243 Figure 6.3 Mean bone-conduction ASSR thresholds at each carrier frequency for young infants, older infants and adults with normal hearing 244 Figure 6.4 Graphical representation of linear regression analyses comparing age in weeks to ASSR thresholds at each of the carrier frequencies... 245 Figure 6.5 Mean bone-conduction ASSR amplitudes at each carrier frequency for young infants, older infants and adults with normal hearing 246 Figure 6.6 Mean bone-conduction ASSR phase delays at each carrier frequency for young infants, older infants and adults with normal hearing.. 247 Chapter 7 Figure 7.1 Representative polar plots for air- (60 dB HL) and bone-conduction (40 dB HL) ASSRs elicited by a 2000-Hz stimulus for an individual infant and adult 286 xix Figure 7.2 Grand mean (vector aberage) polar plots for air- (60 dB HL) and bone-conduction (40 dB HL) ASSRs elicited by a 2000-Hz stimulus for an individual infant and adult 287 Figure 7.3 Cumulative percent occurrence of ipsilateral/contralateral asymmetries in ASSRs to bone- and air-conduction stimuli pooled across carrier frequency for infants and adults 288 Figure 7.4 Air- and bone-conduction ASSR mean amplitudes in the E E G channel ipsilateral and contralateral to the stimulus test ear across frequency in infants and adults 289 Figure 7.5 Air- and bone-conduction ASSR mean phase delays in the E E G channel ipsilateral and contralateral to the stimulus test ear across frequency in infants and adults 290 xx List of Abbreviations ABR = auditory brainstem response A C = air-conduction A/D = analog-to-digital conversion rate A N O V A = analysis of variance A L T = alternated stimulus polarity A M = amplitude modulated ANSI = American National Standards Institute ASSR = auditory steady-state response BC = bone conduction D/A = digital-to-analog conversion rate dB HL = decibels hearing level dB re: 1 u.N = decibels re: 1 micro Newton dB SPL = decibels sound pressure level df = degrees of freedom EEG = electroencephalogram F= Fisher's F ratio FFT = Fast Fourier Transform F M = frequency modulated Hz = Hertz INV = inverted stimulus polarity M M - mixed modulated ms = millisecond xxi n= sample size NON = non-inverted stimulus polarity nV = nanovolt p = probability r = Pearson's product moment correlation RETFL = reference equivalent threshold force level RETSPL = reference equivalent threshold sound pressure level SD = standard deviation t = Student's /-statistic VEMP= vestibular-evoked myogenic potentials VsEP= vestibular-evoked potentials uV= microvolt xxii Preface Auditory steady-steady responses (ASSRs) are one of a group of auditory evoked potentials that can be used to estimate hearing threshold. The body of research that will be discussed in the subsequent chapters will focus on the 80-Hz ASSR recorded to multiple bone-conduction tonal stimuli presented simultaneously (i.e., multiple ASSRs) in infants and adults. Chapter 1 provides a brief overview of current techniques used to assess bone-conduction hearing in young infants and a more detailed review of research areas that relate to recording and analysing multiple ASSRs, as well as our current understanding of the underlying mechanisms of bone-conduction hearing, bone-conduction thresholds in infants, and maturation of bone-conduction hearing. The overall objectives of the dissertation will also be summarized in the introductory chapter. Chapters 2-7 constitute six separate research studies, each of which include: (i) a brief introduction and specific objectives, (ii) methodology, (iii) results, (iv) discussion of findings, and (v) conclusions and/or clinical implications of the findings. In Chapter 2, experiments which investigated methodologic issues associated with recording multiple ASSRs are described. In Chapters 3-7, experiments which investigated bone-conduction ASSRs in infants and adults with normal hearing are presented. Chapter 8 is the concluding chapter and provides a summary of the findings of this dissertation, their clinical implications for assessing infant bone-conduction hearing, and suggestions for future research. xxin Acknowledgements I would like to thank my supervisor, David Stapells, for his encouragement and guidance throughout my doctoral studies and the preparation of this dissertation. He has enthusiastically offered suggestions for improving research design, and spent many hours over the years discussing data and providing editorial comments about manuscripts in preparation for publication. I would also like to thank Terence Picton and Sasha John who have also provided valuable comments on several of the manuscripts. In particular, I would like to thank Sasha John for providing his time and expertise to discuss ideas, help troubleshoot technical problems, and to custom-design equipment that was needed to complete my final study. I would also like to thank David Stapells, Terence Picton and Navid Shahnaz for their input and support as members of my Dissertation Committee. I would like to acknowledge Dan Black from dB Special Instruments for generously providing equipment for calibration of the bone-conduction stimuli and technical advice over the years. I would also like to say thank you to my colleagues in the Human Auditory Physiology Laboratory who were always willing to help trouble-shoot equipment problems or try out new ideas. Thank you, Maxine Armstrong, Lindsay Bendickson, Andrew Dimitrijevic, Charles Fontaine, Jennifer Hatton, Lingyan Mo, Sue Anne Poh, Elais Ponton, and Noreen Simmons. I would like to acknowledge the generous support of my doctoral studies from the Canadian Institutes of Health Research (CIHR) [Fellowship] and the Michael Smith Foundation for Health Research [Trainee Award]. This research was also funded by grants from CIHR and NSERC-Canada to David Stapells. I would also like to thank the many participants and families who volunteered their time, the staff at the Special Care Nursery at British Columbia's Children's and Women's Hospital, xxiv and to the audiologists who helped recruit subjects, especially Dr. Sipke Pijl, Grace Shyng, Naomi Smith and Mark Hansen. Finally, I would like to thank my parents and my children for their love and cooperation throughout the tenure of my studies and, in particular, during the preparation of this dissertation. xxv Co-Authorship Statement Together with my doctoral supervisor, David R. Stapells, I planned the research program set out in this dissertation. My contributions include: (i) selecting and designing the research studies, in consultation with Dr. Stapells, (ii) recruiting participants and collecting data with minimal supervision (research assistants helped with some of the data collection), (iii) analysing and interpreting data in collaboration with Dr. Stapells, and (iv) preparing drafts of all of the dissertation chapters, with editorial comments from Dr. Stapells. Additionally, Jennifer Hatton helped with the data collection, analysis and write-up for Experiment 2 (Effects of bone-oscillator placement location on infant ASSRs) in Chapter 4. xxvi C H A P T E R 1 INTRODUCTION T O BONE-CONDUCTION AUDITORY STEADY-S T A T E RESPONSES 1 Assessment of Bone-Conduction Hearing in Infants Normal hearing sensitivity within the first few years of life is essential for normal speech and language development. There is a major focus in hearing healthcare internationally to implement universal newborn hearing screening to identify infants who are at risk for hearing loss before the age of three months, in order to diagnose hearing loss and begin intervention by six months of age (Joint Committee on Infant Hearing, 2000). In order to initiate early intervention and to maximize speech and language development, accurate and reliable audiometric information is needed to diagnose the type and configuration of hearing loss in early infancy. Typically, air-and bone-conduction thresholds to frequency-specific stimuli in the speech-frequency range are obtained. In order to achieve this goal, accurate and efficient techniques that are appropriate for very young infants are needed to assess frequency-specific hearing thresholds. It is generally accepted that conventional behavioural measures cannot reliably assess hearing thresholds in infants younger than 5-6 months of age. Moore, Wilson and Thompson were the first researchers to report these findings in 1977. Delaroche, Thiebault and Dauman (2004) recently attempted to show that infants 4-6 months of age could respond reliably to air-and bone-conduction stimuli using a behavioural protocol that replaces the usual visual reinforcer (Liden & Kankkuken, 1969; Moore & Wilson, 1978) with a highly interactive exchange with the sole examiner as reinforcement for the conditioned responses from the infant. Not surprisingly, they obtained reliable responses in only half of the infants that they tested, re-confirming Moore and colleagues' earlier (1977) findings. Auditory evoked potentials, such as the auditory brainstem response (ABR), can be reliably recorded in infants as early as 29 weeks post-conceptional age (Ponton, Moore, Eggermont, Wu & Huang, 1994) and are currently used 2 clinically to estimate hearing threshold in infants. Based on recent research in infants with normal and impaired hearing, auditory steady-state responses (ASSRs) may also provide accurate estimates of air-conduction hearing thresholds; however, more data are needed for larger groups of infants with hearing loss before this technique can be fully adopted for clinical use (Stapells, Herdman, Small, Dimitrijevic & Hatton, 2005). Additionally, there are no bone-conduction ASSR data for either normal or hearing-impaired infants, and there are only three published studies (prior to the studies presented in Chapters 2-7) that have estimated bone-conduction ASSRs, and these investigated only adults with normal hearing. Lins, Picton, Boucher, Durieux-Smith, Champagne, Moran, Perez-Abalo, Martin, & Savio (1996) and Dimitrijevic, John, Van Roon, Purcell, Adamonis, Ostroff, Nedzelski, & Picton (2002) recorded ASSRs to bone-conduction stimuli directly, but assessed responses to stimuli no higher than 20-30 dB above threshold using a forehead placement for the bone oscillator. Cone-Wesson, Rickards, Poulis, Parker, Tan & Pollard (2002) indirectly estimated ASSR bone-conduction thresholds using the "sensorineural acuity level" (SAL) test which uses bone-conducted noise to mask an air-conduction stimulus that is presented just above the subject's threshold (Jerger & Jerger, 1965). There are no published bone-conduction ASSR data for adults with sensorineural, conductive or mixed hearing losses, except for one study that simulated conductive hearing losses in adults and estimated the difference between their air- and bone-conduction ASSR thresholds (Jeng, Brown, Johnson & Vander Werff, 2004). Behavioural hearing threshold estimation routinely assesses air-conduction and bone-conduction pure-tone thresholds to distinguish between sensorineural, conductive and mixed hearing losses. Accurate bone-conduction thresholds are particularly important when assessing infants and young children who have unilateral or bilateral otits media or atresia 3 (Jahrsdoerfer, Yeakley, Hal, Robbins & Gray, 1985; Stapells & Ruben, 1989). The accurate diagnosis of a conductive or mixed hearing loss is necessary in order to plan medical intervention and aural (re)habilitation. If ASSRs are to become a standard clinical tool for assessing infants, it follows that ASSRs should be investigated not only to air-conduction stimuli but also to bone-conduction stimuli in this target population. The brief-tone ABR to air- and bone-conduction stimuli is the current "gold standard" for estimating frequency-specific thresholds in infants and young children (Stapells, 2000a). The ABR is generated in the auditory pathways of the brainstem (Hashimoto, Ishiyama, Yoshimoto & Nemoto, 1981; Starr and Hamilton, 1976) and is unaffected by sleep state (Picton & Hillyard, 1974), making it ideal for assessing infants. One shortcoming of the ABR technique is that only one ear and one frequency can be tested at the same time. Another limitation of the ABR to brief tones is that detection of a response in the waveform usually depends on skilled, subjective assessment of replicated responses, allowing for error in judgement of the presence of responses depending on the experience of the clinician (Stapells, 2000a). The multiple 80-Hz ASSR, which uses amplitude and/or frequency modulated stimuli to evoke a response, is currently of great interest for use with infants, because it can be used to quickly and objectively obtain frequency-specific thresholds in subjects who are awake or in natural sleep (Picton, John, Dimitrijevic & Purcell, 2003). Similar to the ABR, the 80-Hz ASSR is believed to be generated primarily in the brainstem (Herdman, Lins, Van Roon, Stapells, Scherg & Picton, 2002; Kuwada, Anderson, Batra, Fitzpatrick, Teissier & D'Angelo, 2002; Mauer & Doring, 1999) and is minimally affected by sleep. ASSRs can be recorded for single- or multiple-carrier frequencies to one, or both, ears simultaneously. ASSRs are detected objectively using statistical tests (John & Picton, 2000); their detection does not rely on the experience of the clinician. Multiple ASSRs 4 are of considerable interest as an assessment tool because of their objectivity and potential reduction in clinical testing time. What are Auditory Steady-State Responses? The ABR is an example of a transient response which, by definition, means that the response to one stimulus finishes before the next stimulus is presented, i.e., each individual evoked potential recorded is largely independent of any preceding evoked potentials. A steady-state response is "a repetitive evoked potential whose constituent discrete frequency components remain constant in amplitude and phase over an infinitely long time period" (Regan, 1989, p. 35). ASSRs can be generated by a short-duration stimulus (e.g., a brief tone) presented at a sufficiently fast rate so that the transient response to one stimulus overlaps with the transient response to the previous stimulus. ASSRs can also be elicited by tonal stimuli that are amplitude and/or frequency modulated (the neurons "phase-lock" to the envelope of the modulated continuous tone); the ASSRs are measured in the frequency domain at the frequencies used to modulate the carrier tones (John & Picton, 2000a Lins & Picton, 1995). ASSRs have also been evoked by the beating resulting from two tones presented simultaneously as well as by noise stimuli (reviewed in Picton et al., 2003). In the time domain, a steady-state response resembles overlapping sinusoids or complex waveforms depending on the frequency at which the stimuli are presented. Unlike the E E G recordings for an ABR which is comprised of frequencies from approximately 30-1500 Hz (Elberling, 1979; Sininger, 1996; Spivak, 1993; Stapells, 1989; Suzuki, Sakabe & Miyashita, 1982), the energy present in the E E G when recording an ASSR, after sufficient averaging to reduce background noise, is dominated by energy at the stimulus presentation rate and its harmonics (Regan, 1989, p.35). At higher presentation rates, when 5 constituent frequencies are further apart, the waveform is simpler because fewer harmonics contribute to the response. These concepts are illustrated for two different time-domain waveforms in Figure 1.1. The transient waveforms, wave V of the ABR and the components of the middle latency response, are shown in the top of the figure. The steady-state waveforms are shown in the bottom part of this figure; when an 80-Hz stimulation rate is used, wave V of the ABR overlaps and results in an almost sinusoidal steady-state response; similarly, when a 40-Hz presentation rate is used, the components of the ABR and middle-latency response overlap resulting in a steady-state response with a slightly more complex waveform than the 80-Hz response. Figure 1.1 Human steady-state responses to visual stimuli were the first to be investigated (Regan, 1966). Steady-state responses recorded at the scalp to auditory stimuli in humans were first studied in detail by Galambos, Makeig and Talmachoff (1981); however, earlier studies have contributed to our understanding of ASSRs. Geisler (1960) recorded responses at the scalp elicited by clicks presented at 1-120 Hz and found that responses at the higher stimulation rates contained energy that matched the stimulation rate. Schimmel, Rapin and Cohen (1974) found that auditory evoked potentials with peak latencies ranging from 20-40 ms were recorded most efficiently at 40-45 Hz, which Galambos et al. (1981) later confirmed. Campbell, Atkinson, Francis and Green (1977) investigated responses to clicks and different tonal stimuli presented at 8-32 Hz and found that responses were largest at 12-16 Hz and tended to be larger for clicks and 500-Hz tonal stimuli compared to 2000-Hz tonal stimuli. 6 Galambos et al. (1981) showed that large responses to brief-tone stimuli were obtained in adults at rates around 40 Hz and that responses could be detected at levels near threshold. They also found that low-frequency brief tones yielded larger responses than high-frequency brief tones, and that the amplitude of the response was larger for an alert subject compared to a subject who was drowsy or asleep. The researchers proposed that ASSRs were largest at this rate because the prominent peaks of the ABR and middle-latency response which occur at 25-ms intervals (i.e., 1/40 = .025 s = 25 ms) would overlap and add constructively; slower or faster rates would result in destructive interference and much smaller responses. If the 40-Hz response arises from the same neurons as the transient middle-latency response, then it should be possible to predict the ASSR from the superimposition of multiple overlapping transient responses; this theory is known as the "superimposition" theory (Galambos et al., 1981; Hari, Hamalainen & Joutsiniemi, 1989; Plourde, Stapells & Picton, 1991; Stapells, Costello, Galambos, & Makeig, 1988). Others have proposed the "intrinsic rate" theory that suggests that rapid stimulation evokes responses from neurons that are responsible for rhythmic activity and resonate at the frequency of stimulation (Basar, Rosen, Basar-Eroglu and Greitschus, 1987) to explain the discrepancies between actual ASSRs and those predicted by superimposition (Azzena, Conti, Santarelli, Ottaviani, Paludetti & Maurizi, 1995) and the presence of the response after the stimulus is turned off (Santarelli, Maurizi, Conti, Ottaviani, Paludetti, and Pettorossi, 1995). Gutschalk, Mase, Roth, Ille, Rupp, Hahnel, Picton and Scherg (1999) modelled the refractoriness of the nerve fibres (they assumed multiple neural generators) which led them to conclude that superimposition of transient responses was the most likely explanation for the 40-Hz response. Based on source analysis (Herdman et al., 2002; Johnson, Weinberg, Ribary, Cheyne & Ancill, 1988, Kuwada et al., 2002; Makela & Hari, 1987; Mauer & Doring, 1999) and 7 magnetoencephalographic (Ciulla, Takeda, Morabito, Endo, Kumagai & Xiao, 1996; Schoonhoven, Boden, Verbundt, de Munck, 2002) studies, the 40-Hz ASSR is thought to arise from neural generators in the brainstem, auditory cortices and thalamocortical neural circuits. Because the 40-Hz response was detectable at threshold levels, it was thought that this method had enormous potential for assessing hearing sensitivity in infants; in fact, clinical equipment to estimate hearing threshold in infants using the 40-Hz ASSR was designed and marketed as the first "objective hearing screener for infants". Unfortunately, shortly thereafter, it was shown that the 40-Hz response was very difficult to record in sleeping infants (Stapells, Galambos, Costello and Makeig, 1988; Suzuki & Kobayashi, 1984). This serves as a reminder that collecting sufficient data in the target population must precede the recommendation of a new technique for clinical use. Although interest in the 40-Hz ASSR and its clinical application waned at this point, some researchers continued to study ASSRs. Rickards and Clark (1984) found that ASSRs could be recorded at higher presentation rates but the response amplitudes were much smaller compared to the 40-Hz response. Cohen, Rickards and Clark (1991) showed, however, that the amplitude of the background E E G unrelated to the stimulus (i.e., noise) decreases as presentation rate increases, resulting in signal-to-noise ratios at higher rates that were similar to those observed for the 40-Hz response. Cohen et al. (1991) also discovered that ASSRs elicited by presentation rates greater than 70 Hz were unaffected by level of arousal, unlike the 40-Hz ASSR which is substantially attenuated in sleep (Cohen et al., 1991; Dobie & Wilson, 1998; Jerger, Chmiel, Frost & Coker, 1986; Linden, Campbell, Hamel & Picton, 1985; Plourde et al , 1991). The 80-Hz response is thought to be generated primarily by neurons in the brainstem (Herdman et al., 2002; Kuwada et al., 2002; Mauer & Doting, 1999), although small contributions from the 8 auditory cortices have also been proposed. These findings for the 80-Hz ASSR were pivotal in reviving the research community's interest in investigating the potential use of ASSRs to assess hearing in infants. Stimulus and recording parameters for multiple ASSRs The main advantage of the 80-Hz ASSR, compared to the brief-tone ABR, is that the presence of a response can be determined objectively, rather than by the subjective interpretation of replicated waveforms. An additional advantage of the "multiple" ASSR is that response detection is linked to the modulation rate of the carrier frequency, making it possible for responses elicited by different carrier frequencies modulated at different rates to be detected simultaneously (e.g., Lins & Picton, 1995; Lins et al., 1996). An example of four stimuli presented simultaneously to obtain multiple ASSR recordings are shown in Figure 1.2. Because of the steady-state nature of the response, it is most efficiently analysed in the frequency domain using a Fourier analyzer (Stapells et al., 1984) or Fast Fourier Transform (reviewed in Regan, 1989). The Fast Fourier Transform is the more common method and will be the focus of this discussion. The Fast Fourier Transform converts information in the time domain (i.e., changes in amplitude over time) to the frequency domain by generating a set of complex values that vary with frequency; these values can then be converted to amplitudes and phases and describe the energy for each of many frequency bins, including the stimulus frequency and surrounding frequency bins. An example of this conversion from the time domain to the frequency domain for an ASSR elicited by a single stimulus is shown in Figure 1.3. Figure 1.4 illustrates ASSRs elicited by four stimuli presented simultaneously; the responses are shown as polar plots and in the frequency domain as amplitude spectra. The transformation is implemented digitally using a Fast Fourier Transform (FFT) algorithm originally developed by Cooley and Tukey (1965). The 9 FFT calculates complex values for a set of frequency bins ranging from 0 to the Nyquist frequency [defined as half of the analog-to-digital (A/D) conversion rate ( Regan, 1982, p.25)] with a frequency resolution equal to the reciprocal of the duration of the recording sweep. For example, for an ASSR recorded using a 500-Hz A/D rate and a sweep duration of 16.384 s, the FFT would generate a set of values from 0-250 Hz with a resolution of 0.0610 Hz. The FFT also requires that the number of data points collected in the time domain be a power of two to optimize computation time. The number of data points and the A/D conversion rate determine the duration of the input buffer; the time it takes to fill the input buffer is equivalent to the duration of one epoch of data acquisition. A/D and D/A conversion must be synchronized so that the duration of the input and output buffer are the same. For example, if a digital-to-analog (D/A) conversion rate of 32,000 Hz and a D/A buffer of 32,768 points (John & Picton, 2000) are used, it is necessary to adjust the D/A factor relative to the A/D conversion rate that is selected (e.g., for a 1250-Hz A/D conversion rate, a D/A factor of 32 is required). Figures 1.2, 1.3 and 1.4 To prevent aliasing, the minimum A/D conversion rate must be at least twice that of the highest frequency (i.e., the "Nyquist frequency") that is contained in the EEG. Any frequency components in the E E G above the Nyquist frequency can alias, resulting in frequency components that were not in the original signal (Picton, Hink, Perez-Abalo, Linden & Wiens, 1984). For example, for a 500-Hz sampling rate, any energy greater than 250 Hz could potentially alias. The alias frequency can be predicted using the following calculation (National Instruments Lab VIEW ™ Measurement Manual, 2000): 10 Alias frequency = Absolute value (closest integer multiple of sampling frequency minus input frequency) For example, a 500-Hz tone that is amplitude-modulated at 77 Hz would have energy at 423, 500 and 577 Hz. If this energy is present in the E E G being digitized at 500 Hz, the alias frequency would be: (500 Hz - 423 Hz) = 77 Hz, which is the same as the modulation rate for this 500-Hz carrier frequency. As shown in Table 1.1, when standard audiometric frequencies (500, 1000, 2000 and 4000 Hz) are used as carrier frequencies to elicit ASSRs, this calculation predicts aliasing will be a potential problem for all of the carrier frequencies when using a 500-Hz A/D rate. Similarly, using a 1000-Hz A/D rate, aliasing will be a potential problem for 1000-, 2000-and 4000-Hz carrier frequencies but not for a 500-Hz carrier frequency. Table 1.1 It is well known that large amplitude electromagnetic artifact is present in the E E G when presenting bone-conduction stimuli at moderate to high levels for ABR recordings. Stimulus artifact in an ABR is easily distinguished from an actual response because it is visible early in the time-domain waveform. Alternating the stimulus polarity is a common technique used to remove or reduce stimulus artifact when recording ABRs (e.g., Hall, 1992, pg. 319). When stimulus polarity is alternated, the response polarity is unchanged but the stimulus artifact inverts in polarity; the inverted and non-inverted artifact effectively cancels itself out, or at least is greatly reduced. A significant problem with bone-conduction stimulus artifact in the EEG, when recording ASSRs, is that this energy can potentially alias to exactly the same frequency as the 11 ASSR modulation rate, and be interpreted as a response because the stimulus and response overlap in time, as shown in Figure 1.5. The top panel in Figure 1.5 clearly shows that large stimulus artifact is present in the unfiltered EEG and is not removed by a 250-Hz low-pass filter when multiple bone-conduction amplitude-modulated tones are presented. This is of particular concern for ASSRs recorded using relatively slow A/D conversion rates, such as 500 and 1000 Hz, which have commonly been used (e.g., Dimitrijevic et al., 2002; Herdman & Stapells, 2001, 2003). Selection of an A/D rate for which the carrier frequencies are not integer multiples, such as 1250 Hz, results in alias frequencies that do not coincide with the modulation rates of the carrier frequencies, thereby substantially reducing the potential for confusing stimulus artifact with an actual response. Dimitrijevic et al. (2002) used a 1000-Hz A/D rate and reported ASSRs elicited by bone-conduction stimuli presented at 1000-Hz that were not completely eliminated by masking; they suggested that their findings may be explained by stimulus artifact from the bone oscillator contributing to the ASSR but did not pursue the issue further. Dimitrijevic et al. (2002) and Lins et al. (1996) also found differences between amplitude/phase measures for bone-versus air-conduction ASSRs, particularly for 500- and 1000-Hz carrier frequencies, which may also be related to stimulus artifact in the ASSR. Concerns about stimulus artifact contributing to bone-conduction ASSRs were also raised after our first attempt to record ASSRs to bone-conduction stimuli. We found responses whose amplitudes were far too large, even for subjects with normal hearing; these responses were 5 to 15 times larger than those to air-conduction stimuli (Picton et al., 2003). The experiments in Chapters 2 and 3 were designed to investigate the possibility of artifactual responses when recording ASSRs using different E E G recording and stimulus parameters. 12 Figure 1.5 Analysis of multiple ASSRs There are a number of methods for analysing the amplitude and phase characteristics of ASSRs; however, only the method used in Chapters 2-7 will be discussed in detail. The ASSRs presented in these chapters were recorded using the Multiple Auditory Steady State Response system (MASTER), a Windows-based data acquisition system developed by John and Picton (2000b). MASTER simultaneously generates auditory stimuli designed to evoke ASSRs and acquires, averages and analyses the resultant E E G activity (response plus background noise). MASTER generates multiple sinusoidal tones, referred to as "carrier frequencies", that are amplitude (AM) and/or frequency (FM) modulated; these carrier frequencies are presented at different modulation rates in order to detect responses to different carrier frequencies simultaneously. For example, the bone-conduction stimuli in Chapters 2 and 3 were sinusoidal tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% A M at 77.148, 84.961, 92.773 and 100.586. Hz, respectively. However, Cohen et al. (1991) showed that combining A M and F M , or "mixed" modulation (MM), results in larger response amplitudes compared to either A M or F M alone using single stimuli. John, Dimitrijevic, van Roon & Picton (2001) showed that M M also results in larger ASSR amplitudes when multiple stimuli are presented simultaneously compared to A M or F M alone. To maximize the ASSR amplitude using M M , the phase of the F M component of the stimulus is adjusted so that the maximum frequency occurs just after the maximum amplitude (John et al., 2001). In Chapters 4-7, the sinusoidal tones were 100% A M plus 25% FM, and the phase of F M for each carrier frequency 13 was adjusted to optimize ASSR amplitudes. Frequency modulation was not used in the first two studies in order to use the least complex stimuli (i.e., A M only) while investigating the optimal E E G recording parameters and stimulus polarity. MASTER (John & Picton, 2000) identifies responses at the modulation rates for each of the carrier frequencies in the recorded E E G activity and calculates a noise estimate in the surrounding frequency bins where no stimulation occurred. The results of the FFT can be shown as amplitude spectra or presented graphically as a polar plot (Figures 1.3 and 1.4). On a polar plot, a vector extends from the origin; the length of the vector indicates the amplitude of the response, the angle of the vector from the x-axis determines the onset phase of the response and the circle centred on the end of the vector (i.e., the circle radius) represents the 95% confidence interval for the noise (Figure 1.3). If this circle does not include the origin, there is only a 5% chance that the response occurred by chance (i.e., that it is a false-positive). F-ratios are calculated that compare the baseline-to-peak amplitude of the energy at the modulation rates (i.e., response) to the baseline-to-peak amplitude of the energy within ±60 bins of the modulation rate (i.e., noise) to estimate the probability that the response amplitude is contained within the distribution of the amplitude of the noise. If the p value based on the F-statistic is less than .05, the response is considered to be significantly larger than the noise and, therefore, deemed to be present. To be considered absent, the p value must be greater than or equal to .05; however, an estimate of the E E G noise present in the recording must be also be considered. If E E G noise is large, a response may be judged as absent based on the p value, when in fact a response is present but is obscured by large background E E G noise. Very large E E G noise is automatically rejected and not included in the FFT analyses; in these studies, the artifact rejection level was set for amplitudes exceeding ± 40 uV. Lower level noise, which is not large enough to be rejected, can 14 also make it difficult to judge the presence or absence of a response. To handle this problem, it is possible to base the absence of a response on meeting a pre-determined noise criterion, in addition to a non-significant p value. The MASTER circle radius, which is updated for every sweep of data collected, can be used as an on-line estimate of E E G noise for each of the modulation rates. For all of the experiments in Chapters 2-7, unless otherwise specified, ASSRs with p>.05 were recorded for a minimum of seven sweeps until a circle radius of <20 nV (equivalent to a mean E E G noise < 11 nV) was obtained to ensure that the recording was quiet enough to judge a response as absent. An additional way to reduce the E E G noise in a recording is to use "weighted averaging" (John, Dimitrijevic & Picton, 2001), which was incorporated as an option in a later version of MASTER, as an alternate to standard averaging. Weighted averaging helps reduce recording time by assigning less weight to the epochs that contain greater E E G noise and more weight to the epochs that contain less E E G noise. Standard averaging was used for data acquisition in Chapters 2 and 3; weighted averaging was used in subsequent chapters. The components of an auditory evoked potential, for example, wave V of the ABR, are commonly described in terms of their amplitude and latency characteristics. In the time domain, peak latencies are linear in nature and reflect the transmission time for neural activity time in response to auditory stimuli. Similarly, measuring the onset phase of an ASSR relative to the onset phase of the stimulus yields a measure of phase delay which reflects neural transmission time. A notable difference between peak latency and phase measures is that phase data are circular; the conversion of a phase measure in degrees to a latency measure in milliseconds (ms) is not straightforward. The first step is to adjust the phase value of the response generated by MASTER to an "onset phase". Because the stimulus is modulated using a sine function, and the 15 amplitude spectrum for each response is calculated using as a cosine function, 90 degrees must be added to the MASTER phase to yield the actual onset phase (John & Picton, 2000b). Onset phase is then converted to "phase delay", a measure that resembles latency (i.e., phase delay becomes longer as the response becomes more delayed relative to the onset of the stimulus). Phase delay is calculated by subtracting onset phase from 360 degrees (Rodriguez et al., 1986). The conversion of phase delay in degrees to milliseconds depends on the modulation rate; the following formula describes this relationship: Phase delay (ms) = Phase delay (degrees)/ (360 x modulation rate) There are two ambiguities that arise with phase delay measures that relate to: (i) the circularity of the data, and (ii) the steady-state nature of the response. The possible range of phase delay values is 0-360 degrees. Phase delay values that occur within one cycle are not difficult to handle, for example, averaging 271, 269 and 275 degrees would result in an average phase 272 degrees. However, how would.the average phase be calculated for 359, 355 and 3 degrees be calculated? Without accounting for the circularity of the data, the average phase delay would be 239 degrees which clearly does not accurately represent the data. John and Picton (2000b) suggest "unwrapping" measurements that cross a cycle boundary, i.e, add one cycle (or 360 degrees) to the measurements that are more than 180 degrees lower than the adjacent measurements. So, for the example given, rather than an average phase delay of 239 degrees, the average phase delay would be (359 + 355 + 363)/3 = 359 degrees which is more sensible. In the following chapters, ASSR phase are expressed in terms of phase delay in degrees or phase delay converted to milliseconds where specified, and where appropriate, measurements were 16 unwrapped. The second ambiguity relates to the steady-state nature of the response; it is not possible to know how many cycles of the stimulus have occurred before the onset of the response because the response is cyclical. It is possible to derive an "apparent latency" or "group delay" but the measure is difficult to relate to physiologic delay (reviewed in John & Picton, 2002b) and will not be used in the following chapters. Bone-Conduction Mechanisms There are a number of theories about the underlying mechanisms of bone-conduction hearing in mature skulls, which have yet to be thoroughly examined, and very little is known about transmission of bone-conducted sound in infants. For many years, three osseous pathways ("inertial", "compressional" and "osseotympanic") were described to explain how vibratory energy was transmitted from the bone oscillator, situated firmly at the mastoid or forehead, to the cochlea. The "inertial" mode theorizes that there is an inertial lag set up between the ossicular chain and the bony shell of the cochlea when the skull vibrates that results in relative motion of the stapes footplate in the oval window; movement of the stapes footplate results in fluid displacement within the cochlea and stimulation of the basilar membrane, particularly for low-frequency stimuli (Barany, 1938; Kirikae, 1959; Wever & Lawrence, 1954). The "compressional" mode proposes that skull vibrations propagate as transverse waves to the temporal bone and cause distortion of the bony cochlear shell, thereby causing fluid displacement in and out of the oval window, particularly for high-frequency stimuli. The "osseotympanic" mode proposes that vibratory energy from the skull radiates into the ear canal and is transmitted to the cochlea by the same pathway as an air-conducted stimulus. Tonndorf (1968, 1976) also 17 hypothesized a "fluid" mode by which a minor amount of energy would travel along the soft tissues of the head and pass through the contents of the skull to act on the cochlea directly. Theoretically, the waves of energy from each source would then sum vectorially in the cochlea to produce a response. Recent human and animal studies have both challenged and added to these theories. Stenfelt and Goode (2005) identified five components that they considered most likely to be important contributors to bone-conduction hearing, and attempted to evaluate how much each one contributes to bone-conduction hearing in the normal ear. These five components include: Component 1: sound radiated in the ear canal as a bone-conduction stimulus (including the role of the mandible and the occlusion effect); Component 2: middle-ear cavity radiation and ossicle inertia; Component 3: inertia of the cochlea fluids; Component 4: compression of the cochlear walls; and Component 5: pressure transmission from the cerebrospinal fluid. Component 1 refers to the air-conducted sound in the external ear canal that arises when a bone oscillator presents a signal that vibrates the skull (comparable to osseotympanic pathway). Skull vibration produces motion relative to the surrounding air molecules, in particular, this motion deforms the ear canal wall and generates sound pressure that can be transmitted to the cochlea. Von Bekesy (1960) was the first to show that it was possible to cancel out a bone-conducted tone (400-Hz) by simultaneously presenting an air-conducted tone of the same frequency with appropriate amplitude and phase adjustments. Karma, Tonndorf and Queller (1976) also showed that an air-conducted tone could be subjectively cancelled by a bone-conducted tone, and specifically found that the radiated sound in the ear canal could be cancelled at frequencies below 700 Hz. Most studies agree that the sound pressure produced in the ear canal is greatest for low-frequency bone-conducted stimuli (<500-700 Hz) and is generated 18 primarily in the cartilaginous part of the ear canal (Huizing, 1960; Stenfelt, Wild, Hato & Goode, 2003 Tonndorf, 1966); little sound radiates at low frequencies in the bony part of the ear canal because the skull and the bony part move in unison with minimal resultant deformation of the bony canal wall (Hakansson, Brandt, Carlsson & Tjellstrom, 1994). Despite the evidence that shows that low-frequency sounds can be generated in the ear canal to a low-frequency bone-conducted stimulus, there are data that suggest that this "osseotympanic" pathway does not play a significant role in transmitting bone-conducted sound, at least in an open ear canal. For example, threshold data from patients with otosclerosis reveal significantly elevated air-conduction thresholds but normal bone-conduction thresholds, even though sound from the ear canal is significantly attenuated by the middle-ear system. Also, bone-conduction thresholds can be within normal limits in patients with large perforations in the tympanic membrane which would also affect transmission of energy from the ear canal (Brinkman, Marres and Tolk, 1965). More recently, Stenfelt et al. (2003) reported that the level of the air-radiated sound in the external ear canal is 10 dB below threshold for the bone-conducted stimulus, based on sound pressure and malleus umbo velocity measures, which is consistent with the idea that radiated sound is not contributing significantly to the bone-conducted stimulus. When the ear is occluded, the contribution of the sound pressure in the ear canal to a bone-conducted stimulus is thought to be more important (Huizing, 1960; Tonndorf, 1966). It is well known, at least for adults, that occluding the ear canal while estimating bone-conduction thresholds, significantly improves bone-conduction thresholds below 1000 Hz, a change known as the "occlusion effect" (Tonndorf, 1966). Huizing (1960) theorized that this improvement in low-frequency bone-conduction hearing resulted from a change in the resonance properties of the ear canal, i.e., an open tube vs a closed tube. Tonndorf (1966) suggested that the unoccluded ear 19 acts as a high-pass filter due to the mass effect of the air in the ear canal and the compliance of the air and the tympanic membrane, i.e., low-frequency energy is lost through the open ear. In an occluded ear, the high-pass filter is no longer in effect, thus the low frequencies (up to 1000 Hz) are enhanced and bone-conduction thresholds improve. This theory has proven to be correct for low frequencies and supports that the sound pressure in the ear canal, for an occluded ear, is a major contributor to bone-conducted hearing. Huizing's theory is probably correct at high frequencies but the effect is minor (Senfelt & Goode, 2005). Some of the earlier studies suggested that the mandibular joint could also play a minor role in transmitting vibrations from the skull via the "osseotympanic" pathway (Tonndorf, 1966; von Bekesy, 1960) because there is relative motion between the jaw and the skull above the resonant frequency of this joint, which is 110-180 Hz (Franke, von Gierke, Grossman & von Wittern, 1952; Howell, Williams & Dix, 1988). Several studies have shown no differences in ear canal sound pressure to a bone-conducted stimulus when the mandible was removed; therefore, the mandibular joint is unlikely to contribute (Allen & Fernandez, 1960; Howell & Williams, 1989; Stenfelt et al., 2003). Component 2 involves the middle ear. Skull vibrations can potentially act on the middle ear and contribute to bone-conduction hearing, either by generating sound radiation in the middle-ear cavity, or by acting on the ossicles. Groen (1962) suggested that skull vibrations generate sound pressure in the middle ear, which then acts on the tympanic membrane and transmits energy to the cochlea, resulting in a significant contribution to bone-conduction hearing at 2500 Hz. Tonndorf (1966) and Stenfelt, Hato and Goode (2002) studied the influence of the middle-ear cavity in cats and human temporal bones, respectively, and found no evidence to support Groen's theory. However, skull vibrations can affect ossicular inertia in a frequency-20 dependent manner based on the mechanical properties of both the ossicular chain, and the ligaments and tendons which connect the chain to the bony wall of the middle-ear cavity, tympanic membrane and oval window (Stenfelt & Goode, 2005). Barany (1938) found that increasing middle-ear inertia by adding mass to the tympanic membrane improved bone-conduction hearing to a 435 Hz tone and that changing the static pressure of the middle-ear cavity decreased bone-conduction sensitivity. Subsequent studies obtained the same results but determined, using a range of test frequencies, that these changes in bone-conduction hearing occur only below 2000 Hz (Brinkman et al., 1965; Huizing, 1960; Kirikae, 1959; Legouix & Tarab, 1959; Tonndorf, 1966). Stenfelt et al. (2002) found that the resonant frequencies of the stapes footplate (1500 Hz) and the malleus umbo (1800 Hz) were lowered when the mass at the tympanic membrane was increased. They concluded that the inertia of the ossicles affects bone-conduction hearing the most between 1000 and 3000 Hz. Despite this evidence, M0ller (2000) found that when ossicles were removed, the actual change in bone-conduction threshold was small. Similarly, Stenfelt and Goode (2002) found that fixation of the ossicles to the temporal bone, which eliminated the middle-ear pathway [component (ii)], actually improved bone-conduction ABR thresholds by 5-10 dB instead of elevating threshold. These studies support that the middle-ear component is not one of the main contributors to bone-conduction hearing. Component 3 concerns inertial forces arising from skull vibrations that can affect the movement of the fluid in the cochlea (comparable to the "inertial" mode discussed earlier). Cochlear fluids are incompressible but movement is possible at the oval and round windows. A pressure gradient exists between the two windows which results in fluid movement between the scala vestibuli and scala tympani and the generation of a travelling wave across the basilar 21 membrane. In otosclerotic ears with stapes fixation, there are normal bone-conduction thresholds at most frequencies and no more than a 15-25 dB loss at 2000 Hz, despite greater hearing loss via air conduction (Carhart, 1962), which suggests that the bone-conduction stimuli is not transmitted through the middle-ear system, but to the cochlea directly. A third avenue for fluid movement is present and is known as the "third window". The third window is comprised of compliant structures, including: the cochlear and vestibular aqueducts, nerve fibres, veins, amd micro channels that enter the cochlea (Kucuk, Abe, Ushiki, Inuyama, Fukuda & Ishikawa, 1991; Stenfelt & Goode, 2005). The presence of a third window is supported by Yoshida and Uemura's (1991) finding that changes in static pressure in cerebrospinal fluid are transmitted to the cochlea through the cochlear aqueduct. Also, Groen, Rosowski and Merchant (1997) found patent cochlear aqueducts in 93% of the human temporal bones they examined, which provides anatomical evidence for such a mechanism. It was determined that the third window only contributes when the stimulus is presented by bone-conduction and not via air-conduction (Stenfelt & Goode, 2005; Tonndorf, 1966). Stenfelt and Goode (2005) suggest that inertia of the cochlear fluid in a normal ear is most likely to be the largest contributor to bone-conduction hearing, particularly at frequencies below 1000 Hz. Component 4, compression of the cochlear walls resulting from skull vibration, may also influence the movement of cochlear fluid and contribute to bone-conducted sound (von Bekesy, 1932; Tonndorf, 1966). Comparable to the "compressional" mode discussed above, this component involves compression and expansion of the cochlear wall when a transverse wave produced by the bone-conducted sound vibrates the skull. When the cochlear walls are compressed, the oval and round windows bulge outward; because the round window is more compliant than the oval window, and the scala vestibuli has a larger volume than the scala 22 tympani, the pressure gradient shifts from the scala vestibuli to the scala tympani. This idea is not likely to be a major contributor at low and mid-frequencies, however, based on what we know about otosclerotic ears. With stapes fixation, there should be greater bulging of the round window and a greater shift in pressure towards the scala tympani, which should result in better bone-conduction hearing; this is certainly not the case for otosclerotic ears (Carhart, 1962). Compression of the cochlear wall is, therefore, not thought to be a significant contributor to bone-conduction sensitivity below 4000 Hz (Stenfelt & Goode, 2005). The fifth component (comparable to the "fluid" mode suggested by Tonndorf) involves transmission of pressure from the cerebrospinal fluid to the cochlear fluid, probably via the cochlear aqueduct as discussed earlier (Yoshida & Uemura, 1991). Seaman (1991) reported that changes in intracranial pressure in amphibians can be transmitted to the inner ear and result in stimulation of sensory receptors. Sohmer et al. (2000) found similar results when they measured click-ABR thresholds to bone-conduction stimuli presented at the forehead, mastoid and directly on the fontanelle of young infants. They found that ABR thresholds were similar whether they were recorded to stimuli presented at the fontanelle or on the bone next to the fontanelle which suggests that there is a major "fluid" pathway in bone-conduction hearing that results from skull vibrations communicating directly with cochlear fluids via the cochlear/vestibular aqueducts and perineural and perivascular spaces. They also found that ABR thresholds in neonates were better at the temporal bone where the skull was thinner compared to the frontal bone where the skull was thicker. In addition, they found that the magnitude of the vibration decreased as the distance from the oscillator location on the frontal bone increased. Sohmer et al. (2000) suggest that the vibratory energy penetrates the skull at the location of the oscillator and decreases as the distance from the source increases, at least in the skull of a neonate. In rats, Freeman, Sichel and Sohmer 23 (2000) found that bone-conduction ABR thresholds to clicks were 15 dB better when the skull was intact compared to rats with craniotomies. In intact skulls, a decrease in cerebral spinal fluid increased the ABR thresholds by 7 dB in rats, which further supports that the amount of fluid contained within the skull affects bone-conduction thresholds. These results also suggest that the difference in volume of the skull contents between infants and adults would affect the amount of acoustic energy that reaches the cochlea and affect absolute thresholds. It is not clear, however, how differences in skull structure and fluid volume interact in the transfer of bone-conduction energy in infants and adults. These results provide compelling evidence for a significant