Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A comparison of the auditory steady-state and auditory brainstem responses to air- and bone-conducted… Valeriote, Hope 2015

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

Item Metadata

Download

Media
24-ubc_2015_may_valeriote_hope.pdf [ 3.07MB ]
Metadata
JSON: 24-1.0167692.json
JSON-LD: 24-1.0167692-ld.json
RDF/XML (Pretty): 24-1.0167692-rdf.xml
RDF/JSON: 24-1.0167692-rdf.json
Turtle: 24-1.0167692-turtle.txt
N-Triples: 24-1.0167692-rdf-ntriples.txt
Original Record: 24-1.0167692-source.json
Full Text
24-1.0167692-fulltext.txt
Citation
24-1.0167692.ris

Full Text

	  A COMPARISON OF THE AUDITORY STEADY-STATE AND AUDITORY BRAINSTEM RESPONSES TO AIR- AND BONE-CONDUCTED STIMULI IN INFANTS WITH HEARING LOSS by  Hope Valeriote   B.A., McGill University, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Audiology and Speech Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   February, 2015  © Hope Valeriote, 2015 	   ii	  Abstract   This study investigates how well the air- (AC) and bone-conduction (BC) auditory steady-state response (ASSR) detects conductive hearing loss compared to the gold standard method, the auditory brainstem response (ABR). Similar studies using infants with sensorineural hearing loss have suggested that the ASSR is an effective method for assessing hearing thresholds and detecting hearing loss in infants. This study compares AC and BC ASSR and ABR thresholds in infants with normal hearing and conductive loss.   Twenty-three normal hearing infants and 15 infants with conductive hearing loss (0-6 months) were assessed using the ABR and 80-Hz ASSR elicited to AC and BC stimuli (AM2). Mean thresholds for normal hearing infants were: 500 Hz (i) AC ABR: 25 dB nHL, (ii) BC ABR: 10 dB nHL, (iii) AC ASSR: 30 dB HL , (iv) BC ASSR: 17 dB HL and 2000 Hz (i) AC ABR: 18 dB nHL, (ii) BC ABR: 15 dB nHL, (iii) AC ASSR: 20 dB HL, (iv) BC ASSR: 26 dB HL. For infants with confirmed conductive hearing loss, 500 Hz thresholds air-conduction ABR thresholds increased to approximately 48 dB nHL, while bone-conduction ABR thresholds were approximately 12 dB nHL. Air-conduction ASSR thresholds for infants with conductive hearing loss increased to approximately 37 dB HL and bone-conduction thresholds were approximately 15 dB HL. Overall, mean bone-conduction thresholds were similar between groups, while there was a trend for mean air-conduction thresholds to be higher for infants with conductive hearing loss than infants with normal-hearing for both ABR and ASSR testing methods.   Previously suggested “normal levels” in the literature appear to be too high to detect mild conductive hearing loss at 500 Hz. Normal levels of 40 and 30 dB HL is suggested for air- and bone conduction 500 Hz ASSR, respectively, to be more accurate in detecting mild hearing losses. Even with an adjusted “normal level”, it appears to be difficult to use the ASSR to differentiate between normal hearing and mild conductive hearing loss. More research is needed using infants with varying degrees of hearing loss at multiple frequencies.    	   iii	  Preface  With guidance and assistance from Dr. Susan Small, I designed the research study described in chapter 2, collected subject data, analyzed the data and prepared this thesis. Dr. Charlotte Douglas aided in obtaining the ethics approval from the University of Saskatchewan, recruitment of participants and supervised data collection. Ethics approval for the project entitled “Comparison of air- and bone-conduction auditory brainstem and multiple auditory steady-state responses in infants with hearing loss” was obtained from the University of British Columbia’s Clinical Research Ethics Board (certificate # H13-02445) and the University of Saskatchewan’s Biomedical Research Ethics Board (certificate # 13-319).   	   iv	  Table of Contents  Abstract ................................................................................................... ii	  Preface .................................................................................................... iii	  Table of Contents ................................................................................... iv	  List of Tables ......................................................................................... vii	  List of Figures ........................................................................................ ix	  List of Symbols and Abbreviations ....................................................... x	  Acknowledgements ................................................................................ xi	  INTRODUCTION ...................................................................................................1	  MECHANISMS FOR HEARING ..........................................................................2	  Air-Conduction .................................................................................................2	  Bone-Conduction ..............................................................................................2	  THE AUDITORY BRAINSTEM RESPONSE .....................................................5	  Physiology of the ABR and its Generators .....................................................5	  Stimuli ................................................................................................................7	  Rate and Intensity ............................................................................................................... 8	  Effect of Sleep ....................................................................................................8	         Detection and Measurement of a Response ....................................................9	  Electroencephalogram ......................................................................................................... 9	  Averaging and Artifact Rejection ....................................................................................... 9	  Noise and Signal-to-Noise Ratio ...................................................................................... 10	  Response Amplitude and Latency .................................................................................... 11	  Visual Replicability .......................................................................................................... 11	  Morphology....................................................................................................................... 11	  Electrode Montage and Two-Channel Recordings ........................................................... 12	  Occlusion Effect ..............................................................................................13	  	   v	  Threshold Measurement ................................................................................13	  Estimating Hearing Loss with the ABR .......................................................14	  THE AUDITORY STEADY-STATE RESPONSE ............................................16	  Anatomy and Physiology of the 80-Hz ASSR and its Generators ..............16	  Stimuli ..............................................................................................................18	  Type .................................................................................................................................. 18	  Modulation Frequency (rate) ............................................................................................ 19	  Carrier Frequency ............................................................................................................. 21	  Stimulus Intensity and Single vs. Multiple Stimuli .......................................................... 21	  Subject Parameters : Age and Sleep .............................................................23	  Measuring and Detecting a Response ...........................................................24	  EEG ................................................................................................................................... 24	  Noise and Signal-to-Noise Ratio ...................................................................................... 25	  Phase and Phase Coherence .............................................................................................. 25	  Response Amplitude and the F-Ratio ............................................................................... 26	  Stopping Criteria ............................................................................................................... 27	  Electrode Montage ..........................................................................................27	  Two-Channel Recordings ...............................................................................28	  Stimuli and ASSR Systems ............................................................................28	  Noise Measurement .......................................................................................................... 29	  Threshold Measurement ................................................................................30	  ASSR vs. ABR .................................................................................................32	  MATURATION .....................................................................................................36	  Air-Conduction ...............................................................................................36	  Bone-Conduction ............................................................................................37	  Air-Bone Gap ..................................................................................................39	  Stimulus Conduction Considerations ...........................................................41	  Artifact .............................................................................................................................. 