fluid pathway, in addition to the osseous modes, for bone-conduction hearing. One point of agreement regarding the mechanisms of bone-conduction hearing for a mature skull is that there is likely to be more than one factor contributing to bone-conduction hearing. As mentioned earlier, very little is known about the bone-conduction mechanisms in infants; however, we do know that the infant skull undergoes substantial changes in size and structure during the postnatal years. The infant skull is much smaller at birth compared to an adult skull and the cranial bones are not fused as they are in the adult skull. The cranial bones grow and are joined by sutures that are complete at approximately one year (Anson & Donaldson, 1981a). Histologic and radiographic data indicate that the mastoid length, width and depth increase rapidly from birth to two years of age (Eby & Nadol, 1986). The mechanical properties of solid bone, sutures and membranous connections between cranial bones, such as the fontanelle, have very different biomechanical properties. Sohmer, Freeman, Geal-Dor, Adelman, & Savion (2000) found that acceleration of vibratory energy across the fontanelle was 14 dB less than across the temporal bone in infants. These results strongly suggest that the infant skull, in the first two years of life, is certain to have different mechanical properties from that of an older 24 child or adult, and it is probable that these maturation differences would affect an infant's sensitivity to bone-conducted sound. Bone-Conduction Threshold Estimation in Infants It is a relatively straightforward procedure to deliver a calibrated air-conduction stimulus to an infant or adult via earphones, although the sound pressure level at the tympanic membrane may vary (e.g., Ranee & Tomlin, 2006; Sininger, Abdala & Cone-Wesson, 1997, Voss & Hermann, 2005). For bone-conduction testing, however, producing a predictable output of energy at the skull, and ultimately to the cochlea, is more complicated because of a number of procedural factors, including: (i) oscillator coupling force and method (steel headband, elastic-band or hand-held), (ii) oscillator placement location on the head (temporal or frontal bone), and (iii) whether the bone-conduction testing is performed with ears occluded or unoccluded. Procedures are relatively standardized for estimating bone-conduction behavioural thresholds in adults. In contrast, there are no standardized methods for bone-conduction testing in infants, only recommended "best practices", many of which are based on assumptions rather than systematic investigation. Because the bone oscillator sits on the surface of the skull, a sufficient and constant force is required to couple it to the head and to produce a calibrated output from the transducer. Amount of force and coupling method are, therefore, important issues. Also, the bone oscillator can be positioned in different locations on the skull, potentially affecting the intensity of the signal reaching the cochlea (Stuart, Yang & Stenstrom, 1990; Yang et al., 1987). It is also well known, at least for adults, that occluding the ear canal while estimating bone-conduction thresholds significantly improves bone-conduction thresholds in the low frequencies, as 25 discussed above. For bone-conduction behavioural testing in adults, the bone oscillator is positioned on the head using a steel headband which applies a constant force to the oscillator, although the exact amount of force applied (approximately 400-500 g) will vary depending on individual head size. To accommodate the smaller head size for older children, a smaller steel headband is used; the force generated is less than for larger adult heads and will also vary somewhat with the size of the child's head (Harrell, 2002; Wilber, 1979a). The coupling force is limited to some degree by patient comfort; however, behavioural thresholds to bone-conduction stimuli in older children and adults can usually be obtained in approximately 10-15 minutes with only minor discomfort from the steel headband. Estimation of bone-conduction hearing thresholds in young sleeping infants using auditory evoked potentials poses unique challenges and the testing methods are not standardized. Infants have smaller heads than adults, precluding the use of the standard adult or child steel headband. Also, infants must remain asleep during testing; disturbing the infant as little as possible while positioning the oscillator is critical, as is minimizing the discomfort during a potentially longer testing time. One concern for bone-conduction testing is that the amount of force applied to the oscillator is consistent. Coupling the oscillator to the infant's head using an elastic band is currently the clinical method suggested by many because a known force can be applied to the elastic band and the amount of force verified using a spring scale (Yang & Stuart, 1990). It is important to keep in mind, however, that in many clinical settings the verification step is often omitted. The aim when testing sleeping infants is always to use sleep time efficiently and not to wake up the infant before the testing is complete. Holding the oscillator in place by hand is also commonly done clinically because it is faster and is far less likely to wake 26 up the infant than positioning an elastic band (which requires more manipulation of the infant's head). The hand-held method is also more comfortable for the infant because it can be removed and replaced easily between test conditions. Despite the practical advantages of the hand-held method, its use has been discouraged (e.g., Yang & Stuart, 2000; Yang, Stuart, Stenstrom, & Hollett, 1991). This preference for one coupling method over the other, however, is based on assumptions. The two main assumptions arguing against the hand-held method are: (i) there is potential for the applied force to vary during testing, resulting in an inconsistent output from the transducer (Yang et al., 1991), poorer thresholds and greater variability in thresholds, and (ii) pressing down on the superior surface of the oscillator by hand, i.e., mass loading it, will dampen the response characteristics of the bone oscillator (Wilber, 1979b). However, bone-conduction responses in infants have never been compared using the elastic-band and hand-held methods. There are adult calibration values for bone-conduction stimuli for different bone-oscillator placement locations on the skull (i.e., mastoid and forehead); on average, adult behavioural thresholds measured at the forehead are higher than at the mastoid by 14, 8.5, 11.5, and 8.0 dB at 500, 1000, 2000, and 4000 Hz, respectively (ANSI, 1996). Bone-oscillator placement location has also' been raised as an issue (Stuart, Yang & Strenstom, 1990; Yang & Stuart, 2000) when estimating bone-conduction thresholds in sleeping infants because of the potential for thresholds to differ depending on where the oscillator is placed on the temporal (mastoid vs upper region) or frontal bone. When an elastic band is used to couple the oscillator to the skull, it is typically placed on the infant's temporal bone, posterior to the upper portion of the pinna so that the elastic band does not cover the infant's eyes or slide off the head; when the hand-held method is used, oscillator placement is not restricted to one area on the temporal bone. 27 There are some auditory brainstem response (ABR) data, albeit incomplete, which have been interpreted to suggest that bone oscillator location does affect the response in infants. Yang, Rupert and Moushegian (1987) investigated the effect of oscillator placement (frontal, occipital, and temporal) on wave V ABR latencies to bone-conduction clicks for neonates, 1 -year-old infants and adults. Latency results were found to vary with both age of subject and oscillator placement. The temporal bone yielded significantly shorter latencies than either the occipital or frontal placements in the neonatal group and 1-year-old group (Yang et al., 1987). Based on these latency differences, Yang et al. (1987) estimated that signal attenuation in the neonates from the temporal to frontal placement ranged from 30 to 35 dB. In the 1-year-old group, signal attenuation between the frontal and temporal placements was estimated to be between 20-25 dB. Attenuation from occipital to temporal was 15-20 dB for both infant groups. In contrast, in adults, the latency differences were not significant, and the attenuation between oscillator placement was judged to be no more than 5-10 dB (Yang et al., 1987). This attenuation estimate in adults is smaller than the ANSI (1996) standards for mastoid and forehead placements discussed above, and the changes in signal attenuation as a function of location suggested by Yang et al. (1987) must be interpreted with caution for several reasons. First, they used click stimuli rather than frequency-specific stimuli, so they were unable to separate out any frequency-dependent effects of oscillator placement on attenuation. Second, they did not directly estimate ABR thresholds for the different placements or report ABR amplitudes, which are more related to threshold than latency measures. Third, the attenuation estimates were derived using wave V latency-intensity functions; inaccurate attenuation estimates may result because click-evoked wave V latency is not linearly related to intensity (Picton, Stapells, & Campbell, 1981). 28 Stuart, Yang, and Stenstrom (1990) further investigated the effect of oscillator placement on wave V latencies in infants by comparing different areas on the temporal bone (superior, supero-posterior and posterior placements). They concluded that changing the location of the oscillator on the temporal bone produced significantly different wave V latencies in the neonate at both 15 and 30 dB nHL. Specifically, the posterior position (similar to the "mastoid placement" in Chapter 4) yielded the shortest wave V latency, whereas the superior position yielded the longest latency. The supero-posterior position (similar to the "temporal placement" in Chapter 4) yielded intermediate latency values. These latency differences were attributed to reflect greater signal attenuation as a function of distance from the cochlea (Stuart et al., 1990). Specifically, it was suggested that signal intensity reduces as one moves the bone oscillator farther away from the cochlea. Again, these findings were based solely on click-ABR latency data, and the same limitations that are discussed above apply. Neither of these two placement studies compared bone-conduction threshold in infants at different oscillator placements on the skull, which is the important measure for clinical applications. Correcting for the occlusion effect when estimating bone-conduction with occluded ears is also well described in adults ( Dirks & Swindeman, 1967 Elpern & Naunton, 1963; Hodgson & Tillman, 1966; Small & Stapells, 2003). For insert earphones, behavioural thresholds to brief tones improve by 3-5 dB at 500-1000 Hz (Small & Stapells, 2003), whereas pure-tone behavioural thresholds in the low frequencies (250-1000 Hz) improve by as much as 17 dB, depending on the insertion depth (Dean & Martin, 2000). As discussed earlier, any test protocols used to estimate thresholds in sleeping infants must be designed to minimize the possibility of waking the infant. Air-conduction thresholds are typically assessed using insert earphones, followed by bone-conduction testing (Stapells, 2000a). Should the insert earphones be removed 29 before assessing bone-conduction thresholds? It is important to know whether occluding the ears increases, decreases or has no effect on bone-conduction thresholds in infants. There are no studies that have investigated the occlusion effect in infants. Maturation of Bone-Conduction Hearing There are no psychoacoustic studies that have investigated the development of bone-conduction absolute thresholds. The little that is known about maturation of bone-conduction hearing is taken from anatomic and physiologic studies. The acoustic energy that originates from the bone oscillator is transformed by the properties of the skull and its contents, and possibly the outer- and middle-ear structures, before it reaches the cochlea; bone-conduction thresholds are likely influenced by developmental changes in these structures. Comparisons across infant studies of ABR thresholds to bone-conduction stimuli are less consistent than air-conduction results. Cornacchia, Martini, and Morra (1983) reported no difference in bone-conduction click-ABR thresholds between toddlers (16-20 months of age) and adults, whereas Stuart et al. (1993, 1994) found bone-conduction click-ABR thresholds in adults that were 17 dB higher (i.e., poorer) compared to infants (aged 0-96 days). Of more interest, in terms of frequency-specific threshold information, are the bone-conduction ABR thresholds to brief tones in infants. As shown in Table 1.2, bone-conduction ABR thresholds to 500-Hz brief tones are better in infants than adults; for 5-6-month-old infants, this difference is 2-7 dB. Cone-Wesson & Ramirez (1997) reported 500-Hz ABR thresholds for neonates that are 20-25 dB better than ABR thresholds for older infants (Foxe & Stapells, 1993; Stapells & Ruben, 1989) and 16-27 dB better than adults (Cone-Wesson & Ramirez , 1997; Foxe & Stapells, 1993). Foxe and Stapells (1993) and Stapells and Ruben (1989) reported 2000-Hz ABR thresholds for 6-30 month-old infants that were 2-4 dB poorer than adults, whereas, Cone-Wesson and Ramirez (1997) found that 4000-Hz ABR thresholds were 5 dB better in neonates compared to adults. There are no data for bone-conduction ABR data at 1000 Hz There are also frequency-dependent differences in bone-conduction ABR thresholds that are different for infants compared to adults. Bone-conduction ABR thresholds (in dB nHL) to 500-Hz brief tones are better than for 2000-Hz brief tones in infants.(Foxe & Stapells, 1993; Stapells & Ruben, 1989). In adults, the opposite is observed: bone-conduction ABR thresholds (in dB nHL) to 2000-Hz brief tones are better than thresholds to 500-Hz brief tones (Foxe & Stapells, 1993). Figure 1.2 The pattern of amplitudes and latencies of wave V for bone-conduction ABRs in infants as a function of frequency are also different from adults. Infant low-frequency bone-conduction ABR amplitudes are larger than those to high frequencies (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989). Direct comparison of ABR amplitude differences between infants and adults cannot be made because of the additional contribution of the 40-Hz response to the adult ABR (when using a 40/second rate) that is not present in the infant ABR (Foxe & Stapells, 1993). There are also infant-adult differences in bone-conduction ABR wave latencies; infants have only slightly longer wave V latencies for 2000-Hz brief tones compared to adults but much shorter wave V latencies (1.02 ms shorter) for 500-Hz brief tones compared to adults (Foxe & Stapells, 1993; Nousak & Stapells, 1992). These results suggest that a bone-conducted stimulus may be more effective at 500 Hz, but not at 2000 Hz, for an infant compared to an adult. 31 Why are ABRs to low-frequency bone-conduction stimuli better compared to those to high-frequency stimuli for infants, and when do bone-conduction thresholds become adult-like? The possibility that bone-conducted 500-Hz stimuli are less frequency-specific than 2000-Hz stimuli in infants as an explanation for why thresholds are better at 500 Hz was ruled out by high-pass noise/derived-band studies (Kramer, 1992; Nousak & Stapells, 1992). Another theory to explain better bone-conduction thresholds to 500-Hz stimuli in infants is that the bone-conducted stimuli at this frequency are more effective in infants due to the structure of an immature skull. Infant-adult wave V latency differences support this theory. The absolute wave V latencies are significantly shorter and the amplitudes are larger for 500-Hz bone-conduction brief tones compared to 2000-Hz brief tones in infants up to at least one year of age (Foxe & Stapells, 1993; Stapells & Ruben, 1989). Based on expected latencies differences in infant and adult air-conduction ABR thresholds, Foxe and Stapells (1993) predicted that 500-Hz bone-conduction tones were as much as 17 dB more effective for infants. The anatomy of the infant skull also supports this theory. Infants have a smaller temporal bone which is not yet fused to the other bones of the cranium. The sutures between the bony plates could essentially limit the dispersion of vibratory energy from the bone oscillator to the infant's smaller temporal bone and stimulate the cochlea, which is full size and largely adult-like in function. The end result would be a more effective stimulus; however, this theory does not address why high-frequency stimuli would not also be more effective. Cone-Wesson and Ramirez (2001) suggested an alternate explanation. They found that bone-conduction stimuli transmit 5-21 dB SPL more acoustic energy for 500-Hz stimuli into the external ear canal (i.e., the osseotympanic pathway for a bone-conducted signal) when presented to an infant's head than when presented to an adult's head. They hypothesized that 32 bone-conduction hearing is enhanced via this pathway by 5-21 dB at this frequency and that better bone-conduction thresholds can be explained by the overall level difference of the stimulus presented to the infant ear canal. The 9-17 dB more effective stimulus (Foxe & Stapells, 1993) is in the same range as the 5-21 dB boost in energy calculated by Cone-Wesson and Ramirez. This hypothesis has not been investigated further; however, it is possible to test this theory using bone-conduction ABR threshold data for infants with conductive hearing losses. If an infant has a conductive loss, any low-frequency bone-conducted energy that is transmitted via air conduction will be blocked by the abnormal middle ear and cannot enhance the effectiveness of low-frequency bone-conducted sound at the cochlea. Stapells and Ruben (1989) reported better bone-conduction thresholds at 500 Hz in infants compared to adults both for infants with normal hearing and for those with conductive losses; therefore, Cone-Wesson and Ramirez's hypothesis must be incorrect. It appears that an air-conduction pathway for bone-conducted sound is not a significant contributor to bone-conduction hearing in infants, at least in unoccluded ears, and that a different mechanism must explain the greater effectiveness of bone-conducted stimuli in the low frequencies. Based on the limited bone-conduction ABR data, it is appears that there are maturational differences in bone-conduction hearing sensitivity at 500, 2000 and possibly at 4000 Hz. Further data are needed across a range of frequencies and for infants of different ages to understand the frequency-dependent changes that occur in the post-natal period and to establish the age at which bone-conduction hearing thresholds become adult-like. ASSRs are well suited to this task because they can be recorded to multiple simultaneous stimuli in sleeping infants. Maturational changes are also seen in ABRs recorded in the EEG channels ipsilateral and contralateral to air- (Edwards, Durieux-Smifh & Picton, 1985; McPherson, Hirasugi, & Starr, 33 1985; Salamy, Eldredge & Wakeley, 1985 Stapells & Mosseri, 1991) and bone-conduction stimuli (Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stuart et al., 1996). A number of studies in young infants using monaural air-conducted click stimuli (i.e., one cochlea stimulated) have shown that the morphology of the ABR waveform in the E E G channel contralateral to the stimulated ear differs from that in the E E G channel ipsilateral to the stimulated ear (Edwards, Durieux-Smith & Picton, 1985; Furune, Watanabe, Negoro, Yamamoto, Aso & Takesu, 1985; McPherson et al., 1985; Stapells & Mosseri, 1991). In young infants, the ABR waveform in the ipsilateral E E G channel has recognizable waves I, III and V; yet, in the contralateral E E G channel, the peaks in the ABR waveform differ in polarity, amplitude and latency to the degree that a new labelling scheme using the letters A through E was proposed (Edward et al., 1985). Stapells & Mosseri (1991) reported that the peak-to-peak amplitudes of A-B and D-E' in the contralateral E E G channel are smaller than the I'-III and V - V in the ipsilateral channel for infants up to 18 months of age, which is consistent with the results reported earlier for neonates by Edwards et al. (1985). Edwards et al. also reported that only 35% of neonates had contralateral responses to 30 dB nHL air-conducted clicks. Interaural attenuation in adults for bone-conducted stimuli is 10 dB on average at 250-4000 Hz (Nolan & Lyon, 1981). For air-conducted stimuli, interaural attenuation is 40-50 dB on average at 250-4000 Hz for supra-aural earphones (Goldstein & Newman, 1994, p. 117), and 69-94 dB on average at 500-4000 Hz for insert earphones with deep insertion: (Sklare & Denenberg, 1987). Because interaural attenuation in adults is much less for bone- vs air-conducted stimuli, it is difficult to determine how much each cochlea contributes to the resulting ABR to a bone-conducted stimul. Previous estimates of attenuation of bone-conducted signals reported by Yang et al. (1987) for infants up to one year of age suggest that a bone-conducted signal is attenuated 34 approximately 15-35 dB by an infant skull based on differences in click-ABR wave V latencies when the bone oscillator is placed in different locations on the infant skull. For infants, a bone-conducted signal may or may not cross over to stimulate the opposite cochlea depending on the stimulus presentation level. Presenting masking noise to the ear contralateral to the test ear is often used to isolate the test ear; however, masking may not be practical when assessing infants because of the extra time required for testing, the difficulties of earphone placement and the uncertainty about how much masking noise to use. Previous studies have shown that two-channel recordings (i.e., E E G channels ipsilateral and contralateral to stimulus ear) of the bone-conduction ABR also show maturational differences (Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stuart, Yang & Botea, 1996). The top left graph in Figure 1.6 shows that contralateral ABR results to brief-tone bone-conduction stimuli in infants have longer wave V latencies for 500- and 2000-Hz stimuli compared to ipsilateral recordings. The bottom graph in Figure 1.6 indicates that the amplitude of the contralateral wave V is also much smaller in amplitude in infants for 500 Hz compared to the ipsilateral wave V; the same trend is seen for 2000 Hz; however, at 2000 Hz, there is much less difference between the amplitude of the contralateral and ipsilateral wave V (Foxe & Stapells, 1993; Stapells & Ruben, 1989). In contrast, adults have similar wave V amplitudes and latencies at 500 and 2000 Hz. These latency and amplitude asymmetries in infants are present at all intensity levels and persist beyond one year of age. The asymmetries in the ipsilateral and contralateral bone-conduction recordings provide valuable ear-specific diagnostic information based on the assumption that the cochlea closest to the channel which produces the largest and earliest wave V is the cochlea that contributed most to the ABR (Stapells & Ruben, 1989). Because the ipsilateral/contralateral asymmetries are robust at low levels, they can be used to 35 assess the laterality of the response concurrently with threshold estimation in young infants'. There is only one published study that has recorded two-channel ASSRs to air- conducted stimuli (van der Reijden, Mens & Snik, 2005) and there are no published studies that have recorded two-channel (ipsilateral/contralateral) ASSRs to bone-conduction stimuli. Figure 1.6 Potential clinical application of bone-conduction ASSRs Accurate and efficient techniques are needed to assess air-conduction and bone-conduction hearing thresholds in very young infants who have been identified as being at risk for hearing loss. There is considerable interest in implementing multiple ASSRs as a clinical tool because they are objective, frequency-specific and potentially more efficient than current assessment tools, such as the brief-tone ABR (Stapells, 2000a). There are only a few studies that have recorded bone-conduction ASSRs in adults with normal hearing and no studies done in infants. No study has directly investigated stimulus artifact issues when recording bone-conduction or high-intensity air-conduction ASSRs. It is critical that this technique be systematically investigated. Assessing adults using multiple bone-conduction ASSRs allows a direct comparison to bone-conduction behavioural thresholds that is not possible in young 1 The presence of wave I for a click-ABR can also be used to determine which cochlea is responding to the stimulus; however, wave I can be small and difficult to detect in infants with normal hearing and disappears if hearing loss is present (Jahrsdoerfer et al., 1985). Electrocochleography (NI) using a transtympanic or tympanic electrode can also identify which cochlea is responding (e.g., Ferraro & Ferguson, 1989); the difficulty with this technique is that infants can only be tested using this method under general anaesthesia. 36 infants. The next step is to collect bone-conduction ASSR data in infants with normal hearing across a range of ages, with the goal of establishing "normal levels" for bone-conduction ASSRs in infants of different ages. It would also be beneficial to determine whether two-channel ASSR recordings could be used to determine the cochlea of origin when assessing bone-conduction hearing in young infants. The overall goal of this dissertation was to use multiple ASSRs to estimate bone-conduction thresholds in infants and adults with normal hearing, and to investigate the maturation of bone-conduction hearing sensitivity. To accomplish this goal, six studies were completed. The objectives and hypotheses tested for each of these studies are as follows: [Study 1] Objective: To assess the possibility of spurious ASSRs to high-intensity air- and bone-conduction stimuli in subjects with severe-to-profound sensorineural hearing loss who cannot hear the stimulus. Hypothesis: Artifactual ASSRs will result if there is large-amplitude stimulus artifact in the E E G and the A/D conversion rate is an integer multiple of the carrier frequencies [Chapter 2]. [Study 2] Objective: To compare multiple ASSR amplitude and phase characteristics for air- and bone-conduction stimuli using single- and alternated stimulus polarities, and to determine ASSR thresholds to bone-conduction stimuli in adult participants with normal hearing. Hypothesis: There will be no differences in amplitude, phase or threshold for multiple ASSRs elicited by single- or alternated-polarity bone-conducted stimuli, and bone-conduction ASSRs will have similar characteristics to ASSRs elicited to diotic air-conducted stimuli [Chapter 3]. 37 [Study 3] Objective: To investigate the effects of testing methods used to assess bone-conduction hearing in infants: elastic-band vs. hand-held bone-oscillator coupling methods; oscillator placement location (temporal, mastoid and forehead: infant ASSRs); occluded vs. unoccluded ears (infant ASSRs). Hypothesis: Bone-oscillator coupling method will not affect bone-conduction thresholds (adult or infant); bone-conduction ASSR thresholds and response characteristics will not differ for the temporal and mastoid locations but ASSRs will be smaller and ASSR thresholds will be elevated for the forehead location; bone-conduction ASSR thresholds will be better in the low frequencies with ears occluded compared to ears unoccluded, similar to adults [Chapter 4]. [Study 4] (iv) Objective: To investigate bone-conduction ASSR thresholds in infants from a Neonatal Intensive Care Unit and in young infants (0-8 months) with normal hearing and to compare these to adult bone-conduction ASSR thresholds. Hypothesis: There will be infant-adult differences in bone-conduction ASSR thresholds due to the differences in the size and structure of the infant and adult skull [Chapter 5]. [Study 5] Objective: To compare multiple ASSR thresholds to bone-conducted stimuli in infants of different ages (0-11 months; 12-24 months) with adults, all with normal hearing, to investigate the time course of maturation of bone-conduction thresholds across frequency, and to establish preliminary "normal levels" for bone-conduction ASSRs in young and older infants. Hypothesis: Infant bone-conduction thresholds increase in the low frequencies and decrease in the high frequencies as they mature, and infant-adult differences are greater for young infants compared to older infants. [Chapter 6]. 38 [Study 6] Objective: To determine whether ipsilateral/contralateral asymmetries are present in the air- and bone-conduction "brainstem" (80-Hz) multiple ASSRs of infants. Hypothesis: Ipsilateral/contralateral asymmetries are present in air- and bone-conduction 80-Hz multiple ASSRs, and the asymmetries in the bone-conduction ASSRs can be used to isolate which cochlea is the primary contributor to the response [Chapter 7]. The major findings of this dissertation and their clinical significance will be summarized and discussed in Chapter 8. Future directions for investigating bone-conduction ASSRs in infants will also be outlined in the concluding chapter. 39 Table 1.1 Aliasing is expected at the modulation frequencies for the specified carrier frequencies, assuming the A/D converter sees energy originating from stimulus artifact above the Nyquist frequency. Carrier frequency (Hz) 500 1000 2000 4000 500-Hz A/D yes yes yes yes 1000-Hz A/D no yes yes yes 1250-Hz A/D no no no no 40 Table 1.2 A comparison across studies of brief-tone ABR thresholds in dB re: 1 pN to bone-conducted stimuli in infants and adults. Stimulus frequency (Hz) Study Age n 500 1000 2000 4000 Foxe & Stapells (1993) Adults 13 79 59 -Cone-Wesson & Ramirez (1997) Adults 3 68 - 58 Foxe & Stapells (1993) 5 months 8-9 77 - . 63 -Stapells & Ruben (1989) 6 months 28 72 61 -Cone-Wesson & Ramirez (1997) 2 days 22 52 - 53 41 Figure 1.1 Examples of transient-evoked potentials and how they are related to steady-state potentials. In A, with a modulation frequency of 80 Hz, there are brainstem responses occurring every 12.5 ms, the same period as the modulated (tone) stimulus. In B, with a stimulus modulation frequency of 40 Hz, the responses occur every 25 ms, the same period as the stimulus modulation. (Figure provided by Andrew Dimitrijevic) A B t r ans i en t steady-state f ' f 80 Hz 40 Hz 42 Figure 1.2 Stimuli used to obtain multiple ASSRs. The left side of the figure represents the acoustic waveforms seen in the time domain over the first 30 ms of the stimulus buffer. Each carrier frequency is amplitude modulated by its signature modulation frequency. The four different signals are then added together to form the combined stimulus that is presented to the subject. On the right of the figure are shown the amplitude spectra for these signals (measured over the full stimulus buffer) and displayed between 0-5000 Hz. For each stimulus, the spectrum shows energy at the carrier frequency and at two side bands separated from the carrier frequency by the modulation frequency. The spectrum of the combined stimulus represents the sum of the four different amplitude spectra. (Reprinted with permission from Picton, Durieux-Smith, Champagne, Whittingham, Moran, Giguere & Beauregard, 1998) Carrier Modulation Time Frequency 500 30.9 2000 96.9 Combined Stimulus MW IOOO 88 .9 i^iP^^rJifH^—^ffl^ 4000 104.8 30 ms 0 5000 Hz 43 Figure 1.3 This figure respresents the ASSR of one subject to a 1000-Hz 60 dB SPL carrier tone modulated at a rate of 91 Hz an with a depth of 100%. The response was recorded using a sweep of 8192 ponts lastinf for 11.3 s and the sweep wsa averaged over 64 replications. The full sweep was divided in 16 sections each containg th response to 64 cycles of stimulation.The T2 95% confidence limit was calculated using the X and 7 FFT components of these 16 sections of the response sweep. The polar plot in the upper right of the figure represents these results. The point at the centre of the circle is the vector average of the 16 responses. The average amplitude of the response is the distance between this point and the origin. The phase of the response is the angle formed between the line joining the origin to the average and the positive X axis. The circle indicates the T2 95% confidence limit for the average response. As the confidence limit does not include the origin, a response is considered to be present. The response waveform was calculated for four cycles of modulation by collapsing each of the 16 sections of the recording down to four cycles. On the left are the 16 replicated waveforms superimposed and in the centre is shown the average of these wavforms. The FFT of the complete 8192-point sweep is shown at the bottom of the figure. Responses can be identified as significantly larger than the adjacent frequency at the fundamental frequency (/0) stimulation (91 Hz) and at the second (2 JO) and third (3J0) harmonics. The FFT-/ 95% confidence limit was calculated using the FFT amplitudes at 60 points at each side of the response. (Reprinted with permission from Lins, Picton, Picton, Champagne & Durieux-Smith, 1995) Intensity: 80 dB SPL Carrier: 1000 Hz Modulation: 91 Hz Amplitude. 0.259 j t V Phase: 65 deg Confidence limits (p<0.05) T : 0.021 |iV T 2 : 0.016 j i V 0.5 jiV s fo 44 m s 2fo 0.2 p-V n 91 Hz 44 Figure 1.4 Auditory steady-state responses to four simultaneous bone-conducted stimuli (40 dB HL) with different carrier frequencies (fc) from a typical infant subject. Carrier frequencies of 500, 1000, 2000 and 4000 Hz were modulated (fin) at 77, 85, 93 and 100 Hz, respectively, and presented simultaneously to one ear. The top portion of the figure represents the responses in terms of their polar coordinates; the radius of the circle shown in the plots provides an estimate of E E G background noise (if the circle does not include the origin, the response is present). The bottom portion of the figure represents the responses as amplitude spectra resulting from FFT analyses (the presence of a response at the modulation frequency of interest is marked with an asterisks). fc: 500 Hz 1000 Hz 2000 Hz 4000 Hz fm: 77 Hz 85 Hz 93 Hz 100 Hz P O L A R P L O T S EEG i 1 1 1 1 1 i fm: 75 80 85 90 95 100 105 Hz 45 Figure 1.5 The first line of the top panel shows E E G when no stimulus is presented. The second line shows large stimulus artifact in the E E G when a 50 dB HL 500-Hz bone-conduction stimulus is presented. The first line in the lower panel shows that stimulus artifact in the E E G is substantially reduced, but still present, when a 12 dB/oct 250-Hz low pass filter is used. UNFILTERED BC 500 LOW-PASS FILTERED Low Pass:250 Hz (12 dB/oct) 46 Figure 1.6 Auditory brainstem responses recorded in a 3-month-old infant with bilateral atresia using two-channel E E G recordings (Cz-Ml, left; Cz-M2, right) for left-side and right-side bone conducted stimulation. Ear contralateral to bone oscillator placement could not be masked. Reversal in response asymmetry with change in bone oscillator placement indicates bilateral cochlear sensitivity ro these stimuli. Filled arrow - approximate ipsilateral wave V latency, open arrows - approximate contralateral wave V latency. Left figure: 500-Hz tones. Right figure: 2000-Hz tones. (Figure provided by David Stapells) 500 Hz 20 dB nHL 2000 Hz 30dB nHL L E F T RIGHT LEFT EEG Cz - M1 RIGHT EEG Cz-M2 25 ms 25 ms 47 REFERENCES Allen, G., Fernandez, C. (1960). The mechanism of bone conduction. Annals of Otology, Rhinology and Laryngology 69: 5-28. ANSI. 1996. American Standard Specifications for Audiometers (ANSI S3.6-1996). New York, ANSI. Anson, B. J., & Donaldson, J. A. (1981). The ear: Developmental anatomy. In B. J. Anson & J. A. Donaldson (Eds.), Surgical anatomy of the temporal bone Philadelphia: W. B. Saunders Company. Azzena, G. B., Conti G., Santarelli R., Ottaviani F., Paludetti G., & Maurizi M . (1995). Generation of human auditory steady-state responses (SSRs). I: stimulus rate effects. Hearing Research 83: 1-8. Barany, E. (1938). A contribution to the physiology of bone conduction. Acta Otolaryngologica, Supplement 26: 1-221. Basar, E., Rosen B., Basar-Eroglu C , & Greitschus F. (1987). The associations between 40-Hz and the middle latency response of the auditory evoked potential. Journal of Neuroscience 33: 103-117. Brinkman, W., Marres, E. , Tolk, J. (1965). The mechanism of bone conduction. Acta Otolaryngologica 59: 109-115. Campbell, F. W., Atkinson J., Francis M . R., & Green D. M . (1977). Estimation of auditory thresholds using evoked potentials. A clinical screening test. Progress in Clinical Neurophysiology 2: 68-78. Carhart, R. (1962). Effect of stapes fixation on bone conduction response. In Otosclerosis, ed. H. Schuknecht, pp. 175-197. Boston: Little, Brown and Company. Ciulla, C , Takeda, T., Morabito, M. , Endo, H., Kumagai, T., Xiao, R. (1996). M E G measurements of 40 Hz auditory evoked response in human brain. Electroencephalography Clinical Neurophysiology Supplement 47: 121-127. Cohen, L. T., Rickards F. W., & Clark G. M . (1991). A comparison of steady-state evoked potentials to modulated tones in awake and sleeping humans. Journal of the Acoustical Society of America 90: 2467-2479. Cone-Wesson, B. (1995). Bone-conduction ABR tests. American Journal of Audiology 4: 14-19. Cone-Wesson, B. & Ramirez, G.M. (1997). Hearing sensitivity in newborns estimated from ABRs to bone-conducted sound. Journal of the Academy of Audiology 8:299-307. Dean, M . S., & Martin F. N. (2000). Insert earphone and the occlusion effect. American Journal of Audiology 9: 131-134. Dirks, D., & Swindeman J. G. (1967). The variability of occluded and unoccluded bone-conduction thresholds. Journal of Speech and Hearing Research 10: 232-249. Delaroche, M. , Thiebaut R., & Dauman R. (2004). Behavioral audiometry: protocols for measuring hearing thresholds in babies aged 4—18 months. International Journal of Pediatric Otorhinolaryngology 68: 1233-1243. Dimitrijevic, A., John, M . S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., Nedzelski, J. M . , & Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology 13: 205-224. Dobie, R. A., & Wilson M . J. (1998). Low-level steady-state auditory evoked potentials: Effects of rate and sedation on detectability. Journal of the Acoustical Society of America 104: 3482-3488. 49 Eby, T. L., & Nadol J. B. (1986). Postnatal growth of the human temporal bone. Annals of Otology, Rhino logy and Laryngology 95: 356-364. Elberling, C. (1979). Frequency composition of brain-stem auditory evoked potentials. Scandinavian Audio logy 8: 51-55. Elpern, B., & Naunton R. F. (1963). The stability of the occlusion effect. Archives of Otolaryngology 77: 376-384. Ferraro, J. A., Ferguson, R. (1989). Tympanic ECochG and conventional ABR: a combined approach for the identification of wave I and the 1-V interval. Ear & Hearing 3: 161-166. Foxe, J. J., & Stapells, D. R. (1993). Normal infant and adult auditory brainstem responses to bone-conducted tones. Audiology 32: 95-109. Franke, E. , von Gierke, H., Grossman, F., von Wittern, W. (1952). Response of the human skull to mechanical vibrations. Journal of the Acoustical Society of America 24: 142-146. Freeman, S., Sichel J.-Y., & Sohmer H. (2000). Bone conduction experiments in animals-evidence for a non-osseous mechanism. Hearing Research 146: 72-80. Furune, S., Watanabe K., Negoro T., Yamamoto N., Aso K., & Takaesu E. (1985). Auditory brainstem response: a comparative study of ipsilateral versus contralateral recording in neurological disorders of children. Brain Development 7: 463-469. Galambos, R., Makeig S., & Talmachoff P. (1981). A 40-Hz auditory potential recorded from the human scalp. Proceedings of the National Academy of Science of the United States of America 78: 2643-2647. Geisler, C. D. (1960). Average response to clicks in man recorded by scalp electrodes. MJ.T. Technical Report 380: 1-158. 50 Gelfand, S. A. (1981). Hearing: An introduction to psychological and physiological acoustics. pp. 66. New York, NY: Marcel Dekker, Inc. Groen, J. 1962. The value of the Weber test. In Otosclerosis, ed. H. Schuknecht, pp. 165-74. Boston: Little, Brown and Company. Groen, J., Rosowski, J., Merchant, S. (1997). Anatomy of the normal human cochlear aqueduct with functional implications. Hearing Research 107: 9-22. Gutschalk, A., Mase, R., Roth, R., Ille, N., Rupp, A., Hahnel, S., Picton, T. W., & Scherg M . (1999). Deconvolution of 40 Hz steady-state fields reveals two overlapping source activites of the human auditory cortex. Clinical Neurophysiology 110: 856-868. Hakansson, B., Brandt, A., Carlsson, P., Tjellstrom. (1994). Resonance frequency of the human skull in vivo. Journal of the Acoustical Society of America 95: 1474-1481. Hall, J. W. (1992). Handbook of auditory evoked responses. Needham Heights: Allyn & Bacon. Harrell, R. W. 2002. Puretone Evaluation. In Handbook of Clinical Audiology, Fifth edition, ed. J. Katz, pp. 71-87. Philidelphia: Lippincott, Williams & Wilkins. Hari, R., Hamalainen M . , & Joutsiniemi S. (1989). Neuromagnetic steady-state responses to auditory stimuli. Journal of The Acoustical Society of America 86: 1033-1039. Hashimoto, I., Ishiyama, Y. , Yoshimoto, T., & Nemoto, S. (1981). Brain-stem auditory-evoked potentials recorded directly from human brain-stem and thalamus. Brain 104: 841-859. Herdman, A., Lins, O., Van Roon, P., Stapells, D., Scherg, M . , & Picton, T. (2002). Intracerebral sources of human auditory steady-state responses. Brain Topography 15: 69-86. Herdman, A. T., & Stapells, D. R. (2001). Thresholds of auditory steady-state responses in normal-hearing adults. Scandinavian Audiology 30: 41-49. 51 Herdman, A. T., & Stapells, D. R. (2003). Auditory steady-state response thresholds of adults with sensorineural hearing impairment. International Journal of Audiology. 42: 237-248. Hodgson, W. R., & Tillman T. W. (1966). Reliability of bone conduction occlusion effects in normals. Journal of Auditory Research 6: 141-151. Howell, P., Williams, M . (1989). Jaw movement and bone conduction in normal listeners and a unilateral hemi-mandibulectome. Scandinavian Audiology 18: 231-236. Howell, P., Williams, M. , Dix, H. (1988). Assessment of sound in the ear canal caused by movement of the jaw relative to the skull. Scandinavian Audiology 17: 93-98. Huizing, E. H. (1960). Bone-conduction - The influences of the middle ear. Acta Otolaryngologica, Supplement 155: 1-99. Jahrsdoerfer, R. A., Yeakley, J. W., Hall, J. W., Robbins, K. T., & Gray, L. C. (1985). High-resolution CT scanning and auditory brain stem response in congenital aural atresia: Patient selection and surgical correlation. Otolaryngology-Head and Neck Surgery 93: 292-298. Jeng, F.-C., Brown, C.J., Johnson, T.A., Vander Werff, K.R. (2004). Estimating air-bone gaps using auditory steady-state responses. Journal of the American Academy of Audiology 15: 67-78. Jerger, J., Chmiel R., Frost J. D., & Coker N. (1986). Effect of sleep on the auditory steady state evoked potential. Ear and Hearing 7: 240-245. John, M . S., Dimitrijevic A., van Roon P., & Picton T. W. (2001). Multiple auditory steady-state responses to A M and F M stimuli. Audiology and Neurootology 6: 12-27. 52 John, M.S., & Picton, T.W. (2000). MASTER: A Windows program for recording multiple auditory steady-state responses. Computing Methods and Programs in Biomedicine 61: 125-150. Johnson, B.W., Weinberg H., RibaryU., Cheyne D.O., & Ancill R. (1988). Topographic distribution of the 40 Hz auditory evoked-related potential in normal and aged subjects. Brain Topography 1: 117-121. Joint Committee on Infant Hearing. (2000). Year 2000 position statement: Principles and guidelines for early hearing detection and intervention programs. American Journal of Audiology 9: 9-29. Khanna, S. M . , Tonndorf, J., Queller, J. (1976). Mechanical parameters of hearing by bone conduction. Journal of the Acoustical Society of America 60: 139-154. Kirikae, I. (1959). An experimental study on the fundamental mechanism of bone conduction. Acta Otolaryngologica, Supplement 145: 97-107. Kramer, S. J. (1992). Frequency specific auditory brainstem responses to bone-conducted stimuli. Audiology 31: 61-71. Kucuk, B., Abe, K., Ushiki, T., Inuyama, Y. , Fukuda, S., Ishikawa, K. (1991). Microstructures of the bony modiolus in the human cochlea: A scanning electron miscroscope study. Journal of Electron Microscopy 40: 193-197. Kuwada, S., Anderson J.S., Batra R., Kirkpatrick D.C., Teissier N., & D'Angelo W.R. (2002). Sources of the scalp-recorded amplitude-modulation following response. Journal of the American Academy of Audiology 13: 188-204. Legouix, J. P., Tarab, S. (1999). Experimental study of bone conduction in ears with mechanical impairment of the ossicles. Journal of the Acoustical Society of America 31: 1453-1457. 53 Liden, G., & Kankkuken, A. (1969). Visual reinforcement audiometry. Acta Otolaryngologica 67: 281-291. Linden, R. D., Campbell K. B., Hamel. G., & Picton T. W. (1985). Human auditory steady-state evoked potentials during sleep. Ear and Hearing 6: 167-174. Lins, O.G., & Picton T.W. (1995). Auditory steady-state responses to multiple simultaneous stimuli. Electroencephalography and clinical Neurophysiology 96: 420-432. Lins, O.G., Picton, T.W., Boucher, B.L., Durieux-Smith, A., Champagne, S. C , Moran, L. M . , Perez-Abalo, M . C , Martin, V. , & Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear and Hearing 17: 81-96. Lins, O. G., Picton P. E., Picton T. W., Champagne S. C , & Durieux-Smith A. (1995). Auditory steady-state responses to tones amplitude-modulated at 80-110 Hz. Journal of the Acoustical Society of America 97: 3051-3063. Mackersie, C. L., & Stapells D. R. (1994). Auditory brainstem response wave I prediction of conductive component in infants and young children. American Journal of Audiology 3: 52-58. Makela, J. P., & Hari R. (1987). Evidence for cortical origin of the 40-Hz auditory evoked response in man. Electroencephalography and clinical Neurophysiology 66: 539-546. Mauer, G., & Doring, W. H. (1999). Generators of amplitude modulation following response (AMFR). Paper presented at the XVI Biennial Meeting of the International Evoked Response Audiometry Study Group, Tromso, Norway. McPherson, D. L., Hirasugi, Y. , Starr, A. (1985). Auditory brain stem potentials recorded at different scalp locations in neonates and adults. Annals of Otology, Rhinology and Laryngology 94: 236-243. 54 Moore, J. K., & Guan Y. -L . (2001). Cytoarchitectural and axonal maturation in human auditory cortex. Journal of the Association for Research in Otolaryngology 2: 297-311. Moore, J. K., Guan, Y. -L. , & Shi, S.-R. (1997). Axogenesis in the human fetal auditory system, demonstrated by neurofilament immunochemistry. Anatomy and Embryology 195: 15-30. Moore, J. K., Perazzo, L. M . , & Braun, A. (1995). Time course of axonal myelination in the human brainstem auditory pathway. Hearing Research 87: 21-31. Moore, J. K., Ponton, C. W., Eggermont, J. J., Wu, J.-C., & Huang, J. Q. (1996). Perinatal maturation of the auditory brain stem response: Changes in path length and conduction velocity. Ear and Hearing 17(5): 411-418. Moore, J. M . , Wilson W. R., & Thompson G. (1977). Visual reinforcement of head-turn responses in infants under 12 months of age. Journal of Speech and Hearing Disorders 42: 328-334. Moore, J. M . , & Wilson W. 1978. Visual reinforcement audiometry (VRA) with infants. New York: Grune and Stratton. National Instruments. (2000). Lab VIEW™ Measurements Manual. Austin, Texas: National Instruments. Nousak, J. K., & Stapells D. R. (1992). Frequency specificity of the auditory brain stem response to bone-conducted tones in infants and adults. Ear and Hearing 13: 87-95. Picton, T. W., Durieux-Smith A., Champagne S. C , Whittingham J., Moran L. M . , Giguere C , & Beauregard Y. (1998). Objective evaluation of aided thresholds using auditory steady-state responses. Journal of the American Academy of Audiology 9:315-31. Picton, T., John, M . , Dimitrijevic, A., & Purcell, D. (2003). Human auditory steady-state responses. International Journal of Audiology 42: 177-219. 55 Picton, T. W., Hillyard, S. A., Krausz, H. I., & Galambos, R. (1974). Human auditiory evoked potentials. I: Evaluation of components. Electroencephalography and clinical Neurophysiology 36: 179-190. Picton, T. W., Hink, R. F., Perez-Abalo, M . , Linden, R. D., & Wiens, A. S. (1984). Evoked potentials: How Now? Journal of Electrophysiological TechnologylO: 177-221. Picton, T. W., Stapells, D. R., & Campbell, K. B. (1981). Auditory evoked potentials from the human cochlea and brainstem. Journal of Otolaryngologica 10 ( Supplement 10): 1-41. Plourde, G., Stapells D. R., & Picton T. W. (1991). The human auditory steady-state evoked potentials. Acta Otolaryngology (Stockholm) Suppl. 491: 153-160. Ponton, C. W., Moore, J. K., Eggermont, J. J., Wu, B. J . - C , & Huang, J. Q. (1994). Relationship between auditory brainstem size and auditory brainstem response (ABR) latency in human development. Association for Research Otolaryngology Abstracts: 57. Ponton, C. W., Moore, J. K., & Eggermont, J. J. (1996). Auditory brainstem response generation by parallel pathways: Differential maturation of axonal conduction time and synaptic delay. Ear and Hearing 17: 402-410. Ranee, G., & Briggs, R. S. J. (2002). Assessment of hearing in infants with moderate to profound impairment: the Melbourne experience with auditory steady-state evoked potential testing. Annals of Otology, Rhinology and Laryngology, 111 (Supplement 189): 22-28. Ranee, G., Dowell, R. C , Rickards, F. W., Beer, D. E., & Clark, G. M . (1998). Steady-state evoked potential and behavioural hearing thresholds in a group of children with absent click-evoked auditory brain stem response. Ear and Hearing, 19(1): 48-61. Ranee, G., Tomlin, D. (2006). Maturation of auditory steady-state responses in normal babies. Ear & Hearing 27: 20-29. 56 Rodriguez, R., Picton T. W., Linden D., Hamel G. & Laframboise G. (1986). Human auditory-steady state responses: effects of intensity and frequency. Ear and Hearing 7: 300-13. Rickards, F. W., & Clark G. M . (1984). Steady-state evoked potentials to amplitude-modulated tones. In Evoked Potentials II, ed. R. H. Nodar, & C. Barber, pp. 163-8. Boston: Butterworth, Sanders, J.W. (1991). Clinical masking. In Rintlemann. W.F. (Ed.), Perspectives in Audiology Series: Hearing Assessment (pp. 141-172). Austin: PRO-ED, Inc. Santarelli, R., Maurizi M . , Conti G., Ottaviani F., Paludetti G., & Pettorossi V. E. (1995). Generation of human auditory steady-state responses (SSRs). II: Addition of responses to individual stimuli. Hearing Research 83: 9-18. Schimmel, H., Rapin, I., & Cohen, M . M . (1974). Improving evoked response audiometry with special reference to the use of machine scoring. Audiology 13: 33-65. Schoonhoven, R., Boden C. J. R., Verbunt J. P. A., & de Munck J. C. (2002). A whole-head M E G study of the amplitude modulation following response: phase coherence, group delays and dipole source analysis. Clinical Neurophysiology 114: 2096-106. Seaman, R. L. (1991). Method to record evoked potentials from the frog eighth nerve. Hearing Research 51: 301-306. Sininger, Y. S. (1996). Filtering and spectral characteristics of averaged auditory brainstem response and background noise in infants. Journal of the Acoustical Society of America 98: 2048-2055. Sininger, Y . S., Abdala C , & Cone-Wesson B. (1997). Auditory threshold sensitivity of the human neonate as measured by the auditory brainstem response. Hearing Research 104: 27-38. 57 Sohmer, H., Freeman, S. Geal-Dor, M . , Adelman, C , & Savion, I. (2000). Bone-conduction experiments in humans- a fluid pathway from bone to ear. Hearing Research 146:81-88. Small, S., & Stapells, D. (2003). Normal brief-tone bone-conduction behavioural thresholds using the B-71 transducer: Three occlusion conditions. Journal of the American Academy of Audiology 14(10): 556-562. Small, S. A., & Stapells D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear and Hearing 25(6): 611-623. Small, S. A., & Stapells, D.R. (2006). Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing. Ear and Hearing 27: 219-228. Spivak, L. G. (1993). Spectral composition of infant auditory brainstem responses: implications for filtering. Audiology 32: 185-194. Stapells, D. R. (1989). Auditory brainstem response assessment of infants and children. Seminars in Hearing 10: 229-251. Stapells, D. R. (2000a). Frequency-specific evoked potential audiometry in infants. In R. C. Seewald (Ed.), A Sound Foundation Through Early Amplification (pp. 13 -31). Basel: Phonak A G . Stapells, D. R. (2000b). Threshold estimation by the tone-evoked auditory brainstem response: A literature meta-analysis. Journal of Speech-Language Pathology and Audiology, 24(2): 74-83. Stapells, D. R., Galambos R., Costello J. A., & Makeig S. (1988). Inconsistency of auditory middle latency and steady-state responses in infants. Electroencephalography and clinical Neurophysiology 71: 289-295. 58 Stapells, D. R., Herdman, A., Small, S.A., Dimitrijevic, A. Hatton, J. 2005. Current status of the auditory steady-state responses for estimating an infant's audiogram. In A Sound Foundation Through Early Amplification 2004: Proceedings of an International Conference, ed. R. C. Seewald, pp. 43-59. Stafa, Switzerland: Phonak, A G . Stapells, D. R., & Oates P. (1997). Estimation of the pure-tone audiogram by the auditory brainstem response: a review. Audiology & Neuro Otology 2: 257-280. Stapells, D. R., & Ruben R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. Annals of Otology, Rhinology and Laryngology 98: 941-949. Stapells, D. R., Linden D., Suffield J. B., Hamel G., & Picton T. W. (1984). Human auditory steady state potentials. Ear and Hearing 5: 105-113. Starr, A., & Hamilton, A. E. (1976). Correlation between confirmed sites of neurologic lesions and abnormalites of far-field auditory brainstem responses. Electroencephalography and clinical neurophysiology 41: 595-608. Stenfelt, S. & Goode, R.L. (2005). Bone conducted sound: Physiologic and clinical aspects. Otology Neurotology 26: 1245-1261. Stenfelt, S., Hato, N. & Goode, R.L. (2002). Factors contributing to bone conduction: the middle ear. Journal of the Acoustical Society of America 111: 947-959. Stenfelt, S., Wild,Y., Hato, N. & Goode, R.L. (2003). Factors contributing to bone conduction: the outer ear. Journal of the Acoustical Society of America 113: 902-913. Stuart, A., Yang E. Y., & Botea M . (1996). Neonatal auditory brainstem responses recorded from four electrode montages. Journal of Communication Disorders 29: 125-139. Stuart, A., Yang, E. Y., & Stenstrom, R. (1990). Effect of temporal area bone vibrator placement on auditory brain stem response in newborn infants. Ear and Hearing 11: 363-369. 59 Suzuki, T., Sakabe N., & Miyashita Y. (1982). Power spectral analysis of auditory brain stem responses to pure tone stimuli. Scandinavian Audiology 11: 25-30. Tonndorf, J. (1966). Bone conduction: Studies in experimental animals. Acta Otolaryngologica, Supplement 213: 1-132. Tonndorf, J. (1968). A new concept of bone conduction. Archives of Otolaryngology 87: 595-600. Tonndorf, J. (1976). Bone Conduction. In W. D. Neff (Ed.), Handbook of Sensory Physiology. Volume V, Auditory System, Part 3 (Clinical and Special Topics). Berlin: Springer-Verlag. van der Reijden, C. S., Mens, L.H. & Snik, A.F. (2005). E E G derivations providing auditory steady-state responses with high signal-to-noise ratios in infants. Ear & Hearing 26: 299-309. von Bekesy, G. (1932). Zur theorie des horens bei der schalluafnahme durch knochenleirung. Annals Physiologic 111-136. von Bekesy, G. (1960). Experiments in Hearing, ed. E.G. Wever, p. 129. New York, NY: McGraw-Hill. Voss, S. E. & Herrmann, B.S. (2005). How does the sound pressure generated by circumaural, supr-aural, and insert earphones differ for adult and infant ears? Ear & Hearing 26: 636-650. Wever, E. G., & Lawrence M . 1954. Physiologic Acoustics. Princeton, New Jersey: Princeton University Press. Wilber, L. A. (1979). Pure-tone audiometry: Air and bone conduction. In Hearing assessment, ed. W. F. Rintelmann, pp. 27-42. Baltimore, MD: University Park Press. 60 Yang, E. Y., & Stuart A. (1990). A method of auditory brainstem response testing of infants using bone-conducted clicks. Journal of Speech Language-Pathology and Audiology 14: 69-76. Yang, E. Y., & Stuart, A. (2000). The contribution of the auditory brainstem responses to bone-conducted stimuli in newborn hearing screening. Journal of Speech-Language Pathology and Audiology 24: 84-91. Yang, E. Y., Stuart, A., Stenstrom, M . A., & Hollett, S. (1991). Effect of vibrator to head coupling force on the auditory brain stem respone to bone-conducted clicks in newborn infants. Ear and Hearing 12: 55-60. Yang, E. Y. , Rupert A. L., & Moushegian G. (1987). A developmental study of bone conduction auditory brainstem responses in infants. Ear and Hearing 8: 244-251. Yoshida, M . , & Uemura, T. (1991). Transmission of cerebrospinal fluid pressure changes to the inner ear and its effect on cochlear microphonics. European Archives of Otorhinolaryngology 248: 139-143. 61 CHAPTER 2 ARTIFACTUAL RESPONSES WHEN RECORDING AUDITORY STEADY-STATE RESPONSES A version of this chapter has been published. Small, S. A., & Stapells, D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear and Hearing 25(6): 611-23. 62 INTRODUCTION Auditory steady-state responses (ASSRs) are becoming the centre of much interest as a possible clinical tool for estimating hearing threshold. Currently, the auditory brainstem response (ABR) is the most widely used auditory-evoked potential for assessing infants, young children, and multiply handicapped individuals who cannot be tested using conventional behavioural measures. ABRs elicited by air- and bone-conduction brief-tone stimuli provide reasonably accurate estimates of threshold at audiometric frequencies in sleeping and relaxed subjects (Foxe & Stapells, 1993; Stapells, 2000b; Stapells & Ruben, 1989). Detection of a response in the ABR waveform, however, requires skilled, subjective assessment of replicated responses. As yet, there are no objective measures that have been proven to work to detect tone-evoked ABRs (Hyde, Sininger & Don, 1998). The two main disadvantages of the tone-ABR are: (i) it is time consuming because only one ear and one frequency at a time can be tested, and (ii) response detection is subjective, allowing for error in judgement of the presence/absence of responses depending on the experience and skill of the clinician (Stapells, 2000a). ASSRs elicited by modulation frequencies greater than 70 Hz are well suited for assessing individuals who need to be tested in natural sleep or under sedation because state of arousal has little effect on these responses (Cohen, Rickards & Clark, 1991). ASSRs can be recorded for single- or multiple-carrier frequencies to one ear, or both ears, simultaneously, and the presence of a response is detected using an objective statistical test (Picton, John, Dimitrijevic & Purcell, 2003). The clinical application of multiple ASSRs is of great interest because of potential time savings, compared to tone-ABR, and reduced variability in interpretation of responses. ASSRs to air-conduction stimuli have been found to provide accurate prediction of hearing sensitivity at the audiometric frequencies (Aoyagi & Furuse, 1994; 63 Dimitrjevic, John, Van Roon, Purcell, Adamonis, Ostroff, Nedzelski, & Picton, 2002; Herdman and Stapells, 2001, 2003; Lins, Picton, Boucher, Durieux-Smith, Champagne, Moran, Perez-Abalo, Martin, & Savio,1996; Perez-Abalo, Savio, Torres, Martin, Rodriguez & Galan, 2001; Picton, Durieux-Smith, Champagne, Whittingham, Moran, Giguere, & Beauregard, 1998; Ranee & Briggs, 2002; Ranee, Dowell, Rickards, Beer, & Clark, 1998; Ranee, Rickards, Cohen, De Vidi, & Clark, 1995). It has also been suggested that ASSRs can be used to assess residual hearing at levels greater than 80-100 dB nHL, which is the upper limit for ABR testing in patients with severe-to-profound hearing loss (Ranee et al., 1998). Behavioural techniques for threshold estimation routinely assess air- and bone-conduction pure-tone thresholds to distinguish among sensorineural, conductive and mixed hearing losses. This diagnostic information is used to plan medical intervention and aural (re)habilitation. Accurate bone-conduction thresholds are needed for patients with conductive and mixed losses, particularly for children who have unilateral or bilateral otitis media or atresia (Jahrsdoerfer, Yeakley, Hall, Robbins, & Gray, 1985; Stapells & Ruben, 1989). Only four studies have used ASSRs to estimate bone-conduction thresholds. Cone-Wesson et al. (2002) indirectly estimated ASSR bone-conduction thresholds using the "sensorineural acuity level" (SAL) test which uses bone-conduction noise to mask an air-conduction stimulus that is presented just above the subject's threshold (Jerger & Jerger, 1965). Three other studies have directly recorded ASSRs to bone-conduction stimuli but only for adults with normal hearing (Dimitrijevic et al., 2002; Lins et al., 1996; Small & Stapells, 2005: Chapter 3). Dimitrijevic et al. and Lins et al. used a forehead placement for the bone oscillator; our recent study (Small & Stapells, 2005: Chapter 3) positioned the bone oscillator at the mastoid which is more typical of clinical testing. In our first attempt to record steady-state responses to bone-conduction stimuli, we found responses whose 64 amplitudes were far too large, even for subjects with normal hearing. These responses were 5 to 15 times larger than those to air-conduction stimuli (Aoyagi et al., 1994; Dimitrjevic et al., 2002; Herdman & Stapells, 2001, 2003; Lins et al., 1996; Perez-Abalo et al., 2002; Picton et al, 1998; Ranee & Briggs, 2002; Ranee et al., 1995, 1998). Lins et al. (1996) and Dimitrijevic et al. (2002) assessed responses to stimuli no higher than 20-30 dB above threshold, and found differences between amplitude/phase measures for bone versus air-conduction ASSRs, particularly for 500-and 1000-Hz carrier frequencies. Small and Stapells (2005: Chapter 3) recorded bone-conduction ASSRs from 0 to 50 dB HL and found that the slope of the amplitude-intensity function was steeper for the 500-Hz carrier frequency compared to higher carrier frequencies. This is not the case for air-conduction stimuli. The larger than expected bone-conduction response and the differences between air- and bone-conduction ASSRs noted in the previous studies, led us to suspect that large stimulus artifact seen with bone conduction was causing spurious responses, particularly for lower carrier frequencies. A significant problem with bone-conduction stimulus artifact in the E E G is that this energy can be aliased to exactly the same frequency as the ASSR modulation rate, and be interpreted as a response. This is of particular concern for ASSRs because the stimulus and response overlap in time, and relatively slow analog-to-digital (A/D) conversion rates (such as 500 and 1000 Hz) have commonly been used (e.g., Dimitrijevic et al., 2002; Herdman & Stapells, 2001, 2003). These rates are available and, as of July, 2003, were even the default settings for the Bio-Logic (Mundelein, Illinois, USA) MASTER commercial instrument. For air-conduction stimuli, controlling for artifact in the E E G has not been considered to be a major issue because the electromagnetic artifact is very low in amplitude. If there is large stimulus artifact in the EEG, the choice of carrier frequency, modulation rate, E E G filter characteristics 65 and A/D rate become very important to avoid aliasing of this energy to modulation rates that are the same as those used to elicit the ASSR. The sampling rate must be more than twice the maximum frequency that is present in the EEG. The Nyquist frequency is equal to half of the sampling rate. Any frequency components above the Nyquist frequency can alias, resulting in frequency components that were not in the original signal (Picton et al., 1984). For example, for a 500-Hz sampling rate, any energy greater than 250 Hz could potentially alias. The alias frequency can be predicted using the following calculation (National Instruments Lab VIEW ™ Measurement Manual, 2000): Alias frequency = Absolute value (closest integer multiple of sampling frequency minus input frequency) For example, a 500-Hz tone that is amplitude-modulated at 77 Hz would have energy at 423, 500 and 577 Hz. If this energy is present in E E G being digitized at 500 Hz, an alias frequency would be: 500 Hz minus 423 Hz = 77 Hz, which is exactly the same as the modulation rate for this 500-Hz carrier frequency. When standard audiometric frequencies (500, 1000, 2000 and 4000 Hz) are used as carrier frequencies to elicit ASSRs, this calculation predicts aliasing will be a potential problem for all of the carrier frequencies when using a 500-Hz A/D rate. Similarly, using a 1000-Hz A/D rate, aliasing will be a potential problem for 1000-, 2000- and 4000-Hz carrier frequencies but not for a 500-Hz carrier frequency. Selecting an A/D rate for which the carrier frequencies are not integer multiples, such as 1250 Hz, results in alias frequencies that do not coincide with the modulation rates of the carrier frequencies, thereby substantially reducing 66 the potential for confusing stimulus artifact with an actual response1. Alternating the stimulus polarity is a common technique used to remove or reduce stimulus artifact when recording ABRs (e.g., Hall, 1992, pg. 319) and could also be used to reduce the effects of stimulus artifact that is present in the E E G when recording ASSRs. For ASSR recordings, this can be accomplished by inverting the stimulus, then averaging offline the responses to the inverted and non-inverted stimuli to obtain a response representing the "alternated stimulus polarity". Alternating the stimulus polarity, in itself, does not degrade the ASSR, as shown recently by Small and Stapells (2005: Chapter 3) who investigated the effect of "alternating" the stimulus polarity on air-conduction ASSRs recorded for a group of participants with normal hearing. In that study, an air-conduction stimulus was presented at an audible but low intensity (40 dB HL) so that stimulus artifact was not likely to be present in the EEG. ASSRs were recorded to stimuli that had non-inverted, inverted and alternated stimulus polarities. No differences in the amplitude or phase values of the ASSRs were found for either stimulus polarity compared to the alternated stimulus polarity (Small & Stapells, 2005: Chapter 3). Our initial problems with ASSRs to bone-conduction stimuli led us to the present study. The purpose of this study was to assess the possibility of spurious ASSRs to air- and bone-conduction stimuli by obtaining recordings from subjects with severe-to-profound sensorineural 1 There is a another problem to consider when selecting an A/D rate. Just as using a 1000-Hz A/D rate can increase the size of an artifactual response, using a 1250-Hz A/D rate might also increase the size of the noise estimate, thus making a true response fail to reach significance. The noise bins associated with various permutations of the A/D, D/A, carrier frequency and modulation frequencies should be removed from the noise estimate. This concern is a larger issue if only 5 or 10 bins are used in the noise estimate because a single extreme value would have a greater effect. Some researchers and instruments still use this type of statistic. In the MASTER system, 120 bins are used in the noise estimate, consequently, increased noise in a few bins would have a smaller effect on the overall noise estimate. 67 hearing loss who could not hear the stimuli. Because subjects could not hear the stimuli, any ASSRs which were recorded were necessarily artifactual. All bone-conduction ASSRs were recorded with unoccluded ears. Using the Rotman MASTER research system, we investigated three different A/D rates, two of which, 500- and 1000-Hz, have been used in many recent ASSR studies (e.g., Dimitrijevic et al., 2002; Herdman & Stapells, 2001, 2003). We used a bandpass filter of 30-250-Hz to filter the E E G for all A/D rates but added in a steep anti-aliasing filter for the 1250-Hz A/D rate to further minimize the likelihood of aliasing. We obtained recordings at several stimulus intensities, as well as to single vs alternated stimulus polarities. GENERAL MATERIALS AND METHODS The current study was divided into two experiments which investigated (i) ASSRs elicited by bone-conduction stimuli in participants with severe-to-profound sensorineural hearing loss, and (ii) ASSRs to high-level air-conduction stimuli in participants with severe-to-profound hearing loss. This section describes the methodology common to both experiments. The specific details of each experimental design are included for clarity in the description of the results. Participants The primary requirement for this study was that participants were not able to hear the stimuli used to elicit the ASSR. Before recording ASSRs, the participant was asked to listen to the stimuli and to describe whether they could be heard or felt. Results for any stimuli that were audible to a participant were omitted from the data analyses. Table 2.1 summarizes the subject characteristics for the hearing-impaired adults who participated in this study. A total of 17 participants were tested. Sixteen subjects participated in at least one of the test conditions for Experiment 1. Fourteen subjects participated in at least one condition for Experiment 2. Fifteen 68 participants had severe-to-profound sensorineural hearing losses and two had moderately-severe hearing losses (one of these had a mixed loss). The majority of the subjects used either hearing aids or cochlear implants; however, three subjects used sign language (two of these were hearing-impaired and blind). Hearing aids and cochlear implants were turned off throughout the test session. Table 2.1 Stimuli All stimuli were sinusoidal amplitude-modulated (AM) tones with carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. The stimuli were presented alone ("single") or simultaneously ("multiple") depending on the test condition. Single carrier frequencies were presented for bone-conduction stimuli at 60 dB H L and air-conduction stimuli at 120 dB HL (500 and 1000 Hz only) because presentation of multiple stimuli at these levels exceeded the maximum allowable amplitude for the output buffer in the MASTER system and/or saturated the stimulus amplifier. Air- and bone-conduction stimuli were generated by MASTER then routed through Tucker-Davis Technologies SM3 and HB6 modules to allow presentation of non-inverted and inverted stimuli. In order to obtain ASSRs to alternated stimuli, equal numbers of ASSR sweeps, elicited by non-inverted and inverted stimuli, were later averaged offline. The air- and bone-conduction stimuli were then attenuated using an audiometer (Interacoustics AC40) and presented to an EAR-3A insert earphone or B-71 bone oscillator, respectively. The transducer box for the insert earphone was placed just below the nape of the neck, usually on the subject's collar. The bone 69 oscillator was held in position on the temporal bone within 2 cm of the pinna with a wide elastic headband fastened with Velcro (Universal Facial Band #210, Design Veronique, Oakland, CA) with 450- 550 g offeree. Calibration The bone-conduction stimuli were calibrated in dB HL (ANSI, 1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 550 g of force. Linearity, lack of distortion, and symmetry of oscillation when stimulus polarity was inverted, were verified up to 60 dB H L for the B-71 transducer. Air-conduction stimuli were calibrated in dB HL (ANSI, 1996) using a Quest Model 1800 sound level meter with a Briiel and Kjaer DB0138 2-cc coupler. Recording ASSRs were recorded using the Rotman MASTER research system (John & Picton, 2000). Three gold-plated electrodes were used to record the electrophysiologic responses; the non-inverting electrode was placed at Cz, the inverting electrode was positioned at the nape of the neck, just below the hairline, and an electrode placed at the forehead acted as ground. All inter-electrode impedances were below 3kOhms at 10 Hz. For all A/D rate conditions, the responses were filtered using a bandpass of 30 to 250 Hz (12 dB/octave) and amplified 80,000 times (Nicolet HGA-200A and Nic501 A). For the 1250-Hz A/D condition, a 300-Hz lowpass anti-aliasing filter (Stanford Research System; 115dB/oct slope) was added in. Each E E G recording sweep was made up of 16 epochs. For the 500-Hz A/D rate, each epoch was made up of 512 data points and lasted a total of 16.384 seconds. For the 1000-Hz A/D rate, each epoch was made of 1024 points and also lasted 16.384 seconds. For the 1250-Hz A/D rate, each epoch was made up 1024 points and lasted a total of 13.1072 seconds. The digital-to-analog (D/A) rate 70 was 32,000 Hz for the 500- and 1000-Hz A/D rates and 31,250 Hz for the 1250-Hz A/D rate. In order for the input and output buffers to remain synchronized, the D/A rate had to be a multiple of the A/D rate (John & Picton, 2000). Artifact rejection was set to eliminate epochs containing amplitudes greater than ± 40 uV. ASSR sweeps of data were averaged in the time domain and then analysed on-line in the frequency domain using a Fast Fourier Transform (FFT). Amplitudes were measured baseline-to-peak and expressed in nanovolts (nV). An F-ratio was calculated by MASTER and estimated the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ± 60 bins of the modulation frequency ("noise"). A response was considered to be present if the F-ratio, compared to the critical values for F (2,240), was significant at a level ofp < .05. Consistent with previous studies (e.g., Dimitrijevic et al., 1993; Herdman & Stapells, 2001, 2003), recordings without significant responses were continued until one of the following stopping criteria was met. A response could be considered absent if p > .05 and the average amplitude of the noise was less than 11 nV. A response was also considered to be absent if the response amplitude was < 10 nV and if p >.30. However, the presence or absence of a response was not determined for cases where p > .05 and the amplitude of the noise was > 11 nV. Mean amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant. Phase values from MASTER were adjusted by adding 90° to yield the onset phase. Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Phase-delay values were reported for significant responses only. Individual amplitude values were also reported for significant responses only. 71 Procedure Testing was performed in a double-walled sound-attenuated booth. Participants reclined in a comfortable chair and were instructed to relax or sleep during the ASSR test session. Test sessions were 1.5-2.0 hours in duration. Participants signed a consent form before commencing any of the experiments and were paid an honorarium at the end of each session. Data Analyses A response was considered to be present for the single-polarity conditions if there were a response (p < .05) in one or both of the non-inverted or inverted stimulus polarity conditions. Using an alpha level ofp < .05, one would expect a 5% false-positive response rate. Thus, ASSRs were judged to be "artifactual" when the number of subjects showing responses present was significantly greater than the 5% expected by chance as determined by the binomial exact test of statistical significance (Zar, 1984, pp. 383-386). For 8-9 subjects, the presence of three or more responses are considered artifactual; for 3-7 subjects, the presence of two or more responses are considered artifactual. RESULTS Experiment 1 - Artifactual responses to bone- conduction stimuli The purpose of this experiment was to record ASSRs to bone-conduction stimuli in adults who could not hear the stimuli to investigate the possibility of spurious responses. The subjects who participated are described in Table 2.1. A total of 16 hearing-impaired adults were tested. The right mastoid was used as the test ear for bone-conduction ASSRs for all participants, except for those who had cochlear implants in the right mastoid. For these individuals, the left mastoid was used as the test ear. Non-inverted and inverted bone-conduction multiple stimuli were 72 presented at 40 and 50 dB HL for all A/D rates. Multiple ASSRs were also recorded to both stimulus polarities at 20 dB H L for a 500-Hz A/D rate (n = 8) and at 20-30 dB H L for a 1000-Hz A/D rate for a small number of subjects (total of 4 subjects). For the 1250-Hz A/D rate, intensities below 40 dB HL were not tested based on results for the 1000-Hz A/D rate (i.e., artifactual responses were not found below this intensity at lower A/D rates). Results (n = 3-8) for 40, 50 and 60 dB HL were obtained at this A/D rate. Single stimuli were presented at 60 dB H L for 500- and 1000-Hz carrier frequencies, but only for the 1000- and 1250-Hz A/D rates (n = 6 per rate). Results for 60 dB HL were not obtained for the 500-Hz A/D rate. Figure 2.1 presents results for an individual subject (#8) who had responses to 50 dB HL 500- and 1000-Hz bone-conduction stimuli recorded using a 1000-Hz A/D rate. For the 500-Hz stimuli, significant responses were present for all three stimulus polarities with similar amplitude (419-523 nV) and phase (141-154°) values. For the 1000-Hz stimuli, the subject had significant responses to non-inverted and inverted stimuli that were 183 degrees out of phase and similar in amplitude (51-62 nV); no significant response was present to alternated stimuli (amplitude = 6 nV; p = .727), thus this result is not plotted. The results for this subject for the 1000-Hz stimuli are consistent with the presence of stimulus artifact in the E E G which is removed by alternating the stimulus polarity. In contrast, the results for the 500-Hz stimuli show artifactual responses that are not affected by alternating the stimulus polarity. Figure 2.1 73 Figure 2.2 presents results for all subjects (who could not hear the stimuli), showing polar plots indicating the amplitude and phase of the significant ASSRs recorded to 50 dBHL bone-conduction stimuli. Responses that were not significant are not shown in these plots. Clearly, there were artifactual ASSRs to high-level bone-conduction stimuli, as far more than 5% of subjects show responses, especially for 500- and 1000-Hz carrier frequencies. Carrier frequency of the stimulus and the A/D rate used to record the ASSRs had a substantial effect on the number of significant responses that were detected. There were many responses for 500- and 1000-Hz stimuli, and only a few responses for 2000- and 4000-Hz carrier frequencies. There were also many more spurious responses using the 500- and 1000-Hz A/D rates compared to the 1250-Hz A/D rate, particularly for the 1000-, 2000- and 4000-Hz carrier frequencies. Responses to the 500-Hz stimuli were present at all A/D rates, but were considerably larger in amplitude when using the 500- and 1000-Hz A/D rates (as indicated by the polar plot radius). Figure 2.2 For the 500-Hz stimulus and the 500-Hz A/D rate, the responses to non-inverted and inverted stimuli were very large - as much as 1558 nV - and inverted with inversion of the stimulus polarity. Neural responses should not invert; therefore, these responses appear to be due to stimulus artifact2. Alternating the stimulus polarity should reduce or eliminate responses 2 The inverting of the stimulus by 180 degrees causes the artifacts in the recorded E E G data to invert 180 degrees. The addition of these two out-of-phase artifacts will cause them to cancel each other out. The physiologic ASSR, however, will not be affected by flipping the phase of the stimulus (i.e., of the carrier), because it is a neural response determined by the modulation envelope of the stimulus. 74 arising from stimulus artifact. This was, in fact, what happened for the 1000- and 2000-Hz stimuli across conditions. Artifactual responses to alternated stimuli, however, were still clearly present to 500-Hz stimuli at all A/D rates. Figure 2.3 summarizes the results for ASSRs to single-polarity bone-conduction stimuli, showing the percent occurrence of responses for all carrier frequencies and A/D rates presented at 40, 50 and 60 dB HL. For 500-Hz stimuli, artifactual responses were present (i.e., significantly more than expected by chance) for all A/D rates except at 40 dB HL using the 1000- and 1250-Hz A/D rates. For 1000-Hz stimuli, artifactual responses were present at all intensities for the 500- and 1000-Hz A/D rates but only at 60 dB HL for the 1250-Hz A/D rate. For 2000-Hz stimuli, artifactual responses were present at 50 dB HL for the 500- and 1000-Hz A/D rates. No artifactual responses were recorded for 4000-Hz stimuli for any condition. Figure 2.3 Figure 2.4 presents the percent occurrence of ASSRs to alternated-polarity bone-conduction stimuli, showing that artifactual responses were not present to alternated stimuli for 1000-, 2000-, and 4000-Hz stimuli for any A/D rate. Artifactual responses were present to 500-Hz stimuli for all intensities for the 500-Hz A/D rate and for 50 and 60 dB HL for the 1000- and 1250-Hz A/D rates. The artifactual responses to 500-Hz stimuli for the 1000-Hz and 1250-Hz A/D rates are not easily explained by aliasing and, because the responses did not invert with inversion of the stimulus polarity, stimulus artifact is not likely the origin of these responses. 75 Figure 2.4 Artifactual responses were less likely to be present to lower stimulus intensities. For a 500-Hz A/D rate, there were artifactual responses to 500-Hz stimuli at 20 dB HL for the single-polarity condition but no artifactual responses for the alternated stimuli. For the 1000-Hz A/D rate, there were artifactual responses to 1000-Hz stimuli at 20 and 30 dB HL for the single-polarity condition but no artifactual responses for the alternated stimuli. For the 1250-Hz A/D rate, there were no responses to single or alternated stimuli presented at 40 dB HL; lower intensities were thus not tested. Experiment 2 - Artifactual responses to high-intensity air- conduction stimuli Stimulus artifact is clearly problematic when recording ASSRs to bone-conduction stimuli. These findings led to the second experiment which investigated whether high-level air-conduction stimuli can result in artifactual responses. The subjects who participated in Experiment 2 are described in Table 2.1. Four to six subjects were tested for each A/D rate, resulting in a total of 14 hearing-impaired adults participating. In one subject, both ears were tested (separately) using a 500-Hz A/D rate, but only the results for one ear are included in the figures and data analyses (see below). In the first phase of this experiment, using a 500-Hz A/D rate, the insert earphone transducer was placed near the nape of the neck and the foam tip was inserted into the ear canal. Because of the difficulties finding deaf participants who could not hear at high-intensity air-conduction stimuli, subsequent phases of this experiment which used 1000- and 1250-Hz A/D rates were set up so that the tube between the transducer and the foam tip was clamped and the foam tip was not placed in the ear canal, in order to ensure that the 76 stimulus could not be heard. The main purpose of this experiment was to determine if stimulus artifact could result in spurious results. Unfortunately, any information about artifactual responses due to sources other than stimulus artifact was lost by clamping the tube. To record ASSRs for a 500-Hz A/D rate, non-inverted and inverted multiple stimuli were presented at 116, 119, 116 and 114 dB HL, for 500-.1000-, 2000- and 4000-Hz carrier frequencies, respectively. To record ASSRs using 1000- and 1250-Hz A/D rates, non-inverted and inverted air-conduction single stimuli were presented at 116 and 119 dB HL for 500- and 1000-Hz carrier frequencies, respectively. ASSRs were also obtained for alternated stimuli for all conditions. The polar plots in Figure 2.5 show that artifactual responses were obtained to high-intensity air-conduction stimuli at 500 Hz using the 500-Hz A/D rate and at 1000 Hz using the 1000- and 1250-Hz A/D rates, despite the absence of an audible signal at the ear. As shown in Figure 2.6, alternating the stimulus polarity eliminated these responses. The number of responses at 2000 and 4000 Hz for the 500-Hz A/D rate were not significantly greater than expected by chance. The spurious responses seen for the 1000-Hz rate likely resulted from stimulus artifact in the EEG, considering that there was no stimulus presented to the ear. The few responses seen for the 1250-Hz rate may be due to chance. Two responses were present but these were from two different subjects and for only one stimulus polarity. In both cases, a small response averaged with no response in the opposite polarity resulted in a non-significant ASSR to the alternated stimulus condition (i.e., not a true cancellation due to alternated stimuli). In one subject (#10), different results were obtained depending upon which ear was stimulated using the 500-Hz A/D rate. When the right ear was stimulated, there were no responses to the air-conduction stimuli. When the subject's left ear was stimulated, however, responses were present for the 500-Hz carrier frequency for both stimulus polarities. These responses were similar in phase for each 77 polarity, and thus were not removed by alternating the stimulus polarity. At the same time as stimulation of this ear, the subject reported feeling nauseous and slightly dizzy. Figures 2.5 and 2.6 DISCUSSION The results of these experiments clearly indicate that high-intensity air- or bone-conduction stimuli can produce artifactual (or spurious) ASSRs, especially to 500- and 1000-Hz carrier frequencies. Soon after the results of these experiments were presented, other researchers reported spurious ASSRs to bone- (Jeng, Brown, Johnson & Vander Werff, 2004) and high-intensity air-conduction (Gorga, Neely, Hoover, Dierking, Beauchaine & Manning, 2004) stimuli in subjects who had severe-to-profound hearing losses and could not hear the stimuli. Overall, the greater number of artifactual responses for the 500- and 1000-Hz A/D rates compared to the 1250-Hz rate are consistent with our hypothesis that stimulus artifact in the E E G aliases to exactly the modulation rates for specific carrier frequencies. Specifically, energy from the stimulus artifact can result in spurious responses at all carrier frequencies when using a 500-Hz A/D rate and at 1000-, 2000- and 4000-Hz carrier frequencies when using a 1000-Hz A/D rate. These results show that artifactual responses do, in fact, occur at the frequencies predicted. As shown in Figures 2.1, 2.2, and 2.4, in many cases the phase delay of the responses to non-inverted and inverted stimulus polarities differ by approximately 180 degrees and alternating the stimulus polarity substantially reduces the number of artifactual responses present. These changes with stimulus polarity further support our hypothesis that stimulus artifact is the underlying cause of many of the spurious responses as one would expect stimulus artifact to 78 invert with inversion of the stimulus polarity. There were very few responses at 2000- and 4000-Hz, although they were predicted to occur. These results are consistent with the presence of lower amplitude stimulus artifact to alias at these frequencies compared to lower frequencies. Higher carrier frequencies are further away from the E E G low-pass filter cut-off frequency and thus stimulus artifact would be much smaller in amplitude. Also, for ANSI H L calibration much less energy (7-27 dB) is required to drive the oscillator at 2000-4000 Hz compared to 500-1000 Hz (ANSI, 1996) which, in turn, reduces the amplitude of stimulus artifact produced at the higher frequencies. Clearly, appropriate choice of E E G filter slope and low-pass cutoff and A/D rate can avoid spurious responses due to aliasing. Oversampling using very high A/D rates will also alleviate aliasing problems (but will also result in very large continuous data files). Also, alternating the stimulus polarity can help reduce the number of artifactual responses. However, artifactual responses due to other causes still occur for bone-conduction stimuli at levels 50 dB HL and higher. In the present study, the stimulus polarity was alternated offline. Currently, there is no clinical equipment available which allows the stimulus polarity to be alternated, offline or online. Future research is required to develop an alternated stimulus that can be presented online to record ASSRs. Based on the findings of the present experiments, Picton and John (2004) recently investigated additional ways to reduce the likelihood of spurious responses by using stimuli with frequency spectra that would not alias back to the modulation frequencies of responses, such as "alternating sinusoidally amplitude-modulated tones". They found that this type of tone, which alternates the polarity of the carrier frequency at every cycle of the modulation, results in response amplitudes that are similar to response amplitudes obtained using the alternated stimuli 79 described in this study. They also recorded ASSRs to "beat" stimuli, but reported response amplitudes significantly reduced compared to alternated stimuli and are thus not suitable for threshold estimation. Other attempts have also been made to reduce or avoid the impact of stimulus artifact in the E E G when using bone-conduction stimuli to elicit ASSRs. One possibility might be to reduce the intensity of the bone-conduction stimuli required to elicit an ASSR by occluding the ear canals. The occlusion effect reduces behavioural pure-tone thresholds in adults by an average of 16 and 8 dB for 500- and 1000-Hz long-duration bone-conduction stimuli (Goldstein & Hayes, 1965; Dirks & Swindeman, 1967). Dimitrijevic et al., (2002) estimated bone-conduction ASSR thresholds with occluded ears; however, because they used a forehead placement for the bone oscillator which increases thresholds by 14 and 8.5 dB (ANSI, 1996), there was no net decrease in intensity of the stimulus. Occluding the ears when recording ASSRs to bone-conduction stimuli at the mastoid might be used to reduce the stimulus level needed to reach threshold and thus reduce any stimulus artifact in the EEG, especially at 500 and 1000 Hz. However, the occlusion effect has not been investigated in infants. An alternate approach might be to use the SAL test to indirectly estimate bone-conduction thresholds (Cone-Wesson et al., 2002). Stimulus artifact may be less of a concern for the SAL test because a bone-conduction noise masker is used to mask an air-conduction ASSR, instead of a bone-conduction stimulus eliciting the ASSR. The electromagnetic artifact of the bone-conduction noise used in the SAL procedure, however, may result in higher ASSR noise levels and thus higher thresholds. An important disadvantage of this technique is that the bone-conduction threshold is derived from two estimated thresholds, instead of one direct estimate of threshold, and is consequently more variable and time consuming. The SAL technique has not been assessed in subjects who cannot 80 hear the stimuli to rule out the possibility of spurious responses. The spurious responses to high-intensity air-conduction stimuli for single-polarity stimuli at 500 and 1000 Hz for a 500-Hz A/D rate and at 1000 Hz only when using a 1000-Hz A/D rate are also consistent with aliasing problems related to stimulus artifact. In Experiment 2, no air-conduction stimulus was actually presented to the subjects' ears when using the 1000- and 1250-Hz A/D rates; it therefore follows that any responses that were detected using these rates must be due to stimulus artifact. As expected, the responses were eliminated when the stimulus polarity was alternated. However, in one of the ears of the six subjects in Experiment 2 who had air-conduction stimuli presented directly to the ear, alternating the stimulus polarity did not eliminate the spurious responses. This is similar to the bone-conduction ASSRs to the 500-Hz carrier frequency that were not cancelled out by alternating the stimulus. The artifactual responses present to a 500-Hz carrier frequency at the 500-, 1000- and 1250-Hz A/D rates cannot be easily explained by aliasing. For the 500-Hz A/D rate, alternating the stimulus polarity should have removed the artifactual responses if they were indeed caused by stimulus artifact. Using the 1000-Hz A/D rate, the alias frequencies are predicted to be the same as the modulation rates used for the 1000-, 2000- and 4000-Hz carrier frequencies, but would not occur for the 500-Hz carrier frequency. As expected, the number of responses at 1000, 2000 and 4000 Hz were substantially reduced or eliminated by alternating the stimuli. Based on aliasing predictions, the responses at 500 Hz were not expected for the 1000-Hz A/D rate; these spurious responses were not reduced in number by alternating the stimulus polarity. Further, using the 1250-Hz A/D rate, aliasing is not predicted to cause artifactual responses at any of the modulation rates. The results for this A/D rate show, however, that artifactual responses are present to the 500-Hz stimuli, even after alternating the stimulus polarity and using an anti-81 aliasing filter with a steep slope (that should also eliminate any aliasing). Thus, the remaining spurious 500-Hz responses are not due to aliasing, and are likely physiologic (though non-auditory) in origin. A number of studies have shown that high-level auditory stimuli can result in vestibular-evoked myogenic responses (VEMPs), particularly recorded from the region of the inion and the sternocleidomastoid muscle3. Von Bekesy (1935) was the first to show that very loud sounds can activate the vestibular system. Subsequently, Geisler, Frishkopf, & Rosenblith (1958) and Bickford, Jacobson, & Cody (1964) recorded responses from the inion muscle to loud clicks. The high-intensity auditory stimuli were thought to activate vestibular afferents believed to arise from the saccule, that, in turn, modulated the tonic electromyogram activity of the head, neck, arm and leg muscles (Bickford et al., 1964). The waveform was comprised of four negative peaks with latencies of approximately 12, 26, 51 and 75 ms, respectively. They found that the amplitude of the response was proportional to the degree of tonic activation of the muscle and that all peaks were present for intensities greater than 90-100 dB SPL. Amplitude of the response decreased with intensity and the earliest peaks were not detectable below 90-100 dB SPL. Townsend and Cody (1971) used brief tones to elicit a response from the inion and found that 250- and 500-Hz stimuli produced the largest amplitudes. The majority of research on VEMPs has focussed on responses from the sternocleidomastoid muscle (SCM). The amplitude of this response directly relates to the amount of tonic activation of the SCM. If the subject is relaxed and there is no load on the muscle, the vestibular-evoked response is not detected 3 Prevec and colleagues investigated whether the somatosensory system contributes to the frequency-following response to high-level bone-conduction stimuli in deaf subjects (Prevec & Ribaric-Jankes, 1996; Ribaric, Prevec & Kozina, 1984). They concluded that it is highly improbable. 82 (Townsend & Cody, 1971). Response amplitudes were maximal at for 300-1000 Hz air-conduction brief tones presented at 95-100 dB nHL (Cheng, Huang, & Young, 2003; Welgampola & Colebatch, 2001; Todd, Cody & Banks, 2000). Sheykholeslami, Kermany & Kaga (2001) found that they could elicit SCM responses to 100-800 Hz bone-conduction brief tones and obtained maximum amplitudes for 200- and 400-Hz stimuli presented at 70 dB nHL. Recently, Nong, Ura and Noda (2000) reported an acoustically evoked short-latency negative response at 3-4 ms when recording ABRs in patients with profound hearing loss. They suggested that the saccule was the sense organ and that the response originated from the vestibular nucleus in the brainstem. It is not likely that VEMPs from the SCM are the source of the non-auditory ASSRs recorded in this study because the electrodes were not placed on or near this muscle and the subjects reclined in a chair in a relaxed state. It is possible, however, that non-auditory responses in this study could arise from VEMPs from the inion muscle because of its close proximity to the inverting electrode on the nape. Future ASSR research should investigate other electrode placements. The high-intensity bone-conduction and air-conduction stimuli used in this study were less intense than those used to elicit optimal VEMPs but may still have been sufficiently loud to elicit a response that could be mistaken for an ASSR 4. It is also possible that the response originates directly from the vestibular system (Nong et al., 2000). It was not possible to verily a vestibular contribution to the spurious ASSRs recorded in the present study; however, further support for this possibility is provided by Subject #10, who reported vestibular effects 4 It is conceivable that forehead placement of the bone oscillator might result in less vestibular stimulation, possibly due to a different mode of stimulation compared to temporal bone placement. However, intensity differences per se cannot explain this, as calibrations for 0 dB HL adjust for forehead-mastoid differences. 83 when presented with high-intensity air-conduction directly to her left ear. The results of this study have a number of general implications: (i) to avoid aliasing, the A/D rate should be selected so that the carrier frequencies are not integer multiples of the A/D rate, (ii) artifactual ASSRs may occur when determining steady-state response thresholds for bone-conduction stimuli, in any subject, (iii) artifactual ASSRs may also occur when assessing thresholds to air-conduction stimuli, especially for individuals with profound hearing loss, and (iv) alternating the stimulus polarity helps reduce artifactual responses arising from stimulus artifact. Finally, this study shows an important additional finding. Because the phases of the 500-Hz bone-conduction responses do not invert with inversion of stimulus polarity, some of these responses may be physiologic but non-auditory in origin. Perhaps the responses we obtained to the 500-Hz stimuli are vestibular in origin? The most important clinical implication that arises from the current study is that stimulus artifact must be considered when recording ASSRs using bone-conduction and high-intensity air-conduction stimuli for threshold estimation. Previous studies have suggested that ASSRs can be used to assess residual hearing in individuals with profound hearing loss because these responses can be elicited by stimuli at higher intensities than can be used to elicit ABRs. The results of this study indicate that high-intensity air- (and bone-) conduction stimuli can produce spurious ASSRs, particularly for low-frequency stimuli. It is common for individuals with profound hearing loss to have residual hearing only in the low frequencies; consequently, if an ASSR is present to a low-frequency high-intensity stimulus, there is no way to know if it is auditory or non-auditory in origin because of the possibility of activating the vestibular system and/or stimulus artifact in the EEG. Alternating stimulus polarity can help rule out stimulus artifact as a contributor, but not a non-auditory physiologic response. Current clinical equipment designed to 84 record multiple ASSRs, however, does not alternate stimulus polarity. It also must be re-iterated that there is a safety issue that must be considered when presenting high-level stimuli to the ear for as long as five minutes (a common length of time required to record threshold ASSRs). Exposure to these levels could cause a temporary threshold shift, tinnitus and, possibly, permanent damage. Based on the results of this study, it is clear that a rate such as 1250 Hz is required to record ASSRs to bone-conduction and high-intensity air-conduction stimuli so that the carrier frequency of the stimulus is not an integer multiple of the A/D rate. ASSRs to single-polarity bone-conduction stimuli using a 1250-Hz A/D rate can be recorded accurately for (i) 500-Hz stimuli up to 40 dB HL, (ii) 1000-Hz stimuli up to 50 dB HL, (iii) 2000-Hz stimuli up to at least 50 dB HL, and (iv) 4000-Hz stimuli up to 60 dB HL. ASSRs to alternated-polarity bone-conduction stimuli using a 1250-Hz A/D rate can be recorded accurately for (i) 500-Hz stimuli up to 40 dB HL, (ii) 1000-Hz stimuli up to 60 dB HL, (iii) 2000-Hz stimuli up to at least 50 (and probably 60) dB HL, and (iv) 4000-Hz stimuli up to at least 50 (and probably 60) dB HL. ASSRs to single-polarity air-conduction stimuli using a 1250-Hz rate can be recorded accurately for all frequencies, except, perhaps at 1000 Hz, up to at least 114-120 dB HL. ASSRs to alternated-polarity bone-conduction stimuli can be recorded accurately for all frequencies up to at least 114-120 dB HL. Even though ASSRs appear promising, bone-conduction ASSRs will not be ready for clinical use until there are normative threshold data for infants of different ages, and sufficient threshold data for infant and adult subjects with impaired hearing. 85 Tab le 2.1 Hear ing character ist ics of the participants Pure-tone behavioural thresholds (dB HL) Air conduction Bone conduction 500 1000 2000 4000 500 1000 2000 4000 # Experiment Device L R L R L R L R UR UR UR UR 1 1. 2 CI NR NR NR NR NR NR NR NR NR NR NR NR 2 1, 2 CI 110 NR NR NR NR NR NR NR NR NR NR NR 3 1, 2 HA 95 95 100 90 90 85 75 70 NR NR NR NR 4 1,2 none NR NR NR NR NR NR NR NR NR NR NR NR 5 1 CI 95 85 100 100 105 105 120 115 NR NR NR NR 6 1 none NR NR NR NR NR NR NR NR NR NR NR NR 7 1, 2 none NR NR NR NR NR NR NR NR NR NR NR NR 8 1. 2 HA 95 80 100 95 100 85 105 110 NR NR NR NR 9 2 CI NR 105 NR 105 NR 110 NR 120 NR NR NR NR 10 1, 2 CI, CI NR NR NR NR NR NR NR NR NR NR NR NR 11 1 HA NR 65 NR 80 NR 85 NR 70 NR NR NR NR 12 1, 2 HA NR NR NR NR NR NR NR NR NR NR NR NR 13 1, 2 HA 65 65 80 70 75 75 75 85 NR NR NR NR 14 1, 2 HA 85 90 100 95 95 95 80 85 55 NR NR NR 15 1. 