41	  	   vi	  EARLY HEARING ...............................................................................................43	  Early Hearing Screening Programs ..............................................................44	  STUDY RATIONALE ..........................................................................................47	  METHODS AND MATERIALS ..........................................................................48	  Stimuli ..............................................................................................................50	  Calibration ......................................................................................................51	  Recording ........................................................................................................52	  Procedure ........................................................................................................53	  DATA ANALYSIS .................................................................................................57	  RESULTS ...............................................................................................................58	  Normal Levels .................................................................................................59	  The Air-Bone Gap ..........................................................................................62	  Difference Scores ............................................................................................67	  DISCUSSION .........................................................................................................68	  Normal Hearing ASSR Thresholds at 500 Hz ..............................................68	  500 Hz ASSR Thresholds in Infants with Conductive Hearing Loss .........69	  500 Hz Accuracy of Classification and Time to Test ...................................70	  Overlap of 500 Hz Thresholds Between Groups ..........................................71	  500 Hz ABR and ASSR Threshold Differences ...........................................75	  CONCLUSION AND CLINICAL APPLICATIONS ........................................78	  FUTURE RESEARCH AND LIMITATIONS ...................................................80	  REFERENCES ......................................................................................................81	  APPENDIX ............................................................................................................98	     	   vii	  List of Tables  Table 1.1. A comparison of the Rotman MASTER Research system and the Clinical Biologic MASTER II system’s stimuli options. .................................................................................. 29  Table 1. 2. A summary of mean and maximum ASSR Thresholds for Normal Hearing Infants in dB HL from the literature. .................................................................................................... 31  Table 2.1. Sample size for 500 and 2000 Hz  air- and bone-conduction stimuli using ASSR and ABR. ..................................................................................................................................... 48	   Table 2.2. 500 Hz Normal Hearing Group: Immittance (tympanometry) and TEOAE screening results for individual participants. ......................................................................................... 49	   Table 2.3. 500 Hz Conductive Hearing Loss Group: Immittance (tympanometry) and TEOAE screening results for individual participants. ........................................................................ 50	   Table 2.4. Normal Hearing and CHL: Air- and bone-conduction 2000 and 500 Hz ASSR and ABR threshold measures of central tendency and dispersion. .............................................. 58	   Table 2.5. All participants: bone-conduction 500 Hz ASSR and ABR threshold measures of central tendency and dispersion. ........................................................................................... 59	   Table 2.6. The cumulative percentage of responses present for AC and BC ABR and ASSR at 500 Hz and 2000 Hz. Normal levels for each condition are shown in bold and represent the intensity at which more than 90% of subjects demonstrated a response. ............................. 60	   Table 2.7. Independent samples t-test comparing the 500 Hz air-bone gap for normal and CHL groups. ................................................................................................................................... 64	   	   viii	  Table 2.8. Independent Samples T-tests: comparing 500 Hz AC ASSR thresholds between groups and BC ASSR thresholds between groups. ............................................................... 67	   Table 2.9. Air- and bone-conduction ASSR-ABR mean difference scores and measures of dispersion. ............................................................................................................................. 67	   Table 2.10. A summary of 500 Hz air- and bone-conduction ASSR “normal levels” from the literature. ............................................................................................................................... 68	   Table 2.11. Accuracy of categorization for 500 Hz AC ASSR using different “cut-off criteria”. Percentage of error is calculated for each group. .................................................................. 70	   Table 2.12. Accuracy of categorization for 500 Hz AC ASSR using different “cut-off criteria”. Percentage of error is calculated for each group. .................................................................. 70	   Table 2.13. Normal Hearing Group: 500 Hz ASSR (dB HL) and ABR (dB nHL) mean thresholds converted to dB ppe SPL and dB re:1µN. ........................................................... 76	     	   ix	  List of Figures  Figure 1.1. A graphical representation of Intelligent Hearing Systems 500 Hz brief-tone ABR stimulus in the time (left) and frequency (right) domain. ....................................................... 7	   Figure 1.2. A graphical representation of AM2 stimuli for carrier frequencies of 500, 1000, 2000 and 4000 Hz. ......................................................................................................................... 19	  	  Figure 2.1. Representative ABR response series to 500 Hz air- and bone-conduction stimuli.. . 55	   Figure 2.2. A comparison of 500 and 2000 Hz AC and BC ABR and ASSR thresholds by group. Normal levels for normal-hearing participants for air- and bone-conduction represents the intensity at which all subjects demonstrated a response and is shown by the dotted lines. .. 61	   Figure 2.3. A comparison of the 500 Hz air-bone gap for normal and CHL groups. The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses. ........................................................................................ 63	   Figure 2.4. A Comparison of AC and BC ABR (dB nHL) and ASSR (dB HL) thresholds at 500 Hz by participant. Participant numbers are shown on the x-axis. ......................................... 65	   Figure 2.5. A comparison of 500 Hz air- conduction and bone-conduction ASSR thresholds by group. The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses. .................................................................. 66	   Figure 2.6. A comparison of AC ASSR thresholds at 500 Hz using normal data from Casey and Small (2014). The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses. ............................................... 74	  	   x	  List of Symbols and Abbreviations  Abbreviation/Symbol Definition ABR Auditory brainstem response ABG Air-Bone Gap AC Air-conduction AM Amplitude modulated AM/FM Amplitude and frequency modulated ASSR Auditory steady-state response BC Bone-Conduction BCEHP British Columbia Early Hearing Program CHL Conductive hearing loss dB Decibel dB HL Decibels hearing level dB SPL Decibels sound pressure level dB ppe SPL Decibels peak-to-peak equivalent sound pressure level dB eHL Decibels estimated hearing level dB nHL Decibels normal hearing level EEG Electroencephalography or electroencephalogram F Fisher’s F ratio FFT Fast fourier transform Hz Hertz MLR Middle-latency response NH Normal hearing nV Nanovolt OAE Otoacoustic emissions OIHP Ontario Infant Hearing Program SNHL Sensorineural hearing loss µV Microvolt 	   xi	  Acknowledgements  I wish to thank my supervisor, Dr. Susan Small for her guidance and knowledge throughout this project. I wish to also thank the members of my committee, Drs. Navid Shahnaz and Anna Van Maanen for their comments and all their time spent reading this thesis. Thank you to Dr. David Stapells for your very useful comments on this thesis. Thank you to the Audiology Department at the Royal University Hospital – especially Dr. Charlotte Douglas – for the endless support during this process. I am incredibly grateful to Dan Black from dB Special Instruments for the hours and hours of troubleshooting and technical support. Without him, this project would not have happened. Also, thank you for my team of classmates for the support and encouragement through the program. I couldn’t have done this with a better group of people! Thank you to my family - words cannot express how thankful I am for your endless love and support through my degrees. I have one fantastic cheer squad. Last but not least, I would like to thank Alex Gascon for his support, patience and incredible words of encouragement. I couldn’t have done this without you.	   1	    INTRODUCTION   With the increased awareness and appreciation for the benefits of early identification of hearing loss, constant quality assurance and improvements to early hearing programs continue to be major driving forces for research on infant hearing. Within the context of the provision of universal early hearing programs, finding more efficient methods of evaluating hearing in infants is also becoming a priority. This project was undertaken within the framework of supporting early diagnosis of hearing loss in infants within a universal hearing program supported by public health (British Columbia Early Hearing Program [BCEHP], 2012)   . The goal was to add to the existing literature comparing the current gold-standard test method, the auditory brainstem response (ABR), to an alternative, but less established method, the auditory steady-state response (ASSR). The study has a very clinical focus that will be described later. The following sections will present background information that is considered relevant for the