2 HA 50 55 50 55 70 60 65 60 55 55 65 65 16 1.2 HA 65 NR 90 NR 90 NR 90 NR NR NR NR NR 17 1, 2 HA 70 60 65 70 70 70 65 75 NR NR NR NR CI - cochlear implant; HA = hearing aid(s); none= used sign language (2 of participants also blind); 1 = no response at highest intensity (110-120 dB HL) = Experiment 1 (bone-conduction stimuli); 2 = Experiment 2 (air-conduction stimuli); NR Figure 2.1 Polar plot for an individual subject (#8) indicating ASSRs elicited by 500- and 1000-Hz bone-conduction stimuli presented at 50 dB HL recorded using a 1000-Hz A/D rate. The polar coordinates of each point represent the amplitude and phase delay for significant ASSRs. Responses that did not meet the criterion for significance are not included in this figure. The number of responses for each polarity condition are shown to the left of the polar plots. The radius of the polar plot, "r", indicates the amplitude scale and is shown to the left of each plot. Non-inverted, inverted and alternated stimulus polarities were used to elicit responses and are denoted by a closed circle (•), open circle (°) and plus sign (+), respectively. The response at 1000 Hz for the alternated stimulus polarity was not significant (amplitude = 6 nV; p = .727), and thus not plotted. BONE CONDUCTION 500 Hz 1000 Hz 90° 90° • n on-inverted o inverted + alternated 87 Figure 2.2 Polar plots indicate ASSRs elicited by multiple bone-conduction stimuli presented at 50 dB HL recorded using three different A/D rates. Each point represents the amplitude and phase delay for significant ASSRs. Responses that did not meet the criterion for significance are not included in this figure. The radius of the polar plot, "r", indicates the amplitude scale and is shown to the left of each plot. Phase delay is shown in degrees. Non-inverted, inverted and alternated stimulus polarities were used to elicit responses and are denoted by a closed circle (•), open circle (°) and plus sign (+), respectively. A total of 16 subjects were tested (8-9 subjects were tested for each A/D rate). The total number of responses for each polarity condition are shown to the left of each polar plot. If the number of responses present per test condition were greater than the number expected by chance, the responses were considered "artifactual". For 8-9 subjects, the presence of three or more responses are considered artifactual. Artifactual results are denoted by an asterix (*). B O N E CONDUCTION 500 Hz 1000 Hz 2000 Hz 4000 Hz 1250 Hz A/D 1000 Hz A/D 500 Hz A/D • non-inverted o inverted + alternated 88 Figure 2.3 Percent occurrence of responses elicited by single-polarity bone-conduction stimuli recorded for different A/D rates, carrier frequencies and intensities. Carrier frequencies are shown in each panel. Within each panel, the results for intensity and A/D rate are indicated. The asterix indicates that the number of responses is significantly greater than chance and, therefore, artifactual. 'DNT' denotes 'did not test'. A total of 16 subjects participated (8-9 subjects at 50 dB HL and 3-8 at 40 dB HL). 0 o c <D 0 3 c O o O Q. o a> c o o CD CL 100 80 60 40 20 0 100 80 60 40 20 0 Bone-conduction: Single polarity 500 Hz 2000 Hz 0 0 0 O D D 40 50 60 1000 Hz * H 4000 Hz 500-Hz AD 1000-Hz AD 1250-Hz AD 0 0 0 40 50 h l - K z z z Q Q Q 1 60 Intensity (dBHL) 89 Figure 2.4 Percent occurrence of responses elicited by "alternated" stimuli bone-conduction recorded for different A/D rates, carrier frequencies and intensities. Carrier frequencies are shown in each panel. Within each panel, the results for intensity and A/D rate are indicated. The asterix indicates that the number of responses was significantly greater than expected by chance and, therefore, artifactual. 'DNT' denotes 'did not test'. A total of 16 subjects participated (7-8 subjects at 50 dB HL and 3-8 at 40 dB HL). Bone conduction: Alternated polarity 40 50 60 1000 Hz oo 4000 Hz 500-Hz AD 1000-Hz AD 1250-Hz AD H | - h - H z z z z 0 0 Q 0 0 0 Q Q Q 40 50 60 Intensity (dBHL) 90 Figure 2.5 Polar plots indicate significant ASSRs elicited by air-conduction stimuli presented at 116 (500 Hz) and 119 dB HL (1000 Hz) at three different A/D rates. Each point represents the amplitude and phase delay for significant ASSRs. Responses that did not meet the criterion for significance are not included in this figure. The radius of the polar plot, "r", indicates the amplitude scale and is shown to the left of each polar plot. Phase delay is shown in degrees. Non-inverted, inverted and alternated stimulus polarities were used to elicit responses and are denoted by a closed circle (•), open circle (°) and plus sign (+), respectively. For each A/D rate, 6 subjects were tested. A total of 14 subjects participated. The total number of responses for each polarity condition are shown to the left of each polar plot. If the number of responses present per test condition were greater than the number expected by chance, the responses were considered "artifactual". For six subjects, the presence of two or more responses is considered artifactual. Artifactual results are denoted by an asterix (*). Results do not include those for left ear of subject #10 (who had both ears tested, 500-Hz A/D), which showed responses to both polarities and to alternated stimuli. AIR CONDUCTION 500 Hz 1000 Hz 90° 90° • n on- inverted o inverted + alternated 91 Figure 2.6 Percent occurrence of responses elicited by single and alternated air-conduction stimuli presented at 116 and 119 dB HL and recorded for different A/D rates and carrier frequencies. Results for single and alternated polarity are also shown. The asterix indicates that the number of responses was significantly greater than expected by chance and, therefore, artifactual. For each A/D rate, 6 subjects were tested. A total of 14 subjects participated. Left-ear results for subject #10 are not included. AIR CONDUCTION 0 o C 0 3 O O o c 0 o 0 Q-100 0 (/> 80 c O 60 a CO 40 •I o 20 0 I 1 500-Hz AD I 1 1000-Hz AD mmm 1250-Hz AD 0 0 0 0 0 000 500 500 1000 1000 single alt single alt Carrier Frequency (Hz) 92 REFERENCES ANSI (1996). American National Standard Specifications for Audiometers (ANSI S3.6-1996). New York, ANSI. Aoyagi, M . , Kiren, T. & Furuse, H. (1994). Pure-tone threshold prediction by 80-Hz amplitude-modulated following response. Acta Otolaryngolica Supplement 511:7-14. Bickford, R. G., Jacobson, J. L. & Cody, D. (1964). Averaged potentials to sound and other stimuli in man. Annals of the New York Academy of Science 112: 201-223. Cheng, P.-W., Huang, T.-W. & Young, Y. -H. (2003). The influence of clicks versus short tone bursts on the vestibular evoked myogenic potential. Ear and Hearing 24: 195-197. Cohen, L. T., Rickards, F. W. & Clark, G. M . (1991). A comparison of steady-state evoked potentials to modulated tones in awake and sleeping humans. Journal of the Acoustical Society of America 90: 2467-2479. Cone-Wesson, B., Rickards, F., Poulis, C , Parker, J., Tan, L.. & Pollard, J. (2002). The auditory steady-state response: clinical observations and applications in infants and children. Journal of the American Academy of Audiology 13: 270-282. Dimitrijevic, A., John, M . S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., Nedzelski, J. M . & Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology 13: 205-224. Dirks, D., & Swindeman, J. G. (1967). The variability of occluded and unoccluded bone-conduction thresholds. Journal of Speech and Hearing Research, 10: 232-249. Foxe, J. J. & Stapells, D. R. (1993). Normal infant and adult auditory brainstem responses to bone-conducted tones. Audiology 32: 95-109. 93 Geisler, C. D., Frishkopf, L . S. & Rosenblith, W. A. (1958). Extracranial responses to acoustic clicks in man. Science 128: 1210-1211. Goldstein, R., & Hayes, C. (1965). The occlusion effect in bone conduction hearing. Journal of Speech and Hearing Research, 8: 137-148. Gorga, M.P., Neely, S.T. Hoover, B.M. Dierking, D.M. Beauchaine, K.L. & Manning, C. (2004). Determining the upper limits of stimulation for auditory steady-state response measurements. Ear & Hearing, 25: 302-307. Hall, J.W. (1992). Handbook of Auditory Evoked Responses, Needham Heights: Allyn & Bacon. Herdman, A. T. & Stapells, D. R. (2001). Thresholds determined using the monotic and dichotic multiple auditory steady-state response technique in normal-hearing subjects. Scandinavian Audiology 30: 41-49. Herdman, A. T. & Stapells, D. R. (2003). Auditory steady-state response thresholds of adults with sensorineural hearing impairment. International Journal of Audiology, 42: 237-248. Hyde, M.L., Sininger, M.L. & Don, M . (1998). Objective detection and analysis of auditory brainstem response: An historical perspective. Seminars in Hearing 19(1): 97-114. Jahrsdoerfer, R. A., Yeakley, J. W., Hall, J. W., Robbins, K. T. & Gray, L. C. (1985). High-resolution CT scanning and auditory brain stem response in congenital aural atresia: Patient selection and surgical correlation. Otolaryngology-Head and Neck Surgery 93: 292-298. Jeng, F - C , Brown, C.J., Johnson, T.A. & Vander Werff, K.R. (2004). Estimating air-bone gaps using auditory steady-state responses. Journal of the American Academy of Audiology 75:67-78. 94 Jerger, J. & Jerger, S. (1965). Critical evaluation of SAL audiometry. Journal of Speech and Hearing Research 8: 103-127. John, M . S. & Picton, T. W. (2000). MASTER: a Windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine 61: 125-150. Lins, O. G., Picton, T. W., Boucher, B. L., Durieux-Smith, A., Champagne, S. C , Moran, L. M . , Perez-Abalo, M . C , Martin, V. & Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear and Hearing 17: 81-96. National Instruments. (2000). LabVIEW™ Measurements Manual. Austin, Texas: National Instruments. Nong, D.X., Ura, M . & Noda, Y. (2000). An acoustically-evoked short latency negative response in profound subjects. Acta Otolaryngologica 128(8): 960-966. Perez-Abalo, M . C , Savio, G., Torres, A., Martin, V., Rodriguez, E. & Galan, L. (2001). Steady state responses to multiple amplitude-modulated tones: an optimized method to test frequency-specific thresholds in hearing-impaired children and normal-hearing subjects. Ear and Hearing 22: 200-211. Picton, T. W., Durieux-Smith, A. & Moran, L. M . (1994). Recording auditory brainstem responses from.infants. International Journal of Pediatric Otorhinolaryngology 28: 93-110. Picton, T. W., Durieux-Smith, A., Champagne, S. C , Whittingham, J., Moran, L. M . , Giguere, C. & Beauregard, Y. (1998). Objective evaluation of aided thresholds using auditory steady-state responses. Journal ofAmerican Academy of Audiology 9: 315-331. 95 Picton, T.W. & John, M.S. (2004). Avoiding electromagnetic artifacts when recording auditory steady-state responses. Journal of American Academy of Audiology 15: 541-554. Picton, T. W., Hink, R. F., Perez-Abalo, M . , Linden, R. D. & Wiens, A. S. (1984). Evoked potentials: HowNow? Journal of Electrophysiological Technology 10: 177-221. Picton, T.W., John, M.S., Dimitrijevic, A. & Purcell, D. (2003). Auditory steady-state responses. International Journal of Audiology 42: 177-219. Prevec, T. S., & Ribaric^Jankes, K. (1996). Can somatosensory system generate frequency following response? Pfliigers Archives-European Journal of Physiology, 437(Supplement), R301-302. Ribaric, K., Prevec, T. S., & Kozina, V. (1984). Frequency-following response evoked by acoustic stimuli in normal and profoundly deaf subjects. Audiology, 23, 388-400. Sheykholeslami, K., Kernmany, M.K. & Kaga, K. (2000). Bone-conducted vestibular-evoked myogenic potentials in patients with congenital atresia of the external auditory canal. International Journal of Pediatric Otorhinolaryngology 57: 25-29. Small, S.A. & Stapells, D.R. (2005). Multiple auditory steady-state responses to bone-conduction stimuli in adults with normal hearing. Journal of the American Academy of Audiology 16:172-183. [Chapter 3] Stapells, D. R. (2000a). Frequency-specific evoked potential audiometry in infants. In R.C. Seewald (ed.) A Sound Foundation through Early Amplification: Proceedings of an International Conference (pp. 13-31). Chicago: Phonak A G . Stapells, D. R. (2000b). Threshold estimation by the tone-evoked auditory brainstem responses: A literature meta-analysis. Journal of Speech-Language Pathology and Audiology 24(2): 74-83. 96 Stapells, D. R. & Ruben, R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. Annals of Otology, Rhinology and Laryngology 98: 941-949. Todd, N., Cody, P. & Banks, P. (2000). A saccular origin of frequency tuning in myogenic vestibular evoked potentials. Hearing Research 141: 180-188. Townsend, G. L. and Cody, D. (1971). The averaged inion response evoked by acoustic stimulation: its relation to the saccule. Annals of Otology 80: 121-132. von Bekesy, G. (1935). Uber Akustische Reizung des Vestibularapparates. Archiv fur die gemaste Physiologie des Menschen und der Tiere 236: 59-76. Welgampola, M . S. and Colebatch J.G. (2001). Characteristics of tone burst-evoked myogenic potentials in the sternocleidomastoid muscles. Otology and Neurotology 22: 796-802. Zar, J.H. (1984). Biostatistical analysis. Englewood Cliffs, New Jersey: Prentice-Hall. 97 CHAPTER 3 MULTIPLE AUDITORY STEADY-STATE RESPONSES TO BONE-CONDUCTION STIMULI IN ADULTS WITH NORMAL HEARING A version of this chapter has been published. Small, S. A., & Stapells, D. R. (2005). Multiple auditory steady-state responses to bone-conduction stimuli in adults with normal hearing. Journal of the American Academy of Audiology 16: 172-183. 98 INTRODUCTION Auditory evoked potentials (AEPs) are required to estimate hearing thresholds at audiometric frequencies for individuals who cannot be tested using conventional behavioural measures. Infants, young children, and multiply handicapped individuals who are difficult to test are typically the individuals who benefit most from evoked potential audiometry. There are also adults who are assessed using AEPs in cases where hearing loss is a condition under consideration for monetary compensation. Auditory brainstem responses (ABRs) are the AEPs that are currently used clinically for threshold assessment in infants and young children. ABRs may be evoked by air- and bone-conduction brief-tone stimuli to obtain frequency-specific audiometric information in sleeping and relaxed subjects (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stapells, 2000a, b). One shortcoming of the ABR technique is that only one ear and one frequency can be tested at the same time. Another limitation of the ABR to brief tones is that detection of a response in the waveform depends on skilled, subjective assessment of replicated responses, allowing for error in judgement of the presence of responses depending on the experience of the clinician (Stapells, 2000a). Auditory steady-state responses (ASSRs) use amplitude and/or frequency modulated stimuli to evoke AEPs, and are currently of great interest because they can quickly and objectively obtain frequency-specific thresholds (for review, see Picton et al., 2003). ASSRs can be recorded for single- or multiple-carrier frequencies to one, or both, ears simultaneously. ASSRs are detected objectively using statistical tests (John & Picton, 2000); their detection does not rely on the experience of the clinician. Multiple ASSRs are thus of considerable interest as an assessment tool because of their objectivity and potential for reducing clinical testing time. 99 ASSRs to air-conduction stimuli have been found to provide reasonably accurate prediction of hearing sensitivity at the audiometric frequencies for adults and young children (for review, see Picton et al., 2003). In order to distinguish among sensorineural, conductive and mixed hearing losses, AEP techniques must provide thresholds for both air- and bone-conduction stimuli, as is routinely done in behavioural audiometry. Frequency-specific thresholds and identification of type of hearing loss are necessary to make decisions regarding medical intervention and planning aural (re)habilitation. Accurate bone-conduction thresholds are particularly important when assessing children who have unilateral or bilateral otitis media or atresia (Jahrsdoerfer et al., 1985; Stapells & Ruben, 1989). ASSRs to bone-conduction stimuli have not been thoroughly investigated. At the outset of the present study, only three previous studies had reported findings for ASSRs elicited by bone-conduction stimuli to elicit ASSRs. Two of these studies used bone-conduction stimuli presented at the forehead to elicit ASSRs in adults with normal hearing (Lins et al., 1996; Dimitrijevic et al., 2002); a third study, conducted in our lab, recorded ASSRs to bone-conduction stimuli presented at the mastoid of adults with severe-to-profound hearing loss (Small & Stapells, 2004: Chapter 2). All three studies that recorded ASSRs to bone-conduction stimuli reported results that are different for 500 Hz compared to higher carrier frequencies. Lins et al. (1996) and Dimitrijevic et al. (2002) assessed responses to bone-conduction stimuli no more than 20-30 dB above threshold and found differences in amplitude/phase measures for bone- vs air-conduction ASSRs, particularly, for 500- and 1000-Hz carrier frequencies. In a previous study, we recorded ASSRs in individuals who could not hear the stimuli and clearly showed that bone- and high-intensity air-conduction stimuli can produce spurious "responses", especially for 500- and 1000-Hz carrier frequencies (Small & Stapells, 2004: Chapter 2). Based on these findings, it is possible that 100 some of the results reported by Lins et al. and Dimitrijevic et al. may have been contaminated by stimulus artifact. The presence of spurious responses in subjects who cannot hear the stimulus is explained, in part, by the presence of high-amplitude stimulus artifact in the E E G produced by the bone oscillator. A significant problem with bone-conduction stimulus artifact in the E E G is that this energy can alias to exactly the same frequency as the ASSR modulation rate of the stimulus, and be interpreted as a response. "Alternating" the stimulus polarity is a common technique used to remove or reduce stimulus artifact when recording ABRs (e.g., Hall, 1992, pg. 319) and can also be used to reduce the effect of stimulus artifact that is present in the E E G when recording ASSRs. For ASSR recordings, this can be accomplished by inverting the stimulus, then averaging offline the responses to the inverted and non-inverted stimuli to obtain a response representing the "alternated stimulus polarity". In our previous study (Small & Stapells, 2004: Chapter 2), we investigated ways to reduce or eliminate stimulus artifact. We found that the use of 500-Hz or 1000-Hz A/D rates and single-polarity stimuli, which have been commonly used to record ASSRs (Dimitrijevic et al., 2002; Herdman & Stapells, 2001, 2003), resulted in significant artifactual responses for 500-, 1000- and 2000-Hz carrier frequencies. Use of an alternated stimulus polarity significantly reduce the number of spurious responses at these rates by cancelling out artifact in the ASSRs. Use of a 1250-Hz A/D rate and insertion of a 300-Hz steep lowpass anti-aliasing E E G filter helped avoid spurious responses by preventing aliasing. In some of the subjects with severe-to-profound hearing loss, spurious responses to 500 Hz remained even after changing the A/D rate and adding in an additional anti-aliasing filter, and may be non-auditory physiologic responses (perhaps vestibular in nature). 101 Based on the results of our previous research in individuals with severe-to-profound sensorineural hearing loss (Small & Stapells, 2004: Chapter 2), we concluded that artifactual responses to bone-conduction stimuli are present for the following conditions. For single-polarity 500- and 1000-Hz stimuli, artifactual responses are present at: (i) 20-40 dB HL and higher for a 500- and 1000-Hz A/D rate, and (ii) 50-60 dB HL and higher for a 1250-Hz A/D rate. For single-polarity 2000-Hz stimuli, artifactual responses are present at: (i) 50 dB HL and higher for 500-Hz and 1000-Hz A/D rates, and (ii) not present for a 1250-Hz A/D rate. Artifactual responses are not present to single-polarity 4000-Hz stimuli at any A/D rate. Based on these findings, it is likely that Lins et al. (1996) and Dimitrijevic et al. (2002) used stimuli that were in the intensity and frequency range that result in artifactual responses. The present study had three purposes. First, to determine ASSR thresholds to multiple bone-conduction stimuli presented at the mastoid, in adult participants with normal hearing, using stimulus and recording parameters selected to minimize stimulus artifact in the E E G and its resultant spurious responses. Second, to investigate the effects of changing stimulus polarity on the ASSRs in normal-hearing subjects. Third, to compare amplitude and phase characteristics of ASSRs to bone-conduction, presumed to reflect the response of both cochleae, to ASSRs elicited by air-conduction stimuli presented binaurally (diotically), known to reflect the response of both cochleae. M A T E R I A L S AND M E T H O D S Participants Two groups of individuals with normal hearing participated. ASSRs to air-conduction stimuli were recorded for a group ("AC-ASSR") of 10 adults aged 19 to 37 years. ASSRs to 102 bone-conduction stimuli were recorded in another group ("BC-ASSR") of 10 adults aged 20 to 48 years. Initially, we obtained A C - and BC-ASSR recordings for the same group of adults but realized that we had stimulus artifact in our bone-conduction results. We then obtained BC-ASSRs using a higher A/D rate in a different group of adults; stimulus artifact was not an issue for the AC-ASSRs using the lower rate; thus the air-conduction recordings were not repeated. All participants had normal hearing, with behavioural air- and bone-conduction thresholds of 20 dB HL (ANSI, 1996) or better in both ears from 250 to 8000 Hz. The mean behavioural pure-tone air- and bone-conduction thresholds, shown in Table 3.1, are similar for these two groups. Table 3.1 Stimuli All stimuli were sinusoidal tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. The stimuli were presented simultaneously for all conditions tested. Air- and bone-conduction stimuli were generated by the Rotman MASTER research system (John & Picton, 2000), routed through Tucker-Davis Technologies SM3 and HB6 modules to allow presentation of non-inverted and inverted stimuli, and attenuated through a clinical audiometer (Interacoustics AC40). Bone-conduction stimuli were presented to a Radioear B-71 bone oscillator which was held in position on the temporal bone within 2 cm of the pinna with a wide elastic headband fastened with Velcro™ (Universal Facial Band #210, Design Veronique, Oakland, CA) with 450 - 550 g of force. Bone-conduction stimuli were presented using 10-dB steps at: (i) 0 to 50 103 dB HL for non-inverted stimuli, (ii) 30 to 50 dB HL for inverted stimuli, and (iii) 30 to 50 dB HL for "alternated" stimuli. ASSRs to "alternated" stimuli were obtained by averaging offline the ASSR waveforms to non-inverted and inverted stimuli offline (Small & Stapells, 2003: Chapter 2). ASSRs to alternated stimuli were not tested at levels lower than 30 dB HL because stimulus artifact and aliasing is considered negligible for bone-conduction stimuli less than 30 dB HL, provided a 1250-Hz A/D rate is used (Small & Stapells, 2003: Chapter 2); that is, alternated stimuli were not needed to help reduce the effects of stimulus artifact at these levels. Air-conduction stimuli were presented using EAR-3A insert earphones. Air-conduction stimuli were presented binaurally (diotically) at 30, 40 and 60 dB H L to be comparable to the bone-conduction stimuli that were presented across a range of intensities. Stimulus artifact is not expected to be a problem for air-conduction stimuli presented at moderate levels; therefore, only single-polarity stimuli ("non-inverted") were used. Additionally, as a check of whether "alternating" stimuli degraded the ASSR, monaural air-conduction stimuli were presented at 40 dB HL using non-inverted and inverted stimuli. A 40 dB HL air-conduction stimulus was used to ensure that stimulus artifact was not an issue; monaural presentation avoided possible additional complexities of stimulating both cochleae. Calibration The bone-conduction stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re:lpN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 550 g of force. Air-conduction stimuli were calibrated in Reference Equivalent Threshold Sound Pressure Levels (RETSPL) corresponding to 0 dB HL (ANSI S3.6-1996) using a Quest Model 1800 sound level meter and a Briiel and 104 Kjaer DB0138 2-cc coupler. ASSR Recording ASSRs were recorded using the Rotman MASTER system. Three gold-plated electrodes were used to record the electrophysiologic responses; the non-inverting electrode was placed at Cz, the inverting electrode was positioned at the nape of the neck, just below the hairline, and an electrode placed at the high forehead acted as ground. All inter-electrode impedances were below 3kOhmsat 10 Hz. For ASSRs elicited by bone-conduction stimuli, the E E G was filtered using 30-250 Hz filter (12 dB/oct) and amplified 80,000 times (Nicolet HGA-200A and Nic501A). The E E G was further filtered using a 300-Hz lowpass anti-aliasing filter (Stanford Research Systems; 115dB/oct), and the E E G was then processed using a 1250-Hz A/D conversion rate. The digital-to-analog (D/A) rate was 31,250 Hz. Each E E G recording sweep was made up of 16 epochs of 1024 data points and lasted a total of 13.107 seconds. For ASSRs elicited by air-conduction stimuli, the E E G was filtered using a 30-250 Hz (12 dB/oct) filter, amplified 80,000 times (Nicolet HGA-200A and Nic501 A) and processed using a 500-Hz A/D rate (Herdman & Stapells, 2001, 2003). The D/A rate was 32,000 Hz. Each E E G recording sweep was made up of 16 epochs of 1024 data points and lasted a total of 16.384 seconds. Artifact rejection for bone- and air-conduction stimuli was set to eliminate epochs of electrophysiologic activity that exceeded ± 40 | iV in amplitude in order to reduce contributions to the E E G due to muscle artifact. ASSRs were averaged in the time domain and then analysed online into the frequency domain using a Fast Fourier Transform (FFT). The FFT resolution was 0.060 Hz over a range of 0 to 250 Hz for the 500-Hz A/D rate and 0.076 Hz over a range of 0 to 625 Hz for the 1250-Hz 105 A/D rate. Amplitudes were measured baseline-to-peak and expressed in nV. An F-ratio was calculated by MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ± 60 bins of the modulation frequency ("noise") (John and Picton, 2000). A response was considered to be present if the F-ratio, compared to the critical values for F (2, 240), was significant at a level of/? < .05. A response was considered to be absent if p > .05 and the amplitude of the noise was less than 11 nV. Alternatively, a response was also considered to be absent when response amplitude was < 10 nV and if the p value > .30. Amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant. The phase values from MASTER were adjusted by adding 90° to yield the onset phase (John & Picton, 2000). Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Any phase-delay values that differed > 180° from an adjacent measure were "unwrapped" by adding 360° to their value (John & Picton, 2000). Phase values for ASSRs that were not significant were excluded from mean phase-delay calculations. Phase-delay values were averaged across subjects. Results are reported only if at least five subjects contributed to the mean. Procedure Testing was performed in a double-walled sound-attenuating booth. Participants reclined in a comfortable chair and were instructed to relax or sleep during the ASSR test session. Behavioural thresholds for pure-tone air- and bone-conduction stimuli were obtained at the beginning of the session to establish normal hearing thresholds. ASSRs were elicited to bone-conduction stimuli in descending order of intensity. ASSRs were elicited to air-conduction 106 stimuli in a randomized order of intensity. The total testing time was approximately 1.5 hours, including the time to obtain behavioural thresholds. Participants signed a consent form before commencing any of the experiments and were paid an honorarium at the end of each session. Data Analyses Changes in ASSR amplitude and phase-delay were compared across stimulus polarity condition ("non-inverted vs inverted", "non-inverted vs alternated" and "inverted vs alternated") and carrier frequency (500, 1000, 2000 and 4000 Hz) and between mode of stimulus presentation (bone- vs binaural air-conduction). Mean amplitude and phase-delay values were calculated for ASSRs elicited by monaural air-conduction stimuli. The slope of the amplitude and phase-delay functions with intensity were calculated from 30-50 dB HL for bone-conduction stimuli, and 30-60 dB HL for binaural air-conduction stimuli. In order to compare bone-conduction to air-conduction (binaural), phase delay was also converted to latency (in milliseconds), after correcting for the insert earphone delay (-0.92 ms). Comparisons across stimulus polarity and carrier frequency for amplitude and phase-delay measures were made for ASSRs to monaural air-conduction stimuli presented at 40 dB HL using a two-way repeated-measures analysis of variance (ANOVA). Comparisons across stimulus polarity and carrier frequency for the slopes of the intensity-amplitude and intensity-phase-delay functions for ASSRs elicited by bone-conduction stimuli were also made using a two-way repeated-measures ANOVA. Comparisons between mode of presentation and across carrier frequency for the slopes of amplitude- and phase-delay-intensity functions for ASSRs elicited by binaural air- and bone-conduction stimuli were made using a mixed ANOVA. Only alternated bone-conduction ASSR phase-delay results were included in the mixed ANOVA. Huynh-Feldt epsilon-adjustments for repeated measures were made when appropriate. Newman-107 Keuls post-hoc comparisons were performed for significant main effects and interactions. The criterion for statistical significance was p < .05 for all analyses. RESULTS Normal bone-conduction ASSR thresholds The mean (1 SD) bone-conduction ASSR thresholds in these subjects with normal hearing were 22.0 (11.4), 26.0 (13.5), 18.0 (7.9) and 18.0 (11.4) dBHL for 500, 1000, 2000 and 4000 Hz, respectively. The percentage of participants who had thresholds < 20 dB HL were 60, 40, 90 and 60% at 500, 1000, 2000 and 4000 Hz, respectively. The percentage of participants who had thresholds <30 dB HL were 90, 70, 100 and 100% at 500, 1000, 2000 and 4000 Hz, respectively. Monaural air-conduction stimuli (single- and alternated stimulus polarities) As shown in Table 3.2, the mean amplitude and phase-delay values for ASSRs elicited by monaural air-conduction stimuli presented at 40 dB HL were similar for the non-inverted, inverted and alternated stimulus polarity conditions. The mean amplitude values across frequency were 46.5, 40.0 and 42.7 nV for non-inverted, inverted and alternated stimulus polarities, respectively. Results of an A N O V A comparing ASSR amplitudes across polarity and carrier frequency revealed that there was no effect of frequency nor an interaction between frequency and polarity, but there was a small but significant effect of polarity (p=.030) (Table 3.3). Post hoc comparisons indicated that the significant polarity main effect was due to a small but significant difference between amplitudes for "non-inverted vs inverted" stimulus polarities. There were no significant differences between ASSR amplitudes to "non-inverted vs alternated" nor to "inverted vs alternated" stimulus polarities. 108 Tables 3.2 and 3.3 Results of an A N O V A comparing ASSR phase delay across polarity and carrier frequency revealed no interaction between frequency and polarity but significant main effects of polarity (p = .035) and frequency (p = .007). The mean phase-delay values across frequency were 125.6, 138.8 and 129.2 degrees for non-inverted, inverted and alternated stimulus polarities, respectively. Post hoc comparisons indicated the small difference in ASSR phase delay for "non-inverted vs inverted" stimulus polarities was significant, but no significant difference was present for "non-inverted vs alternated" nor to "inverted vs alternated" stimulus polarities. Post hoc comparisons also indicated that phase delay was significantly longer for 1000 Hz compared to 2000 and 4000 Hz; no differences were found for 500 Hz compared to higher carrier frequencies. Bone-conduction stimuli (single- and alternated stimulus polarities) As shown in Table 3.4, there were no differences in the slopes of the intensity-amplitude functions between alternated and single-polarity stimuli for bone-conduction ASSRs. Amplitude slope tended to increase with decreases in frequency, which is also indicated in Figures 3.1 and 3.2 and Table 3.4. A two-way repeated measures A N O V A indicated that there was no effect of polarity nor an interaction between stimulus polarity and carrier frequency for amplitude; however, there were significant differences across carrier frequency, as shown in Table 3.5. Post-hoc comparisons indicated that the mean slope for amplitude was significantly steeper for the 500-Hz carrier frequency compared to higher carrier frequencies. 109 Tables 3.4 and 3.5 and Figures 3.1 and 3.2 As shown in Table 3.4, there were also no differences in the slopes of the intensity-phase-delay functions between alternated and single-polarity stimuli for bone-conduction ASSRs. Figures 3.3 and 3.4 and Table 3.4 indicate that the phase-delay slopes for ASSRs to 500, 2000 and 4000-Hz stimuli tended to be similar; the phase-delay slopes for ASSRs to 1000-Hz stimuli tended to be steeper than the other carrier frequencies. A two-way repeated measures A N O V A , shown in Table 3.5, indicated that, for phase delay, there was no effect of polarity nor an interaction between stimulus polarity and carrier frequency; however, there were significant differences across carrier frequency. Post-hoc pair-wise comparisons indicated that the mean slope for phase-delay for the 1000-Hz carrier frequency was significantly steeper compared to the 500-, 2000- and 4000-Hz carrier frequencies. Figures 3.3 and 3.4 Bone-conduction vs binaural air-conduction ASSRs Figures 3.1 and 3.2 and Table 3.4 show that amplitudes for ASSRs to binaural air-conduction stimuli (30, 40 & 60 dB HL) and bone-conduction stimuli (30, 40 & 50 dB HL) had the same slope with increased intensity for 1000-, 2000- and 4000-Hz carrier frequencies. For 500 Hz, the increase in amplitude with intensity was steeper for bone-conduction ASSR amplitudes (single- and alternated-stimulus polarities) compared to air-conduction ASSR amplitudes (Table 3.4). Results of a two-way mixed A N O V A for amplitude showed that ASSRs 110 to bone-conduction stimuli had steeper intensity-amplitude slopes than ASSRs to air-conduction stimuli. There was also a significant main effect for frequency and a significant interaction between mode of stimulus and carrier frequency (Table 3.6). Post-hoc pair-wise comparisons for the interaction between frequency and mode of stimulus indicated that the slope for amplitude for ASSRs to the 500-Hz bone-conduction stimulus was significantly steeper than any of the ASSRs to bone-conduction stimuli at higher frequencies or air-conduction stimuli at any frequency. Table 3.6 Figures 3.3 and 3.4 and Table 3.4 show that the intensity-phase-delay functions across carrier frequencies for air- and bone-conduction ASSRs tended to be steeper at 1000 Hz compared to the other carrier frequencies. The increase in phase delay with intensity was steeper for bone-conduction ASSR phase delay compared to air-conduction ASSR phase delay. Results of a two-way mixed A N O V A showed that ASSRs to bone-conduction stimuli had steeper slopes for phase delay than those to air-conduction stimuli and that there were differences in slope across carrier frequencies. There was no significant interaction between mode of stimulus and carrier frequency (Table 3.6). Post-hoc comparisons of carrier frequency (pooled across air- and bone-conduction mode) indicated that the slope for phase delay was significantly steeper for 1000-Hz stimuli compared to 2000- or 4000-Hz but no difference was found between 500- and 1000-Hz stimuli. Figures 3.3 and 3.4 also show that bone-conduction ASSRs had longer phase delay than binaural air-conduction ASSRs. Phase delay was converted to milliseconds by averaging the phase delay of the ASSRs elicited by 30 and 40 dBHL stimuli and dividing by (360 x modulation 111 frequency). The mean differences in phase delay in milliseconds (ms) between bone- (non-inverted) and air-conduction ASSRs were 2.49, 3.52, 3.48 and 3.81 ms at 500, 1000, 2000 and 4000 Hz, respectively. Similarly, the mean differences in phase delay in milliseconds (ms) between bone- (alternated) and air-conduction ASSRs were 2.42, 3.32, 3.48 and 3.91 ms at 500, 1000, 2000 and 4000 Hz, respectively. DISCUSSION The bone-conduction ASSR thresholds in the present study range from 18-26 dB HL for 500-, 1000-, 2000- and 4000-Hz stimuli, respectively. As shown in Table 3.7, comparison of the bone-conduction ASSR thresholds to previously reported results (Lins et al., 1996; Herdman & Stapells, 2001; Dimitrijevic et al., 2002; Jeng et al., 2004) indicates values with substantial variability across studies. All bone-conduction thresholds in Table 3.7, except for the thresholds obtained in the present study, are reported in dB HL at the mastoid and corrected for the occlusion effect (16, 8, 0 and 2 dB at 500, 1000, 2000 and 4000, respectively). Averaged across frequencies, thresholds in the present study are approximately 5 dB higher than those reported by Dimitrijevic et al. (2002), 4 dB lower than those of Lins et al. (1996), and 18 dB lower those reported by Jeng et al. (2004). Much of the variability in the threshold values may be due to: (i) differences in the placement of the bone oscillator (forehead vs mastoid), (ii) occluded vs unoccluded ears, (iii) method of stimulus calibration and (iv) inherent variability in bone-conduction calibration, and (v) small sample sizes. The application of correction factors to allow comparisons across studies also introduce variability. Differences in ASSR thresholds to bone-conduction stimuli across studies should therefore be interpreted cautiously because of these sources of variability in its measurement. 112 Bone-conduction ASSR thresholds in the present study show similar values compared to air-conduction ASSR thresholds reported by other studies. For example, as shown in Table 3.7, Herdman and Stapells (2001) reported air-conduction thresholds (converted to dB HL) that are within 3-7 dB of our bone-conduction ASSR thresholds (in dB HL). Table 3.7 Bone-conduction ASSRs have longer latencies compared to air-conduction ASSRs. The direction of these results is consistent with those reported by Boezeman et al. (1983) and Gorga et al. (1993) for ABR wave V. However, the ASSR latency differences in the present study (2.42-3.91 ms) are substantially greater than the differences reported by Boezeman et al. (0.88 ms for a 2000-Hz brief tone) and Gorga and colleagues (0.16-0.41 ms for 250- to 4000-Hz brief tones). Possible explanations for these results are: (i) latency differences derived from ASSR phase delays may not be exactly the same as ABR latencies, and/or (ii) bone conduction at the mastoid may not be equivalent in effective level to binaural air conduction. Stenfelt and Hakansson (2002) found a 6-10 dB difference in loudness between air- and bone-conduction stimuli for 250- to 4000-Hz stimuli presented at 30-80 dB HL. They attributed this result to distortion from the bone transducer, multi-modal stimulation, and differences in transmission path affecting the perceived loudness. The differences in amplitude between air- and bone-conduction ASSRs elicited in the present study are similar to the results reported by Dimitrijevic et al. (2002). In both studies, ASSR intensity-amplitude functions were steeper for 500-Hz bone-conduction stimuli compared to air-conduction stimuli. Dimitrijevic et al. (2002) used bone-conduction white-noise masking 113 in an attempt to eliminate the multiple ASSRs to bone-conduction stimuli in order to confirm that the responses were physiologic, not artifactual. They found that not all ASSRs to bone-conduction stimuli presented at 30 dB SL were eliminated by masking; responses were present in 17.5% of the subjects, which was significantly greater than the 5% expected by chance. The majority of these responses were to 1000-Hz stimuli (A. Dimitrijevic; personal communication). Dimitrijevic et al. (2002) suggested that this result was either due to undermasking or that a small amount of stimulus artifact was present in the response. Dimitrijevic et al. used a 1000-Hz A/D rate, a rate which we have shown can result in aliasing of stimulus artifact in the E E G at higher bone-conduction intensities (Small & Stapells, 2004: Chapter 2). Lins et al. (1996) also found differences between air- and bone-conduction ASSR amplitudes to 500- and 1000-Hz carrier frequencies compared to higher carrier frequencies, but they were unable to explain their results. They used a 679-Hz A/D rate to digitize the EEG. Although aliasing would likely have occurred, artifact in the E E G would not have aliased at exactly the modulation rate of the carrier frequency. Although aliasing may not have been an issue, "non-auditory" physiologic responses to 500- and 1000-Hz stimuli may account for their unexpected results (Small & Stapells, 2004: Chapter 2). In the present study, we found that ASSR phase delay systematically decreased with increasing intensity for air- and bone-conduction stimuli (Figures 3.3 and 3.4) except for 500 Hz air-conduction stimuli. Dimitrijevic et al. (2002) reported similar results for air-conduction stimuli. In contrast to the present study, Dimitrijevic et al. (2002) found no clear relationship between onset phase and intensity for bone-conduction stimuli. The presence of aliasing may have confounded their bone-conduction results. The lack of change in phase delay with intensity for the air-conduction 500-Hz results in the present study is not well understood. 114 The possibility of artifactual responses (whether non-auditory physiologic or stimulus artifact) in the ASSR is demonstrated by the presence of spurious responses to 50 and 60 dB HL 500-Hz bone-conducted stimuli in participants with severe-to-profound hearing losses who cannot hear the bone-conduction stimuli (Small & Stapells, 2004: Chapter 2). Our results for subjects with normal hearing (present study) show that alternating the stimulus polarity has no effect on the slope of the intensity-amplitude or intensity-phase-delay functions for ASSRs to 500-Hz or any other carrier frequencies. Because alternating polarity does not alter the responses of normal subjects, the findings of the present study lend support to the possibility that these responses are physiologic rather than the result of stimulus artifact. Previous studies have shown that vestibular-evoked myogenic potentials can be elicited by high intensity clicks and tone bursts (Colebatch, 2001). Recently, Nong, Ura and Noda (2000) reported an acoustically evoked short-latency negative response at 3-4 ms when recording ABRs in patients with profound hearing loss. The results of the present study also confirm that an "alternating" stimulus polarity does not, in itself, distort the amplitude or phase characteristics, as indicated by the results for the ASSRs to a 40 dB HL monaural air-conduction stimulus. Summary Normal ASSR thresholds to amplitude-modulated bone-conduction stimuli are 22, 26, 18 and 18 dB HL for 500, 1000, 2000 and 4000 Hz, respectively. Alternating the polarity of the stimulus does not significantly change the amplitude or phase of the ASSRs, and thus can be used to reduce the contribution of spurious responses in bone-conduction ASSRs resulting from stimulus artifact in the EEG. Spurious responses are also avoided by using a 1250-Hz A/D rate and a steep anti-aliasing filter. In adults with normal hearing, ASSR amplitudes are larger compared to higher carrier frequencies for bone-conduction stimuli, despite alternating the 115 stimulus polarity. Bone-conduction amplitude measures for ASSRs.to 500-Hz stimuli are also larger compared to binaural air-conduction ASSRs (all carrier frequencies). The change in ASSR phase delay with intensity is steeper for 1000 Hz compared to 500, 2000 and 4000 Hz. The change in phase-delay with intensity is steeper for bone-conduction ASSRs compared to binaural air-conduction ASSRs. The steeper intensity-amplitude slope for bone-conduction ASSRs to 500 Hz may be related to our findings that artifactual responses are present to high-intensity bone-conduction stimuli for single- and alternated- stimulus polarities (Small & Stapells, 2004: Chapter 2). Conclusions and clinical implications Multiple ASSRs to bone-conduction stimuli can be used to assess threshold at 500, 1000, 2000 and 4000 Hz in adults with normal hearing. Multiple ASSRs can potentially be used to assess bone-conduction thresholds in individuals with mild-to-moderate hearing loss. For 500-Hz stimuli, however, the possibility of non-auditory physiologic responses for levels 50 dB HL or greater limits recording ASSRs at this frequency to intensities 40 dB HL and lower (Small & Stapells, 2004: Chapter 2). Based on this study's results, this gives a maximum range of approximately 18 dB (i.e., 40 dB HL minus normal hearing threshold of 22 dB HL) at 500 Hz, possibly limiting clinical conclusions at 500 Hz to determine whether thresholds are "normal" or "elevated". For higher carrier frequencies, stimulus artifact and non-auditory physiologic responses are not likely to be an issue and the testing range is limited only by the output characteristics of the transducer. It must be noted, however, that there are no normative threshold data for infants and no threshold data from any subjects with impaired hearing (infants or adults). Bone-conduction ASSRs are, therefore, not yet ready for clinical use. 116 Table 3.1 M e a n (1 S D ) pure-tone behavioural air- and bone-conduct ion thresholds (dB HL) for two groups of adult participants with normal hearing (n = 10 per group). Group M O D E 500 H z 1000 H z 2000 H z 4000 H z A C - A S S R A IR 4 .5(8 .3) 3.5(5.8) 4.0 (7.0) 1.0 (5.7) B O N E 1.0(5.2) -3.5 (6.4) 1.0 (6.3) -7.5 (4.3 B C - A S S R A IR 4.4 (5.8) 2.8 (4.4) 0 .0(7.9) 3 .3(7.5) B O N E 4.9 (7.7) -3.1 (4.9) 4.4 (6.3) -3.6 (6.3) 117 Table 3.2 Air -conduct ion: M e a n (1 S D ) ampli tude and phase-de lay va lues for multiple A S S R S obtained in participants with normal hearing (n=10) to single- and alternated-st imulus polarity elicited by 40 dB HL stimuli presented monaural ly. P O L A R I T Y 500 H z 1000 H z 2000 H z 4000 H z A M P L I T U D E (nV) N O N 46 (32) 45 (25) 4 6 ( 1 2 ) 49 (20) INV 33 (26) 43(29) 4 6 ( 1 2 ) 3 7 ( 1 3 ) A L T 38 (28) 43 (28) 45 (12) 4 4 ( 1 5 ) P H A S E D E L A Y (degrees) N O N 134 (32) 162 (63) 108 (31) 103 (17 ) INV 120 (49) 164(57) 121 (29) 126 (58) A L T 125 (42) 167 (62) 114(28) 107 (19 ) N O N = non-inverted st imulus polarity INV = inverted st imulus polarity A L T = alternated st imulus polarity 118 Table 3.3 Air-conduct ion stimuli: Two-way repeated-measures A N O V A showing compar isons of mean ampli tude and phase-de lay va lues for three st imulus polarit ies and four carrier f requencies for A S S R s to monaural stimuli presented at 40 d B HL in normal-hearing participants. Sou rce df F ef Pb Ampl i tude Polarity 218 6.37 0.526 .030* Frequency 327 0.26 0.698 0.834 Polarity x Frequency 654 1.79 0.112 P h a s e delay Polarity 218 5.49 0.637 .032* Frequency 327 5.25 0.693 .007* Polarity x Frequency 654 0.57 0.754 a Huyhn-Feldt epsi lon (e) correction factor for degrees of f reedom b Probabil i ty reflects corrected degrees of f reedom * significant (p < .05) 119 Tab le 3.4 M e a n (1 S D ) s lope for ampli tude- and phase-de lay for bone- and binaural air-conduct ion stimuli for s ingle- and alternated-stimulus polarities. Group Mode Polarity 500 Frequency (Hz) 1000 2000 4000 Ampl i tude B C - A S S R B C N O N 5.41 3.02 2.38 2.03 (nV/dB) -4.66 -1.92 -2.49 -1.18 B C - A S S R B C INV 4.27 3.58 2.16 2.12 -2.61 -1.95 -1.39 -1.55 B C - A S S R B C A L T 4.63 3.36 2.29 2.12 -2.46 -1.88 -1.83 -1.18 A C - A S S R A C N O N 1.57 1.9 1.66 2.11 -1.77 -1.26 -1.06 -1.06 P h a s e delay B C - A S S R B C N O N -3.93 -5.06 -2.58 -2.63 (degrees/dB) -2.08 -3.55 -1.2 -1.63 B C - A S S R B C INV -1.77 -3.97 -2.25 -2.53 -2.59 -2.52 -1.04 -1.43 B C - A S S R B C A L T -2.56 -3.65 -2.5 -2.67 -2.42 -2.79 -0.84 -1.27 A C - A S S R A C N O N -0.18 -2.84 -2.29 -1.47 -1.94 -1.21 -0.68 -0.56 B C = bone conduct ion A C = binaural air-conduction N O N = non-inverted st imulus polarity INV = inverted st imulus polarity A L T = alternated st imulus polarity 120 Table 3.5 Bone-conduct ion stimuli: Two-way repeated measures A N O V A for bone-conduct ion A S S R s compar ing s lopes of ampli tude- and phase-delay- intensi ty functions for three st imulus polarities and four carrier f requencies. Source df F ef Pb Ampl i tude Polarity 218 0.24 0.516 0.642 Frequency 327 6.21 0.588 .012* Polarity x Frequency 354 0.61 0.72 P h a s e delay Polarity 216 1.49 0.609 0.259 Frequency 324 4.04 0.622 .042* Polarity x Frequency 648 1.00 0.438 a Huynh-Feldt epsi lon (e) correction factor for degrees of f reedom b Probabil i ty reflects corrected degrees of f reedom * significant (p < .05) 121 Table 3.6 Two-way mixed A N O V A showing compar isons of the s lope of intensity-ampli tude function and intensity-phase-delay functions for two stimulus modes (bone-vs binaural air-conduction st imulus presentation) and four carrier f requencies (within-subjects factor). The bone-conduct ion A S S R s were elicited using an alternated stimulus polarity. Source df F ef Pb Ampl i tude Mode 118 6.04 .024* Frequency 354 3.06 0.884 .043* Mode x Frequency 354 5.06 .004* P h a s e delay Mode 117 11.14 .004* Frequency 351 3.97 0.697 .026* Mode x Frequency 351 1.38 0.261 a Huynh-Feldt epsi lon (e) correction factor for degrees of f reedom b Probabil i ty reflects corrected degrees of f reedom * significant (p < .05) 122 Table 3.7 Across-s tudy compar ison of A S S R thresholds in d B HL (corrected to the mastoid; unoccluded). S T U D Y M O D E 5 0 0 H z 1000Hz 2 0 0 0 H z 4 0 0 0 H z P R E S E N T S T U D Y BONE ' 22 26 18 18 Lins e ta i . ,1996 BONE* 31 29 20 19 Dimitrijevic et a l . , 2002 1 BONE* 32 18 10 13 J e n g et al . , 2004 BONE* 47 33 40 37 Herdman & Stapel ls , 2001 AIR" 17 19 15 15 • Multiple A M tones; measured at mastoid, unoccluded * Multiple A M tones; measured at fo rehead, occ luded t Multiple A M / F M tones; measured at forehead, occ luded • Multiple A M tones; measured at forehead, occ luded .°° Multiple A M tones; single ear (insert earphone) 1 Thresho lds were calculated from A S S R ( A M tones)-behavioural ( A M tones) threshold di f ferences (personal communicat ion, A . Dimitrijevic) 123 Figure 3.1 Mean (±1 SD) amplitudes for bone- and air-conduction (binaural) multiple ASSRs for participants with normal hearing. Responses shown were elicited by single-polarity (non-inverted) stimuli. The response noise floor is denoted by a dotted line. Two groups of 10 subjects each participated. 