present study. As the primary topic of this study involves air- and bone- conduction testing using the auditory steady-state response and auditory brainstem response in the context of an early hearing program, the following major topic areas will be discussed: • Mechanisms for Hearing (Air- and Bone-conduction) • The Auditory Steady-state response • The Auditory Brainstem response • Early Hearing Programs and Early Identification    	   2	  MECHANISMS FOR HEARING  Air-Conduction  The typical or most common route for sound to travel into the auditory system is via air-conduction. Simplistically, sound travels through the air into the ear canal and reaches the tympanic membrane. The pressure of a sound wave causes the tympanic membrane to vibrate, which conducts the waves through the middle ear system to the fluid-filled cochlea. When the basilar membrane is displaced maximally by the travelling sound wave near the characteristic place for the frequency of the stimulus, stereocilia are deflected based on the amplitude of the stimulus, and hair cells may be depolarized. This results in the release of neurotransmitters that bind to the auditory nerve to initiate action potentials that carry this information through the brainstem to the ultimate destination in the auditory cortex. These action potentials leak electrical potential that can be measured by sensitive electrodes on the surface of the head. It is this process that allows us to use electrophysiological methods of testing hearing, like the auditory brainstem response and auditory steady-state response. For a more detailed description, see Pickles (2013).  Clinically, air-conduction testing is used to assess hearing through the most natural route into the system, and to determine the maximal degree of hearing impairment that is caused by a combination of pathologies in the external, middle and inner ear systems. In other words, air-conduction thresholds obtained through any method of testing will be affected regardless of the type of hearing loss. When used in conjunction with bone-conduction testing, it is possible to determine contributions from the external and middle ear in isolation.   Bone-Conduction  Bone-conduction hearing is arguably more complicated than air-conduction hearing. By way of vibrating the temporal bone, sound is able to bypass the outer- and middle-ear systems. As Dauman (2013) describes, despite the numerous mechanisms in bone-conduction hearing, they all lead to a final common pathway - the basilar membrane. There may be several routes through 	   3	  which the vibration from a bone oscillator stimulates the system, but regardless of the route taken, the travelling wave eventually reaches the basilar membrane via bone-conduction; this initiates the process of hearing. There have been several proposed mechanisms for bone-conduction hearing that will be briefly reviewed below. There is not complete agreement in the field for all of these proposed mechanisms, and it is beyond the scope of this thesis to discuss this evidence in depth. For a more complete review, see Dauman (2013) and Stenfelt and Goode (2005a).  (1) Sound radiated into the external ear canal: When the skull vibrates, the sound pressure is radiated in the air. This results in sound being transmitted through the external ear canal in the same way as air-conducted sound. This mechanism has been termed “osseotympanic”(Azzena et al., 1995; Santarelli et al., 1995).While the cartilaginous portion of the ear canal is responsible for most of the sound radiation for the low frequencies, this mechanism does not appear to be effective in the radiation of high frequency sounds (Stenfelt & Goode, 2005; Khanna et al., 1976). The osseotympanic mechanism is responsible for the occlusion effect where sound pressure in the external ear canal is elevated below 1000 Hz when the ear canal is occluded (John, Dimitrijevic, & Picton, 2001; John, Dimitrijevic, van Roon, & Picton, 2001; John & Picton, 2000a; Rickards et al., 1994; Stenfelt & Goode, 2005a, 2005b). When the ear canal is unoccluded, the majority of the vibrations will be leaked out of the external ear and will not be transmitted. Researchers have argued that this mechanism is not a dominant one for bone-conduction hearing in young infants (ages 0-7 months) as the occlusion effect is negligible until 10-22 months of age (John & Picton, 2000b; Small & Hu, 2011).  Inertia of the (2) middle-ear ossicles and (3) cochlear fluids: The vibration of the temporal bone of the human skull that reaches the bony labyrinth can propagate an inertial force to the fluids in the cochlea (Stenfelt, 2011; Stenfelt & Goode, 2005a). The inertial force displaces the fluids in the cochlea and results in basilar membrane displacement. Similarly, the temporal bone vibration can propagate an inertial force to the middle ear; this vibrates the tympanic membrane and middle ear ossicles (the stapes) which transfers the force to the oval window, and finally, fluid displacement within the cochlea occurs. This should be noted as it is overly simplistic to suggest 	   4	  that bone-conduction isolates the inner ear. It is more accurate to claim that bone-conduction tests primarily the inner ear with small contributions from the outer and middle ear.   (4) Compression of cochlear walls: The travelling wave is suggested to force the bones of the skull to undergo compression and distention such that the structural change of the bone affects the fluid in the cochlea via a pressure gradient. Compression of the cochlear walls results in fluid travelling toward the round window, and the opposite is true of distention, which then stimulates the basilar membrane (John, Purcell, Dimitrijevic, & Picton, 2002; Stenfelt & Goode, 2005a, 2005b; Vander Werff, Brown, Gienapp, & Schmidt Clay, 2002). (5) Pressure transmission from the cerebrospinal fluid: The transmission of the sound wave through cerebral spinal fluid was investigated using animal studies (Herdman, Picton, & Stapells, 2002; Vander Werff et al., 2002). It was thought that the transmission did not necessarily have to go through the skull bones, but could be transmitted by the fluids and soft tissues within (John, Dimitrijevic, van Roon, et al., 2001; Stenfelt, Hato, & Goode, 2002; Stenfelt, Wild, Hato, & Goode, 2003). Electrophysiological studies in humans undergoing surgery has confirmed this as well as animal studies (Dauman, 2013).  Clinically, bone-conduction thresholds are primarily used to determine the type of hearing loss, to assess the need for medical intervention, and in some cases to predict candidacy for bone-conduction hearing devices. As was discussed above, it is clear that mechanisms for air-conduction and bone-conduction hearing are different. For this reason, studying air- and bone-conduction responses separately is warranted.  The following sections will provide background knowledge regarding the test methods used for this study using both air- and bone-conducted stimuli.      	   5	  THE AUDITORY BRAINSTEM RESPONSE   The auditory brainstem response is an electrophysiologic response to transient stimuli that is characterized by a series of five identifiable waves (Cone-Wesson, Dowell, Tomlin, Rance, & Ming, 2002; Picton, Dimitrijevic, & John, 2002). Clinically, the response is recorded as a “far-field” response, meaning it is recorded at a distance from its generators and is the result of the activity of many neurons. This electrical current generated by the action potential must travel through brain tissue, fluid, the skull and skin to be measured by the electrodes so recording equipment must be sophisticated and powerful enough to detect a small signal in the presence of noise. The ABR has several clinical applications including threshold measurement in populations that are impossible or difficult to test behaviorally, diagnosis of auditory neuropathy spectrum disorder, and detection of pathologies in the auditory pathway from the inner ear to brainstem (Starr & Hamilton, 1976). The ABR can be recorded in infants as early as 28 weeks gestational age (though this is not recommended until at least 37 weeks gestational age) (Perez-Abalo et al., 2001; Picton, John, Purcell, & Plourde, 2003; Rodriguez, Picton, Linden, Hamel, & Laframboise, 1986) and is the current “gold-standard” measure for threshold estimation in infants too young to complete behavioural testing (BCEHP, 2012; Joint Committee on Infant Hearing, 2007). It can be evoked using both air- and bone-conducted stimuli, making it particularly useful to assess both the degree of hearing loss and also the type (i.e., normal hearing sensitivity, sensorineural hearing loss, conductive hearing loss or mixed hearing loss).   Physiology of the ABR and its Generators    The ABR is a brainstem response with components that are evoked within approximately 2-15 milliseconds following the presentation of a stimulus. An electrical field is measured when an action potential is sent from the sensory neurons of the inner ear up the auditory nerve and is propagated up the brainstem. Specifically, the positive current leakage from the extracellular excitatory post-synaptic potential is being measured by electrodes recording the ABR. This is described in more detail in Picton (2011) and Luck (2005).  	   6	   Studies investigating ABR generators began with animal testing (for example, Linden, Campbell, Hamel, and Picton (1985)’s experiment with cats), and eventually, research was done with humans. The following are the generally accepted generators for the human ABR according to wave, but it is important to understand that each wave may have contributions from multiple areas and so the precise generators are not so clear. Wave I has been shown to be generated by the distal portion of the 8th nerve and occurs simultaneously with the initial negative deflection of the compound action potential. This can be partially used to judge cochlear function (Azzena et al., 1995; John, Dimitrijevic, & Picton, 2001; Lins et al., 1995; Lins & Picton, 1995; Maurizi et al., 1990; Plourde, Stapells, & Picton, 1991; Rickards et al., 1994; Santarelli et al., 1995). Wave II is likely generated by the proximal 8th nerve (Moller & Jannetta, 1981, 1989, 1983a,b; Moller et al., 1982; Moller et al., 1995). It is not consistently recorded in newborns, likely because the shortened length of the auditory nerve results in a wave I/II complex being recorded (Moller & Jannetta, 1981, 1989, 1983a,b; Moller et al., 1982; Moller et al., 1995). Wave III has several primary generators including the cochlear nucleus, trapezoid body, superior olivary complex (Starr & Hamilton, 1976, Moller et al., 1983). Wave IV is likely generated in part by the superior olivary complex, but is not entirely clear due to the crossing of auditory fibres in this area (Moller et al., 1995, Starr & Hamilton, 1976). Last, but certainly not least, wave V is generated by both the ipsilateral but primarily contralateral lateral lemniscus and inferior colliculus (Moller et al., 1995, Starr & Hamilton, 1976).  This knowledge of ABR generators has lead to many site of lesion applications in clinical practice (Starr & Hamilton, 1976). It has also been used to confirm a type of loss based on the latency of the responses. For example, in a conductive loss, one would expect an overall increase in latency for each wave, whereas a brainstem lesion would result in interpeak latency increases depending on its location.     	   7	  Stimuli   Stimuli used to evoke an ABR can range from clicks, to chirps to brief-tones. The preferred stimulus is one that is brief in duration, that produces synchronous neural activity, and that has energy concentrated at a specific frequency. This synchronous activity of the neurons then constructively contributes to response amplitudes so that the recording equipment may easily measure them. Furthermore, stimuli that contain energy more concentrated at a specific frequency and that activate a smaller, more precise area of the basilar membrane (i.e., tones) is preferred to clicks, mainly due to their ability to assess frequency-specific hearing sensitivity (John, Dimitrijevic, van Roon, et al., 2001; John & Picton, 2000a; S. J. Kramer, 1992; Lins et al., 1995). For the purposes of this thesis, background on brief-tone stimuli will be the focus as these are the recommended stimuli for threshold estimation in infants and are the stimuli used in this study.  Brief-tone stimuli are typically presented with alternating polarity in an effort to reduce stimulus artifact due to its cancelling effect. The BCEHP uses the Intelligent Hearing Systems (IHS) SmartEP to generate 5 cycle Blackman windowed brief tones or 2-1-2 cycle linearly-gated tones (BCEHP, 2012). Figure 1.1 below illustrates the 500-Hz stimulus used in the present study.    Figure 1.1. A graphical representation of Intelligent Hearing Systems 500 Hz brief-tone ABR stimulus in the time (left) and frequency (right) domain.     	   8	  Rate and Intensity  The chosen stimulus repetition rate is an important parameter to consider, especially when recording in infants. Maturation of the auditory system accounts for the difference in recommended rates for infants when compared to adults; synaptic inefficiency and incomplete myelination have been proposed explanations for this infant-adult difference (John & Picton, 2000b; Vander Werff et al., 2002). As the rate of stimulus presentation increases, latencies increase. This is most pronounced for wave V compared to the other ABR waves (John & Picton, 2000b; Vander Werff et al., 2002), and increases in latency are larger for infants than for adults (John et al., 2002). BCEHP recommends a rate of 39.1/second for infants. This is recommended to maximize efficiency of the recording session while maintaining the ability to record a reliable ABR.  Intensity is an additional parameter that has an effect on the ABR response. An increase in stimulus intensity typically results in a decrease in wave V latencies and an increase in amplitude. This is one of the features of the ABR that allows wave V to be tracked down to threshold for threshold estimations (Cone-Wesson et al., 2002; Herdman et al., 2002).  Effect of Sleep  Campbell et al., (1986) showed that sleep does not significantly affect the morphology of the ABR waves. This was a fortunate finding as threshold ABR recordings in infants are largely dependent on sleep. Because the response amplitudes are very small and require that myogenic and physiologic noise levels are sufficiently low, it is only possible if the infant is relaxed and asleep.  Picton et al. (2002) published a very informative study regarding average sleep time for sedated (children) and non-sedated (infant) assessments. They found that sedated assessments allowed for an average of 58 minutes of testing time with 7.6 threshold measures obtained and non-sedated assessments averaged 49 minutes of sleep with 6.2 measures obtained. This supports the idea that an efficient protocol must be used to obtain the most information about hearing in one session.    	   9	  Detection and Measurement of a Response   Electroencephalogram  In electrophysiologic assessments, we are working with the assumptions that the signal (or response) is time-locked to the stimulus and that the noise is somewhat random. Using these assumptions, it is possible to record the electroencephalogram (EEG) within a specified time-window after the presentation of a stimulus, and with processing such as averaging, artifact rejection, and filtering, these responses and detect a response that is differentiated by the noise.  Averaging and Artifact Rejection  Common mode rejection in the differential amplifiers is key to recording an ABR. This allows for the rejection of a signal that is in phase and is present at both recording sites (i.e., inverting and non-inverting). A signal that is in phase and recorded at both electrodes is assumed to represent unwanted noise. Once this noise energy reaches the two electrodes in phase, one electrode inverts the signal (inverting electrode), and the signals from both electrodes are averaged. This results in a “cancelling out” of the energy common to both electrodes. The end result is an average containing primarily the signal (the ABR) without the noise (Thorton, 2007; Elberling & Don, 2007). Artifact rejection is also a method that aids in extracting the target signal by rejecting signals with amplitudes outside of the range that is expected from the response. BCEHP suggests setting artifact rejection to +/- 25 microvolts (BCEHP, 2012). Filters can then be used to extract the components of interest. Optimal filter settings have the ability to minimize unwanted, nonphysiologic noise from entering the recorded averages. To avoid “throwing the baby out with the bathwater”, it is also necessary to ensure that what is being filtered out is not a response. For this reason, BCEHP recommends a band-pass filter to be set from 30-1500 Hz, which will avoid the response region for the ABR (BCEHP, 2012). If the low-filter setting is too high, wave V amplitude will be decreased thereby making it harder to detect, leading to an elevated threshold.  	   10	  Noise and Signal-to-Noise Ratio  Residual noise is a measure that is not available on all recording equipment. The Intelligent Hearing System (IHS) SmartEP, however, does record this measure and it is a measure that is required by the BCEHP (BCEHP, 2012). Residual noise is used to determine the amount of noise in an averaged recording. This is particularly important when attempting to detect a response (wave V) in less than optimal conditions (i.e., in a noisy environment or with a noisy participant). When residual noise is too high, the relatively small wave V may be “buried” in the noise and undetectable, despite being present. When the residual noise is reduced, and kept below an acceptable level (BCEHP requires <0.08 nV), a clinician may be confident in determining a response is absent if it is not visually detectable.   As per the BCEHP program (BCEHP, 2012), residual noise measurement is absolutely required when using the ABR as a threshold measurement. The program requires that for a response to be considered absent the wave must be visually “flat” (that is, there are no identifiable peaks) and the residual noise value must be less than 0.08 nV. The use of band-pass filters assist in managing non-physiologic noise (i.e., electrical noise). In the present study, a bandpass filter of 30-1500 Hz recommended by BCEHP was used (BCEHP, 2012).  The signal-to-noise ratio (SNR) can also be improved by increasing the number of replications recorded and averaged, up to a point. The BCEHP recommends that at least two replications of 2000 sweeps be recorded to determine if a response is present or absent at a particular intensity. They caution that there may be no benefit to recording more than four separate replications at the same intensity.  SNR and RN measures were used in the present study in the interpretation of ABR responses to determine their presence or absence. The IHS system provides real-time SNR and RN estimations of signals being acquired, and these measures are calculated from the signal and noise estimates using a split-sweep technique. This technique splits sweeps into two separate buffers according to their order of acquisition; the even buffer and the odd buffer. The buffers are added and subtracted from each other to estimate the amount of noise present in the average 	   11	  waveform as well as the size of the response present. The IHS SNR measure differs slightly from the standard deviation ratio (SDR) as described in Picton et al., (1983); it is equivalent to the SDR divided by two.    Response Amplitude and Latency  Response amplitude varies with intensity and hearing sensitivity. That is, near threshold, wave V amplitudes will be smaller than at suprathreshold. Typical amplitudes range from 0.1-0.5 microvolts (Cone-Wesson et al., 2002; Herdman et al., 2002; Stapells & Ruben, 1989).    Response latency also varies with stimulus intensity. While the absolute latency of the ABR waves is clinically utilized during neurodiagnostic and auditory neuropathy spectrum disorder evaluations, the absolute latencies of wave V is not strictly monitored for threshold evaluations. At most, approximate latencies are used to give the clinical an idea of where to look for the response. The exception to this rule is when recording responses to bone-conducted stimuli when latency asymmetries between channels are used. See “Two-Channel Recordings” section for more detail. In general, wave V latency tends to decrease with an increase in stimulus intensity. In newborns, the change is approximately 30-40 microseconds per 1 dB increase (Perez-Abalo et al., 2001; Picton, John, Purcell, et al., 2003).  