200 150 100 > c 50 G> "O 3 0 +^ "5. 200 E < 150 Single-polarity stimulus (BC & AC) 2000 Hz 1000 Hz 4000 Hz BC • AC noise floor 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Intensity (dBHL) 124 Figure 3.2 Mean (±1 SD) amplitudes for bone- and air-conduction (binaural) multiple ASSRs for participants with normal hearing. Responses shown were elicited by alternated stimuli. The response noise floor is denoted by a dotted line. Two groups of 10 subjects each participated. 200 150 100 > c 50 ~o 3 0 +^ Q. 200 E < 150 Alternated-polarity stimulus (BC & AC) 1000 Hz 100 50 0 4000 Hz 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Intensity (dBHL) 125 Figure 3.3 Mean (±1 SD) phase delay for bone- and air-conduction (binaural) multiple ASSRs for participants with normal hearing. Responses shown were elicited by single-polarity (non-inverted) stimuli. Two groups of 10 subjects each participated. A minimum of 5 subjects contributed to each mean value plotted. 360 ^ 270 $ 180 g> 90 "D > 0 CO "5> 360 "D <D 270 (/) <0 -C 180 OL 90 0 Single-polarity stimulus (BC & AC) 1 1 i i i i r 500 Hz J L J I I L I I I I I I i r 1000 Hz J I I I L J I I L 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Intensity (dBHL) 126 Figure 3.4 Mean (±1 SD) phase delay for bone- and air-conduction (binaural) multiple ASSRs for participants with normal hearing. Responses shown were elicited by alternated stimuli. Two groups of 10 subjects each participated. A minimum of 5 subjects contributed to each mean value plotted. 360 ^ 270 $ 180 ? 90 "O > 0 CB 0) 360 "O <D 270 (/) CO -C 180 CL 90 0 Alternated-polarity stimulus (BC & AC) i i i i i i i 1000 Hz i r i 1 i r 500 Hz J L J L "i 1 i 1 1 1 r 2000 Hz J I I I I I L "i r B C A A C J L I I I I I I 1 1 r 4000 Hz J I L 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Intensity (dBHL) 127 REFERENCES ANSI. (1996). American National Standard Specification for Audiometers. Vol. S3.6-1996. New York: ANSI. Boezeman, E.H.J.F., Kapteyn T.S., Visser S.L,.& Snel. A . M . (1983). Comparison of the latencies between bone and air conduction in the auditory brainstem evoked potential. Electroencephalography and clinical Neurophysiology 56: 244-247. Colebatch J.G. (2001). Vestibular evoked potentials. Current Opinions in Neurology 14 (l):21-26. Cone-Wesson, B. & Ramirez, G.M. (1997). Hearing sensitivity in newborns estimated from ABRs to bone-conducted sounds. Journal of the American Academy of Audiology 8:299-307. Dimitrijevic, A., John M . S., Van Roon P., Purcell D. W., Adamonis J., Ostroff J., Nedzelski, J. M . , & Picton T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology 13: 205-24. Foxe J.J. & Stapells, D.R. (1993). Normal infant and adult auditory brainstem responses to bone-conducted tones. Audiology 32:95-109. Gorga, M.P., Kaminski, J.R., Beauchaine, K.L., .& Bergman, B.M. (1993). A comparison of auditory brain stem response thresholds and latencies elicited by air- and bone-conducted stimuli. Ear and Hearing 14: 85-93. Herdman, A. T. & Stapells D. R. (2001). Thresholds determined using the monotic and dichotic multiple auditory steady-state response technique in normal-hearing subjects. Scandinavian Audiology 30: 41-49. 128 Herdman, A. T. & Stapells D. R. (2003). Auditory steady-state response thresholds of adults with sensorineural hearing impairment. International Journal of Audiology 42: 237-248. Jahrsdoerfer, R. A., Yeakley J. W., Hall J. W., Robbins K. T., & Gray L. C. (1985). High-resolution CT scanning and auditory brain stem response in congenital aural atresia: Patient selection and surgical correlation. Otolaryngology-Head and Neck Surgery 93: 292-298. Jeng, F . - C , Brown, C.J., Johnson, T.A., & Vander Werff, K.R. (2004). Estimating air-bone gaps using auditory steady-state responses. Journal of the American Academy of Audiology 15: 67-78. John, M.S & Picton, T.W. (2000). MASTER: a Windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine 61:125-150. Lins, O. G., Picton T. W., Boucher B. L., Durieux-Smith A., Champagne S. C , Moran L. M . , Perez-Abalo M . C , Martin V., & Savio G. (1996). Frequency-specific audiometry using steady-state responses. Ear and Hearing 17: 81-96. Nong, D.X., Ura, M . , & Noda, Y. (2000). An acoustically-evoked short latency negative response in profound subjects. Acta Otolaryngologica 128(8): 960-966. Picton, T.W, John, M.S., Dimitrijevic, A., & Purcell, D. (2003). Human auditory steady-state responses. International Journal of Audiology 42:177-219. Small, S.A, & Stapells, D.R. (2004). Stimulus artifact issues when recording auditory steady-state responses. Ear and Hearing 25: 611-623. 129 Stapells, D.R. (2000a). Frequency-specific evoked potential audiometry in infants. In: Seewald RC, ed. A Sound Foundation Through Early Amplification. Basel: Phonak A G , 13-31. Stapells, D.R. (2000b). Threshold estimation by the tone-evoked auditory brainstem response: A literature meta-analysis. Journal of Speech-Language Pathology and Audiology 24 (2):74-83. Stapells, D.R. & Ruben, R.J. (1989). Auditory brainstem responses to bone-conducted tones in infants . Annals of Otology, Rhinology and Laryngology 98:941-949. Stenfelt S. & Hakansson, B. (2002). Air versus bone conduction: an equal loudness investigation. Hearing Research 167:1-12. 130 C H A P T E R 4 E F F E C T S OF BONE O S C I L L A T O R COUPLING M E T H O D , P L A C E M E N T L O C A T I O N , AND OCCLUSION ON BONE-CONDUCTION AUDITORY S T E A D Y - S T A T E RESPONSES IN INFANTS A version of this chapter has been published. Small, S. A., Hatton, J. L., & Stapells, D. R. (2007). Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants. Ear and Hearing 28(l):83-98. 131 INTRODUCTION It is important to obtain air- and bone-conduction thresholds in infants to distinguish between sensorineural, conductive and mixed hearing losses, similar to assessment of adults. It is a relatively straightforward procedure to deliver a calibrated air-conduction stimulus to the infant or adult ear via earphones. For bone-conduction testing, however, producing a predictable output of energy at the skull, and ultimately to the cochlea, is more complicated because of a number of procedural factors. Important procedural factors include: (i) oscillator coupling force and method (steel headband, elastic-band or hand-held), (ii) oscillator placement location on the head (temporal or frontal bone), and (iii) whether the bone-conduction testing is performed with ears occluded or unoccluded. Procedures are relatively standardized for estimating behavioural thresholds to bone-conducted stimuli in adults and children; the bone oscillator is positioned on the head of an adult or child using a steel headband which applies a constant force to the bone oscillator. In contrast, there are no standardized methods for bone-conduction testing in infants, only recommended "best practices", many of which are based on assumptions rather than systematic investigation. Because the bone oscillator sits on the surface of the skull, a sufficient and constant force is required to couple it to the head and to produce a calibrated output from the transducer. Amount of force and coupling method are, therefore, important issues. Also, the bone oscillator can be positioned in different locations on the skull, potentially affecting the intensity of the signal reaching the cochlea (Stuart, Yang & Stenstrom, 1990; Yang, Rupert & Moushegian, 1987). It is also well known, at least for adults, that occluding the ear canal while estimating bone-conduction thresholds significantly improves bone-conduction thresholds in the low frequencies, a change known as the "occlusion effect" (Tonndorf, 1966). 132 For behavioural testing with bone-conducted stimuli, the oscillator is coupled to the head using a steel headband to apply a constant force to the bone oscillator; the exact amount of force applied (approximately 400-500 g) will vary depending on individual head size. To accommodate the smaller head size for older children, a smaller steel headband is used; the force generated is less than for larger adult heads and will also vary somewhat with the size of the child's head (Harrell, 2002; Wilber, 1979). The coupling force used is limited to some degree by patient comfort; however, behavioural thresholds to bone-conducted stimuli in older children and adults can usually be obtained in approximately 10-15 minutes with only minor discomfort from the steel headband. Estimation of bone-conduction hearing thresholds in young sleeping infants using auditory evoked potentials poses unique challenges. Infants have smaller heads than adults, precluding the use of the standard adult or child steel headband. Also, infants must remain asleep during testing; disturbing the infant as little as possible while positioning the oscillator is critical, as is minimizing the discomfort during a much longer testing time. One concern for bone-conduction testing is that the amount of force applied to the oscillator is consistent. Coupling the oscillator to the infant's head using an elastic band is currently the clinical method suggested by many because a known force can be applied to the elastic band and the amount of force can be verified using a spring scale (Yang & Stuart, 1990). It is important to keep in mind, however, that although many clinical settings use an elastic band to couple the oscillator to the head, for practical reasons, the amount of force applied to the elastic band is not verified (i.e., the clinician does not want to use any "sleep time" to perform the verification step and does not want to risk waking the infant in the process of verifying the force level and re-adjusting the elastic band). Holding the oscillator in place by hand is also commonly done 133 clinically because it is faster and is far less likely to wake up the infant than positioning an elastic band (which requires more manipulation of the infant's head). Another benefit of the hand-held method is that it is more comfortable for the infant because it can be easily removed and replaced between test conditions. Despite the practical advantages of the hand-held method, its use has been discouraged (Yang & Stuart, 2000; Yang, Stuart, Stenstrom, & Hollett, 1991). This preference for one coupling method over the other, however, is based on assumptions. The two main assumptions made are: (i) for the hand-held method, there is potential for the applied force to vary during testing resulting in an inconsistent output from the transducer (Yang et al., 1991) and greater variability in threshold, and (ii) pressing down on the superior surface of the oscillator by hand (i.e., mass loading it) will dampen the response characteristics of the bone oscillator (Wilber, 1979). However, bone-conduction responses in infants have never been compared using the elastic-band and hand-held methods. There are adult calibration values for bone-conducted stimuli for different bone-oscillator placement locations on the skull (i.e., mastoid and forehead). On average, adult behavioural thresholds measured at the forehead are higher than at the mastoid by 14, 8.5, 11.5, and 8.0 dB at 500, 1000, 2000, and 4000 Hz, respectively (Amercian National Standards Institute [ANSI], 1996). Bone-oscillator placement location has also been raised as an issue (Stuart et al., 1990; Yang & Stuart, 2000) when estimating bone-conduction thresholds in sleeping infants because of the potential for thresholds to differ depending on where the oscillator is placed on the temporal (mastoid vs. upper region) or frontal bone. When an elastic band is used to couple the oscillator to the skull, it is typically placed on the infant's temporal bone, posterior to the upper portion of the pinna; using the hand-held method, a variety of positions are possible. 134 There are some auditory brainstem response (ABR) data which have been interpreted to suggest that bone oscillator location does affect the response in infants (Yang et al., 1987). Yang et al. (1987) investigated the effect of oscillator placement (frontal/forehead, occipital, and temporal) on wave V ABR latencies to bone-conduction clicks for neonates, one-year-old infants and adults. Latency results were found to vary with both age of subject and oscillator placement. The temporal bone yielded significantly shorter latencies than either the occipital and frontal placements in the neonatal group and one-year-old group. Based on these latency differences, Yang et al. estimated that signal attenuation in the neonates from the temporal to frontal placement ranged from 30 to 35 dB. In the one-year-old group, signal attenuation between the frontal and temporal placements was estimated to be between 20-25 dB. Attenuation from occipital to temporal was 15-20 dB for both infant groups. In contrast, the latency differences in adults were not significant, and the attenuation between oscillator placement was judged to be no more than 5-10 dB. This attenuation estimate in adults is smaller than the ANSI-1996 standards for mastoid and forehead placements discussed above, and the changes in signal attenuation as a function of location suggested by Yang et al. must be interpreted with caution for several reasons. First, they used click stimuli rather than frequency-specific stimuli, so they were unable to separate out any frequency-dependent effects of oscillator placement on attenuation. Second, they did not directly estimate ABR thresholds for the different placements or report ABR amplitudes, which are more related to threshold than latency measures. Third, the attenuation estimates were derived using latency-intensity functions; inaccurate attenuation estimates may result because click-evoked ABR latency is not linearly related to intensity (Picton, Stapells, & Campbell, 1981). 135 Stuart et al. (1990) further investigated the effect of oscillator placement on wave V latencies in infants by comparing different areas on the temporal bone (superior, supero-posterior and posterior placements). They concluded that changing the location of the oscillator on the temporal bone produced significantly different wave V latencies in the neonate at both 15 and 30 dB nHL. Specifically, the posterior position (similar to the "mastoid" in the current study) yielded the shortest wave V latency, whereas the superior position yielded the longest latency. The supero-posterior position (similar to the "temporal placement" in the current study) yielded intermediate latency values. These latency differences were attributed to reflect greater signal attenuation as a function of distance from the cochlea (Stuart et al., 1990). Specifically, it was suggested that signal intensity reduces as one moves the bone oscillator farther away from the cochlea. Again, these findings were based solely on click-ABR latency data and the same limitations that are discussed above apply. Neither of these two placement studies compared bone-conduction threshold in infants at different oscillator placements on the skull, which is the important measure for clinical applications. Correcting for the occlusion effect when estimating bone-conduction thresholds with occluded ears is also well described in adults (Dirks & Swindeman, 1967; Elpern & Naunton, 1963; Hodgson & Tillman, 1966; Small & Stapells, 2003). For insert earphones, behavioural thresholds to brief-tones improve by 3-5 dB at 500-1000 Hz (Small & Stapells, 2003), whereas pure-tone behavioural thresholds in the low frequencies (250-1000 Hz) improve by as much as 17 dB, depending on the insertion depth (Dean & Martin, 2000). As discussed earlier, any test protocol used to estimate thresholds in sleeping infants must be designed to minimize the possibility of waking the infant. Air-conduction thresholds are typically assessed using insert earphones, followed by bone-conduction testing (Stapells, 2000). Should the insert earphones 136 be removed before assessing bone-conduction thresholds? It is important to know whether occluding the ears increases, decreases or has no effect on bone-conduction thresholds in infants. There are no studies that have investigated the occlusion effect in infants. The objectives of this study were: (i) to investigate the variability in the amount of force applied using the two common bone-oscillator coupling methods used in infants, and to determine whether oscillator coupling method affects estimation of bone-conduction threshold in infants and adults, (ii) to examine the effects of bone-oscillator placement location on bone-conduction ASSR thresholds and amplitudes in young infants, and (iii) to determine whether the occlusion effect is present in infants by comparing infant bone-conduction ASSR thresholds for unoccluded and occluded ears. GENERAL MATERIALS AND METHODS The current study was divided into three main experiments which investigated the effects of: (1) bone-oscillator coupling method on the variability in the amount of force applied to the oscillator and its impact on estimation of bone-conduction thresholds in infants and adults, (2) bone oscillator placement location on ASSR amplitudes and thresholds in infants, and (3) occluding the ear canal on bone-conduction ASSR amplitudes, phase delays and thresholds in infants. The General Methods section describes the methodology common to all experiments including: stimuli, calibration of stimuli, ASSR recording parameters and ASSR data analyses. The specific details of the participants, experimental design, statistical analyses and description of the results for each experiment are reported in separate sections. 137 Stimuli All stimuli were bone-conducted tones presented to a Radioear B-71 bone oscillator which was held by hand or by elastic headband with approximately 400-450 g of force. The bone oscillator was placed on the temporal bone (high or low) or on the forehead depending on the experiment. For the temporal-bone placement location, the left or right side was selected at random. Bone-conducted stimuli were presented at intensities 50 to -10 dB HL. For all ASSR experiments, stimuli were sinusoidal bone-conducted tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude and 25% frequency modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. The stimuli were presented simultaneously for all conditions tested (i.e., multiple). All ASSR stimuli were generated by the Rotman MASTER research system (John & Picton, 2000a) using a digital-to-analog (D/A) rate of 31,250 Hz and then attenuated through an Interacoustics AC40 clinical audiometer (laboratory) or Tucker-Davis Technologies HB6 and SM3 modules (laboratory) or MeD/Associates ANL-918 attenuator [(Neonatal Intensive Care Unit (NICU)]. Before the stimuli were attenuated, they were routed through a Stanford Research Systems Model SR650 to increase the gain of the stimulus by 10 dB. To estimate behavioural thresholds in adults in Experiment IB, bone-conduction pure-tones were presented using an Interacoustics AC40 clinical audiometer. Calibration The bone-conducted stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re:luN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 550 g of force. 138 ASSR Recordings ASSRs were recorded using the Rotman MASTER system. Three electrodes were used to re'cord the electrophysiologic responses; the non-inverting electrode was placed at the forehead, the inverting electrode was positioned midline at the nape of the neck, just below the hairline, and an electrode placed at the high forehead acted as ground. All inter-electrode impedances were below 3 kOhms at 10 Hz. The electroencephalogram (EEG) was filtered using 30-250 Hz filter (12 dB/oct) and amplified 80,000 times (Nicolet HGA-200A and Nic501 A). The E E G was further filtered using a 300-Hz lowpass anti-aliasing filter [NICU: Stanford Research Systems (115dB/oct); laboratory: Wavetek Rockland Model 852 (48 dB/oct)]. The E E G was then processed using a 1250-Hz A/D conversion rate (Small & Stapells, 2004: Chapter 2). Each E E G recording sweep was made up of 16 epochs of 1024 data points each (0.8192 seconds per epoch) and lasted a total of 13.107 seconds. Artifact rejection was set to eliminate epochs of electrophysiologic activity that exceeded ± 40 [iV in amplitude in order to reduce contributions to the E E G due to muscle artifact. ASSRs were averaged in the time domain and then analyzed online in the frequency domain using a Fast Fourier Transform (FFT). Weighted averaging was used (John, Dimitrijevic & Picton, 2001). The FFT resolution was 0.076 Hz over a range of 0 to 625 Hz. Amplitudes were measured baseline-to-peak and expressed in nV. An F-ratio was calculated by MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ±60 bins of the modulation frequency ("noise") [John & Picton, 2000a]. A minimum of seven sweeps were recorded for each test 139 condition. A response was considered to be present if the F-ratio, compared to the critical values for F (2, 240), was significant at a level of p < .05 for at least two consecutive sweeps. A response was considered to be absent if p > .05 and the mean amplitude of the noise was less than 11 nV. Alternatively, a response was also considered to be absent when response amplitude was < 10 nV and thep value > .30. ASSR data analyses Mean amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant (except for part of the data analyses in Experiment 2). The phase values from MASTER were adjusted by adding 90° to yield the onset phase (John & Picton, 2000a). Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Any phase-delay values that differed > 180° from an adjacent measure were "unwrapped" by adding 360° to their value (John & Picton, 2000b). Phase values for ASSRs that were not significant were excluded from mean phase-delay calculations. Phase-delay values were averaged across subjects. Results are reported only if at least five subjects contributed to the mean. Statistical Analyses Comparisons of experimental conditions were made using 2-way repeated-measures analyses of variance (ANOVA); case-wise deletion was applied when values were missing from the data sets1. The factors and levels for the ANOVAs performed for each experiment are 1 For all data sets which had missing values, statistical analyses using case-wise deletion to handle missing data were compared to statistical analyses using two different methods for replacing the missing values [mean substitution and imputation based on adjacent values (Little & Rubin, 2000, pp. 3-40)]. The purpose of these comparisons was to ensure that case-wise deletion did not yield different statistical results compared to other strategies for handling missing data. No differences in the statistical results were found for any of the analyses, consequently, the statistical analyses which employed case-wise deletion were considered unbiased. 140 reported in each section. Huynh-Feldt epsilon adjustments for repeated measures were made when appropriate. Newman-Keuls post-hoc comparisons were performed for significant main effects and interactions. The criterion for statistical significance was p<,Q5 for all analyses. E X P E R I M E N T 1: B O N E - O S C I L L A T O R C O U P L I N G M E T H O D ( E L A S T I C - B A N D V S H A N D - H E L D ) The purposes of this experiment were to investigate (i) how much the amount of force applied to the bone oscillator varied with coupling method, and (ii) to determine whether bone-conduction thresholds obtained in infants and adults were different between coupling methods. This experiment was divided into three sub-experiments: (1 A) effect of coupling method on variability in force level applied to oscillator, (IB) effect of coupling method on adult behavioural thresholds, and (1C) effect of coupling method on infant ASSRs. The participants, procedures and a description of the results specific to each sub-experiment will be presented separately. 1A. Effect of coupling method on variability in force level applied to oscillator The purpose of Experiment 1A was to measure the variability in the amount of force applied to the oscillator by four different individuals using the elastic-band and hand-held coupling methods. M E T H O D S Participants and procedure For this experiment, four adults (age range: 23-39 years) were trained to couple the bone oscillator to the head using the elastic-band and hand-held coupling methods and are referred to 141 as the "assistants". The bone oscillator was placed on the temporal bone slightly posterior to the upper part of the pinna. The anterior longitudinal edge of the oscillator was oriented at an approximately 45° angle to the anterior/posterior plane so that the wire connecting the oscillator to the attenuator lay in a posterior/inferior direction. Assistants were instructed to apply the same force (425±25 g) on the superior surface of the bone oscillator on an adult head, whether they used an elastic band or held the oscillator by hand. During a training period, each assistant was instructed to apply the target force to the oscillator in a group of 10 practice trials. After each practice trial, the assistant received feedback about the actual amount of force applied. For the elastic-band method, a spring scale was used to measure the applied force. For the hand-held coupling method, applied force was measured by placing the bone oscillator on a compressive spring scale and pressing down on the transducer with one or two fingers until the desired force was achieved. Feedback was provided for 10 consecutive practice trials for each of the coupling methods. During the testing period, after completing the training, the assistants applied the target force, without feedback, for an additional 10 "test" trials for each coupling method. For each of these test runs, the actual force levels applied were recorded and compared for each assistant and coupling method. Statistical Analyses A 2-way repeated measures A N O V A was performed comparing force levels produced across 10 test trials and between the two coupling conditions (elastic-band & hand-held). Four cases (i.e., the four assistants) were included in the analysis. 142 RESULTS As shown in Table 4.1, there were some differences noted in the force levels produced by the assistants. For Assistants #1 and #4, the mean force applied tended to be more than the target force when using the elastic band method but less than the target force when using the hand-held method. The mean forces applied by Assistants #2 and #3 tended to be more similar across coupling method. Using the elastic-band, assistants produced a wider range of forces and larger standard and mean absolute deviations. Across assistants, the mean force produced was essentially the same for the elastic-band and hand-held methods. Table 4.1 Figure 4.1 shows the errors in the force applied to the oscillator produced by each assistant as a cumulative percent of test runs for the two coupling methods. Using the hand-held method, most of the test runs (70%) were within 50 g of the target, and all were within 100 g of the target. Using the elastic band, only 60% of the test runs were within 50 g of the target and only 88%o were within 100 g; 100% was only obtained within 175 g of the target. Results of an A N O V A comparing force levels across test trials and coupling method revealed no significant main effect of coupling method [F(l,3) = 0.511, p = .503] or test trial [F(9,27) = 1.649, p = .151,] and no significant interaction between test trial and coupling method [F (9,27) = 2.016, p = .011]. Figure 4.1 143 IB. Effect of coupling method on adult behavioural thresholds The purpose of Experiment IB was to compare adult behavioural thresholds obtained for 500, 1000, 2000 and 4000 Hz using the elastic-band and hand-held coupling methods. M E T H O D S Participants and procedure Ten adults (age range: 20-39 years) with normal hearing (pure-tone air-conduction behavioural thresholds < 25 dB H L for 500-4000 Hz) participated. Testing was conducted in a double-walled sound-attenuated booth in the Human Auditory Physiology Laboratory at the University of British Columbia. On average, the noise levels in the sounD/Attenuated booth for one-octave-wide bands centred at 0.5, 1, 2 and 4 kHz were 12, 10, 10 and 12 dB SPL, respectively. Behavioural thresholds were obtained using the elastic-band and hand-held coupling methods (Assistants #1, #2 & #3) with ears unoccluded. The bone-oscillator placement location was the same as that used for Experiment 1 A. The test order for coupling method was randomized. Bone-conducted pure-tones (500-4000 Hz) were presented in 2-dB steps using a bracketing technique. The starting intensity was randomly selected as 20, 30 or 40 dB HL. The lowest level at which 3/5 stimuli were detected was considered threshold. Behavioural thresholds across frequency and coupling methods were compared. Statistical Analyses A 2-way repeated measures A N O V A was performed comparing behavioural thresholds at 500, 1000, 2000 and 4000 Hz between coupling methods (elastic-band and hand-held). Ten subjects were included in the analysis. 144 R E S U L T S Adult mean behavioural bone-conduction thresholds for the elastic-band and hand-held coupling methods are shown in Table 4.2. Overall, the mean elastic-band minus handheld threshold difference was only -1.4 dB. Results of an A N O V A indicated no significant main effect of coupling method [F (1,9) = .432, p = .527] and no interaction between coupling and frequency [F (3,27) = 3.275, p = .092]. The main effect for frequency [F (3,27) = 8.444, p = .0004] was significant and was explained by slightly poorer thresholds at 2000 Hz compared to the other frequencies as indicated by post hoc comparisons. Table 4.2 1 C . Effect of coupling method on infant ASSRs In Experiment 1C, ASSRs to bone-conducted stimuli were recorded in infants using the elastic-band and hand-held coupling methods to determine whether coupling method affects estimation of bone-conduction ASSR threshold in young infants. M E T H O D S Participants and procedure Ten normal-hearing post-term infants (age range of 0.5-38 weeks; mean age of 17 weeks) were recruited from the community. By parent report, all infants were healthy, normal infants with no history of middle-ear disease. Three of the infants were screened using an automatic auditory brainstem response (AABR) hearing screening test at 35 dB nHL. The hearing of the other infants was screened using a distortion-product otoacoustic emissions (DPOAE) screening 145 test. The pass criterion for the DPOAE screening was a signal-to-noise ratio > 5 dB at 2000, 3000 and 4000 Hz in both ears. Infants passed the AABR or DPOAE hearing screening in both ears and were considered to be at low risk for significant hearing loss and thus included in the study. Testing was conducted in the same sound-attenuating booth described above. The oscillator was coupled to the infant's head using the elastic-band (Assistants #1, #2 & #3) and hand-held coupling methods (Assistants #1 & #2)2. The bone-oscillator placement location was the same as that used for Experiment 1A. The test order for coupling method was randomized. Multiple ASSR stimuli (500-4000 Hz carrier frequencies) were presented in 10-dB steps using a bracketing technique. The starting intensity was randomly selected as 10, 20, 30 or 40 dB HL. The lowest level at which a response was present was considered threshold. Bone-conduction ASSR thresholds were compared across frequency and oscillator coupling method. Bone-conduction ASSR amplitudes at 30 dB HL for 500 and 1000 Hz and at 40 dB HL for 2000 and 4000 Hz were compared across frequency and coupling condition. A greater intensity was selected for 2000 and 4000 Hz than for 500 and 1000 Hz to compensate for poorer thresholds (approximately 10 dB) at these higher frequencies (Small & Stapells, 2006: Chapter 5). Statistical Analyses A 2-way repeated-measures A N O V A was performed comparing ASSR thresholds for 10 infants at 500, 1000, 2000 and 4000 Hz between coupling methods (elastic-band & hand-held). ASSR amplitudes at 30 dB HL for 500 and 1000 Hz and 40 dB H L at 2000 and 4000 Hz were also compared between coupling conditions using a 2-way repeated-measures A N O V A ; however, as a result of case-wise deletion, only five infants were included in the analysis. An additional A N O V A comparing ASSR amplitudes at 30 dB H L at 500 and 1000 Hz only between 2 Assistant #4 was not available to participate in this study. 146 coupling conditions was also performed; eight infants were included in the A N O V A as a result of case-wise deletion. R E S U L T S Representative ASSR results to bone-conducted stimuli using the elastic-band and hand-held coupling methods are shown for a typical infant in Figure 4.2. For this 8-week-old infant, there was no consistent effect of coupling method on ASSR thresholds. ASSR threshold was 30 dB poorer at 500 Hz and 10 dB better at 2000 Hz for the hand-held compared to the elastic-band coupling condition, and no difference in ASSR thresholds was found for the two coupling conditions at 1000 and 4000 Hz. Table 4.3 shows the mean bone-conduction ASSR thresholds for all infants. There was no significant difference between mean ASSR thresholds using the two coupling methods [F(l,9) = .192,/? =.67]; however, there was a trend for ASSR thresholds at 4000 Hz to be slightly higher using the hand-held method compared to the elastic-band coupling method [coupling method x frequency interaction: F (3,27) = 2.791,/?= .061]. Figure 4.2 and Tables 4.3 As shown in Table 4.4, infant mean ASSR amplitudes at 30 dBHL (500 and 1000 Hz) and 40 dB HL (2000 and 4000 Hz) were not significantly different for the elastic-band and hand-held coupling methods [F (1,4) = .027,/? = .878]. However, there was a significant interaction between coupling method and frequency [F (3,12) = 3.951, /? = .036];post hoc comparisons indicated no elastic-band vs hand-held differences in ASSR amplitudes at the same carrier frequency. Mean ASSR amplitudes at 30 dB HL for 500 and 1000 Hz were not significantly 147 different for coupling method [F (1,7)= 2.23\,p = between coupling method and frequency [F(l,7)= .179] and there was no significant interaction 0.002,/? = .967]. Table 4.4 EXPERIMENT 2: EFFECT OF BONE-OSCILLATOR PLACEMENT LOCATION ON INFANT ASSRS The purpose of this experiment was to determine if different placement locations of the bone oscillator on the skull affect ASSRs recorded to bone-conducted stimuli in pre-term infants. The three oscillator placements investigated in this study included: (i) the temporal bone slightly posterior to the upper part of the pinna, (ii) the lower portion of the temporal bone which will later develop into the mastoid bone, and (iii) the middle of the forehead. The placements are referred to as "temporal", "mastoid" and "forehead". METHODS Participants and procedures Fifteen pre-term infants from a NICU participated. They ranged in age from 32 to 43 weeks post-conceptional age (PCA) with a mean age of 34.5 weeks PCA. All infants were medically stable at time of test and had a mean (1SD) APGAR score of 7.4 (2.6) at birth. All of these infants were screened using an automatic auditory brainstem response (AABR) hearing screening test at 35 dB nHL. Infants passed the AABR hearing screening in both ears and were considered to be at low risk for significant hearing loss and thus included in the studies. Testing of pre-term infants was performed at the bedside in the NICU. Infants were tested with ears unoccluded. Ambient acoustic noise was likely to affect the results when recording 148 ASSRs in the NICU. Average ambient noise levels in the NICU measured at two different time points using 1/3-octave-wide bands were 52, 51, 60, 58, 50, 36 and 47 dB SPL at 0.125, 0.5, 1, 2, 4, and 8 kHz, respectively. The overall A-weighted noise level in the NICU was 65 dB SPL. To reduce the impact of high ambient noise in the NICU, recording of ASSRs was paused when the background acoustic noise was excessively high (e.g., when sink taps were running near the infant's bassinet). ASSRs were first elicited to bone-conducted stimuli presented at the temporal bone using the hand-held coupling method at a starting intensity of 40 dB HL (30 dB HL for 2/15 infants). ASSRs were then recorded, in random order at the mastoid and forehead, to stimuli presented at 40 dB HL. ASSR threshold was then determined, always beginning with the temporal placement, using a bracketing technique with a final step size of 10 dB. The order of threshold search for the mastoid and forehead placements was randomized. If a response did not reach significance at the highest intensity (50 dB HL), threshold was arbitrarily assigned a value of 60 dB HL. Statistical Analyses Comparison of ASSR threshold for 500, 1000, 2000 and 4000 Hz and for the three bone-oscillator placement locations (temporal, mastoid and forehead) was done using a 2-way repeated measures A N O V A . After case-wise deletion, nine pre-term infants were included in the analysis of ASSR threshold. ASSR amplitudes (all responses) for 500, 1000, 2000 and 4000 Hz were compared for three bone-oscillator placement locations for 13 pre-term infants using a 2-way repeated measures ANOVA. For the second analysis of ASSR amplitudes (significant responses only), ASSR amplitudes for 1000 and 4000 Hz and for the three bone-oscillator placement locations were compared for seven pre-term infants.. 149 RESULTS Representative ASSR results to bone-conducted stimuli at the three bone oscillator placements are shown in Figure 4.3 for a typical pre-term infant (35 weeks PCA). ASSR thresholds at the temporal bone and mastoid for this infant were the same at 500 and 1000 Hz and differed only by 10 dB at 2000 and 4000 Hz. In contrast, thresholds at the forehead were 10-40 dB HL poorer across frequencies compared to the other placements. As shown in Figure 4. 4, there were many fewer significant responses at the forehead compared to the temporal and mastoid placements at the maximum intensity tested for 2000 and 4000 Hz. Table 4.5 shows mean bone-conduction ASSR thresholds for the pre-term infants for three bone-oscillator placement conditions. Mean ASSR thresholds for the temporal bone and mastoid placements were similar, whereas the mean forehead thresholds were elevated compared to the other two placements. The results of an A N O V A revealed a significant effect of placement [F (2,14) = 11.124, p = .005,] and frequency [F(3,21) = 12.328,/? = .0001], but no significant interaction between placement and frequency [F (6,42) = 0.826,p = .556]. Post hoc comparisons revealed that thresholds for the forehead placement were significantly elevated in comparison to the temporal bone (p = .0016) and mastoid (p = .003) bone oscillator placements. No significant differences were found for ASSR threshold between the temporal bone and mastoid bone oscillator placements. Consistent with our results reported in another study (Small & Stapells, 2006: Chapter 5), bone-conduction ASSR thresholds in pre-term infants were significantly lower (i.e., better) at 500 and 1000 Hz compared to 2000 and 4000 Hz. Figures 4.3 and 4.4 and Table 4.5 150 Mean bone-conduction ASSR amplitudes for stimuli presented at 40 dB HL at the three bone vibrator placements are shown in Table 4.6. Mean ASSR amplitudes obtained when the bone oscillator was placed at the forehead were smaller compared to the other two placements. Results of an A N O V A comparing ASSR amplitude at 40 dB H L revealed a significant effect of placement [F(2,24) = 15.327,/? = .0001]. Post hoc comparisons indicated that amplitudes for the forehead placement were significantly smaller compared to the temporal and mastoid placements; there was no significant difference in mean amplitudes for the temporal and mastoid placements. The main effect of frequency [F (3,36) = 24.524,/? = .0001] was also significant, and explained by significantly larger ASSR amplitudes at 500 and 1000 Hz compared to 2000 and 4000 Hz. There was no significant interaction between placement and frequency [F (6,72) = 1.607,/? = .193]. Table 4.6 The above analysis included amplitudes for non-significant results because not all of the pre-term infants had significant responses at 40 dB HL for all frequencies, particularly at 2000 Hz. A second analysis of the 40 dB HL ASSR amplitude data was thus done to include only significant responses (i.e., not biased to small non-significant response amplitudes); ASSR amplitudes for 1000 and 4000 Hz were selected to represent the low and high frequencies, respectively. Mean amplitudes for the forehead placement (31 nV for 1000 Hz, 14 nV for 4000 Hz) remained smaller compared to the temporal (52 nV for 1000 Hz, 22 nV for 4000 Hz) and mastoid (57 nV for 1000 Hz, 26 nV for 4000 Hz) placements. The results of an A N O V A on this smaller data set also revealed a significant effect of placement (/? = .002); post hoc comparisons 151 again indicated that this significant difference was explained by smaller mean ASSR amplitudes for the forehead compared to the other two placements which were not significantly different. There was also no significant interaction between placement and frequency (p = .161) for this smaller data set. EXPERIMENT 3: EFFECT OF OCCLUDING THE EAR CANAL ON INFANT ASSRS The purpose of this experiment was to determine whether the "occlusion effect", which results in improvement in bone-conduction thresholds at 500 and 1000 Hz in occluded adult ears is also present for young infants. M A T E R I A L S AND M E T H O D S Participants and procedure Bone-conduction ASSR thresholds were obtained in 12 post-term infants (age range of 1 27 weeks; mean age 15 weeks) with the ear canal unoccluded and occluded. As described in Experiment 1C, all infants were screened using a DPOAE screening test. By parent report, all infants were normal healthy babies with no history of middle-ear disease. The bone-oscillator placement location was the same as that used for Experiment 1 A. For the unoccluded condition, no insert earphones were placed in either ear canal. For the occluded condition, a pediatric-size insert earphone foam tip was inserted into the ear canal that was on the same side as the bone oscillator. The foam tip was attached to the ER3A tubing that was plugged at the end distal to the ear (i.e., the end of the tubing normally attached to the transducer). Unoccluded ASSR thresholds were always obtained before occluded ASSR thresholds. The oscillator was hand-held for three of the infants; the elastic-band was used for 152 the remaining nine infants. Multiple ASSR stimuli (500-4000 Hz carrier frequencies) were presented in 10-dB steps using a bracketing technique. The starting intensity varied from 10-40 dB HL. The lowest level at which a response was present was considered threshold. Bone-conduction ASSR thresholds were compared across frequency and occlusion condition. Statistical Analyses A 2-way repeated measures A N O V A was performed to compare ASSR thresholds at 500, 1000, 2000 and 4000 Hz for ears occluded and unoccluded; the same analysis was completed separately for ASSR amplitudes and phase delays at 40 dB H L across frequency and occlusion condition. After case-wise deletion, there were 10, 10 and 7 infants remaining in the analyses of ASSR threshold, amplitude and phase delay, respectively. R E S U L T S Representative ASSR results to bone-conducted stimuli for unoccluded and occluded ears are shown for a typical post-term infant in Figure 4.5. In this example, occluding the ear canal of this 6-week-old infant had no effect on bone-conduction ASSR thresholds for 500, 1000 and 2000 Hz, and only a 10-dB increase in ASSR threshold at 4000 Hz. As indicated by the means shown in Table 4.7, occluding the ear canal had no significant effect on mean bone-conduction ASSR thresholds [F (1,8) = .075, p = .791] and there was no significant interaction between frequency and occlusion condition [F(3,24) - .530,p = .666]. There was a significant effect of frequency [F(3,24) = 15.770,/? = 0001] as expected (Small & Stapells, 2006). Similarly, as shown in Figure 4.6, mean ASSR amplitudes were not different for the unoccluded and occluded conditions Results of an A N O V A comparing ASSR amplitudes at 40 dB H L indicated no significant main effect of occlusion condition [F (1,7) = 2.908,/? = .132] and no significant 153 interaction between occlusion condition and frequency [F(3,21) = 1.294,/? = .303]. There was also no difference in mean ASSR phase delay between occlusion conditions, as shown in Figure 4.7. Results of an A N O V A comparing ASSR phase delays at 40 dB HL also revealed no significant effect of occlusion condition [F(l,6) = 0.016,/? = .904] and no significant interaction between occlusion condition and frequency [F (3,16) = 0.058,/? = .981]. Figures 4.5, 4.6 and $.7 and Table 4.7 DISCUSSION Bone oscillator coupling method For a number of years, it has been assumed that good clinical practice requires the use of an elastic band to couple the bone oscillator to an infant's head when estimating bone-conduction thresholds using auditory-evoked potentials. That is, it has been suggested that it is not appropriate to hold the oscillator in place by hand (Cone-Wesson, 1995; Stapells, 2000; Stapells & Oates, 1997; Yang & Stuart, 2000; Yang et al , 1991). This is the first experiment that has actually compared the force levels applied to the bone oscillator when using the elastic-band and hand-held coupling methods. It is also the first experiment to compare frequency-specific bone-conduction thresholds in infants and adults using both of these coupling methods. The results for Experiment 1 showed no significant difference in the mean force applied between coupling methods, although there is some variability in the amount of force applied by individual assistants using each of the coupling methods and a tendency for mean absolute deviations from the target force to be slightly larger for the elastic-band compared to the hand-154 held method. Notably, threshold variability (i.e., standard deviations) was not greater for the hand-held method. For bone-conduction testing, the main concern is that a consistent amount of force be applied to the oscillator to produce a predictable output of energy at the skull. Yang et al. (1991) measured wave V latencies and amplitudes for bone-conducted click-ABRs using force levels of 225, 325, 425 and 525 g applied using an elastic band to determine the most efficient amount of force to apply to the oscillator. They found that wave V latency decreased significantly when force levels increased from 225 to 325 g but did not change from 325 to 425 g or from 425 to 525 g. They also noted that coupling with a low force level (225 g) caused the oscillator to slip out of the elastic band and that a high force level (525 g) resulted in the elastic band slipping off of the oscillator. Based on their data, Yang et al. concluded that 400-450 g was the most efficient amount of force to apply. They further suggested that the practice of holding the oscillator by hand not be encouraged because a constant force could not be applied; they also conjectured that the stimulus output could be dampened using the hand-held method (see Wilber, 1979). They did not investigate either of these assumptions. It is also possible that the frequency response of the bone oscillator could be dampened using an elastic band, although it does seem unlikely that such a dampening effect would be identical for both the elastic-band and the hand-held methods, which would have to be the case as we found no difference in thresholds across frequency for these two methods. The results of Experiment 1 suggest that individuals can be trained to apply a targeted amount of force to the oscillator by hand to within a reasonable amount of variability (33 g mean absolute deviation) and accuracy (425 ± 100 g for 0% errors produced). It is also noteworthy that when a trained assistant couples the oscillator to the head using an elastic band, without verifying the force level with a spring scale, which is in fact most often the case clinically, the force 155 applied is actually more variable (325-600 g) and the accuracy is poorer (425 ± 175 g for 0% errors produced) for the elastic-band coupling compared to the hand-held method. The advantage of the elastic-band method is that you can directly measure the force applied; however, the disadvantage of verifying the force is that it takes extra time and may involve undoing the Velcro™ band several times, which will often wake the baby. One of the main objectives of this experiment was to determine whether bone-conduction thresholds are affected by coupling method. The adult head is a relatively rigid structure with fused sutures so it is assumed that coupling method is the only factor being tested when adult behavioural thresholds to bone-conducted stimuli are compared using the elastic band and hand-held coupling methods. The bone-conduction behavioural thresholds for adults obtained in this experiment were not significantly different for the two coupling methods (< 3 dB difference across frequencies), suggesting that pressing down on the oscillator casing by hand, at least with approximately 425 g of force, does not significantly dampen the response characteristics of the oscillator (Yang et al., 1991). Comparison of bone-conduction ASSRs in young infants obtained using the two different coupling methods indicated an elastic-band versus hand-held mean difference of less than 1 dB. This small difference was not statistically significant. As noted above, ASSR threshold variability was also not different between the two methods. Similarly, the differences in ASSR amplitudes for the two coupling conditions were small (< lOnV difference) and not statistically significant. The small, albeit non-significant, elevation in threshold for the hand-held condition for 4000 Hz may have some practical importance. The infant ASSR threshold and amplitude results for the elastic-band and hand-held coupling conditions are consistent with the adult threshold results which also showed little difference with coupling method. 