Visual Replicability  Threshold estimation using the ABR ultimately relies on the visual interpretation from an experienced clinician. As programs such as BCEHP describe, the visual replicability of a response detected by an experienced clinician “trumps” other measures like SNR (BCEHP, 2012).   Morphology   The following excerpt is taken from the BCEHP (2008) training manual for audiologists which describes the expected morphology for Wave V: 	   12	  Tone Pip/Threshold: The morphology of the ABR to tone pips is very different from that seen in otoneurologic ABR testing with click stimuli. The waveform changes are most marked near threshold.  1. Typically, the earlier waves of the ABR are absent, and the response is a slow and late V-V’ negative-going transition.  2. There may be no wave V at all, but only a negative V’ peak.  3. There may also be a positive-going deflection following V’, at the end of the analysis interval.  4. The tonepip response at 2 kHz usually shows a wave V that is more clearly defined and sharper than at 0.5 kHz, and the 4 kHz response can be quite similar to conventional click responses.  5. At moderate sensation levels there may even be earlier ABR waves (especially at 2000 and 4000 Hz).   These criteria were used in the present study to aid in response detection.  Electrode Montage and Two-Channel Recordings  The electrode montage that is recommended for two-channel infant ABR threshold evaluation uses four electrodes. The non-inverting electrode is placed at the high forehead as close to Cz as possible, ground electrode is placed to one side of the non-inverting electrode maximizing the distance between the electrodes, and two inverting electrodes are placed on each mastoid (BCEHP, 2012).   ABR recordings can be acquired with one or two channels. The advantage of two-channel recording is best seen using bone-conduction stimuli. The seminal paper on the topic by Stapells and Ruben showed that when both ipsilateral and contralateral responses are recorded in young infants (under one year of age) an asymmetry exists between channels. Stapells and Mosseri then studied the maturation of this asymmetry and showed that it starts diminishing around 9 months of age. In clinical practice, however, many clinicians report the asymmetry being present at one year of age. This asymmetry has been used in infant electrophysiological hearing assessments using ABR to determine the responding cochlea and is most recognized for its ability to eliminate the need for masking (Edwards, Durieux-Smith & Picton, 1985; Foxe & Stapells,1993; Stapells & Mosseri, 1991; Stapells & Ruben, 1989). Specifically, the response from the ipsilateral channel tends to be earlier in latency and larger in amplitude compared to the contralateral channel’s response. In situations where masking is either impossible or difficult, 	   13	  this provides an excellent method of determining the laterality of a response. However, when ipsi/contra asymmetries are not clear, masking is required.  Occlusion Effect  The occlusion effect has been noted predominately in the low frequencies (250, 500 and 1000 Hz) and is a concern for ABR bone-conduction testing in infants (Dirks & Swindeman, 1967; Elpern & Naunton, 1963; Herdman & Stapells, 2003; John, Dimitrijevic, & Picton, 2003). For example, adult data from Small & Stapells showed significant occlusion effects at 500 and 1000 Hz. If an occlusion effect is present and not compensated for, it is possible that infants’ thresholds may be underestimated, leading a clinician to think that the infant’s hearing is better than it is (Herdman & Stapells, 2003). These data are only available for the ASSR and show that the occlusion effect is only present after seven months of age (discussed in the bone-conduction section). Consequently, it was not necessary to remove insert phones when testing bone-conduction for the infants in this study who were between 0-6 months of age.  Threshold Measurement  The application of the ABR for threshold measurement has been utilized clinically for decades. The process of completing a threshold assessment using ABR should be performed by clinicians very skilled in this area as interpretation requires a certain level of expertise. ABR thresholds have been compared to gold-standard behavioral audiometry and have been shown to provide an accurate estimate of behavioral thresholds in both normal hearing patients and in those with hearing loss; that is behavioral thresholds and tone-ABR thresholds are highly correlated and so by knowing one, we can infer the other (Picton, John, Dimitrijevic, & Purcell, 2003; Vander Werff, Prieve, & Georgantas, 2009). As discussed earlier in this section, threshold measurement is accomplished by identifying the lowest intensity at which a wave V is present and identifiable, and the level below at which no response is measured. Only when both of these levels are identified can a clinician confidently identify the threshold. Once thresholds are identified, clinicians are then able to determine if a hearing loss is present and if intervention is required. 	   14	  Estimating Hearing Loss with the ABR   As the ABR is capable of providing a reasonable estimate of hearing thresholds using both air- and bone-conducted stimuli (Stapells et al., 1995, Gorga et al. 1993, Lee et al. 2007; Ribeiro & Carvallo, 2008, Rance, Tomlin & Rickards, 2006; Kramer, 1992, Cone-Wesson & Ramirez, 1997, Hatton et al., 2012), it can be used to determine both the degree and type of hearing loss (Hooks & Weber, 1984; John, Brown, Muir, & Picton, 2004; Stapells & Ruben, 1989; Stuart, Yang, & Stenstrom, 1990). That is, how severe the hearing loss is, and whether it is a conductive, sensorineural or mixed. As explained below, bone-conduction assessment plays a crucial role in differentiating the type of loss.  The bone-conduction ABR is a means of isolating inner ear function from outer and middle ear systems and is particularly informative when trying to determine if an elevated air-conduction threshold is a result of a pathology of the middle or inner ear. In most cases, for infants, this is often the difference between identifying a temporary versus permanent hearing loss. Mechanisms of air- and bone-conduction hearing are reviewed in the air-conduction and bone-conduction” sections and ABR profiles corresponding to different hearing statuses are explained below:  • Normal Hearing ABR Profile : In order to differentiate hearing loss from normal hearing, normal hearing levels for ABR needed to be established. Based on the collection of normal hearing studies (e.g., Foxe & Stapells, 1993, Stapells, Gravel & Martin, 1995, Stapells, 2000) early hearing programs may determine “normal levels” that are used to indicate whether further testing is warranted and whether a hearing loss is present. The choice of normal levels to use may vary depending on the program. For example, BCEHP normal air-conduction levels are 35, 30, 30, and 25 dB nHL for 500, 1000, 2000 and 4000 Hz, respectively, and bone-conduction levels of 30 dB nHL for 500 and 2000 Hz (BCEHP, 2012), while the Ontario Infant Hearing Program (OIHP) uses normal air-conduction levels of 40, 35, 30, and 25 dB nHL for 500, 1000, 2000 and 4000 Hz respectively and bone-conduction levels of 30 dB nHL for 500 and 2000 Hz. These values will vary based on the target hearing loss specified by the program (OIHP, 2008; BCEHP, 2012). 	   15	  • Sensorineural Hearing Loss ABR Profile: Elevated air- and bone-conduction ABR thresholds are indicative of a sensorineural hearing loss. Tone-ABR thresholds and behavioural thresholds have been shown to be highly correlated in infants and young children with sensorineural hearing loss (Stapells, 2000b; Vander Werff, 2009; Stapells, 1984). • Conductive Hearing Loss ABR Profile: (i) Elevated air-conduction thresholds with normal bone-conduction thresholds, and (ii) presence of an “air-bone gap” (ABG) (a gap between responses to air-conducted versus bone-conducted stimuli) confirms the presence of a conductive component to the hearing loss. For the ABR, an ABG must be >10 dB to be considered clinically significant as infants with normal hearing and no conductive hearing loss exhibit some degree of an ABG, especially at low frequencies (Vander Werff, Prieve & Georgantas, 2009). Latency, as mentioned above, is not a useful measure with ABR threshold assessments. It has been suggested for a number of years that ABR absolutely latency delays may be useful in estimating the conductive component of a loss, but D. Swanepoel, Hugo, and Roode (2004) showed that this leads to significant individual errors and is therefore not appropriate. • Mixed Hearing Loss ABR Profile: (i) Elevated air-conduction thresholds, (ii) elevated bone-conduction thresholds, and (iii) presence of a clinically significant ABG. This suggests the pathology is both conductive and sensorineural in nature.     	   16	  THE AUDITORY STEADY-STATE RESPONSE  The auditory steady-state response (ASSR) is an electrophysiological following response that can be used in frequency-specific evaluation of the peripheral (and partially central) auditory system using air- or bone-conducted stimuli. The steady-state response was first assessed in the visual system (Jeng, Brownt, Johnson, & Vander Werff, 2004) and later in the auditory system (Picton, Dimitrijevic, Perez-Abalo, & Van Roon, 2005). Since its inception around the 1960’s (Small & Stapells, 2005), researchers have taken a particular interest in its clinical utility, likely due to its efficiency, ease and objectivity of interpretation. The fundamental difference between the two responses is that the ABR is a transient response (that changes over time), while the ASSR is a response to a continuous stimulus that remains stable over time that is able to be presented using single or multiple stimuli simultaneously (Picton et al., 2003).   There are two extensively researched ASSRs; the 40 and 80-Hz ASSR. These responses are elicited to different modulation rates (or frequencies), and thus, their generation sites (e.g., Herdman et al., 2002; Galambos et al., 1981). Early research tended to focus on the 40-Hz ASSR, while more recent focus has been on the 80-Hz ASSR. In the intervening decades, researchers established that the 40-Hz ASSR was not particularly useful in determining auditory thresholds for infants and children due to age and sleep effects (see subject factors section), and thus, the focus of the research community changed to the 80-Hz ASSR (Han, Mo, Liu, Chen, & Huang, 2006; Petitot, Collett, & Durrant, 2005; Rance & Tomlin, 2006; Van Maanen & Stapells, 2005). The sections to follow will describe the ASSR technique in further detail, with the focus being on the 80-Hz ASSR.  Anatomy and Physiology of the 80-Hz ASSR and its Generators   When measuring an auditory response and making clinical judgments based on this response, it is useful to have knowledge about where the response is coming from (what exactly we are measuring) and how the response is evoked (how we are measuring it). The ASSR can be evoked by a continuous periodic stimulus or a transient stimulus that is repeated at a fast rate (Regan, 1982; Picton et al., 2003).  As summarized in Picton et al., (2003), neurons low in the auditory 	   17	  system respond temporally (follow the envelope of the stimulus), where as higher neurons respond to particular stimulus rates by increasing their rate of discharge. The higher in the system, the lower the range of modulations that the neurons respond to synchronously. Cortical neurons stop synchronous firing to the envelope at rates above 50-100 Hz, whereas brainstem neurons can follow modulations above 1000 Hz. This makes the 80-Hz ASSR within the range of modulations to which a brainstem neuron can respond. When a pure-tone is amplitude modulated by a certain modulation frequency, the stimulus does not contain energy at that modulation frequency, but instead, activates the auditory system (via basilar membrane displacement) according to the carrier frequency. These characteristics are described in more detail in sections to follow. The non-linear characteristics of the cochlea and outer hair cells (compression and rectification) cause the neural response to carry energy at the modulation frequency and the carrier frequency (CF), even though the stimulus itself does not contain energy at the modulation frequency. This is because the inner hair cells cause the release of a neurotransmitter only when they move in one direction. This then causes the stimulus to undergo “half-wave” rectification that provides energy at the modulation frequency (Picton et al., 2003).   Popular hypotheses on ASSR mechanisms include the superimposition (or overlap) theory and the intrinsic rate theory. The superimposition theory simply assumes that the ASSR is a linear superimposition of the transient MLR at a fast rate (Galambos et al., 1981; Gutschalk et al., 1999). The intrinsic rate theory suggests that there are other mechanisms involved that differ from those utilized in transient responses. Specifically, mechanisms involving the refractory or recovery periods of neurons seem to play a role, such that these neurons have an “intrinsic rate” of firing that acts as a tuned neural oscillator that responds best to an input tuned to 40-Hz. (Leigh-Paffenroth & Fowler, 2006; Small, Hatton, & Stapells, 2007; Small & Stapells, 2006). The superimposition theory was first queried by Galambos et al., (1981), who showed that steady-state responses were very prominent at stimulus rates near 40 Hz. They suspected that the 40-Hz ASSR was simply a superimposition of the transient middle-latency responses at a fast rate. In later studies using magnetoencephalography, the 40-Hz ASSRs were found to have generators in the supratemporal gyrus of the auditory cortex, which are similar to the generators for transient middle-latency responses (Gutschalk et al., 1999).  	   18	  Several years of research on both human and animal subjects have suggested different anatomical sites of generation for the 40-Hz and 80-Hz ASSRs. The 80-Hz ASSR is thought to have generators most closely related to the generators of the ABR; that is, although there may be some cortical contributions to the response, the 80-Hz ASSR is found to be primarily a brainstem response (Herdman et al., 2002; Picton, van Roon, & John, 2007; D. Swanepoel & Erasmus, 2007). Conversely, the 40-Hz response was found to be generated primarily from the cortex, with some contribution from the brainstem (Herdman et al., 2002, Kuwada et al., 2002).   Stimuli  Type  The most commonly utilized stimuli to elicit an ASSR are sinusoidally amplitude modulated, frequency modulated, and mixed modulated tonal stimuli. It is also possible to elicit an ASSR using clicks, chirps and brief tones (Brooke, Brennan, & Stevens, 2009). There is a classic tradeoff that has to be made when recording auditory electrophysiological responses to different stimuli. For threshold estimation, this becomes particularly important to understand the compromises made by selecting a certain stimulus and what effects this selection has on the frequency specificity and accuracy in measuring thresholds. The same tradeoff is encountered during ABR evaluations using clicks vs. tones (Gorga et al., 1989)   ; while clicks elicit larger responses that are easier to detect, they are also less frequency specific, resulting in less frequency-specific hearing information. Tones provide the opposite; much better frequency specificity, but much smaller response amplitudes (Gorga et al., 1989). The same is true of the ASSR. Stimuli that can elicit frequency-specific ASSRs include brief tones, amplitude modulated and frequency-modulated tones, mixed modulation tones, and exponential amplitude modulated tones (John et al., 2002; Ribeiro, Carvallo, & Marcoux, 2010). For threshold estimation in infants, John et al., (2004) showed that of the frequency specific stimuli (AM, MM and exponential AM), MM and exponential AM stimuli produced larger responses than AM stimuli, with exponential AM stimuli having the highest percentage of significant responses. They suggest that using AM2 stimuli will increase the reliability and efficiency of testing in infants, as there are no significant increases in amplitude above this exponential value. Another 	   19	  of the largest response amplitudes are elicited using AM/FM stimuli that add in phase. The issue here is that they may not add always in phase, so there is a risk in using these stimuli and is not the preferred stimuli of choice (Rodrigues, Lewis, & Fichino, 2010). For these reasons, as well as limitations of the recording equipment, the present study used AM2 tones, which contain spectral energy at the carrier frequency separated from the carrier frequency by the modulation frequency. This was used despite the fact that the majority of the existing literature presents thresholds to AM stimuli. See figure 1.2 below for a graphical representation of AM2 stimuli used in the present study.    Figure 1.2. A graphical representation of AM2 stimuli for carrier frequencies of 500, 1000, 2000 and 4000 Hz.  Modulation Frequency (rate)  The modulation frequency refers to how often per second the stimulus repeats or varies in amplitude and/or frequency. The most commonly used steady-state stimuli in the literature are sinusoidally amplitude-modulated tones. These stimuli contain spectral energy at the carrier frequency as well as energy in sidebins on either side of the carrier frequency corresponding to the modulation frequency (Picton et al., 2003). When viewing the stimulus in the frequency domain, sidebins will contain energy at a distance separated by the modulation frequency. For example, if the carrier frequency is 2000 Hz and the tone is 100% amplitude modulated at a modulation rate of 100, there will be energy at 2000 Hz, 1900 Hz and 2100 Hz (Picton et al., 	   20	  2003). The most commonly researched responses, the 40-Hz and 80-Hz ASSRs, are named for their modulation frequencies. The 40-Hz responses have the largest SNRs at modulation frequencies between 40-50 Hz, and between 80-100 Hz for the 80-Hz response (Cohen et al., 1991; Picton et al., 2003).   40-Hz vs. 80-Hz   When selecting which modulation frequency to use, one must consider how they differ in their population of interest. For example, effects of modulation rate differ for adults and infants. For adults, effect of modulation frequency on response amplitude appears to result in a decrease in response amplitude with increasing modulation frequency (Rickards and Clark, 1984). It has also been shown that amplitude measures of the 40-Hz ASSR are significantly smaller in infants than in adults (Suzuki & Kobayashi, 1984) and are unreliable in measuring infant thresholds (Stapells et al., 1988, Maurizi et al., 1990). Stapells et al., (1988) showed no consistent response amplitude peaks for the ASSR between modulation rates of 9-59 Hz for infants. Furthermore, the 40-Hz response amplitude decreased in infants and young children when sleeping (Aoyagi et al., 1993, Stapells et al., 1988, Maurizi et al., 1990). This presents a significant disadvantage for use in infants, as sleep tends to be a quieter participant state to reduce EEG noise (Cohen et al., 1991). Fortunately, 80-Hz ASSRs are easily recorded in sleeping infants, and are the recommended responses to measure hearing thresholds in infants (Hatzopoulos et al., 2010; Rance & Tomlin, 2006; Rodrigues & Lewis, 2010).  Newborn studies suggested optimal modulation frequencies to be used when recording the 80-Hz ASSR in the interest of minimizing recording time required to obtain thresholds. Optimal values of 72 Hz for 500 Hz, 85 Hz for 1500 Hz and 97 Hz for 4000 Hz were suggested for use by Rickards et al., (1994) in an infant study. Purcell and John examined 500 Hz and 2000 Hz carrier frequencies using different modulation rates. They found that for 500 Hz, the largest response was found between 80 and 90 Hz and between 90 and 100 Hz for a 2000 Hz carrier frequency. Evidently, some variability exists among studies. The present study used modulation frequencies of 78, 85, 93 and 101 for 500, 1000, 2000 and 4000 Hz, respectively (Van Maanen & Stapells, 2009). 	   