156 The only published study that compares bone-conduction thresholds using different coupling methods was conducted in dogs by Munro, Paul and Cox (1997). They compared bone-conduction ABR thresholds to click stimuli in two species of dogs that had significantly different head sizes (Dalmations vs. Jack Russell terriers) holding the oscillator both by hand and using a 500 g weight. Similar to our findings, these authors found no difference in bone-conduction thresholds using a hand-held coupling method compared to a method that applies a constant force (i.e., elastic band or 500 g weight). The clinical implications of these findings are that bone-conduction thresholds can be obtained reliably in infants using either an elastic-band or hand-held coupling method, with the caveat that the individual who is coupling the oscillator to the patient's head must have received appropriate training on whichever method is used. There are clinical situations in which one method may be preferred. For example, the hand-held method may be a better choice if putting the elastic band on the infant's head is likely to wake the infant. There are also clinical settings in which holding the oscillator by hand is not practical, for example, when the evoked potential equipment is outside the sound booth and an assistant is not available to hold the oscillator. Bone oscillator placement This was the first infant experiment to assess directly the effects of changing oscillator placement on frequency-specific bone-conduction thresholds. The findings from the current experiment suggest that ASSR thresholds obtained with bone oscillator placement on the forehead are substantially elevated with respect to thresholds found with either temporal or mastoid oscillator placements. Thresholds obtained at the temporal and mastoid oscillator placements did not differ significantly. ASSR thresholds for the NICU infants averaged across the temporal and mastoid placements were 17, 15, 34, and 30 dB HL for 500, 1000, 2000 and 157 4000 Hz, respectively. On average, thresholds for the forehead placement were significantly higher than both the temporal and mastoid placements by at least 14, 11, 18, and 14 dB at 500, 1000, 2000 and 4000 Hz, respectively. These differences may be even greater because absent responses at the maximum intensity were seen more often with the forehead placement compared to the other two placements. Specifically, absent responses at the highest intensity (50 dB HL) were seen for the forehead placement in 18% of the recordings whereas absent responses were seen in only 5% of the mastoid and temporal recordings. This was particularly the case for 2000 and 4000 Hz where absent responses were seen for 37% of the forehead placement results compared to only 5% of the temporal and mastoid placement results. The temporal/mastoid placement should, therefore, be used to maximize the intensity range available to assess thresholds in infants. The findings that infant thresholds differ between the forehead and mastoid are consistent with previous adult behavioural studies (Dirks, 1994). In adults, forehead thresholds are elevated with respect to mastoid thresholds by an average of 14, 8.5, 11.5, and 8.0 dB for 500, 1000, 2000, and 4000 Hz, respectively (ANSI, 1996). At 500 Hz, infants and adults have the same forehead-mastoid threshold differences. The difference in forehead-mastoid threshold for infants, however, increases with increasing frequency: thresholds at 1000, 2000, and 4000 Hz were larger than those of the adult by 2.5, 6.5, and 6.0 dB, respectively. Possible reasons for the larger attenuation differences at higher frequencies exhibited by the infant compared to the adult may result from the membranous sutures surrounding the temporal bone in the infant (Yang et al., 1987). These membranous sutures have the effect of attenuating the vibratory signal before it reaches the cochlea. When the forehead oscillator placement location is used, the vibratory energy must pass through two layers of membranous sutures before reaching the temporal bone 158 to stimulate the cochlea and thus initiate a physiological response (Yang et al., 1987). Consequently, the effective intensity that reaches the cochlea decreases as the distance between the bone oscillator and the cochlea increases (Stuart et al., 1990). Thresholds obtained at the forehead should, therefore, be worse (i.e., higher) than those obtained at either the temporal/mastoid locations which is consistent with the current findings. As noted by Yang et al. (1987), the membranous sutures in the infant skull may act like a low-pass filter, thus allowing low-frequency energy to pass the sutures with minimal attenuation while the high- frequency energy is substantially attenuated. The idea that membranous sutures act like a low-pass filter in infants may explain why infants and adults have the same attenuation between placements at 500 Hz, but demonstrate more attenuation than adults at 1000, 2000, and 4000 Hz. In contrast to the findings in the current experiment, Stuart et al. (1990) concluded that changing the position of the oscillator on the temporal bone produced significant differences in signal attenuation to the cochlea. They recorded ABRs to click stimuli at different intensities using different oscillator placements on the temporal bone. They reported differences in ABR latencies for the different temporal placements and suggested that attenuation within the temporal bone occurs in infants because the temporal bone consists of several unfused components, thereby causing a reduction in signal transmission when areas of the temporal bone farther away from the cochlea are used for oscillator placement. They conjectured that oscillator placements which are closest to the cochlea will transmit the greatest intensities, whereas oscillator placements on the unfused areas, further away on the temporal bone, will result in lower intensities (Stuart et al., 1990). Although ABR latency differences existed between the various temporal placements, the assumption that attenuation would follow the same pattern appears to be incorrect. Latency is not linearly related to signal attenuation (e.g., Mackersie & Stapells, 159 1994; Picton, Stapells, & Campbell, 1981), consequently, latency-intensity functions cannot be used to accurately estimate threshold changes (Mackersie & Stapells, 1994). They also did not report wave V amplitudes, which are better predictors of threshold than latency, did not directly estimate threshold at the different placements on the temporal bone, which would have been the best measure for assessing attenuation, and did not use frequency-specific stimuli. The findings of Experiment 2 confirmed that latency data do not accurately estimate attenuation. Although the results of this experiment show elevated thresholds at the forehead compared to the temporal and mastoid oscillator placements, several limitations exist regarding predicting attenuation of the bone-conducted signal across the skull. One limitation is that ASSR thresholds obtained in pre-term infants may not reflect the ASSR thresholds for normal full-term infants for the different oscillator placements. Another limitation is that threshold was not always reached at the maximum test intensity for each placement, particularly at the forehead, resulting in an underestimation of the differences between placements. It is likely that the elevated (i.e., worse) thresholds occur with the forehead location due to greater signal attenuation as a function of distance from the cochlea and, furthermore, the greater differences in threshold at the higher frequencies between the forehead and mastoid in infants are likely due to the low-pass filtering effect imposed by the sutures of the infant skull. Unoccluded vs occluded ears It is well established that there are significant infant-adult differences in the transfer of acoustic energy through the outer and middle ear, but these differences are not well understood. We know that the infant ear canal is narrower and shorter than the adult ear (Keefe et al., 1994); it has also been shown that the resonant frequency of the ear canal is higher in infants than adults (Keefe et al., 1994; Kruger, 1987; Kruger & Ruben, 1987). We also know that the ear canal wall 160 is thinner and more compliant in infants up to two months of age than in adults (Holte, Margolis & Cavanaugh,1991). Because we do not fully understand the infant-adult differences in the transfer of acoustic energy in the outer/middle ear, we cannot assume that phenomena such as the occlusion effect, which has been studied only in adults, are necessarily present in infants or, if present, follow the same pattern. This is the first experiment to investigate the occlusion effect in infants. The results of this experiment show that bone-conduction ASSR thresholds in infants under six months of age do not change at any of the frequencies tested when the ear canal was occluded. Comparison of unoccluded and occluded bone-conduction mean ASSR thresholds indicated no more than a 4-dB difference across frequency, in contrast to adults whose bone-conduction thresholds at 250-1000 Hz improve as much as 17 dB when the ear canal is occluded using an insert earphone (Dean & Martin, 2000). Bone-conduction mean ASSR amplitudes and phase delays were also not significantly affected by occluding the ear canal in these young infants. There are a number of possible explanations for the absence of an occlusion effect in young infants. In adults, the unoccluded ear acts as a high-pass filter [(i.e., low-frequency energy is lost through the open ear canal (Gelfand, 1981, p. 66; Tonndorf, 1966)]; in an occluded ear, the improvement in the bone-conduction thresholds is due to the enhancement of the low frequencies. In the infant ear canal, it is possible that there is little increase in low-frequency energy when the ear canal is occluded due to its smaller volume or shorter length (Keefe et al., 1994), or that the insert phone takes up most of the small ear canal volume. Alternatively, if there is low-frequency energy trapped in the occluded infant ear (similar to adults), it may be absorbed by the infant's compliant ear canal wall (Keefe et al., 1993, 1994), resulting in no net increase in energy passing through to stimulate the cochlea. 161 These preliminary results suggest that there is no effect of occlusion in infants under six months of age; however, further studies should be conducted in a larger group of infants to confirm these findings. The clinical implications of these occlusion findings are that it may be possible to do bone-conduction testing, at least in infants under six months of age, with ears occluded without applying a correction factor. Also, it is important to investigate whether the occlusion effect is present in older infants and to determine at which age the occlusion effect should be compensated for in clinical testing. Conclusions The results of these experiments have clinical implications for bone-conduction testing procedures used in infants. Our findings support that: (i) either the elastic-band or the hand-held method is appropriate for coupling the bone-oscillator to the head, with the important caveat that adequate training has taken place for the method used, (ii) either a temporal or mastoid placement location can be used (forehead placement should be avoided), and (iii) ears may be unoccluded or occluded during bone-conduction testing without significantly affecting threshold estimation. 162 Tab le 4.1 The range, mean and mean absolute deviat ion (MAD) for force levels produced by four ass is tants trained to use the elast ic-band (EB) and hand-held (HH) bone-osci l lator coupl ing methods. Coupl ing E B H H Assistant Range M e a n S D M A D R a n g e M e a n S D M A D #1 425-575 483 49 58 360-450 400 32 33 #2 325-490 409 61 54 340-500 436 52 41 #3 350-525 440 54 40 375-550 479 51 64 #4 400-600 490 70 75 360-460 405 35 35 All 325-600 455 66 57 340-550 430 52 43 Target of 425±25g; 10 test trials each S D = 1 standard deviat ion M A D : M e a n absolute deviat ion from target 163 Table 4.2 M e a n behavioural thresholds for 500-, 1000-, 2000- and 4 0 0 0 - H z bone-conduct ion pure tones using elast ic-band (EB) and hand-held (HH) coupl ing method in adults with normal hearing (n=10). Coupl ing 500 H z 1000 H z 2000 H z 4000 H z Threshold E B Mean 1.7 -0.3 11.5 -1.7 d B HL S D 3.7 6.6 7.5 5.5 n 10 10 10 10 H H Mean 1.5 2.3 8.7 0.1 S D 4.9 4.1 5.4 5.0 n 10 10 10 10 Difference Mean 0.2 -2.6 2.8 -1.8 S D 4.2 3.9 4.5 5.6 n 10 10 10 10 Difference = E B minus H H S D = 1 standard deviation 164 Tab le 4.3 M e a n A S S R thresholds for 500-, 1000-, 2000- and 4000 -Hz bone-conduct ion carrier f requencies using elast ic-band (EB) and hand-held (HH) coupl ing method in infants with normal hearing (n=10; 2-10 months). Coupl ing 500 H z 1000 H z 2000 H z 4000 H z Threshold E B Mean 14.0 6.0 26.0 13.0 dB HL S D 14.3 8.43 9.7 11.6 n 10 10 10 10 H H Mean 14.0 1.0 25.0 22.0 S D 14.3 7.4 7.1 7.9 n 10 10 10 10 Difference Mean 0.0 5.0 1.0 -9.0 S D 15.6 5.3 7.4 12.9 n 10 10 10 10 Difference = E B minus H H S D = 1 standard deviat ion 165 Table 4.4 A S S R mean ampl i tudes elicited by bone-conduct ion stimuli presented at 30 dB HL for 500- and 1000-Hz and at 40 d B HL for 2000- and 4000- carrier f requencies using elast ic-band (EB) and hand-held (HH) coupl ing method in infants with normal hearing (n=6-9; 2 to 10 months). Coupl ing 500 H z 1000 H z 2000 H z 4000 H z Ampl i tude E B M e a n 4 1 . 0 5 9 . 2 2 4 . 5 3 6 . 3 nV S D 2 1 . 3 19.5 12.6 15.8 n 9 9 6 6 H H M e a n 3 0 . 5 5 4 . 5 3 6 . 9 3 2 . 8 S D 16 . 3 18.5 21.6 17.1 n 9 9 7 7 Difference M e a n 5 . 8 5 . 4 - 1 2 . 6 9 . 0 S D 15 . 3 17.1 16 . 3 11.2 n 8 8 5 5 Difference = E B minus H H S D = 1 standard deviat ion 166 Tab le 4.5 Bone-conduct ion A S S R mean thresholds for 500-, 1000-, 2000-and 4000 -Hz carrier f requencies for three bone-osci l lator p lacements in pre-term infants with normal hearing (n=9-15). P lace 500 H z 1000 H z 2000 H z 4000 H z Threshold Tempora l Mean 16.0 16.7 34.6 33.3 dB HL S D 11.8 9.0 15.1 15.0 n 15 15 13 15 Mastoid Mean 17.3 14.0 32.3 26.0 S D 13.3 9.1 19.6 12.9 n 15 15 13 15 Forehead Mean 30.7 26.7 51.1 44.0 S D 16.2 13.5 13.6 9.7 n 15 15 9 10 S D = 1 standard deviation 167 Table 4.6 Bone-conduct ion A S S R mean ampl i tudes at 40 dB HL for 500-, 1000-, 2000- and 4000 -Hz carrier f requencies for three bone-osci l lator p lacements in pre-term infants with normal hearing (n=13-15). P lace 500 H z 1000 H z 2000 H z 4000 H z Ampl i tude Tempora l Mean 47.8 58.2 18.5 17.8 nV S D 19.6 24.1 10.2 9.0 n 13 13 13 13 Mastoid Mean 48.9 66.0 15.3 21.8 S D 30.1 32.4 9.3 9.9 n 15 15 15 15 Forehead Mean 29.9 41.7 8.7 9.7 S D 17.6 23.4 4.2 5.6 n 13 13 13 13 S D = 1 standard deviat ion 168 Table 4.7 M e a n A S S R thresholds for 500-, 1000-, 2000- and 4 0 0 0 - H z bone-conduct ion carrier f requencies for unoccluded and occ luded ears in infants with normal hearing (n=10-13; 0 to 6 months). 500 H z 1000 H z 2000 H z 4000 H z Threshold Unocc luded Mean 18.5 3.1 30.0 16.2 dB HL S D 14.1 8.6 5.8 9.6 n 13 13 13 13 Occ luded Mean 14.6 3.9 30.0 13.6 S D 13.9 8.7 13.3 9.2 n 13 13 10 11 Difference Mean 3.9 -0.8 1.0 0.9 S D 11.2 10.4 14.5 8.3 n 13 13 10 11 Difference = unocc luded minus occ luded S D = 1 standard deviat ion 169 Figure 4.1 Cumulative percent of trials that resulted in errors produced, i.e., 25, 50, 100, 150 and 175 g differences from the target force of 425±25 g, for each individual assistant (solid & striped bars) and across assistants (n=4; cross-hatched bars). 1 0 0 Elastic band J2 co 80 60 *- 40 20 0 100 c o o L . CD Q. CJ) > £ 60 3 O 40 20 S Assistants Assistant #2 J Assistant#3 Assistant #4 • Total 25 50 100 150 175 Difference from target (g) 170 Figure 4.2 Representative bone-conduction ASSRs for an individual post-term infant (8 weeks) for the elastic-band and hand-held coupling methods. Shown are amplitude spectra resulting from FFT analyses (70-101 Hz) of the ASSRs. Filled triangles indicate responses which differ significantly (p<.05) from the background noise. Open triangles indicate no response (p>.05 and E E G noise<l 1 nV). Threshold is defined as the lowest intensity that produced a significant response. ELASTIC BAND F R E Q U E N C Y (kHz) Intensity (dBHL) 0.5 1.0 T 2.0 4.0 T I T • 40 i l l . . . T T V T iklllllll r T V T 70 r T V V 10 • T V V 0 ll • V V V V -10 Threshold 85 90 95 100 105 0 40 20 HAND HELD F R E Q U E N C Y (kHz) 0.5 1.0 2.0 4.0 20 n V ] I " " T V T V V V k L — " -V V v v v V v V 75 BO 85 90 95 100 105 30 0 30 20 171 Figure 4.3 Representative bone-conduction ASSRs for an individual pre-term infant (35 weeks PCA) for three bone oscillator placements. Shown are amplitude spectra resulting from FFT analyses (70-101 Hz) of the ASSRs. Filled triangles indicate responses which differ significantly (p<.05) from the background noise. Open triangles indicate no response (p>.05 and E E G noise<l 1 nV). Threshold is defined as the lowest intensity that produced a significant response. Intensity (dBHL) 50 40 30 20 10 T E M P O R A L T • V • V v • T V T V v MASTOID F R E Q U E N C Y (kHz) 0.5 1.0 2.0 4.0 • 20nV] T • V v V v u u t . J . L ^ , . F O R E H E A D T • • hLiJiiM I V I v • v v v v v jjlttlhlllf I -10 v v v v V V V v V V V v i M r t l i M i m m 75 80 85 90 95 100 105 75 80 85 90 95 100 105 75 80 85 90 95 100 105 Threshold 10 0 30 30 10 0 20 20 50 10 50 40 172 Figure 4.4 Cumulative percent occurrence of pre-term subjects with significant responses across frequency for temporal (black bars), mastoid (unfilled bars) and forehead (cross-hatched bars) placements (n=15). o o c Q) 100 80 60 40 20 0 500 Hz X X X X X X X X X X X X X 1000 Hz JDJ X X X X X X X X X X X X X X X X X X X X X X X X V 10 20 30 40 50 10 20 30 40 50 Intensity (dB HL) 173 Figure 4.5 Representative bone-conduction ASSRs for an individual post-term infant (6 weeks) for test ear unoccluded and occluded. Shown are amplitude spectra resulting from FFT analyses (70-101 Hz) of the ASSRs. Filled triangles indicate responses which differ significantly (p<.05) from the background noise. Open triangles indicate no response (p>.05 and E E G noise<l 1 nV). Threshold is defined as the lowest intensity that produced a significant response. Intensity ( d B H L ) 40 30 20 10 O C C L U D E D F R E Q U E N C Y (kHz) 0.5 1.0 2.0 4 .0 • T I V T IiiiiiikiiiiLiiiiii I n • l .L, J .^L i i v V v U N O C C L U D E D F R E Q U E N C Y (kHz) 0.5 1.0 2.0 4 .0 20 nV] • T • V • „ U J w ^ , .,! Al^l i iL.HuU.*.. V V V v -10 T h r e s h o l d v v v v 75 80 85 90 95 100 105 75 80 85 90 95 100 105 10 10 30 20 10 10 30 10 174 Figure 4.6 Mean bone-conduction ASSR amplitudes (±1 SD) across frequency for ears unoccluded (filled circles) and occluded (open circles) for 13 post-term infants. i 1 1 1 1 1 r 150 h500 Hz 100 CD 5 0 I ° = 150 % 100 < 50 0 i 1——i 1 1 1 r 2000 Hz T 1 1 1 1 1——r 1000 Hz J I L i 1 1 1 1 r 4000 Hz unoccluded occluded J I I L -10 0 10 20 30 40 50 -10 0 10 20 30 40 50 I n t e n s i t y ( d B H L ) 175 Figure 4.7 Mean bone-conduction ASSR phase delays (±1 SD) across frequency for ears unoccluded (filled circles) and occluded (open circles) for 13 post-term infants. 0) CD D) CD <D "D CD (7) 360 -270 -180 90 0 360 270 180 90 0 I I 1 1 1 1 1 500 Hz - i i i i i i i -i i i i i i i —#— unoccluded 1 000 HZ ~ O - occluded - I l l l l l l -1 1 1 1 1 1 1 2000 Hz - i i i i i i i -l l l l l l l 4000 Hz - -•10 0 10 20 30 40 50 -10 0 10 20 30 40 50 Intensity (dB HL) 176 REFERENCES ANSI. (1996). American National Standard Specifications for Audiometers (ANSI S3.6-1996). New York: ANSI. Cone-Wesson, B. (1995). Bone-conduction ABR tests. American Journal of Audiology 4: 14-19. Dean, M . S., & Martin F. N. (2000). Insert earphone and the occlusion effect. American Journal of Audiology 9: 131-134. Dirks, D., & Swindeman J. G. (1967). The variability of occluded and unoccluded bone-conduction thresholds. Journal of Speech and Hearing Research 10: 232-249. Elpern, B., & Naunton R. F. (1963). The stability of the occlusion effect. Archives of Otolaryngology 11: 376-384. Gelfand, S. A. (1981). Hearing: An introduction to psychological and physiological acoustics. pp. 66. New York, NY: Marcel Dekker, Inc. Harrell, R. W. 2002. Pure tone Evaluation. In Handbook of Clinical Audiology, Fifth edition, ed. J. Katz, pp. 71-87. Philidelphia: Lippincott, Williams & Wilkins. Hodgson, W. R., & Tillman T. W. (1966). Reliability of bone conduction occlusion effects in normals. Journal of Auditory Research 6: 141-151. Holte, L. A., Margolis R. H., & Cavanaugh R. M . , Jr. (1991). Developmental changes in multifrequency tympanometry. Audiology 30: 1-24. John, M. S., Dimitrijevic A., & Picton T. W. (2001). Weighted avearging of steady-state responses. Clinical Neurophysiology 112: 555-562. 177 John, M . S., & Picton T. W. (2000a). MASTER: A Windows program for recording multiple auditory steady-state responses. Computing Methods and Programs in Biomedicine 61: 125-150. John, M . S., & Picton T. W. (2000b). Human auditory steady-state responses to amplitude-modulated tones: phase and latency measurements. Hearing Research 141: 57-79. Keefe, D., Bulen J., Arehart K , & Burns E. (1993). Ear-canal impedance and reflection coefficient in human infants and adults. Journal of the Acoustical Society of America 94: 2617-2638. Keefe, D. H., Bulen J. C , Campbell S. L., & Burns E. M . (1994). Pressure transfer function and absorption cross section from the diffuse field to the human infant ear canal. Journal of the Acoustical Society of America 95: 355-71. Kruger, B. (1987). An update on the external ear resonance in infants and young children. Ear and Hearing 8: 333-336. Kruger, B., & Ruben R. J. (1987). The acoustic properties of the infant ear. Acta Otolaryngologica (Stockholm) 103: 578-585. Little, R. J..A, & Rubin, D.B. (2002). Statistical Analysis with Missing Data, 2nd ed., pp.3-40. Hoboken, New Jersey: Wiley & Sons, Inc. Mackersie, C. L., & Stapells D. R. (1994). Auditory brainstem response wave I prediction of conductive component in infants and young children. American Journal of Audiology 3: 52-58. Munro, K. J., Paul, B. & Cox, C L . (1997). Normative auditory brainstem response data for bone conduction in the dog. Journal of Small Animal Practice 38: 353-356. 178 Murphy, K.R., & Myors, B. (2004). Statistical Power Analysis: A Simple and General Model for Traditional and Modern Hypothesis Tests, 2n d ed, pp. 22-97. Mahwah, New Jersey: Lawrence Erlbaum Associates. Picton, T. W., Stapells D. R. & Campbell K. B. (1981). Auditory evoked potentials from the human cochlea and brainstem. Journal of Otolaryngology 10: 1 -41. Small, S.A., & Stapells, D.R. (2003). Normal brief-tone bone-conduction behavioural thresholds using the B-71 transducer: Three occlusion conditions. Journal of the American Academy of Audiology, 14: 556-562. Small, S. A., & Stapells D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear and Hearing 25(6): 611-623. [Chapter 2] Small, S. A., & Stapells, D.R. (2006). Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing. Ear and Hearing 27: 219-28. [Chapter 4] Stapells, D. R. 2000. Frequency-specific evoked potential audiometry in infants. In A Sound Foundation Through Early Amplification, ed. R. C. Seewald, pp. 13 -31. Basel: Phonak A G . Stapells, D. R., & Oates P. (1997). Estimation of the pure-tone audiogram by the auditory brainstem response: a review. Audiology & Neuro Otology 2: 257-280. Stuart, A., Yang E. Y. , & Stenstrom R. (1990). Effect of temporal area bone vibrator placement on auditory brain stem response in newborn infants. Ear and Hearing 11: 363-369. Tonndorf, J. (1966). Bone conduction: Studies in experimental animals. Acta Otolaryngologica, Supplement 213:1-132. 179 Wilber, L. A. (1979). Pure-tone audiometry: Air and bone conduction. In Hearing assessment, ed. W. F. Rintelmann, pp. 27-42. Baltimore, MD: University Park Press. Yang, E. Y., & Stuart A. (1990). A method of auditory brainstem response testing of infants using bone-conducted clicks. Journal of Speech Language-Pathology and Audiology 14: 69-76. Yang, E. Y., & Stuart, A. (2000). The contribution of the auditory brainstem responses to bone-conducted stimuli in newborn hearing screening. Journal of Speech-Language Pathology and Audiology 24: 84-91. Yang, E. Y., Stuart, A., Stenstrom M . A., & Hollett S. (1991). Effect of vibrator to head coupling force on the auditory brain stem respone to bone-conducted clicks in newborn infants. Ear and Hearing 12: 55-60. Yang, E. Y., Rupert A. L., & Moushegian G. (1987). A developmental study of bone conduction auditory brainstem responses in infants. Ear and Hearing 8: 244-251. 180 CHAPTER 5 MULTIPLE AUDITORY STEADY-STATE RESPONSE THRESHOLDS TO BONE-CONDUCTION STIMULI IN YOUNG INFANTS WITH NORMAL HEARING A version of this chapter has been published. Small, S. A., & Stapells, D. R. (2006). Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing. Ear and Hearing 27(3): 219-228. 181 INTRODUCTION Universal newborn hearing screening programs are identifying infants at risk for hearing loss in the first few days and weeks of life. These infants are referred for diagnostic hearing assessments to determine the severity and type of hearing loss so that aural habilitation can begin before the age of six months (Joint Committee on Infant Hearing, 2000). Accurate and efficient techniques that are appropriate for very young infants are needed to assess frequency- specific hearing thresholds in order to achieve this goal. Conventional behavioural measures cannot reliably assess hearing thresholds in infants younger than 5-6 months of age (Moore, Wilson & Thompson, 1977). The auditory brainstem response (ABR) can be recorded in infants as early as 29 weeks post-conceptional age (Ponton, Moore, Eggermont, Wu & Huang, 1994), and obtaining ABR thresholds to air- and bone-conduction brief-tone stimuli is currently the clinical "gold standard" for estimating frequency-specific hearing threshold in young infants (Stapells, 2000, Stapells, Herdman, Small, Dimitrijevic & Hatton, 2005). A number of recent studies have shown that the 70-110 Hz auditory steady-state responses (ASSR), which use amplitude and/or frequency modulated stimuli to evoke a response, may also provide accurate frequency-specific estimates of air-conduction hearing thresholds. ASSRs can be elicited by single stimuli or by multiple stimuli presented simultaneously (reviewed in Picton, John, Dimitrijevic & Purcell, 2003). The 80-Hz ASSR is currently of great interest because it is minimally affected by sleep (Cohen, Rickards, & Clark, 1991) and possible to record in infants. Similar to the ABR, the 80-Hz ASSR is believed to be generated primarily in the brainstem (Herdman, Lins, Van Roon, Stapells, Scherg & Picton, 2002; Mauer & Doling, 1999). One limitation of the ABR to brief tones is that detection of a response in the waveform depends on skilled, subjective assessment of replicated responses, allowing for error in judgement of the presence of responses depending on 182 the experience of the clinician (Stapells, 2000). ASSRs are detected objectively using statistical tests (John & Picton, 2000a) and, thus, their detection does not rely on the experience of the clinician. Another shortcoming of the ABR technique is that only one ear and one frequency can be tested at the same time; recording multiple ASSRs in both ears simultaneously has the potential to reduce clinical testing time. Behavioural hearing threshold estimation routinely assesses air-conduction and bone-conduction pure-tone thresholds to distinguish among sensorineural, conductive and mixed hearing losses. Accurate bone-conduction thresholds are particularly important when assessing children who have unilateral or bilateral otits media or atresia (Jahrsdoerfer, Yeakley, Hal, Robbins & Gray, 1985; Stapells & Ruben, 1989). The accurate diagnosis of a conductive or mixed hearing loss is necessary in order to plan appropriate intervention. Few bone-conduction ASSR data exist, and before ASSRs can be used as a standard diagnostic clinical tool, normative bone-conduction ASSR threshold data are needed for infants of different ages. This is particularly important for infants younger than six months, who will be the population referred from hearing screening programs for diagnostic testing. There are only three studies that have recorded bone-conduction ASSRs in adults with normal hearing (Dimitrjevic, John, Van Roon, Purcell, Adamonis, Ostroff, Nedzelski, & Picton, 2002; Lins, Picton, Boucher, Durieux-Smith, Champagne, Moran, Perez-Abalo, Martin, & Savio, 1996; Small & Stapells, 2005: Chapter 3) and no bone-conduction ASSR data for adults with hearing loss, except those in the severe-to-profound range. There are also no published data for bone-conduction ASSRs for normal or hearing-impaired infants. The main purpose of this study is to determine, bone-conduction ASSR thresholds in pre-and post-term infants who are younger than eight months of age with normal hearing, and to 183 compare these to bone-conduction ASSR thresholds in adults with normal hearing. MATERIALS AND METHODS Participants Two groups of infants who passed a hearing screening in both ears participated. ASSRs to bone-conduction stimuli were recorded in 29 pre-term infants in a Neonatal Intensive Care Unit (NICU) [age range of 32-43 weeks post-conceptional age (PCA); mean age of 34.5 weeks PCA] and 14 post-term infants recruited from the community [age range of 0.5-27 weeks; mean age of 17 weeks]. Data from 10 adults with normal hearing from Small & Stapells (2005) [Chapter 3] were also included [see Small & Stapells (2005: Chapter 3) for methods used to record the ASSRs]. All of the infants from the NICU and four of the infants recruited from the community were screened using an automatic auditory brainstem response (AABR) screening test at 35 dB nHL. The hearing of the other infants was screened using a distortion-product otoacoustic emissions (DPOAE) screening test. Infants who passed the A A B R or DPOAE hearing screening test in both ears were considered to be at low risk for significant hearing loss and thus included in the study. Stimuli All stimuli were sinusoidal bone-conduction tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude and 25% frequency modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. The stimuli were presented simultaneously for all conditions tested (i.e., multiple). All stimuli were generated by the Rotman MASTER research system (John & Picton, 2000a) using a buffer length of 25,600 points and a digital-to-analog 184 (D/A) rate of 31,250 Hz, which is an integer sub-multiple of the 20 MHz clock rate, but not an integer multiple of the carrier frequencies. The stimuli were then attenuated through either an Interacoustics AC40 clinical audiometer (laboratory) or Med-Associates ANL-918 attenuator (NICU). Before the stimuli were attenuated, they were routed through a Stanford Research Systems Model SR650 to increase the gain of the stimuli by 10 dB. The bone-conduction stimuli were presented to a Radioear B-71 bone oscillator which was held by hand for the infant groups and by elastic headband for the adults with 400-450 g of force (Small, Hatton & Stapells, 2007: Chapter 4)1. The bone oscillator was placed on the temporal bone within 2 cm of the pinna for all infants tested. Bone-conduction stimuli were presented using 10-dB steps from 50 to -10 dB HL [intensities greater than 50 dB HL for multiple stimuli resulted in non-linearities in the oscillator output (Small & Stapells, 2004: Chapter 2)]. ASSRs for all infants and adults were elicited by a non-inverted stimulus polarity (Small & Stapells, 2004: Chapter 2)2. In a separate study, we have shown that estimation of bone-conduction ASSR thresholds in infants and bone-conduction behavioural thresholds in adults was not significantly affected by a hand-held or elastic-band coupling method (Small & Stapells, 2007: Chapter 4). 2 To rule out the methodologic concern that smaller heads (i.e., compared to adults) may result in spurious responses due to the potentially larger electromagnetic stimulus artifact in the EEG aliasing to the modulation rates of the carrier frequencies (Small & Stapells, 2004), bone-conduction ASSR amplitudes and phase delays for non-inverted, inverted and alternated stimulus polarities were compared in pre-term infants. No differences were found in the amplitude and phase characteristics of the ASSR for these infants at the highest intensity tested (50 dB HL) between ASSRs recorded to single-polarity or alternated stimuli. These results confirm that ASSRs to bone-conduction stimuli can be recorded using single-polarity stimuli up to 50 dB HL in young infants. It should be noted that if the inverting electrode were placed at the mastoid, stimulus artifact could be worse. 185 Calibration The bone-conduction stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re:luN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 550 g of force. ASSR Recordings ASSRs were recorded using the Rotman MASTER system. Three electrodes were used to record the electrophysiologic responses; the non-inverting electrode was placed midline at the high forehead, the inverting electrode was positioned midline at the nape of the neck, just below the hairline, and an electrode placed at the low forehead acted as ground. All inter-electrode impedances were below 3 kOhms at 10 Hz. The E E G was filtered using a 30-250 Hz filter (12 dB/oct) and amplified 80,000 times (8000X in Nicolet HGA-200A and Nic501A; 10X in NIDAQ card). The E E G was further filtered using a 300-Hz lowpass anti-aliasing filter [NICU: Stanford Research Systems (115dB/oct); laboratory: Wavetek Rockland Model 852 (48 dB/oct)]. The E E G was then processed using a 1250-Hz A/D conversion rate (Small & Stapells, 2004: Chapter 2). Each E E G recording sweep was made up of 16 epochs of 1024 data points (0.819 seconds per epoch) and lasted a total of 13.107 seconds. Artifact rejection was set to eliminate epochs of electrophysiologic activity that exceeded ± 40 |IV in amplitude in order to reduce contributions to the E E G due to muscle artifact. ASSRs were averaged in the time domain and then analyzed online in the frequency domain using a Fast Fourier Transform (FFT). Weighted averaging (John, Dimitrijevic & Picton, 2001) was used. The FFT resolution was 0.076 Hz over a range of 0 to 625 Hz. 186 Amplitudes were measured baseline-to-peak and expressed in nV. An F-ratio was calculated by MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ± 60 bins of the modulation frequency ("noise") (John & Picton, 2000a). A minimum of seven sweeps were recorded for each test condition. A response was considered to be present if the F-ratio, compared to the critical values for F (2, 240), was significant at a level of p < .05 for at least two consecutive sweeps. A response was considered to be absent if p > .05 and the mean amplitude of the noise was less than 11 nV. Alternatively, a response was also considered to be absent when response amplitude was < 10 nV and thep value > .30. Procedure Testing of pre-term infants was performed at the bedside in the NICU; testing of the post-term infants recruited from the community was conducted in a double-walled sound-attenuated booth in the Human Auditory Physiology Laboratory at the University of British Columbia. Both infant groups (and adults) were tested with ears unoccluded. Ambient acoustic noise was likely to affect the results when recording ASSRs in the NICU. Average ambient noise levels in the NICU measured at two different time points using 1/3-octave bands were 52, 51, 60, 58, 50, 36 and 47 dB SPL at 0.125, 0.250, 0.5, 1, 2, 4, and 8 kHz, respectively. The overall A-weighted noise level in the NICU was 65 dB SPL. On average, the noise levels at 500, 1000, 2000 and 4000 Hz were 48, 48, 40 and 24 dB higher in the NICU than in the sound-attenuated booth. To reduce the impact of high ambient noise in the NICU, recording of ASSRs was paused when the background acoustic noise was excessively high (e.g., when sink taps were running near the infant's bassinet). 187 Participants were only tested when quiet and asleep. Hearing screening was performed in both ears at the beginning of the test session to establish that the participants were unlikely to have hearing loss. Multiple ASSRs were elicited to bone-conduction stimuli beginning at a randomized starting intensity. Threshold for each carrier frequency was determined using a bracketing technique adjusting the presentation level using 10-dB steps. The lowest intensity at which a response was present was considered "threshold". In some cases, a response did not reach significance at a level above "threshold"; if only one response did not reach significance, the lowest level at which the response was present was considered threshold. If two or more responses were not significant, the response at the lowest level was considered a "false positive" and the lowest level above the absent responses was deemed "threshold". The total recording time was approximately 1.0-1.5 hours, including the time to obtain screening test results. The parents of the infant participants signed a consent form before commencing any of the experiments; parents of infants tested at the Human Auditory Physiology Laboratory were paid an honorarium at the end of each session. Data Analyses Mean amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant. The phase values from MASTER were adjusted by adding 90° to yield the onset phase (John & Picton, 2000b). Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Any phase-delay values that differed > 180° from an adjacent measure were "unwrapped" by adding 360° to their value (John & Picton, 2000b). Phase values for ASSRs that were not significant were excluded from mean phase-delay calculations. Phase-delay values were averaged across subjects. Results are only reported if at least five subjects contributed to the mean. 188 Bone-conduction ASSR thresholds were compared across frequency and between pre-and post-term infant groups, and compared to bone-conduction ASSR thresholds in adults with normal hearing that were obtained in our previous study (Small & Stapells, 2005: Chapter 3). Bone-conduction ASSR amplitudes and phase delays (ms) at 40 dB HL were also compared across frequency for the two infant groups and adults. Comparisons across age groups were made using a 2-way mixed-model analysis of variance (ANOVA). Huynh-Feldt epsilon adjustments for repeated measures were made when appropriate. Newman-Keuls post-hoc comparisons were performed for significant main effects and interactions. The criterion for statistical significance was p < .05 for all analyses. R E S U L T S Comparison of bone-conduction ASSRs in infants and adult Representative bone-conduction ASSR threshold results for individuals from each age group are illustrated in Figure 5.1. Both of the infants had better thresholds at 500 and 1000 Hz compared to the adult, but poorer thresholds at 2000 and 4000 Hz compared to the adult. Also, both infants had thresholds at 500 and 1000 Hz that were 20-30 dB better than their thresholds at 2000 and 4000 Hz, in contrast to the adult subject whose thresholds were the same at 500, 1000 and 2000 Hz and 10 dB better at 4000 Hz. Figure 5.2 shows the cumulative percent occurrence of significant responses for infants and the adults. Overall, infants had many more responses present for 500 and 1000 Hz compared to adults and many fewer responses at 2000 and 4000 Hz compared to adults. Within each infant group, there were many more responses at 500 and 1000 Hz compared to 2000 Hz. In contrast, the carrier frequencies for adults showed similar results across frequency. For the post-term infants, 90% or more of the subjects had responses present 189 at minimum intensities of 30, 10, 40 and 30 dB H L for 500, 1000, 2000 and 4000 Hz, respectively. The adults needed the same minimum intensity to meet a 90% criterion for 500 and 4000 Hz; however, the minimum intensity needed for adults was higher at 1000 Hz and lower at 2000 Hz (30, 40, 20 and 30 dB HL for 500, 1000, 2000 and 4000 Hz, respectively). For pre-term infants, a higher minimum intensity was required compared to the older infants and adults to meet the 90% criterion at frequencies above 500 Hz (40, 30, >50 and 50 dB HL for 500, 1000, 2000 and 4000 Hz, respectively). Figures 5.1 and 5.2 Threshold The mean bone-conduction ASSR thresholds for the pre- and post-term infants are shown in Table 5.1. Mean bone-conduction ASSR thresholds for adults obtained in our previous study (Small & Stapells, 2005: Chapter 3) are included in Table 5.1 for comparison. Similar to the individual threshold results, mean ASSR thresholds for 500 and 1000 Hz are lower (i.e., better) for both infant groups compared to adults, and mean ASSR thresholds for 2000 and 4000 Hz are higher (i.e., poorer) for infants compared to adults. For both infant groups, mean ASSR thresholds are substantially lower at 500 and 1000 Hz compared to 2000 and 4000, in contrast to adults, who show no significant difference in bone-conduction ASSR threshold across frequency (Small & Stapells, 2005: Chapter 3). Results of an A N O V A comparing bone-conduction ASSR threshold across age and frequency (500-4000 Hz), revealed a significant effect of age [F(2,46)=9.254,/? = .0003], frequency [F(3,46) = 12.890,/? = .0001] and interaction between age and frequency [F (6,138) = 7.773,/? - .0001]. Post-hoc comparisons indicate that the 190 significant effect of age was explained by overall higher (i.e., poorer) thresholds for pre-term infants compared to the older infants and adults. Post-hoc comparisons of the frequency effect indicated no difference between thresholds at 500 and 1000 Hz, and no difference between thresholds at 2000 and 4000 Hz; however, thresholds at the lower frequencies were significantly better than at the higher frequencies. Post-hoc comparisons of the significant interaction between age and frequency revealed no difference between thresholds at 500 and 1000 Hz, and no difference between thresholds at 2000 and 4000 Hz for any of the age groups. To simplify the comparison of thresholds between age groups, the thresholds for low frequencies were pooled and the thresholds for the high frequencies were pooled for each age group. Table 5.1 Pooled thresholds for the low frequencies were better (i.e., lower) for infants compared to adults; however, pooled thresholds for the high frequencies were better for adults compared to infants. Results of an A N O V A comparing bone-conduction ASSR threshold across age and frequency revealed a significant effect of age [F(2,45 )= 8.674,/? = .0007], frequency [F (1,45) = 27.760,/? = .0001] and interaction between age and frequency [F(2,45)=l 5.953, /? = .0001]. Post hoc comparisons indicated that for pooled low frequencies, the post-term infants had significantly better thresholds than both the pre-term infants and adults; the difference in threshold between pre-term infants and adults approached significance (/? = .077). For pooled high frequencies, pre-term infants had significantly poorer thresholds than the post-term infants and the adults; the difference in high-frequency threshold between adults and post-191 term infants approached significance (p = .066). For both infant groups, pooled low-frequency thresholds were significantly better than their pooled high-frequency thresholds. No difference was found between pooled low- and high-frequency thresholds for adults. Amplitude and phase Mean amplitudes and phase delays of the ASSR across intensity are shown in Figures 5.3 and 5.4, respectively. Comparison of mean ASSR amplitudes across age groups showed little difference between the two infant groups but substantial differences between the infant and adult groups, particularly at high frequencies. Adults had much larger ASSRs than infants for 2000 and 4000 Hz. In contrast to the high frequencies, infants and adults had similar mean amplitudes for 500 Hz (except for larger amplitudes for adults at 50 dB HL); for 1000 Hz, mean amplitudes for adults were slightly smaller than for infants at lower intensities and similar to infants at 40-50 dB HL. Comparisons across frequency and intensity within age groups reveal different patterns for infants and adults. Mean ASSR amplitudes for both infant groups were larger for 500 and 1000 Hz compared to 2000 and 4000 Hz. In contrast, adults showed little difference in ASSR amplitude at 40 dB HL across frequency. Figures 5.3 and 5.4 Results of an A N O V A comparing mean bone-conduction ASSR amplitude at 40 dB HL across age and frequency revealed a significant effect of age [F (2,42) = 13.762,/? = .0001], frequency [F (3,126) = 7.001, p= .002] and interaction between age and frequency [F (6,126) = 4.494,/? = .003]. Post hoc comparisons indicated that the significant effect of age was explained by significantly larger mean ASSR amplitudes in adults compared to the infants, 192 and that the significant effect of frequency was explained by larger amplitudes for 500 and 1000 Hz compared to higher frequencies. Post hoc comparisons of the interaction between age and frequency showed no difference in mean ASSR amplitude at 40 dB HL for 500 and 1000 Hz across age groups but revealed mean ASSR amplitudes that were larger at 2000 and 4000 Hz in adults compared to infants. The significant interaction between age and frequency was also explained by the larger mean ASSR amplitudes for pre-term infants for 500 and 1000 Hz compared to 2000 and 4000 Hz, and the larger mean ASSR amplitudes for post-term infants for 500 Hz compared to 4000 Hz. Comparison of mean phase delay across frequency and intensity within and across age groups is more difficult because only phase delay values for responses that reached statistical significance can be included in the means, resulting in limited data at lower intensities and higher frequencies (Figure 5.4). Mean phase delay values at 40 dB HL, however, could be obtained for all frequencies and age groups for a minimum of five subjects per group; these mean phase delay values were converted to milliseconds to allow comparison across frequency and are shown in Table 5.2. For adults, mean phase delays were 1.5-2.9 ms shorter across frequency compared to infants. Comparison between infant groups indicated that mean phase delays were 0.35-1.35 ms longer for pre-term infants compared to post-term infants. Within each age group, phase delay decreased with an increase in frequency with the exception of phase delays for 500 Hz; for all age groups, phase delay at 500 Hz was shorter than for 1000 Hz. Results of an A N O V A comparing mean bone-conduction ASSR phase delay in milliseconds across age and frequency revealed a significant effect of age [F(2,18) = 10.498,/? = .001] and frequency [F(3,54) = 6.496,/? = .001] and no significant interaction between age and frequency [F (6,54) = .299,/? = .922]. Post hoc comparisons indicated that the significant effect of age was 193 explained by shorter mean ASSR phase delay in adults compared to infants. The significant effect of frequency was explained by longer ASSR phase delay for 500 compared to 4000 Hz. Table 5.2 DISCUSSION It is clear from the results of this study that there are age- and frequency-dependent differences in bone-conduction ASSRs when comparing pre-term and young infants to adults. The percent occurrence of responses at low frequencies for infants is much higher than for adults; at high frequencies, this pattern reverses. In infants, the percent occurrence of responses is frequency dependent (more responses at low frequencies than high frequencies), whereas in adults, the percent occurrence of responses is independent of frequency. Bone-conduction ASSR thresholds also show age- and frequency-dependent differences. Bone-conduction ASSR mean thresholds in pre-term infants range from 16 to 33 dB H L at 500-4000 Hz. ASSR mean thresholds in post-term infants under eight months of age range from 2 to 26 dB H L from 500-4000 dB HL. Adult thresholds fall within 16 to 25 dB HL from 500-4000 Hz (Small & Stapells, 2005: Chapter 3). Both infant groups show the same pattern for bone-conduction ASSRs; thresholds are better in the low frequencies (500 and 1000 Hz pooled) and poorer in the high frequencies (2000 and 4000 Hz pooled) compared to adults. For pre-term infants, ASSR thresholds are 7 dB better in the low frequencies and 18 dB poorer in the high frequencies compared to adults. For post-term infants, ASSR thresholds are 15 dB better in the low frequencies and 7 dB poorer in the high frequencies compared to adults. It is likely that the 194 masking effect of high ambient noise in the NICU, i.e., elevation of ASSR threshold across frequency in pre-term infants, has obscured some of the age-dependent differences in the low frequencies and increased the differences in the high frequencies. The results of this study also show that infants and adults have significantly different ASSR thresholds to bone-conduction stimuli across frequency. Clearly, both infant groups had better thresholds to low-frequency stimuli than to high-frequency stimuli (pre-term: 19 dB better; post-term: 17 dB better). This large difference between ASSR thresholds to low and high frequencies is not seen for bone-conduction ASSR thresholds in adults, who show only a 6.0 dB difference between the low- and high-frequency thresholds This is the first study that has (i) recorded bone-conduction ASSRs in infants, (ii) estimated bone-conduction thresholds at four frequencies in the same infant, and (iii) obtained bone-conduction thresholds in infants of different ages. The results of this study support the conclusion that there are age-dependent differences in bone-conduction ASSRs when comparing infants to adults. These findings are not surprising given that there are known developmental changes both in neurologic and anatomic structures that could potentially impact on hearing sensitivity to bone-conducted stimuli. For example, it is well established that the amplitude and latency characteristics of the ABR to air-conduction stimuli are not adult-like until approximately three years of age (Moore et al., 1996; Ponton et al., 1996) reflecting maturation of the auditory brainstem. Also, an obvious anatomic change is that skull size (thickness and surface area) increases as an infant matures. The frequency-dependent infant-adult differences are likely related to differences in the size of the skull, particularly the temporal bone, and the structural differences related to suture closure that occur by one year of age (Anson & Donaldson, 1981; Eby & Nadol, 1986). The mechanics of bone-conduction hearing in adults 195 and infants, however, are currently not well understood (e.g., Freeman, Sichel & Sohmer, 2000). Bone-conduction ASSR thresholds undergo developmental changes similar to bone-conduction ABR thresholds but there are also notable differences. The present study agrees with reports by Stapells and Ruben (1989) and Foxe and Stapells (1993), whose results show that high-frequency bone-conduction ABR thresholds (in dB nHL) tend to improve with maturity. The particularly poorer thresholds for the high frequencies indicated for pre-term infants in this study are likely explained, in part, by the high levels of NICU ambient noise resulting in an elevation of thresholds, but they may also be partially accounted for by maturation. Future studies should record bone-conduction ASSRs in pre-term infants in a quiet sound booth to determine how much of the difference in threshold between pre-term infants and adult is due to masking and how much is related to developmental changes. Our findings that low-frequency bone-conduction ASSR thresholds for both infant groups are better compared to adults differ somewhat from those reported for the ABR. Foxe and Stapells (1993) compared ABR thresholds at 500 Hz for infants (mean age: 4.8 months) and adults at 500 Hz and found no difference between them, a finding which is clearly different from the results of this study. Cone-Wesson and Ramirez (1997) tested younger infants (newborns) and found that bone-conduction ABR thresholds at 500 Hz were 20-25 dB better than the infant thresholds reported by Foxe and Stapells (1993), resulting in infant thresholds that were substantially better than the adult thresholds reported by Foxe and Stapells (1993). The trend in the present study's 500-Hz threshold data is more similar to that reported by Cone-Wesson and Ramirez but we found a smaller difference between infant and adult ASSR thresholds (7-15 dB) at this frequency. 196 The frequency-dependent differences in bone-conduction ASSR threshold within infant groups are similar to what has been previously reported for bone-conduction ABRs. Foxe and Stapells (1993) reported that post-term infants had ABR thresholds (in dB nHL) to 500-Hz brief tones that were significantly better (10.5 dB) compared to 2000 Hz, similar to this study's ASSR results. For pre-term infants, the difference between 500- and 2000-Hz ASSR thresholds (21 dB better at 500 Hz) was larger compared to the older infants. This larger difference seen for pre-term infants is due to either ambient noise masking or maturation, or both. Foxe and Stapells (1993) found no significant difference (in nHL) in adult 500- and 2000-Hz ABR thresholds (<4 dB); similarly, ASSR thresholds in adults also do not differ across frequency (Small & Stapells, 2005: Chapter 3). It is noteworthy that the trend for infants' bone-conduction ASSR thresholds to be better for low frequencies compared to high frequencies is the opposite to what is seen for air-conduction ASSR thresholds. Mean air-conduction ASSR thresholds in infants vary across studies (reviewed in: Cone-Wesson, Rickards, Poulis, Parker, Tan and Pollard, 2002; Picton et al., 2003; Stapells et al., 2005), however, air-conduction ASSR thresholds at 500 Hz are poorer than those at high frequencies. Moreover, threshold and amplitudes changes with maturation are quite different for air vs bone conduction ASSRs, with air-conduction showing all frequencies improving (lower threshold and larger ampitudes) with maturation, with larger changes for the higher frequencies (Savio, Cardenas, Perez-Abalo, Gonzalez & Valdes, 2001). The pattern of amplitudes of bone-conduction ASSRs in infants as a function of frequency are similar to those previously reported bone-conduction ABR amplitudes in infants. Low-frequency bone-conduction ABR amplitudes are larger than those to high frequencies (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989); low-frequency ASSR amplitudes in this study are also larger than those to high-frequency stimuli. 197 Direct comparison of ABR and ASSR amplitude differences across age groups cannot be made because of the additional contribution of the 40-Hz response to the adult ABR (when using a 40/second rate) that is not present in the infant ABR (Foxe & Stapells, 1993). The findings of this study indicate that adults have larger ASSRs than infants at 40 dB HL for 2000 and 4000 Hz but little difference between infants and adults at 500 and 1000 Hz. Comparison of infant and adult phase delay (in ms) for the bone-conduction ASSR and wave V latencies for the bone-conduction ABRs is complicated because direct conversion of phase (circular data) to latency (linear data) can be problematic (John & Picton, 2000). Some of the ASSR phase delay values in this study are similar to ABR latencies but others are markedly different and difficult to explain. For example, mean ASSR phase delays (in ms) for 2000 Hz are longer for infants than adults, similar to ABR results which show slightly longer wave V latencies for 2000-Hz brief tones in infants compared to adults (Foxe & Stapells, 1993), but mean ASSR phase delays (ms) for 500 Hz are much longer for the infants than adults, in contrast to ABR results which show much shorter wave V latencies for 500-Hz brief tones in infants compared to adults (1.02 ms shorter) [Foxe & Stapells, 1993; Nousak & Stapells, 1992]. If an extra modulation cycle (360 degrees) were added to the 500-Hz adult mean phase delay value (John & Picton, 2000), the resulting mean latency would be 17.32 ms, which would make it 9.38 ms longer than for post-term infants. Although "unwrapping" the adult phase data yields longer phase delay values, the new phase value is so much longer than for infants, that is unlikely to be a justifiable correction to make to the adult phase data. The explanation for why adult mean phase delays for low frequencies are so much shorter than for infants, when the opposite trend is well established for the ABR, remains unclear. 198 The comparison of phase delay of bone-conduction ASSRs across frequency in post-term infants and adults are qualitatively similar to wave V latency differences reported by Foxe and Stapells (1993) and Stapells and Ruben (1989). They reported longer wave V latencies for 500 compared to 2000 Hz for both infants and adults. Similarly, for the older infants in this study and for adults, mean ASSR phase delays (in ms) at 40 dB H L are longer for 500 and 1000 Hz compared to 2000 and 4000 Hz, although the differences were much smaller for the ASSRs compared to the ABRs. For the pre-term infants, mean phase delay was longer for 1000 Hz compared to higher frequencies, consistent with the older infants and adults; however, the mean phase delay at 500 Hz was shorter than that at 1000 and 2000 Hz. The inconsistencies in the pre-term infant phase delay results, however, are possibly due to the limited number of significant responses included in the mean (n=5). Conclusions ASSR thresholds to bone-conduction stimuli in infants and adults are different. Overall, the results suggest that low-frequency bone-conduction thresholds worsen and high-frequency bone-conduction thresholds improve with maturation. Bone-conduction ASSR threshold differences between post-term infants and adults are likely due to skull maturation. Differences between premature and older infants are likely explained both by skull changes and a masking effect of high ambient noise levels in the NICU (and possibly to other issues due to prematurity), as both infant groups show the same frequency pattern. Preliminary results indicate threshold norms in the range of 30-40 dB HL at 500-1000 Hz and >50 dB HL at 2000-4000 Hz for pre-term infants, and 10-30 dB HL at 500-1000 Hz and 30-40 dB H L at 2000-4000 Hz for 0-8-month-old infants. Normative bone-conduction ASSR 199 thresholds need to be established for much larger groups of infants before implementing ASSRs clinically. 200 Table 5.1 Bone-conduct ion A S S R mean thresholds for 500-, 1000-, 2000- and 4000-H z carrier f requencies for pre-term infants (A7=29), post-term infants (n=14) and adults (n=10)* with normal hearing. A G E 500 H z 1000 H z 2000 H z 4000 H z Threshold pre-term 16.2 15.5 37.3 32.5 dB HL (10.8) (10.2) (15.9) (12.7) post-term 13.6 2.1 26.4 22.1 (13.4) (7.0) (6.3) (8.0) adult* 21.0 25.0 18.0 16.0 (9.9) (12.7) (7.9) (10.8) Va lues in parentheses = 1 S D * Smal l & Stapel ls (2005) 201 Table 5.2 M e a n phase-de lay results (in ms) for multiple A S S R s elicited by 40 dB HL bone-conduct ion stimuli for pre- (n=29) and post-term (n=14) infants and adults (n=10)* with normal hearing. A G E 500 H z 1000 H z 2000 H z 4000 H z P H A S E D E L A Y ms pre-term 8.29 (2.01) 9.69 (0.65) 8.63 (1.38) 7.92 (1.18) post-term 7.94 (0.76) 8.34 (1.17) 7.76 (1.70) 7.01 (2.26) adult* 5.86 (1.50) 6.81 (1.76) 5.73 (1.40) 5.47 (0.90) Va lues in parentheses = 1 S D * Smal l & Stapel ls (2005) 202 Figure 5.1 Representative bone-conduction ASSR for an individual pre-term infant (35 weeks PCA), a post-term infant (6 months) and an adult (25 years) [Small and Stapells, 2005]. Shown are amplitude spectra resulting from FFT analyses (70-101 Hz) of the ASSRs. Filled triangles indicate responses which differ significantly from the background noise (p<.05). Open triangles indicate no response (p>.05 and E E G noise <11 nV). Threshold is defined as the lowest intensity that produced a significant response. Intensity (dB HL) 50 40 30 20 10 0 20 nV -10 Threshold —> Pre-term (35 weeks PCA) Post-term (6 months) Adult (25 years) FREQUENCY (kHz) . 5 1 2 4 T • T • • T • T V V V V V T v V V v v 75 80 85 90 95 100 105 10 0 30 30 T T u In • T T • T T T V V • T „ | V V V V V V 75 80 85 90 95 100 105 10 10 40 30 T T • w T J T • T • V V V T . • . i . ^ j i ^ ..... V V V V 75 80 85 90 95 100 105 20 20 20 10 203 Figure 5.2 Cumulative percent occurrence of subjects with significant responses for pre-term infants (black bars; n=29), post-term infants (white bars; n=14)and adults (cross-hatched bars; «=10) across frequency. Adult results are from Small & Stapells (2005). 0 10 20 30 40 50 0 10 20 30 40 50 I n t e n s i t y ( d B H L ) 204 Figure 5.3 Mean bone-conduction ASSR amplitudes (±1SD) across frequency for 29 pre-term infants (filled circles), 14 post-term infants (open circles) and 10 adults (filled triangles) with normal hearing (Small & Stapells, 2005). Q. 1 5 0 ^ 100 50 i 1 1 1 1 1 r 2000 Hz 1 1 1 i i i — r 1000 H z i - « - p r e t e r m —O— post-term - A - adult 1 1 1 1 1 1 r 4000 Hz J I I L -10 0 10 20 30 40 50 -10 0 10 20 30 40 50 I n t e n s i t y ( d B H L ) 205 Figure 5.4 Mean bone-conduction ASSR phase delays (±1SD) across frequency for pre-term infants (filled circles), post-term infants (open circles) and adults (filled triangles) with normal hearing (Small & Stapells, 2005). Results only plotted if at least five subjects had significant responses. 0) CD D) 0 > JS "O <D (/> 05 360 h 270 h 180 90 0 y. 360 270 180 90 0 # pre-term —O— post-term —A— adult r- I 500 Hz n r 2000 Hz I I l _ J t _ i . T 1 1 1 1 r 4000 Hz I - j 10 20 30 40 50 10 20 30 40 50 Intensity (dB HL) 206 R E F E R E N C E S ANSI (1996). American National Standard Specifications for Audiometers (ANSI S3.6-1996). New York, ANSI. Anson, B. J., Donaldson J. A. (1981). The temporal bone. In B. J. Anson, J. A. Donaldson (eds.) Surgical anatomy of the temporal bone, (pp. 3-25).Philadelphia: W. B. Saunders Company. Cohen, L. T., Rickards, F. W. & Clark, G. M . (1991). A comparison of steady-state evoked potentials to modulated tones in awake and sleeping humans. Journal of the Acoustical Society of America 90: 2467-2479. Cone-Wesson, B. & Ramirez, G.M. (1997). Hearing sensitivity in newborns estimated from ABRs to bone-conducted sound. Journal of the American Academy of Audiology 8:299-307. Cone-Wesson, B, Rickards, F, Poulis, C, Parker, J, Tan, L. & Pollard, J. (2002). The auditory steady-state response: clinical observations and applications in infants and children. Journal of the American Academy of Audiology 13: 270-282. Dimitrijevic, A., John, M . S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., Nedzelski, J. M . & Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology 13: 205-224. Eby, T. L., & Nadol J. B. (1986). Postnatal growth of the human temporal bone. Annals of Otology, Rhinology and Laryngology 95: 356-64. Foxe, J. J. & Stapells, D. R. (1993). Normal infant and adult auditory brainstem responses to bone-conducted tones. Audiology 32: 95-109. 207 Freeman, S., Sichel, J-Y. & Sohmer, H. (2000). Bone conduction experiments in animals-evidence for a non-osseous mechanism. Hearing Research, 146, 72-80. Herdman, A., Lins, O., Van Roon, P., Stapells, D., Scherg, M . , & Picton, T. (2002). Intracerebral sources of human auditory steady-state responses. Brain Topography 15: 69-86. Jahrsdoerfer, R. A., Yeakley, J. W., Hall, J. W., Robbins, K. T. & Gray, L. C. (1985). High-resolution CT scanning and auditory brain stem response in congenital aural atresia: Patient selection and surgical correlation. Otolaryngology-Head and Neck Surgery 93: 292-298. John, M . S. & Picton, T. W. (2000a). MASTER: a Windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine 61: 125-150. John, M . S. & Picton, T. W. (2000b). Human auditory steady-state responses to amplitude-modulated tones: phase and latency measurements. Hearing Research 141 (1-2): 57-79. Joint Committee on Infant Hearing. (2000). Year 2000 position statement: Principles and guidelines for early hearing detection and intervention programs. American Journal of Audiology 9: 9-29. Lins, O. G., Picton, T. W., Boucher, B. L., Durieux-Smith, A., Champagne, S. C , Moran, L. M . , Perez-Abalo, M . C , Martin, V. & Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear and Hearing 17: 81-96. Moore, J. K., Ponton, C. W., Eggermont, J. J., Wu, J . - C , Huang, J. Q. (1996). Perinatal maturation of the auditory brain stem response: Changes in path length and conduction velocity. Ear and Hearing 17: 411-418. 208 Nousak, J. & Stapells, D.R. (1992). Frequency specificity of the auditory brainstem response to bone-conducted tones in infants and adults. Ear and Hearing 13: 87-95. Picton, T.W., John, M.S., Dimitrijevic, A. & Purcell, D. (2003). Auditory steady-state responses. International Journal of Audiology 42: 177-219. Ponton, C. W., Moore, J. K., Eggermont, J. J., Wu, B. J . - C , & Huang, J. Q. (1994). Relationship between auditory brainstem size and auditory brainstem response (ABR) latency in human development. Association for Research in Otolaryngology Abstracts, 57. Savio, G., Cardenas, J., Perez-Abalo, Gonzalez, A. & Valdes, J. (2001). The low and high frequency auditory steady state responses mature at different rates. Audiology Neuro-Otology 6: 279-287. Small, S. A. & Stapells, D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear and Hearing 25(6): 611-23. [Chapter 2] Small, S. A., Hatton, J. & Stapells, D.R. (2007). Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants. Ear & Hearing 28(1): 83-98. [Chapter 4] Small, S.A. & Stapells, D.R. (2005). Multiple auditory steady-state responses to bone-conduction stimuli in adults with normal hearing. Journal of the American Academy of Audiology 16(3): 172-183. [Chapter 3] Stapells, D. R. (2000). Frequency-specific evoked potential audiometry in infants. In R.C. Seewald (ed.) A Sound Foundation through Early Amplification: Proceedings of an International Conference (pp. 13-31). Chicago: Phonak A G . 209 Stapells, D. R., Herdman, A., Small, S.A., Dimitrijevic, A. & Hatton, J. (2005). Current status of the auditory steady-state responses for estimating an infant's audiogram. . In R.C. Seewald (ed.) A Sound Foundation through Early Amplification: Proceedings of an International Conference (pp. 43-59). Chicago: Phonak A G . Stapells, D. R. & Ruben, R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. Annals of Otology, Rhinology and Laryngology 98: 941-949. 210 CHAPTER 6 MATURATION OF BONE-CONDUCTION MULTIPLE AUDITORY STEADY-STATE RESPONSES A version of this chapter has been submitted for publication. Small, S. A., & Stapells, D. R. (submitted). Maturation of bone-conduction multiple auditory steady-state responses. InternationalJournal of Audiology. 211 INTRODUCTION Elevation of hearing thresholds to air-conduction stimuli in infants may result from a sensorineural, conductive or mixed hearing loss. It is necessary to obtain bone-conduction thresholds, both to distinguish among sensorineural, conductive and mixed hearing losses, and to determine the magnitude of the air-bone gap. This is routinely done in adults and should also be done when testing infants. Although bone-conduction testing is known to be essential, many clinicians continue to use only air-conduction stimuli when estimating thresholds in infants using the auditory brainstem response (ABR) or auditory steady-state responses (ASSRs). Despite the importance of bone-conduction testing, there is only a limited number of published studies in infanta that have recorded bone-conduction thresholds to frequency-specific stimuli using either ABR (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989) or ASSR (Small, Hatton & Stapells, 2007: Chapter 4; Small & Stapells, 2006: Chapter 5; Small & Stapells, submitted: Chapter 7) techniques. Collectively, the ABR studies investigated bone-conduction thresholds for 22 neonates at 500 and 4000 Hz (Cone-Wesson & Ramirez, 1997) and 36 six-month-old infants at 500 and 2000 Hz (Foxe & Stapells, 1993; Stapells & Ruben, 1989). Our earlier ASSR studies estimated bone-conduction thresholds in 29 pre-term infants (Small & Stapells, 2006: Chapter 5) and, across the three studies, a total of 35 infants 0-11 months of age (Small et al., 2007: Chapter 4; Small & Stapells, 2006: Chapter 5; Small & Stapells, submitted: Chapter 7) at 500, 1000, 2000 and 4000 Hz. There are no published frequency-specific bone-conduction threshold data (ABR or ASSR) for infants older than 11 months of age. To establish normal bone-conduction hearing levels for infants and to understand the time course of maturation for bone-conduction hearing, comparisons of bone-conduction hearing thresholds at the same frequencies for large groups of young and older infants and adults are needed. 212 From research to date, we know that infant-adult differences exist, for both ABRs and ASSRs to bone-conduction stimuli, and that these differences are age- and frequency-dependent. Stuart, Yang, Strenstrom and Reindorp (1993) reported that bone-conduction click-ABR (i.e., response elicited from a broad range of frequencies along the cochlear partition) thresholds (in dB HL) in neonates are 17.5 dB better than those obtained for adults, and suggested that the delivery of a bone-conducted signal is more effective (assumed across frequency) in neonates than in adults (Stuart, Yang & Strenstrom, 1990; Stuart et al., 1993; Yang, Rupert & Monshegian,1987). Cone-Wesson and Ramirez (1997) also reported better bone-conduction click-ABR thresholds in neonates than adults. In contrast, Cornacchia and Morra (1983) found no difference for bone-conduction click-ABR thresholds between older infants (16-20 months of age) and adults. For brief-tone stimuli, which are frequency specific, Foxe and Stapells (1993) found that bone-conduction ABR thresholds to brief tones for six-month-old infants were similar to adult thresholds at 500 Hz but 5.5 dB poorer compared to adults at 2000 Hz. They also found that these infants had bone-conduction ABR thresholds to 500-Hz brief tones that were significantly better (10.5 dB) compared to 2000 Hz. Cone-Wesson and Ramirez (1997) reported that bone-conduction ABR thresholds in neonates were 20 dB better at 500 Hz and 5 dB better at 4000 Hz compared to adults (n=3). They also showed that bone-conduction ABR thresholds for neonates were 14 dB better at 500 Hz compared to 4000 Hz [it is also noteworthy that their 500-Hz thresholds for the neonates were 25 dB better than those reported by Foxe and Stapells (1993)]. We found similar results for bone-conduction ASSRs in our previous studies for pre-term and 0-8-month-old post-term infants; bone-conduction ASSR thresholds were significantly better at 500 and 1000 Hz compared to 2000 Hz (Small et al., 2007: Chapter 4; Small & Stapells, 2006: Chapter 5). Foxe and Stapells (1993) found no significant difference in adult 500- and 213 2000-Hz ABR thresholds in adults (in dB nHL); similarly, bone-conduction ASSR thresholds (in dB HL) in adults show little or no difference across frequency (Small & Stapells, 2005: Chapter 3). There are no ABR or ASSR studies that have obtained frequency-specific bone-conduction thresholds in infants older than 11 months of age in order to investigate when bone-conduction thresholds become adult-like. The pattern of amplitudes of bone-conduction ASSRs in young infants as a function of frequency are similar to previously reported bone-conduction ABR amplitudes in infants. Low-frequency bone-conduction ABR amplitudes are larger than those to high frequencies (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989); low-frequency bone-conduction ASSR amplitudes are also larger than those to high-frequency stimuli for pre-and young post-term infants (Small & Stapells, 2006: Chapter 5; Small et al., 2007: Chapter 4). Direct comparison of ABR and ASSR amplitude differences across age groups cannot be made because of the additional contribution of the 40-Hz response to the adult ABR (when using a 40/second rate) that is not present in the infant ABR (Foxe & Stapells, 1993). However, comparisons between infants and adults can be made for 80-Hz ASSR amplitudes; adults have larger bone-conduction ASSRs than infants at 40 dB HL for 2000 and 4000 Hz but show little difference compared to infants at 500 and 1000 Hz (Small & Stapells, 2006: Chapter 5). Similar to Foxe and Stapells (1993) and Stapells and Ruben (1989), who found longer bone-conduction wave V latency differences for 500 compared to 2000 Hz for both infants and adults, phase delays were also longer for bone-conduction ASSRs across frequency in post-term infants and adults. Cone-Wesson and Ramirez (1997) also reported longer wave V latencies for 500 compared to 4000 Hz for neonates. For the post-term infants in Small & Stapells (2006) (Chapter 5) and for adults, mean bone-conduction ASSR phase delays (in ms) at 40 dB HL were 214 also longer for 500 and 1000 Hz compared to 2000 and 4000 Hz, although the differences were much smaller for the ASSRs compared to the ABRs. Similar results were found for pre-term infants, except mean phase delay at 500 Hz was shorter compared to 1000 and 2000 Hz (Small & Stapells, 2006: Chapter 5). Bone-conduction ABR results show much shorter wave V latencies for 500-Hz brief tones in infants compared to adults (1.02 ms shorter) (Foxe & Stapells, 1993; Nousak & Stapells, 1992) and slightly longer wave V latencies for 2000-Hz brief tones in infants compared to adults (Foxe & Stapells, 1993). In contrast, mean ASSR phase delays (ms) for 500 Hz are much longer for infants compared to adults. Mean ASSR phase delays (in ms) for 2000 Hz, similar to bone-conduction ABRs, are longer for infants than adults. The purpose of this study was to compare multiple ASSR thresholds to bone-conduction stimuli in infants of different ages to adults - all with normal hearing - to investigate the time course of maturation of bone-conduction thresholds across frequency. It was also the purpose of this study to establish "normal levels" for bone-conduction ASSRs for young and older infants. To this end, new data and previously published bone-conduction ASSR data for three different age groups were combined and compared in this study. Bone-conduction ASSRs were investigated for a group of "older" infants, a group not previously assessed. A larger group of "young" infants was created for this study by pooling data from the smaller groups of young infants presented in previously published studies (Small et al., 2007: Chapter 4; Small & Stapells, 2006: Chapter 5, Small & Stapells, submitted: Chapter 7). Finally, a larger group of adult subjects was formed in order to compare infant and adult bone-conduction ASSRs by pooling new bone-conduction ASSR threshold data collected in this study with previously published adult data (Small & Stapells, 2005: Chapter 3). 215 M A T E R I A L S AND M E T H O D S Participants Two groups of infants who passed a hearing screening in both ears and a group of adults with normal hearing (pure-tone air- and bone-conduction thresholds < 25 dB HL at 500-4000 Hz) participated. ASSRs to bone-conduction stimuli were recorded in 35 "young" infants [age range of 0.5-44.0 weeks; mean age of 16.0 weeks], 13 "older" infants [age range of 12-24 months; mean age of 18.2 months], and 18 adults [age range of 19-48 years; mean age of 22.9 years] recruited from the community. Data for the 35 young infants were previously reported as small groups in three separate studies (9 infants from Small et al., 2007: Chapter 4; 14 infants from Small & Stapells, 2006: Chapter 5; 12 infants from Small & Stapells, submitted: Chapter 7); these small groups were combined to obtain one large group of young infants. Data for 10 of the 18 adults presented in the present study were also previously reported (Small & Stapells, 2005: Chapter 3) and combined with new data from eight additional adults to form a larger group of adults. Bone-conduction ASSRs in older infants were recorded for the first time in this study. Four of the infants recruited from the community were screened using an automatic auditory brainstem response (AABR) screening test at 35 dB nHL. The hearing of the other infants was screened using a distortion-product otoacoustic emissions (DPOAE) screening test. The pass criterion for the DPOAE screening was a signal-to-noise ratio > 5 dB at 2000, 3000 and 4000 Hz in both ears. Infants who passed the AABR or DPOAE hearing screening test in both ears were considered to be at low risk for significant hearing loss and thus included in the study. 216 Stimuli The stimuli were sinusoidal bone-conduction tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude and 25% frequency modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. All stimuli were presented simultaneously (i.e., multiple). The stimuli were generated by the Rotman MASTER research system (John & Picton, 2000a) using a buffer length of 25,600 points and a digital-to-analog (D/A) rate of 31,250 Hz, which is an integer sub-multiple of the 20 MHz clock rate, but not an integer multiple of the carrier frequencies. The stimuli were then attenuated through Tucker-Davis Technologies HB6 and SM3 modules. Before the stimuli were attenuated, they were routed through the Stanford Research Systems Model SR650 to increase the gain of the stimulus by 10 dB. The bone-conduction stimuli were presented to a Radioear B-71 bone oscillator which was coupled to the head by an elastic headband for the infants and adults with 400-450 g of force, except for four of the young infants who had the bone oscillator held in place by hand. The bone oscillator was placed on the temporal bone slightly posterior to the upper part of the pinna for all infants and adults tested [for a detailed description of placement location, see Chapter 4 (Small et al., 2007)]. Bone-conduction stimuli were presented using 10-dB steps from 50 to -10 dB HL [intensities greater than 50 dB HL for multiple stimuli result in non-linearities in the oscillator output (Small & Stapells, 2004: Chapter 2)]. ASSRs for all infants and adults were elicited by a non-inverted stimulus polarity (Small & Stapells, 2004: Chapter 2). Calibration The bone-conduction stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re:luN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The 217 oscillator was coupled to the artificial mastoid with 550 g of force. ASSR Recordings ASSRs were recorded using the Rotman MASTER system. Three electrodes were used to record the electrophysiologic responses for 23 of the 35 young infants, and for all older infants and adults: the non-inverting electrode was placed midline at the high forehead, the inverting electrode was positioned midline at the nape of the neck, just below the hairline, and an electrode placed at the low forehead acted as ground. For 12 of the 35 young infants, four electrodes were used to record the electrophysiologic responses: two inverting electrodes were placed low on the left and right mastoids instead of one inverting electrode positioned at the midline at the nape of the neck1. All inter-electrode impedances were below 3 kOhms at 10 Hz. The electroencephalogram (EEG) was filtered using a 30-250 Hz filter (12 dB/oct) and amplified 80,000 times (8000X in Nicolet HGA-200A and Nic501A; 10X in NIDAQ card). The E E G was further filtered using a 300-Hz lowpass anti-aliasing filter [Wavetek Rockland Model 852 (48 dB/oct)]. The E E G was then processed using a 1250-Hz A/D conversion rate (Small & Stapells, 2004: Chapter 2). Each E E G recording sweep was made up of 16 epochs of 1024 data points (0.819 seconds per epoch) and lasted a total of 13.107 seconds. Artifact rejection was set to eliminate epochs of electrophysiologic activity that exceeded ± 40 U.V in amplitude in order to reduce contributions to the E E G due to muscle artifact. ASSRs were averaged in the time domain and then analysed online in the frequency domain using a Fast Fourier Transform (FFT). Weighted averaging (John, Dimitrijevic & Picton, 2001) was used. The FFT resolution was 0.076 Hz over a range of 0 to 625 Hz. 1 Results of an A N O V A comparing ASSR thresholds, amplitudes and phase delays for the high-forehead-to-nape and high-forehead-to-ipsilateral-mastoid montages revealed no significant differences. 218 Amplitudes were measured baseline-to-peak and expressed in nV. An F-ratio was calculated by MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ± 60 bins of the modulation frequency ("noise") (John & Picton, 2000a). A minimum of seven sweeps were recorded for each test condition. A response was considered to be present if the F-ratio, compared to the critical values for F(2, 240), was significant at a level of p < .05 for at least two consecutive sweeps. A response was considered to be absent if p > .05 and the mean amplitude of the noise was less than 11 nV. Alternatively, a response was also considered to be absent when response amplitude was < 10 nV and thep value > .30. Procedure Testing was conducted in a double-walled sound-attenuated booth. The ambient noise levels in the sound-attenuated booth for one-octave-wide bands centred at 0.5, 1, 2 and 4 kHz were 12, 10, 10 and 12 dB SPL, respectively. All participants were tested with ears unoccluded. Infant participants were only tested when quiet and asleep; adults were tested when relaxed or asleep while reclined in a comfortable chair. Hearing screening was performed in both ears for the infants at the beginning of the test session to establish that the participants were unlikely to have hearing loss. Behavioural pure-tone air-and bone-conduction thresholds were obtained, for adults at the beginning of the test session to ensure that they had normal hearing. Multiple ASSRs were elicited to bone-conduction stimuli beginning at a randomized starting intensity. Threshold for each carrier frequency was determined using a bracketing technique adjusting the presentation level using 10-dB steps. The lowest intensity at which a response was present was considered "threshold". In some cases, a response did not reach significance at a level above 219 "threshold"; if only one response did not reach significance, the lowest level at which the response was present was considered threshold. If two or more responses were not significant, the response at the lowest level was considered a "false positive" and the lowest level above the absent responses was deemed "threshold". The total recording time was approximately 1.0-1.5 hours, including the time to obtain screening test results. Procedures for this study were approved by the University of British Columbia Behavioural Research Ethics Board. The participants (or parents) signed a consent form before commencing any of the experiments; all participants were paid an honorarium at the end of each session. Data Analyses Mean amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant. The phase values from MASTER were adjusted by adding 90° to yield the onset phase (John & Picton, 2000b). Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Any phase-delay values that differed > 180° from an adjacent measure were "unwrapped" by adding 360° to their value (John & Picton, 2000b). Phase values for ASSRs that were not significant were excluded from mean phase-delay calculations. Phase-delay values were averaged across subjects. Results were only reported if at least five subjects contributed to the mean. Bone-conduction ASSR thresholds were compared across frequency and age groups. Bone-conduction ASSR amplitudes and phase delays (ms) at 30 and 40 dB HL were also compared across frequency for the two infant groups and adults. Comparisons across age groups were made using a 2-way mixed-model analysis of variance (ANOVA). Huynh-Feldt epsilon adjustments for repeated measures were made when appropriate. Newman-Keuls post-hoc comparisons were performed for significant main effects and interactions. The criterion for 220 statistical significance was p<.05 for all analyses. RESULTS Representative ASSR results to bone-conduction stimuli are shown for typical infant (young and older) and adult participants in Figure 6.1. For the 4.5-month-old infant, ASSR thresholds for 500 Hz were 10 dB better compared to the older infant and 20 dB better compared to the adult. For 1000-Hz stimuli, both the young and older infant had thresholds that were 10 dB better than the adult. At 2000 Hz, this young infant's ASSR threshold was 20 dB poorer than both the older infant and the adult. At 4000 Hz, the young infant's threshold was 10 dB poorer than the older infant and 20 dB poorer than the adult. Overall, the 4.5-month-old infant's low-frequency thresholds (500 & 1000 Hz) were 25 dB better than the high-frequency thresholds (2000 & 4000 Hz), a pattern which is not seen for the 24-month-old infant or adult whose ASSR thresholds differ by only 5-10 dB between the low and high frequencies. Figure 6.1 Detectability Figure 6.2 shows the percent occurrence of ASSRs for the groups of young and older infants and adults for each of the carrier frequencies. There are many more ASSRs present at 500 and 1000 Hz for the infants, particularly the younger infants. At 2000 Hz, the adults have a few more responses present than the infants, whereas, at 4000 Hz, there is no difference in detectability for the different groups of infants or adults. 221 Figure 6.2 Threshold The effects of age on ASSR threshold for each carrier frequency are shown in Figure 6.3. For low frequencies, ASSR thresholds increase (worsen) substantially with age, whereas those for high frequencies are unaffected by age except for a slight improvement for 2000 Hz. For the young infants, mean ASSR thresholds at 500 and 1000 Hz were 8 dB better compared to the older infants and 17-19 dB better compared to the adults. In contrast, for young infants, mean ASSR thresholds at 2000 Hz were not significantly different compared to the older infants but were 6 dB poorer compared to adults. For older infants, ASSR mean thresholds were 8-11 dB better at 500 and 1000 Hz compared to adults, whereas, at 2000 Hz, thresholds tended to be slightly poorer compared to adults. Mean ASSR thresholds at 4000 Hz differed by less than 3 dB across age groups. Figure 6.3 Results of an A N O V A comparing mean ASSR thresholds at 500, 1000, 2000 and 4000 Hz across age (young infants, older infants and adults) revealed a significant effect of age [F(2,63) = 10.057,/? =.0002], frequency [F (3,189) = 15.201, /?< .0001] and interaction between age and frequency [F (6,189) = 8.3089,/? < .0001]. Post hoc comparisons revealed that the significant effect of age was explained by overall poorer mean ASSR thresholds in adults compared to young infants (/? = .0007). Post hoc comparisons indicated that the significant 222 interaction between age and frequency was explained, in part, by better ASSR thresholds for young infants compared to adults at 500 and 1000 Hz, and better ASSR thresholds for older infants compared to adults at 1000 Hz. The differences in ASSR thresholds at 500 Hz between the young and older infants approached significance (p = .057); there were no significant differences in ASSR thresholds at 2000 and 4000 Hz for any of the three age groups. Figure 6.3 also shows that there are differences in ASSR thresholds between frequencies within each age group. Post hoc comparisons for young infants revealed that ASSR thresholds at 1000 Hz were significantly better compared to other frequencies and ASSR thresholds at 2000 Hz were poorer compared to all other frequencies; ASSR thresholds at 500 vs. 4000 Hz were not significantly different (p = .931). For older infants, ASSR thresholds at 1000 Hz were significantly better compared to ASSR thresholds at 2000 Hz, and the difference in ASSR thresholds between 500 and 1000 Hz approached significance (p = .057); there were no significant differences for 500 vs. 2000 Hz (p = .647), 500 vs. 4000 Hz (p = .075), and 1000 vs. 4000 Hz (p = 1.00). For adults, ASSR thresholds were significantly worse at 500 Hz compared to 2000 and 4000 Hz; there were no differences in ASSR threshold for 500 vs. 1000 Hz (p = .179), 1000 vs 2000 Hz (p = .465), or 1000 Hz vs. 4000 Hz (p = .085). Figure 6.4 shows the relationship between age and threshold of individual subjects at each of the carrier frequencies. ASSR thresholds decrease (i.e., get poorer) significantly with age at 500 Hz (r = A6,p < .0001) and 1000 Hz (r = .60,p < .0001). There was a slight but significant improvement in ASSR threshold with age at 2000 Hz (r = .28, p = .025). ASSR threshold did not change with age at 4000 Hz (r = .07, p = .584). 223 Figure 6.4 Amplitude ASSR mean amplitudes for each age group and carrier frequency are shown in Figure 6.5. At 500 Hz, there was no difference in ASSR mean amplitudes for the young and older infants and adults. At 1000 and 2000 Hz, there was no difference in mean amplitudes between the two infant groups; however, at 1000 Hz, both infant groups had larger response amplitudes than adults, whereas, at 2000 Hz, they had smaller response amplitudes than adults. At 4000 Hz, young infants tended to have smaller ASSR amplitudes compared to both older infants and adults, whereas, there was no difference in response amplitude between older infants and adults. Figure 6.5 As shown in the top of Table 6.1, results of an A N O V A comparing ASSR mean amplitudes for each age group and carrier frequency at 30 and 40 dB HL indicated significantly larger amplitudes at 40 dB HL compared to 30 dB HL, a significant main effect of frequency, and significant age x frequency and age x intensity interactions. The age effect, intensity x frequency interaction and the intensity x frequency x age interaction were not significant. Post hoc comparisons indicated that the significant effect of frequency, pooled for age, was due to significantly larger amplitudes at 1000 Hz compared to 500 Hz (p = .001), 2000 Hz (p < .0001) and 4000 Hz (p =.001), and significantly smaller amplitudes at 2000 compared to 4000 Hz (p = .020); response amplitudes also tended to be larger at 500 compared to 2000 Hz but the 224 difference did not quite reach significance (p =.053). Post hoc comparisons revealed that the significant age x frequency interaction was explained by larger amplitudes at 1000 Hz for younger (p = .051) and older infants (p=.036) compared to adults, smaller amplitudes at 2000 Hz for young (p = .0006) and older infants (p = .006) compared to adults, and smaller amplitudes at 4000 Hz for younger infants compared to adults (p = .045). Response amplitudes at 4000 Hz for young infants also tended to be smaller compared to older infants but the difference did not reach significance (p = .066). The significant interaction between age and intensity was explained by significantly larger ASSR amplitudes at 30 dB HL for adults compared to young (p = .0004) and older (p = .045) infants, and larger amplitudes for older infants compared to young infants (p = .030). Table 6.1 Phase delay As shown in Figure 6.6, there are age- and frequency-dependent differences in ASSR mean phase delays for young and older infants and adults. At 500 Hz, ASSR mean phase delays tended to be longer for both groups of infants compared to adults. At 2000 Hz, ASSR mean phase delays were similar across age groups up to 30 dB HL; however, at 40 dB HL, ASSR mean phase delays were longer for both infant groups compared to adults. At 1000 and 4000 Hz, ASSR mean phase delay was similar across age groups. With the exception of shorter phase delays at 500 Hz compared to 1000 Hz, the trend was for phase delay to become shorter as frequency increased from 1000 to 4000 Hz for each of the age groups, a pattern that has been reported in our previous studies (Adults: Small & Stapells, 2005: Chapter 3, Infants: Small & 225 Stapells, 2006: Chapter 5). Figure 6.6 The bottom of Table 6.1 shows the results of an A N O V A comparing ASSR phase delays for each age group and each carrier frequency presented at 30 and 40 dB HL. The statistical results revealed significantly longer ASSR phase delays at 30 compared to 40 dB HL (p<.0001) and a significant effect of frequency; the main effect of age was not significant. Post hoc comparisons revealed that the significant effect of frequency was due to shorter phase delays for 4000 Hz compared to lower frequencies (p = .0001 for each comparison), and shorter phase delays for 500 compared to 1000 Hz (p =.0001) and 2000 compared to 1000 Hz (p = .0001); no significant differences in phase delay were found between 500 and 2000 Hz (p = .285). All of the interactions between factors were significant. The significant age x intensity interaction was explained by shorter phase delays at 40 dB HL compared to 30 dB H L for older infants (p = .004) and adults (p = .0001), longer phase delays for 40 dB HL for young (p = .0002) and older (p = .0002) infants compared to adults, and longer phase delays for 30 dB H L for older infants compared to adults (p = .007). Post hoc comparisons for the significant age x frequency interaction revealed significantly longer ASSR phase delays at 500 Hz for older infants compared to adults (p = .019); phase delays for younger infants compared to adults were nearly significantly longer (p = .060); whereas, no significant differences in ASSR phase delay were found at 500 Hz between infant groups (p = .531). For 2000 Hz, ASSR phase delays were significantly longer for younger (p = .007) and older (p = .002) infants compared to adults; no significant differences in ASSR phase delay were found between infant groups (p = .596). For 226 1000 and 4000 Hz, no significant differences in ASSR phase delay were found for any of the age groups. Post hoc comparisons for the significant intensity x frequency interaction revealed that phase delays at 40 dB HL were significantly shorter for 500 Hz compared to 1000 Hz (p = .0001) and 2000 Hz (p = .003); whereas phase delays were significantly longer for 500 compared to 4000 Hz (p = .0005), 1000 Hz compared to 2000 (p = .0002) and 4000 Hz (p = .0001), and 2000 compared to 4000 Hz (p = .0001). The same frequency-dependent differences in ASSR phase delay were seen at 30 dB HL except that the phase delays for 500 and 2000 Hz were not significantly different (p = .895). Post hoc comparisons revealed that the significant age x intensity x frequency interaction was explained, in part, by longer ASSR phase delays at 30 compared to 40 dB HL for older infants at 500 Hz (p = .001) and adults at 1000 Hz (p = .0001). Comparisons of ASSR phase delays at 40 dB H L also revealed longer phase delays for young infants compared to adults at 500 (p = .021) and 1000 Hz (p=.047), and longer phase delays for young (p=.001) and older infants (p=.0001) compared to adults at 2000 Hz. At 30 dB HL, the phase delays tended to be longer for young (p = .064) and older infants (p = .071) compared to adults at 2000 Hz but did not reach significance. DISCUSSION This is the first study to compare bone-conduction ASSRs in young and older infants to bone-conduction ASSRs in adults. The results of this study clearly show that there are frequency-dependent maturational changes in bone-conduction ASSRs and that infant-adult 227 differences remain until at least two years of age. With maturation, bone-conduction ASSR thresholds increase (i.e., get worse) significantly at 500 and 1000 Hz, improve slightly at 2000 Hz but do not change significantly at 4000 Hz. Also, there are frequency-dependent differences in bone-conduction ASSR thresholds that differ with maturation. Young infants tend to have better bone-conduction ASSR thresholds in the low frequencies compared to the high frequencies (with the exception of 500 vs. 4000 Hz). Older infants also have significantly better thresholds for 1000 compared to 2000 Hz but do not have significant differences in bone-conduction ASSR thresholds at 500, 1000 and 4000 Hz. For adults, there was no difference in bone-conduction ASSR thresholds at 1000, 2000 and 4000 Hz; the only differences seen in bone-conduction ASSR thresholds were poorer thresholds at 500 Hz compared to 2000 and 4000 Hz, which is opposite to the trend observed for low frequencies in infants [but similar to air-conduction multiple-ASSR results (for review see Picton et al., 2003)]. In our previous study that compared bone-conduction ASSR thresholds across carrier frequency for small groups of young infants and adults (Small & Stapells, 2006: Chapter 5), we reported that young infants had better mean ASSR thresholds in the low frequencies and slightly poorer thresholds in the high frequencies compared to adults (we also found this pattern for pre-term infants); the results reported in the present study for a larger group of young infants are consistent with these earlier findings, with the exception of a few differences in statistical results at individual carrier frequencies. For the large group of young infants in this study (n= 35), ASSR thresholds were significantly poorer at 500 Hz compared to 1000 Hz, whereas in our earlier study (Small & Stapells, 2006: Chapter 5), for a smaller group of infants (n= 14), the difference in ASSR thresholds for 500 vs. 1000 Hz was not statistically significant. Similarly, bone-conduction ASSR thresholds in young infants were significantly poorer at 2000 compared 228 to 4000 Hz in the present study, but were not significantly different for the smaller group of young infants (Small & Stapells, 2006: Chapter 5). We also found differences for the larger group of adults in this study (n= 18) compared to the smaller group of adults (n= 10) in Small & Stapells (2005) (Chapter 3); in the earlier study, there were no differences in ASSR threshold across frequency, yet we found significantly poorer thresholds at 500 Hz compared to 2000 and 4000 Hz in the present study. The results for the larger groups of young infants and adults probably more accurately represent bone-conduction ASSR thresholds for the population that these groups represent than the results reported previously for the smaller groups. There are also frequency-dependent differences in bone-conduction ASSR amplitudes within and among age groups. Compared to adults, infants had larger response amplitudes at 1000 Hz, smaller response amplitudes at 2000 Hz, and no significant difference in amplitude at 500 Hz. At 4000 Hz, only the young infants had significantly smaller response amplitudes compared to adults. There were no significant differences in ASSR mean amplitudes between infant groups across frequency (with the exception of larger amplitudes at 4000 Hz for older infants which nearly reached significance). Within the group of young infants, ASSR amplitudes were significantly larger at 1000 Hz compared to the other frequencies. For older infants, ASSR amplitudes were also larger at 1000 Hz compared to 500 and 2000 Hz but not compared to 4000 Hz. ASSR phase delays also showed frequency-dependent infant-adult differences. There were no significant differences in phase delay among age groups at 1000 and 4000 Hz; however, there were differences in phase delay at 500 and 2000 Hz for both infant groups. Both infant groups had longer phase delays at 500 and 2000 Hz (40 dB HL only) compared to adults. As discussed in Small & Stapells (2006) (Chapter 5), it is not clear why adults have shorter phase 229 delays compared to infants at 500 Hz, when bone-conduction ABR results show shorter wave V latencies at 500 Hz for infants compared to adults. The longer ASSR phase delays for infants at 2000 Hz are consistent with the longer ABR wave V latencies at 2000 Hz for infants compared to adults (Foxe & Stapells, 1993, Nousak & Stapells, 1992). Within each age group, ASSR phase delays became shorter as frequency increased from 1000 to 4000 Hz; however, phase delays for 500 Hz at 40 dB HL were shorter compared to 1000 Hz and 2000 Hz (40 dB H L only). John & Picton (2000b) suggested adding an extra cycle to the phase delay values for 500 Hz to obtain a phase value that is approximately equivalent to group delay or apparent latency. Addition of one cycle (equivalent to 11.46 ms) to the infant and adult bone-conduction ASSR at 500 Hz would result in a longer mean phase delay at 500 Hz compared to higher frequencies, which is more sensible than the shorter mean phase delay seen at 500 Hz for infants and adults; however, this adjustment does not help explain why phase delays at 500 Hz were shorter for adults compared to infants. These differences in ASSR phase delay and ABR latency trends may be due to difficulty with the circular nature of phase measures (John & Picton, 2000b). Our ASSR results are consistent with the idea that the sensitivity of infants to bone-conducted signals differs from adults. Theories to explain the differences in bone-conduction hearing sensitivity between infants and adults were originally based on bone-conducted click-ABR data. Yang et al. (1987) reported longer ABR wave V latencies for bone-conducted vs. air-conducted clicks for one-year-old infants and adults and agreed with Weber's suggestion that the longer latencies were due to the bone-conducted clicks' lower frequency spectral content compared to an air-conducted click (Weber, 1983). Weber hypothesized that lower frequencies would result in a longer travelling time to reach the apical regions of the cochlea, thus explaining why ABR wave V latencies are longer for bone-conducted clicks compared to air-conducted 230 clicks. However, for neonates, Yang et al. (1987) found that ABR wave V latencies for bone-conducted clicks were shorter than for air-conducted clicks. They proposed that bone-conducted clicks were more effective on the much smaller, more isolated temporal bone of the neonate to explain why their neonate results did not fit the pattern for older infants and adults. The bone-conducted click ABR latency results are complex and difficult to interpret for the following reasons: (i) latency measures are not directly related to threshold, (ii) frequency-dependent changes in latency are obscured due to the broad-band nature of a click stimulus, and (iii) wave V latency decreases with maturation (Ponton, Eggermont, Coupland, & Winkelaar, 1992, 1993; Ponton, Moore, & Eggermont, 1996; Ponton, Moore, Eggermont, Wu, & Huang, 1994) which may add'to or cancel out changes in bone-conduction hearing sensitivity for infants of different ages. It has been shown subsequently, using noise masking, that responses elicited by air- and bone-conducted clicks receive the same contributions from different regions across the cochlear partition (i.e., 400-8000 Hz) (Durrant & Hyre, 1993; Kramer, 1992) which proves that Weber's hypothesis is incorrect. These issues with click stimuli and latency measures highlight the need for estimating threshold directly using frequency-specific stimuli. Air- and bone-conducted brief-tones are frequency specific which has been shown definitively using masking techniques similar to those described above (e.g., Kramer, 1992; Nousak & Stapells, 1992), and are better suited to investigating the maturation of hearing, which is frequency dependent, than click stimuli. [Note: The misconception that low-frequency brief tones are not frequency specific is still held by some despite the evidence to the contrary (for review of this issue see, Stapells & Oates, 1997)]. Other theories have been proposed to explain infant-adult differences in bone-conduction hearing. Cone-Wesson and Ramirez (2001) reported that bone-conduction stimuli transmit 5-21 231 dB SPL more acoustic energy for 500-Hz stimuli to the external ear canal (i.e., the osseotympanic pathway for a bone-conducted signal) when presented to an infant's head than when presented to an adult's head. They hypothesized that bone-conduction hearing is enhanced via this pathway by 5-21 dB at this frequency and that better