21	  Carrier Frequency   The carrier frequency (CF) refers to the frequency at which most of the spectral energy is concentrated. For the ASSR, carrier frequency affects where the stimulus energy is concentrated, which area(s) of the basilar membrane are being activated, and response amplitudes. Clinically, it can be thought of as the “test frequency”. Similar to the ABR, common carrier frequencies used for ASSR are 500, 1000, 2000 and 4000 Hz.   For 40-Hz vs. 80-Hz stimuli, the effects of carrier frequency are different. Galambos et al., (1981) and Picton et al., (1987) showed that for 40-Hz ASSRs, the amplitudes are much larger for lower carrier frequencies. When evoked by single stimuli, the 80-Hz responses are similar in amplitude across frequencies. The 80-Hz response shows slightly larger amplitudes for carrier frequencies in the mid-frequencies (1000-2000 Hz) than for either higher or lower frequencies (John et al., 2001b).   Stimulus Intensity and Single vs. Multiple Stimuli   An increase in stimulus intensity typically results in an increase in steady-state response amplitude and a decrease in latency (Rodriguez et al., 1986)   , reaching a saturation point at high intensities (90 dB SPL) (Picton et al., 2003). This is similar to other evoked responses, like the ABR (Wolfe et al., 1978, Campbell, 1986).   One of the main advantages of the ASSR is its ability to record responses to both stimuli presented individually (single) as well as several stimuli presented simultaneously (multiple) (Regan, 1982). Multiple stimuli can be presented to one ear (monotic) or both ears (dichotic) simultaneously (Purcell & John, 2010; Small & Hu, 2011) and can be at least two to three times faster than using single stimuli (John et al., 2002b). When testing using multiple stimuli, each test frequency or carrier frequency (500, 1000, 2000 and 4000 Hz) is assigned a modulating frequency between 75 and 110 Hz for the 80-Hz response that are separated by at least a half-octave (Choi, Purcell, & John, 2011; Tlumak, Durrant, Delgado, & Boston, 2011). Individual modulating frequencies are required in order to keep the processing of these signals independent 	   22	  through the auditory system (Picton et al., 2003). It is then possible to analyze the responses to each of the individual frequencies in a multiple presentation by using Fourier analyzers for each stimulus carrier frequency (Regan, 1982, Picton et al., 2003). Details will be presented in the “Measuring and Detecting a Response” section.   When using multiple stimuli, the question arose whether stimuli, response amplitudes or thresholds would be affected compared to a single-stimulus presentation mode. Researchers were concerned about interactions such as masking effects, suppression and facilitation when presenting multiple stimuli (Picton et al., 2003). Interactions can be recognized by decreased response amplitudes when using multiple compared to single stimulus presentation (Reviewed in Picton, 2003, Mo & Stapells, 2008, Wood, 2009).      (J. Hatton & Stapells, 2011; Ishida, Cuthbert, & Stapells, 2011; Picton, John, Purcell, et al., 2003) Using an AM stimulus presented at multiple carrier frequencies, John et al., (1998) showed that in adults, significant interactions were avoided when (1) carrier frequencies differed by one octave or more with (2) modulation frequencies between 70 and 110 Hz, and amplitude was unaffected when (3) stimuli were presented at 60 dB SPL or less. More specifically, it was shown that higher intensities (>70 dB SPL) produced greater interactions than did lower intensities (35 dB SPL) (Herdman & Stapells, (2001).   In adults, Hansen and Small (2012) data suggested the potential for masking effects being present when low-frequency carrier frequencies are included. A subset of the participants with hearing loss had more accurate threshold measures for single frequency presentation when compared to multiple presentations for 2000 and 4000 Hz. Herdman & Stapells (2003) explored this issue and found no significant difference in thresholds to 2000 and 4000 Hz when presented as single vs. multiple stimuli. It was therefore concluded that there were no masking effects.  Mo, Zhang, Han, and Zhang (2011) tested adults with sensorineural mild-to-moderate hearing loss and determined that interactions for this population were not significantly greater than results for normal hearing adults at 80 dB HL. This suggests that interactions are no more likely for hearing-impaired compared to normal hearing groups at high intensities, so the same guidelines can reasonably apply to both groups. 	   23	  A fairly recent study showed that while adults show significant changes in response amplitudes between multiple stimuli above 80 dB HL, infants (age 6-38 weeks) show a significant reduction in response amplitude with the addition of stimuli at 60 dB HL. This suggests that interactions at the level of the brainstem or cortex do exist in infants and that they occur at lower levels of intensity than for adults (Korczak, Smart, Delgado, Strobel, & Bradford, 2012). Despite these findings, Hatton and Stapells (2011) also showed that even with the reduction in response amplitude from the interactions, the multiple-ASSR method remains more efficient than the single ASSR method as there was no significant difference between thresholds in single vs. multiple ASSR. They suggest that further research investigating the interactions for multiple ASSRs in infants with normal hearing and hearing loss is needed.  Subject Parameters : Age and Sleep    Understanding the literature on effects of age and sleep is particularly relevant for the current study as it involves testing infants in natural sleep. It is important to acknowledge the effects of a change in participant state, as it requires other modifications to be made in terms of the modulation frequency used.   As discussed earlier, the auditory steady-state response has been found to decrease in amplitude with sleep (Galambos, 1981), but only at modulation frequencies less than 70 Hz. Rickards et al., (1994) showed that better SNRs are found with modulation rates of 70-100 Hz when compared to 40-Hz during sleep. For these reasons, it is now generally accepted that the 80-Hz ASSR is a better choice when evaluating the auditory sensitivity of newborns during sleep (Aoyagi et al., 1993; Levi et al., 1993; Richards et al., 1994).  Furthermore, the modulation frequency that results in the largest amplitudes for adults (40-Hz) is not what is recommended for the largest responses in infants (80-Hz). Picton et al., (2003) suggested that this difference in response amplitudes may be a result of immaturities in the cortical regions of the brain of infants so that a primarily cortical response (40-Hz ASSR) may not be as large or as robust as in adults, while a primarily cortical response would be more 	   24	  developed in infancy. Measuring and Detecting a Response   As described in the Picton (2003) review paper, the ASSR is a measure of frequency components that remain constant in amplitude and phase over an infinite period of time. The response stability can also be measured over a defined time period that is longer than the duration of a single stimulus cycle, and that this may be considered sufficient to measure the (un)steadiness of a response (Picton et al., 2003). The ASSR is typically measured in the frequency domain from the EEG using a Fourier analyzer (Stapells et al., 1984) or a Fourier transform (Rickards and Clark, 1984). This analysis will result in energy at the frequency of the stimulus rate and its harmonics at fast presentation rates. Other components of the response are filtered out by the low-pass filters of the brain and the recording equipment (Picton et al., 2003).   EEG  Measuring auditory steady-state responses only became possible after appropriate averaging techniques were available. This was necessary as it is not possible to recognize an ASSR from the ongoing EEG, as the background noise is too great and responses are too small (Picton, 2003). Two important assumptions when averaging exist: that the response is time-locked to the stimulus, and that the noise is somewhat random. In the Geiser (1960) paper, it was discovered that most of the energy that was measured in the response was concentrated at the frequencies corresponding to the rate of stimulus presentation. Also, the amplitude of the ASSR decreased as rate of presentation increased. Effects of sleep decreased response amplitude significantly below 70 Hz (Cohen et al., 1991; Galambos et al., 1981; Jerger et al., 1986; Picton et al, 2003; Suzuki et al., 1994) but did not have this effect on the 80-Hz ASSR or the peak at 40-Hz (Cohen et al., 1991).   There are several measures that are commonly referred to when describing the evoked potentials. Those that are applicable to the ASSR are reviewed in sections to follow.   	   25	  Noise and Signal-to-Noise Ratio  In order to determine if a response is present or absent, an estimate of the amount of noise contaminating the recording is needed. Separating the signal (or response) from the noise is necessary because the ongoing EEG contains both the response and the noise (physiologic and otherwise) and it is not possible to visually extract the signal without the use of averaging techniques (Picton et al., 2003). Noise levels will decrease with increasing number of sweeps in a recording as well as with increasing modulation frequency (John et al. 2002, John & Picton, 2000), so ensuring an appropriate number of sweeps are recorded may help when the patient is noisy. One obvious way of increasing the SNR is to increase the intensity of the stimulus and use noise reduction and averaging techniques (John, Dimitrijevic et al., 2001). Researchers have questioned which modulation frequency would result in the best SNRs for infants and children. It has been argued that because noise is higher at 40-Hz, it follows that the 80-Hz ASSR would be the best choice for infants and children in terms of minimizing EEG noise (Picciotti, Giannantonio, Paludetti, & Conti, 2012; Vander Werff et al., 2009).   