bone-conduction thresholds can be explained by the overall level difference of the stimulus presented to the infant ear canal. This hypothesis has not been investigated further; however, it is possible to test this theory using bone-conduction ABR threshold data for infants with conductive hearing losses. If an infant has a conductive loss, any low-frequency bone-conducted energy that is transmitted via air conduction will be blocked by the abnormal middle ear, and cannot enhance the effectiveness of low-frequency bone-conducted sound at the cochlea. Stapells and Ruben (1989) reported better bone-conduction thresholds at 500 Hz in infants compared to adults both for infants with normal hearing and for those with conductive losses; therefore, Cone-Wesson and Ramirez's hypothesis must be incorrect. Additionally, our previous research showed that there is no difference in infant bone-conduction ASSR thresholds whether the ear canal was unoccluded or occluded (Small et al., 2007; Chapter 4). The underlying mechanism for the occlusion effect in adults is the osseotympanic pathway of bone-conducted sound (Tondorff, 1966), consequently, the absence of an occlusion effect for young infants further refutes Cone-Wesson and Ramirez's hypothesis that greater low-frequency acoustic energy in the infant ear canal contributes to a more effective bone-conducted stimulus at the cochlea compared to adults. In force levels (dB re: 1 U.N), we know that adults are more sensitive to high- than to low-frequency stimuli transmitted by a bone oscillator, which is reflected in the RETFLs in dB re: 1 U . N that correspond to 0 dB H L at the mastoid. In order to make the spectra of a bone oscillator equal across frequency (i.e., converted to dB HL), force levels need to be 38.0, 22.5, 11.0 and 232 15.5 re: l | i N at 500, 1000, 2000 and 4000 Hz, respectively (ANSI S3.6-1996). Our results show that infants do not require the same boost in the low frequencies as adults to make bone-conducted ASSR stimuli detectable, but require approximately the same force levels as adults to detect high frequencies. The hypothesis that a bone-conducted click is a "lower-frequency stimulus" for infants is thus likely correct, but not for the reasons suggested by Weber (1983) and Yang et al. (1987). In adults, the bone-conducted click does not have greater low-frequency effective spectral content compared to an air-conducted click. Rather, the low-frequency component of the click is more effective via bone conduction on an infant skull than via air-conduction. What are the physiologic mechanisms that contribute to the infant-adult differences in bone-conduction ASSRs? It is likely that the infant-adult differences in bone-conduction ASSR (and ABR) thresholds and response characteristics relate, in part, to the size and structural differences between the infant and adult skull. The infant skull is much smaller than that of an adult. Eby and Nadol (1986) measured the dimensions of the mastoid bone in human infants and found that the width, length and depth increase rapidly in the first two years of life. Also, flexible sutures connect the temporal bone to the other bones of the cranium in the infant skull until bony sutures develop at approximately one year of age (Anson & Donaldson, 1981), in contrast to the adult skull which is a rigid structure with fused bones. Based on better click-ABR thresholds to bone- vs. air-conducted stimuli in neonates, Stuart et al. (1990) suggested that the flexible sutures in the infant skull result in less energy dissipating to the rest of the skull, causing the temporal bone to oscillate more in isolation, thus resulting in a more effective bone-conducted stimulus across frequencies in infants. Foxe and Stapells (1993) suggested that the smaller mass of the temporal bone in infants results in a more intense signal activating the 233 cochlea; they estimated that the bone-conducted stimulus was 9-17 dB more effective at 500 Hz for infants compared to adults and 13 dB better at 2000 Hz (after latency adjustment based on expected infant-adult ABR latency differences for air-conducted stimuli). The idea of a more isolated temporal bone (Stuart et al., 1990) is supported by Sohmer and colleagues who found that the acceleration of vibratory energy across the infant fontanelle was 14 dB less than across bone adjacent to the fontanelle for a click stimulus (Sohmer, Freeman, Geal-Dor, Adelman & Savion, 2000). The presence of the fontanelle appears to contribute to significant interaural attenuation of a bone-conducted signal and, as a result, to reduced dissipation of the bone-conducted signal compared to an adult skull (Stuart et al., 1990). Recently, we estimated at least 13, 13, 15 and 14 dB of interaural attenuation (range of at least 0-30 dB) for a bone-conducted stimulus in young infants at 500, 1000, 2000 and 4000 Hz, respectively (Small et al., submitted: Chapter 7), which is similar to the difference in acceleration of vibratory energy across the temporal bone vs. the fontanelle (Sohmer et al., 2000), and to interaural attenuation estimated by Yang et al. (1987). Our estimation of interaural attenuation of bone-conducted signals presented at the mastoid also suggests that the amount of bone-conducted signal that is transferred across the infant skull does not vary with frequency. The infant-adult differences in skull size and structure, and the frequency-independent interaural attenuation of bone-conducted stimuli in infants may support the idea of a smaller, more isolated temporal bone that is more effective at transmitting bone-conducted stimuli, but these findings fail to explain why low-frequency bone-conducted stimuli are detectable at much lower intensities compared to high frequencies in infants. The frequency-dependent pattern observed for bone-conduction ASSR and ABR thresholds differs from that observed for air-conduction stimuli. Air-conduction ASSR 234 thresholds at 500 Hz are poorer than those at high frequencies (reviewed in: Cone-Wesson, Rickards, Poulis, Parker, Tan & Pollard, 2002; Picton et al , 2003; Stapells et al., 2005) which is opposite to the trend we report for infant bone-conduction ASSRs. ASSR threshold and amplitude changes with maturation also differ substantially for air- and bone-conduction stimuli. With maturation, air-conduction ASSR thresholds improve and amplitudes become larger at all frequencies, with larger changes seen for the higher frequencies (Ranee & Tomlin, 2006; Savio, Cardenas, Perez-Abalo, Gonzalez & Valdes, 2001). Sininger, Abdala & Cone-Wesson (1997) found that brief-tone air-conduction ABR thresholds (in dB SPL in the ear canal) were 3 and 24 dB poorer in infants compared to adults at 500 and 4000 Hz, respectively (i.e., infants required levels that were slightly more intense at 500 Hz and much more intense at 4000 Hz compared to adults). Ranee and Tomlin (2006) also found differences when they compared air-conduction ASSR thresholds (in dB SPL in the ear canal) at 500 and 4000 Hz in neonates and adults; ASSR thresholds in neonates were 28 dB and 38 dB poorer compared to adults at 500 and 4000 Hz, respectively. Ranee and Tomlin (2006) and Sininger et al. (1997) concluded that their infant-adult threshold differences in the neonatal period are the result of neural (auditory brainstem) development. This explanation may be feasible given the ABR work done by Ponton and colleagues that showed that synaptic transmission time continues to shorten until approximately three years of age (Ponton, Eggermont, Coupland, & Winkelaar, 1992, 1993; Ponton, Moore, & Eggermont, 1996; Ponton, Moore, Eggermont, Wu, & Huang, 1994). It is not necessarily the case that longer synaptic transmission times result in higher ABR thresholds; however, it is possible that less efficient neural transmission may require a higher stimulus intensity to elicit a response and result in higher ABR thresholds. 235-If the conclusions drawn from the air-conduction ABR and ASSR data are correct (i.e., hearing thresholds improve across frequency with maturation, and perhaps more so at high frequencies, due to neural maturation rather than differences in acoustic energy that reach the cochlea), to explain our ASSR results, a bone-conducted stimulus must be more intense on an infant skull compared to an adult skull both at low and high frequencies. If immaturities in neural processes do account for 3 and 24 dB of the infant-adult ABR threshold differences at 500 and 4000 Hz, respectively (Sininger et al., 1997), then the transmission of a bone-conducted stimulus must be 20 dB (3+17 dB) better at 500 Hz and 22 dB (24 - 2 dB) better at 4000 Hz. If, as suggested by Ranee and colleagues, maturation of neural pathways account for at least 28 and 38 dB of the infant-adult difference in air-conduction ASSR thresholds, then the transmission of bone-conducted stimuli could be more effective by as much as 45 dB (28+17 dB) at 500 Hz and 36 dB (38 - 2 dB) at 4000 Hz. Our results, using this explanation, would thus indicate that the transmission of bone-conducted energy by the infant skull is boosted in intensity both at low and high frequencies, but especially at low frequencies. The infant-adult differences in neural maturation hypothesized from the air-conduction brief-tone ABR and ASSR data do not take into account possible middle-ear changes with maturation that may affect the signal that actually reaches the cochlea (Keefe, Bulen, Arehart, & Burns, 1993; Keefe, Bulen, Campbell, & Burns, 1994). Keefe et al. (1993) compared input conductance for an air-conducted stimulus for infants 1-24 months of age in order to measure total middle-ear conductance which is an estimate of the power absorbed by the middle ear, i.e., power available to stimulate the cochlea. Their findings showed that, for very young infants (less than one month of age), the middle ear absorbs -1, 5, 4, and 11 dB less power at 500, 1000, 2000 and 4000 Hz, respectively, compared to adults. For infants 6-24 months of age, the amount of 236 power absorbed by the infant middle ear was only 2-4 dB less than adults for 500-4000 Hz (Keefe et al., 1993). These middle ear findings suggest that the contribution of neural maturation to neonate-adult differences in air-conduction ABR and ASSR thresholds may be overestimated by 11 dB at 4000 Hz; however, in older infants, maturation of the middle-ear likely plays a minor role in explaining why infants require greater air-conduction sound pressure levels at the tympanic membrane compared to adults, particularly at high frequencies. Although the effects that maturation of the neural pathways have on hearing thresholds should be the same for air- and bone-conducted stimuli, it is not clear how developmental changes in the middle-ear, and perhaps the outer ear, might impact bone-conduction hearing thresholds. Recent studies of bone-conduction mechanisms (reviewed in Stenfelt & Goode, 2005) suggest that potential contributors to bone-conduction hearing for adults include: sound radiated in the ear canal as a bone-conduction stimulus (including the possible role of the mandible), middle-ear cavity radiation and ossicle inertia, inertia of the cochlea fluids, compression of the cochlear walls, and pressure transmission from the cerebrospinal fluid. Stenfelt and Goode (2005) suggested that most of these possible mechanisms actually contribute very little to bone-conduction hearing in unoccluded ears, and hypothesized that inertial forces arising from skull vibrations, which then affect the movement of the cochlear fluid, are responsible for most of the transmission of vibratory energy that contributes to bone-conduction hearing. Other researchers have suggested that pressure transmission from the cerebrospinal fluid may be a major pathway for bone-conduction stimuli to reach the cochlea (Freeman, Sichel & Sohmer, 2000; Sohmer, Freeman, Geal-Dor, Adleman & Savion, 2000). When adult ears are occluded, the relative importance of different bone-conduction mechanisms change; sound pressure in the ear canal, generated from the vibratory energy of a bone-conducted signal, is 237 thought to be the main contributor to bone-conduction hearing with occluded ear and explains the underlying mechanism for the occlusion effect (Sohmer et al. 2005; Tonndorf, 1966). As discussed earlier, one piece of evidence that indicates that infant and adult bone-conduction mechanisms differ is that bone-conduction ASSR thresholds in young infants appear to be unaffected by occluding the ear canal (Small, Hatton & Stapells, 2007: Chapter 4), a result that is clearly different from adults with occluded ears. The infant skull is certain to have different mechanical properties from that of an older child or adult, as discussed earlier, however; it is not clear how the structure of an infant skull and other factors impact on the mechanisms of bone-conduction hearing. Further investigation of bone-conduction mechanisms in infants of different ages are clearly needed to explain the maturational changes observed for bone-conduction ABRs and ASSRs, and the extent to which the infant skull, middle-ear, and other factors may contribute to bone-conduction hearing sensitivity. Clinical implications Based on the ASSR threshold data from the current and previous ASSR studies, "normal" levels for ASSRs to bone-conduction stimuli are now available for 0-11-month-old andl2-24-month-old infants. These "normal" levels are presented in Table 6.2 for each age group (adults are also included for comparison). "Normal" refers to the lowest level at which greater than 90% of the participants had responses present; to be considered normal, an infant should have responses at the intensities presented in Table 6.2. Normal young infants, less than 11 months of age, should have responses at 30, 20, 40 and 30 dB HL at 500, 1000, 2000 and 4000 Hz, respectively. For infants aged 12 to 24 months, normal levels are 10 dB higher at 500 Hz and 10 dB lower at 4000 Hz. For adults, normal levels are 10 dB higher at 500 Hz compared to older infants, 20 dB higher at 500 Hz compared to young infants, and 20 dB higher at 1000 Hz 238 compared to both infant groups. At 2000 Hz, normal levels are 10 dB better for adults compared to both young and older infants. At 4000 Hz, normal levels are 10 dB lower for adults compared to young infants but are the same as the normal levels for older infants. Table 6.2 Many young infants are referred from newborn hearing screening programs for full diagnostic hearing assessments using auditory evoked potentials. Older infants and young children who are difficult to test behaviourally are also tested using physiologic methods. The results presented in this study show that bone-conduction thresholds change with maturation, thus underlying the importance of establishing normal bone-conduction hearing levels for a range of test frequencies for infants of different ages. The normal levels for bone-conduction multiple ASSRs proposed in this study could be used clinically to screen for normal bone-conduction hearing in young and older infants who have elevated air-conduction thresholds or who fail screening. However, before interpreting elevated bone-conduction ASSR thresholds, it is also important to investigate bone-conduction ASSRs in infants with hearing loss. 239 Table 6.1 Bone-conduct ion A S S R ampli tude and phase delay: Three-way mixed A N O V A s showing compar isons between intensities (30 and 40 dB HL), ac ross age groups (35 young infants*, 13 older infants and 18 adults**) and carrier f requencies (500, 1000, 2000 and 4000 Hz). Sou rce df F ef P b Ampl i tude A g e 2, 45 1.069 0.352 (nV) Frequency 3, 135 8.713 0.722 .0002* Intensity 1, 45 67.078 < 0 0 0 1 * A g e x Intensity 2, 45 5.407 1 .008* Intensity x Frequency 3, 135 2.431 0.932 0.073 A g e x Frequency 6, 135 9.149 0.722 < 0 0 0 1 * A g e x Frequency x Intensity 6, 135 0.586 0.332 0.578 P h a s e delay A g e 2, 23 2.353 0.118 (ms) Frequency 3, 69 40.98 1 < 0 0 0 1 * Intensity 1, 23 48.193 1 < 0 0 0 1 * A g e x Intensity 2, 23 9.913 1 .001* Intensity x Frequency 3, 69 4.379 0.896 .009* A g e x Frequency 6, 69 2.511 1 .030* A g e x Frequency x Intensity 6, 69 2.713 0.896 .025* a Huynh-Feldt epsi lon (e) correction factor for degrees of f reedom b Probabil i ty reflects corrected degrees of f reedom * significant (p < .05) 240 Table 6.2 A preliminary est imate of "normal" A S S R levels in dB HL for 500-, 1000-, 2000- and 4000 -Hz bone-conduct ion stimuli for young infants (n=35)*, older infants (n=13) and adults (n=18)** with normal hearing. A g e 500 H z 1000 H z 2000 H z 4000 H z Normal levels 0-11 months 30 20 40 30 dB HL 12-24 months 40 20 40 20 Adult 50 40 30 20 "Normal" level = the lowest level at which greater than 9 0 % of subjects showed "present" responses 241 Figure 6.1 Representative bone-conduction ASSR for an individual young infant (4.5 months), an older infant (24 months) and an adult (22 years). Shown are amplitude spectra resulting from FFT analyses (75-105 Hz) of the ASSRs. Filled triangles indicate responses which differ significantly from the background noise (p < .05). Open triangles indicate no response (p >.05 and E E G noise <11 nV). Threshold is defined as the lowest intensity that produced a significant response. Intensity (dB HL) Young infant Older infant Adult 4.5 months 24 months 22 years FREQUENCY (kHz) . 5 1 2 4 • • 40 30 20 10 0 T • • • v T • T V V T • . V V V V v V 75 80 85 90 95 100 105 Threshold 10 10 40 30 i ^ .mJ .^ . . . . . . .-...t V V V V • • • V • • V V V T llnilhiilllliilihiniu int II t V V V V 75 80 85 90 95 100 105 75 80 85 90 95 100 105 20 10 20 20 30 20 20 10 242 Figure 6.2 Cumulative percent occurrence of subjects with significant responses for young infants (n =35), older infants (n =13) and adults (n =18) across frequency. no 100 oo 80 60 40 20 0 10 20 30 40 50 10 20 30 40 50 Intensity (dB HL) 243 Figure 6.3 Mean bone-conduction ASSR thresholds (± 1 SD) at each carrier frequency for 35 young infants, 13 older infants and 18 adults with normal hearing. 0-11 12-24 adult months months Age Group 244 Figure 6.4 Graphical representation of linear regression analysis comparing age in weeks to ASSR thresholds at each of the carrier frequencies (age was arbitrarily set at 120 weeks for adults). Regression equations (y= mx + b; where y is ASSR threshold, x is age, m is the slope of the regression and b is the y-intercept) and correlation coefficients (r) are shown in the upper right corner of each graph. Significant correlations are marked with an asterisk. 0 20 40 60 80 100 Adult 0 20 40 60 80 100 Adult Age (weeks) 245 Figure 6.5 Mean bone-conduction ASSR amplitudes (± 1 SD) at each carrier frequency for 35 young infants, 13 older infants and 18 adults with normal hearing. 100 80 > 60 c 40 0 20 "D 0 3 100 • mmm 80 E 60 < 40 20 0 T 1 1 r 500 Hz # 0-11 months -O- 12-24 months - A - adult 2000 Hz J I I I L J L J L •10 0 10 20 30 40 50 -10 0 10 20 30 40 50 Intensity (dB HL) 246 Figure 6.6 Mean bone-conduction ASSR phase delays (± 1 SD) at each carrier frequency for young infants, older infants and adults with normal hearing. JS o <D CO CO 14 12 10 8 6 4 14 12 10 8 6 500 Hz 2000 Hz 1000 Hz 4000 Hz _0_ 0-11 months -—o- 12-24 months adult 10 20 30 40 0 Intensity (dB HL) 10 20 30 40 247 R E F E R E N C E S ANSI (1996). American National Standard Specifications for Audiometers (ANSI S3.6-1996). New York, ANSI. Anson, B. J., Donaldson J. A. (1981). The temporal bone. In B. J. Anson, J. A. Donaldson (eds.) Surgical anatomy of the temporal bone, (pp. 3-25). Philadelphia: W. B. Saunders Company. Cone-Wesson, B. & Ramirez, G.M. (1997). Hearing sensitivity in newborns estimated from ABRs to bone-conducted sound. Journal of the American Academy of Audiology 8:299-307. Cone-Wesson, B, Rickards, F, Poulis, C, Parker, J, Tan, L. & Pollard, J. (2002). The auditory steady-state response: clinical observations and applications in infants and children. Journal of the American Academy of Audiology 13: 270-282. Dimitrijevic, A., John, M . S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., Nedzelski, J. M . & Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology 13: 205-224. Durrani, J. D., & Hyre R. (1993). Relative effective frequency responses of bone versus air conduction stimulation examined via masking. Audiology 32: 175-84. Eby, T. L., & Nadol J. B. (1986). Postnatal growth of the human temporal bone. Annals of Otology, Rhinology and Laryngology 95: 356-364. Foxe, J. J. & Stapells, D. R. (1993). Normal infant and adult auditory brainstem responses to bone-conducted tones. Audiology 32: 95-109. 248 Freeman, S., Sichel, J-Y. & Sohmer, H. (2000). Bone conduction experiments in animals-evidence for a non-osseous mechanism. Hearing Research 146: 72-80. Herdman, A., Lins, O., Van Roon, P., Stapells, D., Scherg, M . , & Picton, T. (2002). Intracerebral sources of human auditory steady-state responses. Brain Topography 15, 69-86. Jahrsdoerfer, R. A., Yeakley, J. W., Hall, J. W., Robbins, K. T. & Gray, L. C. (1985). High-resolution CT scanning and auditory brain stem response in congenital aural atresia: Patient selection and surgical correlation. Otolaryngology-Head and Neck Surgery 93: 292-298. John, M . S. & Picton, T. W. (2000a). MASTER: a Windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine 61: 125-150. John, M . S. & Picton, T. W. (2000b). Human auditory steady-state responses to amplitude-modulated tones: phase and latency measurements. Hearing Research 141 (1-2): 57-79. Joint Committee on Infant Hearing. (2000). Year 2000 position statement: Principles and guidelines for early hearing detection and intervention programs. American Journal of Audiology 9: 9-29. Lins, O. G., Picton, T. W., Boucher, B. L., Durieux-Smith, A., Champagne, S. C , Moran, L. M . , Perez-Abalo, M . C , Martin, V. & Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear and Hearing 17: 81-96. 249 Moore, J. K., Ponton C. W., Eggermont J. J., Wu J . - C , Huang J. Q. (1996). Perinatal maturation of the auditory brain stem response: Changes in path length and conduction velocity. Ear and Hearing 17: 411-418. Nousak, J, Stapells, D.R. (1992). Frequency specificity of the auditory brainstem response to bone-conducted tones in infants and adults. Ear and Hearing 13: 87-95. Picton, T.W., John, M.S., Dimitrijevic, A. & Purcell, D. (2003). Auditory steady-state responses. International Journal of Audiology 42: 177-219. Ponton, C. W., Moore, J. K., Eggermont, J. J., Wu, B. J . - C , & Huang, J. Q. (1994). Relationship between auditory brainstem size and auditory brainstem response (ABR) latency in human development. Association for Research in Otolaryngology Abstracts: 57. Ranee, G., Tomlin, D. (2006). Maturation of auditory steady-state responses in normal babies. Ear & Hearing 27: 20-29. Savio, G., Cardenas, J., Perez-Abalo, Gonzalez, A. & Valdes, J. (2001). The low and high frequency auditory steady state responses mature at different rates. Audiology & Neuro-Otology 6: 279-287. Sininger, Y. S., Abdala C , & Cone-Wesson B. (1997). Auditory threshold sensitivity of the human neonate as measured by the auditory brainstem response. Hearing Research 104: 27-38. Small, S. A., Hatton, J., Stapells, D.R. (2007). Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants. Ear & Hearing 28(1): 83-98. [Chapter 4] Small, S. A., Stapells D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear and Hearing 25(6): 611-623. [Chapter 2] 250 Small, S.A. & Stapells, D.R. (2005). Multiple auditory steady-state responses to bone-conduction stimuli in adults with normal hearing. Journal of the American Academy of Audiology 16(3): 172-183. [Chapter 3] Small, S. A., Stapells, D.R. (2006). Multiple auditory steady-state response thresholds to bone-conduction stimuli in young infants with normal hearing. Ear & Hearing 27: 219-228. [Chapter 5] Sohmer, H., Freeman, S. Geal-Dor, M . , Adelman, C , & Savion, I. (2000). Bone-conduction experiments in humans- a fluid pathway from bone to ear. Hearing Research 146:81-88. Stapells, D. R. (2000). Frequency-specific evoked potential audiometry in infants. In R.C. Seewald (ed.) A Sound Foundation through Early Amplification: Proceedings of an International Conference (pp. 13-31). Chicago: Phonak AG. Stapells, D. R., Herdman, A., Small, S.A., Dimitrijevic, A. & Hatton, J. (2005). Current status of the auditory steady-state responses for estimating an infant's audiogram. In R.C. Seewald & J.M. Bamford (eds.) A Sound Foundation through Early Amplification 2004 (pp. 43-59). Chicago: Phonak A G . Stapells, D. R. & Ruben, R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. Annals of Otology, Rhinology and Laryngology 98: 941-949. Stenfelt, S., Goode, R.L. (2005). Bone conducted sound: Physiologic and clinical aspects. Otology & Neurotology 26: 1245-1261. Stuart, A., Yang E. Y. , & Stenstrom R. (1990). Effect of temporal area bone vibrator placement on auditory brain stem response in newborn infants. Ear and Hearing 11: 363-369. 251 Stuart, A., Yang E. Y., Stenstrom R., & Reindorp A. G. (1993). Auditory brainstem response thresholds to air and bone conducted clicks in neonates and adults. The American Journal ofOtology 14: 17'6-182 Tonndorf, J. (1966). Bone conduction: Studies in experimental animals. Acta Otolaryngologica, Supplement 213: 1-132. Yang, E. Y., Rupert A. L., & Moushegian G. (1987). A developmental study of bone conduction auditory brainstem responses in infants. Ear and Hearing 8: 244-251. 252 CHAPTER 7 NORMAL IPSILATERAL/CONTRALATERAL ASYMMETRIES IN INFANT MULTIPLE AUDITORY STEADY-STATE RESPONSES TO AIR- AND BONE-CONDUCTION STIMULI A version of this chapter has been submitted for publication. Small, S. A., & Stapells, D. R. (in review). Normal ipsilateral/contralateral asymmetries in infant multiple auditory steady-state responses to air- and bone-conduction stimuli. Ear and Hearing. 253 INTRODUCTION Previous studies have shown that two-channel recordings [i.e., electroencephalgraphic (EEG) channels ipsilateral and contralateral to stimulus ear] of the bone-conduction auditory brainstem response (ABR) show maturational differences (Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stuart, Yang & Botea, 1996). ABRs to brief-tone bone-conduction stimuli in infants in the contralateral E E G channel have smaller wave V amplitudes and longer wave V latencies compared to the ipsilateral E E G channel. In contrast, adults have more similar wave V amplitudes and latencies in the two channels. These latency and amplitude asymmetries in infants are present at all intensity levels and persist until at least 1-2 years of age (Stapells & Mosseri, 1991). The ipsilateral/contralateral asymmetries, which relate both to developmental changes in skull structure and to neurophysiology, may be used to provide valuable diagnostic information when assessing bone-conduction hearing in infants by identifying the cochlea that is responding to the stimulus (Foxe & Stapells, 1993; Stapells & Ruben, 1989). This is particularly important in cases of bilateral conductive loss when it is difficult (or impossible) to isolate the stimulus ear using behavioural measures (Jahrsdoerfer, Yeakley, Hall, Robbins, & Gray, 1985; Stapells & Ruben, 1989). Because interaural attenuation in adults is much less for bone-conduction [10 dB on average at 250-4000 Hz (Nolan & Lyon, 1981)] compared to air-conduction [supra-aural: 40-50 dB on average at 250-4000 Hz (Goldstein & Newman, 1994, p. 117); insert earphones with deep insertion: 69-94 dB on average at 500-4000 Hz (Sklare & Denenberg, 1987)] stimuli, it is difficult to determine how much each cochlea contributes to the resulting bone-conduction ABR. Presenting masking noise to the ear contralateral to the test ear is often used to isolate the test ear; however, masking may not be practical when assessing infants because of the extra time 254 required for testing, the difficulties of earphone placement, and the uncertainty about how much masking noise to use (Stapells & Ruben, 1989). The asymmetries in the ipsilateral and contralateral bone-conduction recordings provide valuable ear-specific diagnostic information based on the assumption that the cochlea closest to the channel which shows the largest and earliest wave V contributed to the recorded ABR (Stapells & Ruben, 1989). The brief-tone ABR is currently the "gold standard" for estimating frequency-specific air-and bone-conduction thresholds in infants (Stapells, 2000; Stapells, Herdman, Small, Dimitrijevic & Hatton, 2005); in the near future, the 80-Hz multiple auditory steady-state responses (ASSRs) may be used in conjunction with or replace the brief-tone ABR. The most attractive feature of the ASSR is that a response is detected objectively rather than by subjective visual interpretation (which is the case for the ABR). Also, the multiple ASSR can test as many as four frequencies in one ear simultaneously (as many as eight frequencies in total for air-conduction stimuli) (reviewed in Picton, John, Dimitrijevic, & Purcell, 2003); compared to an ASSR elicited by a single stimulus, testing time is 2-3 times faster (Herdman & Stapells, 2001; Herdman & Stapells, 2003; John, Purcell, Dimitrijevic & Picton, 2002). Our recent studies were the first to estimate bone-conduction thresholds in infants using multiple ASSRs (Small, Hatton & Stapells, 2007: Chapter 4; Small & Stapells, 2006: Chapter 5). The results of these studies showed that it is possible to record bone-conduction ASSRs to multiple stimuli in pre- and post-term infants with normal hearing and revealed differences in bone-conduction thresholds between two infant groups compared to adults. In particular, the results showed that infants had better low-frequency and poorer high-frequency ASSR thresholds compared to adults. Although more studies are needed to obtain bone-conduction ASSRs in infants of different ages and in infants with hearing loss before this technique is ready for clinical 255 use, it clearly has promise as a bone-conduction threshold assessment tool for infants. Based on the previous studies that report ipsilateral/contralateral asymmetries in the bone-conduction ABR in infants, and on source analysis studies that showed that the 80-Hz ASSR is primarily generated in the auditory brainstem (Herdman, Lins, Van Roon, Stapells, Scherg & Picton, 2002; Kuwada, Anderson, Batra, Fitzpatrick, Teissier & D'Angelo, 2002; Mauer & Doring, 1999), we hypothesize that these asymmetries will also be present in the 80-Hz ASSRs of infants to bone-conduction stimuli. The purpose of the present study is to determine whether ipsilateral/contralateral asymmetries are seen for infant bone-conduction ASSRs, similar to those reported for infant bone-conduction ABRs. If there are significant differences in ipsilateral and contralateral ASSR amplitude and phase measures, this technique might be used to assess which ear is responding to the bone-conduction stimulus while simultaneously obtaining bone-conduction thresholds. For comparison, two-channel air-conduction ASSRs will also be recorded. Unlike the bone-conduction condition which may result in stimulation of both cochleae, we can be sure that the air-conduction condition, for stimuli presented at intensities less than the minimum interaural attenuation with insert earphones [69 to 94 dB, depending on frequency (Sklare & Denenburg. 1987)], results in monaural stimulation. Any ipsilateral/contralateral asymmetries in the ASSR elicited by the air-conduction stimuli would result only from neurophysiologic differences in the ipsilateral and contralateral views of the underlying neurogenerators, not asymmetries due to the auditory signal crossing over to the cochlea contralateral to the ear closest to the transducer, which is possible when using bone-conduction stimuli. 256 MATERIALS AND METHODS Participants A total of 11 adults (7 female) and 14 infants (7 female) recruited from the community participated in this study. All adults (age range: 18-40 years; mean age of 23 years) had normal hearing (i.e., pure-tone behavioural thresholds <25 dB HL at 500, 1000, 2000 and 4000 Hz, respectively). All infants (age range of 8-44 weeks; mean age of 21 weeks) who participated passed a distortion-product otoacoustic emissions (DPOAE) screening test in both ears. One of the infants participated in a previous study (Small, Hatton & Stapells, 2007: Chapter 4). Four additional adult participants were excluded from the analyses of the mean data because they had atypically large ASSR amplitudes; it is likely that these ASSRS reflected post-auricular muscle responses rather than responses arising from the auditory brainstem (Sohmer, Pratt & Kinarti, 1977). Stimuli All stimuli were sinusoidal tones with the carrier frequencies 500, 1000, 2000 and 4000 Hz that were 100% amplitude and 25% frequency modulated at 77.148, 84.961, 92.773 and 100.586 Hz, respectively. The stimuli were presented simultaneously for all conditions tested (i.e., multiple). All stimuli were generated using Sasha John's custom-made two-channel version of the Rotman MASTER research system (John and & Picton, 2000a). The system used a buffer length of 25,600 points and a digital-to-analog (D/A) rate of 31,250 Hz, which is an integer sub-multiple of the 20 MHz clock rate, but not an integer multiple of the carrier frequencies (Small & Stapells, 2004: Chapter 2). The stimuli were routed through a Stanford Research Systems Model SR650 to increase the gain of the signal by 10 dB and were then attenuated through Tucker-Davis Technologies HB6 and SM3 modules. 257 The bone-conduction stimuli were presented to a Radioear B-71 bone oscillator which was coupled to the head using an elastic headband with 400-450 g of force (Small, Hatton & Stapells, 2007: Chapter 4). The bone oscillator was placed on the temporal bone slightly posterior to the upper part of the pinna for all infants and adults tested. Bone-conduction stimuli were presented using 10-dB steps from 50 to -10 dB HL [intensities greater than 50 dB H L for multiple stimuli result in non-linearities in the oscillator output (Small & Stapells, 2004: Chapter 2)]. The air-conduction stimuli were presented monaurally at 60 dB H L to an EAR-3A insert earphone. This level was selected to ensure that the air-conduction stimulus would be loud enough to elicit significant responses in the E E G channel ipsilateral to the transducer [median ASSR thresholds to air-conduction stimuli range from 32 to 43 dB HL in infants with normal hearing (Stapells et al., 2005)] but not loud enough to cross over to elicit responses from the cochlea opposite to the insert earphone [minimum interaural attenuation of 69 to 94 dB depending on stimulus frequency (Sklare & Denenburg, 1987]). Calibration The bone-conduction stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re:luN corresponding to 0 dB HL for the mastoid (ANSI S3.6 -1996) using a Briiel and Kjaer Model 2218 sound level meter and Model 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 550 g of force. Air-conduction stimuli were calibrated in dB HL (ANSI, 1996) using a Quest Model 1800 sound level meter with a Briiel and Kjaer DB0138 2-cc coupler. 258 ASSR Recordings ASSRs were recorded using the two-channel version of the Rotman MASTER system. Four electrodes were used to record the electrophysiologic responses; the non-inverting electrode was placed midline at the high forehead, the inverting electrodes were positioned low on the left and right mastoids, and an electrode placed at the low forehead served as a ground. All inter-electrode impedances were below 3 kOhms at 10 Hz. The E E G was filtered using a 30-250 Hz filter (12 dB/oct) and amplified 80,000 times (8000X in Nicolet HGA-200A and Nic501A; 10X in NIDAQ card). The EEG was further filtered using a 300-Hz lowpass anti-aliasing filter [Wavetek Rockland Model 852 (48 dB/oct)]. The E E G was then processed using a 1250-Hz A/D conversion rate (Small & Stapells, 2004: Chapter 2). Each E E G recording sweep was made up of 16 epochs of 1024 data points (0.819 seconds per epoch) and lasted a total of 13.107 seconds. Artifact rejection was set to eliminate epochs of electrophysiologic activity that exceeded ± 40 | iV in amplitude in order to reduce contributions to the EEG due. to muscle artifact. ASSRs were averaged in the time domain and then analyzed online in the frequency domain using a Fast Fourier Transform (FFT). Weighted averaging (John, Dimitrijevic & Picton, 2001) was used. The FFT resolution was 0.076 Hz over a range of 0 to 625 Hz. Amplitudes were measured baseline-to-peak and expressed in nV. An F-ratio was calculated by MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within ± 60 bins of the modulation frequency ("noise") (John & Picton, 2000a). A minimum of 10 sweeps was always recorded for each test condition. A response was considered to be present if the F-ratio, compared to the critical values for F (2, 259 240), was significant at a level of p < .05 for at least two consecutive sweeps. A response was considered to be absent ifp > .05 and the mean amplitude of the noise was less than 11 nV. Alternatively, a response was also considered to be absent when response amplitude was < 10 nV and thep value > .30. Procedure Testing was conducted in a double-walled sound-attenuated booth in the Human Auditory Physiology Laboratory at the University of British Columbia. Hearing screening was performed in both ears at the beginning of the test session to establish that the participants were unlikely to have hearing loss. Infant participants were tested only when quiet and asleep; adults were tested while reclining in a comfortable chair and were either relaxed or asleep. Bone-conduction testing was performed for all participants with ears unoccluded. Multiple ASSRs were elicited to bone-conduction stimuli beginning at a randomized starting intensity until thresholds were obtained at all frequencies in the "test ear" (i.e., the ear for which bone-conduction thresholds are obtained); the bone-oscillator was then moved to the "non-test ear" (i.e., the side opposite to where thresholds were obtained) where bone-conduction ASSRs were recorded only at 40 dB HL. Following bone-conduction testing, multiple ASSRs to air-conduction stimuli presented monaurally via insert earphones at 60 dB HL were elicited in the "test ear" and then the "non-test ear". Threshold for each carrier frequency was determined using a bracketing technique adjusting the presentation level using 10-dB steps. The lowest intensity at which a response was present was considered "threshold". In some cases, a recording did not reach significance at a level above "threshold"; if only one recording did not reach significance, the lowest level at which the response was present was considered threshold. If recordings at two or more 260 consecutive levels were not significant, the significant response at the lowest level was considered a "false positive" and the lowest response above the absent responses was deemed "threshold". The total recording time was approximately 1.5-2.0 hours, including the time to obtain screening test results. All participants (or parents) signed a consent form (University of British Columbia Behavioural Research Ethics Board) and were paid an honorarium at the end of each session. Data Analyses Amplitude values were averaged across subjects, including ASSR amplitudes for responses that were not significant. The phase values from MASTER were adjusted by adding 90° to yield the onset phase in degrees (John & Picton, 2000b). Onset phase values were then converted to phase delay by subtracting the onset phase value from 360°. Any phase-delay values that differed > 180° from an adjacent measure were "unwrapped" by adding 360° to their value. Phase delay in degrees was then converted to phase delay in milliseconds to allow comparisons at different frequencies (John & Picton, 2000b). Phase values for ASSRs that were not significant were excluded from mean phase-delay calculations. Phase-delay values were averaged across subjects. Results are reported only if at least five subjects contributed to the mean. The cumulative occurrence of ipsilateral/contralateral asymmetries was determined for the total number of responses obtained both at 30 and 40 dB HL. The results at 40 dB HL included ASSRs in the E E G channels ipsilateral and contralateral to the "test" and "non-test" ear. Comparisons were made within and between age groups for bone-conduction ASSR thresholds. Within-group comparisons of air- and bone-conduction mean ASSR thresholds were made between E E G recording channels (i.e., ipsilateral vs contralateral) and across carrier frequencies. For between-group comparisons, the differences between ipsilateral and 261 contralateral ASSR thresholds (contralateral minus ipsilateral) were compared between age groups and across carrier frequencies; this measure also provided a minimum estimate of interaural attenuation [(i.e., if a response were present in both E E G channels, both cochleae may have been stimulated (although it is also possible that only one cochlea was stimulated); however, if a response were only present in the ipsilateral E E G channel, we could assume that the signal stimulated only the ipsilateral cochlea]. Air- and bone-conduction ASSR amplitudes and phase delays in milliseconds were also compared between E E G channels and across carrier frequencies within and between age groups. Within-group comparisons were made for bone-conduction mean amplitudes at 10, 20, 30 and 40 dB HL; statistical analyses were limited to stimulus intensities of 10 dB HL or greater because the majority of responses at lower intensities were within the noise floor. Within-group comparisons were made for air-conduction mean ASSR amplitudes at 60 dB HL. Within-group comparisons for bone-conduction mean phase delays (ms) were made at 40 dB H L only because too few subjects had significant responses at all of the carrier frequencies at lower intensities. Comparisons of infant air-conduction mean phase delays (ms) were made at 4000 Hz only; at the other frequencies, there were as few as 2-5 infants who had significant responses in the contralateral channel at each of the frequencies, consequently, the ipsilateral/contralateral differences in phase delay could not be quantified. Comparisons of adult mean air-conduction ASSR phase delays (ms) were made at each of the carrier frequencies. Between group comparisons for ASSR amplitude and phase delay between E E G channels were made using difference scores. Because infant air- and bone-conduction ASSR amplitudes were either smaller or larger than those for adults depending on the carrier frequency, the difference between ASSR amplitudes for the ipsilateral and contralateral E E G channel was 262 normalized relative to the amplitude of the ipsilateral responses at 20, 30 and 40 dB HL (results at 10 dB HL were not included because small differences in amplitude relative to the small amplitude in the ipsilateral E E G resulted in normalized values that were not sensible). For bone-conduction phase delays (in ms), the differences between ipsilateral and contralateral ASSR phase delays (contralateral minus ipsilateral) at 40 dB HL were compared across carrier frequencies. For air-conduction ASSRs, the differences in phase delay were compared at 4000 Hz only. Comparisons of E E G channel, intensity, frequency and age were made using 2- and 3-way repeated-measures or mixed-model analyses of variance (ANOVA) or independent and dependent /-tests, as appropriate. Case-wise deletion was used when there were missing values in the data set. Hunyh-Feldt epsilon adjustments for repeated measures were made when appropriate. Newman-Keuls post-hoc comparisons were performed for significant main effects and interactions. The criterion for statistical significance was p<.Q5 for all analyses. R E S U L T S Representative polar plots of two-channel ASSR recordings to 2000-Hz air- and bone-conduction stimuli are shown for a typical infant and adult in Figure 7.1. For the 4-month-old infant, when the left ear/mastoid was stimulated, the air- and bone-conduction ASSRs recorded in the right E E G channel (i.e., contralateral EEG), had smaller amplitudes and longer phase delays compared to those in the left E E G channel (i.e., ipsilateral EEG). The same pattern of ipsilateral/contralateral asymmetries was also evident when the ear/mastoid was switched (i.e, the right ear/mastoid was stimulated). In contrast, the adult did not show consistent ipsilateral/contralateral asymmetries in the air- or bone-conduction ASSRs. In this adult 263 example, the air- and bone ASSR amplitudes were similar in both the E E G channels ipsilateral and contralateral to the ear/mastoids stimulated, and although the air- and bone-conduction ASSR phase delays were longer in the right E E G channel when the left ear was stimulated, the difference in ASSR phase delays between EEG channels when the right ear was stimulated was negligible. Similar to the individual results, the grand mean (vector average) results for two-channel ASSR recordings to 2000-Hz air- and bone-conduction stimuli for infants and adults, shown in Figure 7.2, indicate consistent ipsilateral/contralateral asymmetries in amplitude and/or phase delay for infants, and inconsistent asymmetries in amplitude and phase delay for adults. Figures 7.1 and 7.2 Occurrence of ipsilateral/contralateral aysymmetries The presence of ipsilateral/contralateral asymmetries was determined using four different criteria, which included: (i) "Amp": ASSR amplitude was smaller in the contralateral E E G channel, (ii) "Phase": ASSR phase delay was longer in the contralateral E E G channel, (iii) "Amp & Phase": ASSR amplitude was smaller in the contralateral E E G channel and ASSR phase delay was longer in the contralateral E E G channel, and (iv) "Amp or Phase": ASSR amplitude was smaller in the contralateral E E G channel or ASSR phase delay was longer in the contralateral E E G channel. Criterion (iii) is the most strict, whereas criterion (iv) is the least strict. Figure 7.3 shows the cumulative occurrence (in percent) of ipsilateral/contralateral asymmetries (i.e., smaller ASSR amplitudes and longer phase delays in the E E G channel contralateral to the stimulus ear) pooled across frequency for air- (60 dB HL) and bone- (30 and 40 dB HL) conduction stimuli for infants and adults. Overall, substantially more infants showed these 264 ipsilateral/contralateral asymmetries for both air- and bone-conduction stimuli compared to adults using any of the four criteria. For infants, the greatest number of responses that showed ipsilateral/contralateral asymmetries was found for air-conduction stimuli. Using the strictest criterion (amplitude and phase), 86% of these air-conduction responses showed ipsilateral/contralateral asymmetries, whereas for the least-strict criteria (amplitude or phase), 92-100% of infant responses showed these asymmetries. In infants, the occurrence of ipsilateral/contralateral asymmetries for air-conduction stimuli was slightly higher when amplitude alone was used as the criterion versus phase alone. For adults, ipsilateral/contralateral asymmetries seen for air-conduction stimuli were found for only 32% of the responses for the most-strict criterion and ranged from 52-77% for the least-strict criteria. Figure 7.3 For bone-conduction stimuli, ipsilateral/contralateral asymmetries occurred more frequently in infants than in adults, and these asymmetries occurred slightly more often at 30 dB HL compared to 40 dB HL. For infants, using the most-strict criterion, the cumulative percent occurrence of asymmetries for bone-conduction stimuli was 78% for 30 dB HL and 66% for 40 dB HL; for the least-strict criteria, the cumulative percent occurrence increased to 96% for 30 dB HL and 89%) for 40 dB HL. In infants, the occurrence of ipsilateral/contralateral asymmetries for bone-conduction stimuli, in contrast to the air-conduction stimuli, was slightly lower when amplitude alone was used as the criterion versus phase alone. In adults, there was only a 44% occurrence of asymmetries in the ASSRs to bone-conduction stimuli using the most-strict 265 criterion and, at most, a 59-81% occurrence using the less-strict criteria. The percentage of contralateral responses that were > 10% smaller than ipsilateral responses and > 1.6 ms later than ipsilateral responses for infants and adults are also shown in Figure 7.3. Infants also had a greater number of occurrences of large amplitude (contralateral responses> 10% smaller than ipsilateral responses) and phase (contralateral > 1.6 ms later than ipsilateral responses) differences between EEG channels for air- and bone-conduction stimuli when asymmetries were present compared to adults (Figure 7.3). Large ipsilateral/contralateral amplitude differences for air- and bone-conduction stimuli were 13-14% more prevalent for infants compared to adults Large ipsilateral/contralateral phase differences were 76 and 34% more prevalent for air- and bone-conduction stimuli in infants compared to adults. ASSR threshold The mean ASSR thresholds estimated from responses recorded in the E E G channels ipsilateral and contralateral to the "test ear" are shown in Table 7.1. For infants, mean ASSR thresholds were 13-15 dB poorer for ASSRs recorded in the contralateral E E G channel compared to those in the ipsilateral E E G channel; the channel effect was statistically significant [F(l,10) = 34.468,/? = .0002]. There was a significant effect of frequency, similar to results reported in a previous study (Small & Stapells, 2006); however, no significant interaction between E E G channel and frequency [F (3,30) = 0.759, p = .526] was found. For adults, there was no difference between mean ASSR thresholds for responses recorded in the contralateral or ipsilateral E E G channels [F (1,14) = 0.028, p = .869]. There was a significant effect of frequency [F(3,42) = 12.175,/? < .0001] but no significant interaction between E E G channel and frequency [F (2,27) = 0.060, p =. 981 ]. Post hoc comparisons revealed that the significant effect of 266 frequency was explained by significantly poorer thresholds at 500 Hz compared to higher frequencies and significantly poorer ASSR thresholds at 1000 and 2000 Hz compared to 4000 Hz. The results also show that 34% percent of infants did not have significant bone-conduction ASSRs in the contralateral E E G channel when responses were present in the ipsilateral E E G channel, whereas, only 12% of adults did not have significant ASSRs in the contralateral E E G channel when responses were present in the ipsilateral E E G channel. Table 7.3 Tables 7.2a and 7.2b show the differences in bone-conduction ASSR threshold between E E G channels for each individual infant and adult at each carrier frequency. For infants, the difference between ASSR thresholds in the ipsilateral and contralateral E E G channels was significantly larger than for adults [F(l,16) = 27.762,/? <.0001]. There was no significant effect of frequency [F (3,48) = 0.121,/? = .947] and no significant interaction between age and frequency [F (3,48)= 0.196,/? = .899]. From these data, one can estimate the minimum amount of interaural attenuation for a bone-conducted stimulus and the variability in interaural attenuation for infants and adults. If a response is present in the contralateral E E G channel, it may be due to (i) crossover stimulation of t