The present study used a noise criterion of <15 nV to determine that no response was present in an average. This is comparable to similar to other studies that used <11 nV as a noise criterion (i.e., Casey & Small, 2014; Small & Stapells, 2008; Van Maanen & Stapells, 2009,2010)  Phase and Phase Coherence  It has been shown that steady-state responses are less variable in phase across sweeps, whereas the phase noise tends to be more variable and random. In disentangling a true response from noise, phase can provide an idea as to whether the data is showing a true response, or if it is noise (a more variable phase) (Picton et al., 2003).  Methods of measuring phase can vary. Phase-coherence represents one possible measure. It is related to the SNR and is measured with phase-coherence2. The possible values range from 0.0 to 1.0. The closer that the value is to 1.0, the more likely a response is present and able to be 	   26	  separated from the noise. This method uses the FFT results to form a polar plot. The length of the vector represents the amplitude of the response and the phase/time delay is represented by the vector angle. This technique can be uses instead of the F-test (Brooke et al., 2009) and has been the chosen method for some studies (i.e., Swanepoel et al., 2008). Some systems do not directly use phase to determine presence or absence of a response, but rather focus on amplitude (e.g., Natus MASTER II). As phase and amplitude of responses are both contributors to overall size of a response, irrespective of what measure the system formally uses for analysis, it is indirectly taking both measures into account. (Picton et al., 2003, Rance, 2008).    Response Amplitude and the F-Ratio  Response amplitude can be quantified using measures like peak-to-peak amplitude, root-mean-square (rms) amplitude, and baseline-to-peak amplitude. The amplitude of the response is what is commonly used to determine if a response is present and the robustness of a response. This is the measure that was used in the current study. Using amplitude as the criterion for the presence or absence of a response is accomplished by determining if the response amplitude for a certain carrier frequency is significantly different than the noise at surrounding frequency bins. If there is a significant difference, a response is likely present (Picton et al., 2003; Korczak et al., 2012). This is then computed using the F-Ratio at 2 and 240 degrees of freedom. (Lins et al., 1996, Korczak et al., 2012, Picton et al., 2003). Using the typical confidence interval of p<0.05, this requires the amplitude of a response to be 1.75 times greater than the noise measured. An alpha level is set by the clinician, which determines when a response is considered significant. The present study considered a response present if p<0.05 and absent if the response was p≥0.5 and the noise value was <15 nV for a minimum of 10 sweeps.  Whether response amplitude or phase-coherence is used as the formal measurement, they have been shown to have similar sensitivity and specificity (J. L. Hatton & Stapells, 2013; Mijares, Perez Abalo, Herrera, Lage, & Vega, 2013; Papakonstantinou, Kollmeier, & Riedel, 2013).   	   27	  Stopping Criteria   Stopping criteria may involve measures of noise, F-ratios or even number of sweeps or length of recording. The present study used <15nV as the stopping criterion for EEG noise. Although several studies utilize a more conservative criterion <11nV (Nagashima et al., 2013; Van Maanen A, 2010; Van Maanen & Stapells, 2009), <15nV was used as a compromise to see that infant sleep time would allow for completion of testing. A minimum of ten sweeps was collected to increase the probability that the signal is stable. Also, four consecutive sweeps of a present response was required in order to be confident that a response is present and that we were not recording a false positive response. Casey and Small (2014) studied the effects of repeated testing on the reliability of the responses detected. They suggest that when eight consecutive sweeps were required to determine a present response, error rates decreased to 5%. While they suggest using eight consecutive sweeps, the present study compromised with four consecutive sweeps, as we believed this was more reasonable for testing infants with limited testing time.  Electrode Montage   Due to the clinical orientation of this study, we used current standard electrode placement required by The BC Early Hearing Program (2013). The typical electrode montage used for ASSR recordings is similar to that which is used for ABR. Four electrodes are used in two-channel recordings with the inverting electrode placed on the high forehead, the ground placed to one side of the inverting, and one non-inverting each on the lower portion of each mastoid. These recommendations are consistent with the findings of Van der Reijden and Colleagues (2005), which showed that in infants, electrode placement at Cz and the mastoid of the test ear references to Cz resulted in the best SNR. They discouraged the use of Cz and inion placement in infants as high SNRs were not found using this montage (Small, Smyth, & Leon, 2014).    	   28	  Two-Channel Recordings   Similar to ABR recordings, ASSR recordings can be acquired with one or two channels. The advantage of two-channel recordings that record both ipsilateral and contralateral responses can be seen when an asymmetry exists between channels. This asymmetry has been used in infant electrophysiological hearing assessments using ABR for decades and is most recognized for its ability to eliminate the need for masking (Stapells & Ruben, 1989). The current body of ASSR literature is sparse. Mijares, Baez, Cabrera, Perez-Abalo, and Torres-Fortuny (2014) showed that ipsilateral/contralateral asymmetries are not consistently present in infants for 1000- and 2000-Hz BC ASSRs, so masking would be necessary at these frequencies to isolate the test ear. For 500- and 4000-Hz, the asymmetries were present, but further research is needed on populations with hearing loss. Although both ipsilateral and contralateral channels were recorded, only the ipsilateral channel was analyzed in this study.  Stimuli and ASSR Systems   The following table compares the Rotman MASTER Research system versus Clinical Biologic MASTER II system’s options for stimuli.              	   29	   Research MASTER Biologic MASTER II Amplitude Modulated (Root Mean Square measure or Peak-to-peak SPL levels maintained for different modulation depths) Yes Yes Frequency Modulated  (Root Mean Square measure or Peak-to-peak SPL levels maintained for different modulation depths) Yes Yes Mixed Modulated (Root Mean Square measure or Peak-to-peak SPL levels maintained for different modulation depths) Yes No Exponential envelopes for AM  (Root Mean Square measure (not maintained) or Peak-to-peak SPL levels maintained for different modulation depths). Exponental values may be changed. Yes Yes Exponential envelopes for FM Yes No Noise stimuli (Broad-band noise, High-pass noise, Low-pass noise, complex noise). The user can define the span of the noise in Hz. Yes No. White noise only available for masking. 125 microsecond click Yes No 1 millisecond Burst (Broadband noise, Low-pass noise, high-pass noise) Yes No Change phase of AM stimulus Yes No Change of phase for AM stimuli and inverse carrier Yes No  Table 1.1. A comparison of the Rotman MASTER Research system and the Clinical Biologic MASTER II system’s stimuli options.  Noise Measurement   For the research version, The Multi-Master v1a software system specifications show that noise floor (NF) is measured, and the values in each column of this row are the size of the 95% confidence limits of the noise multiplied by 1000 (in nV). A good noise estimate "N" of background levels of noise is the root mean square of the 120 adjacent amplitude spectra bins which are above and below the frequency of modulation. This can be computed as N=sqrt(sum(a^2)/120), where “a” is each of the adjacent 120 noise bins which are summed together. The circle radius on the Multi-Master v1 program is computed as CR=N * 1.74, or the .95% confidence levels for the noise (1.74 is sqrt of F value 3.04, at .05 for 2, 240). The Biologic system does not display circle radius, but rather translates circle radius to residual EEG noise using the same technique (John, 2003).   	   30	  Threshold Measurement  The obvious hope for the ASSR is to eventually employ this method for threshold evaluation in infants instead of, or in combination with the ABR. The ultimate utility of electrophysiological thresholds is to estimate behavioural thresholds in patients who are unable to provide behavioural thresholds. Once accurate electrophysiological thresholds are obtained, a conversion is usually applied to estimated hearing level (eHL). BCEHP uses the subtraction of correction factors to arrive at an estimated behavioural hearing level (BCEHP, 2012). At this time, there are no generally agreed upon correction factors for the ASSR, although some authors have proposed normal levels (Van Maanen & Stapells, 2009; Casey & Small, 2014). Currently, some clinics in Canada are using the ASSR as a screening tool but it is not yet used as a routine diagnostic measure of threshold. To be able to use the ASSR to determine if an infant has normal hearing, sensorineural, conductive or mixed hearing loss (and if hearing impaired, to what degree) with confidence, normal levels are needed and research involving participants with different degrees and types of hearing loss is needed. Research teams have been working diligently for almost two decades to establish normal threshold levels for infants (summarized in Table 1.2 below). While there are a reasonably large number of studies reporting air-conduction data, only a few studies exist that report bone-conduction thresholds. Additional issues are, even with the studies that do exist, are the stimuli and protocol used to detect the responses, the methodology, the recording equipment used, the age at which infants are tested and the stopping and response criteria vary. This makes study comparisons difficult and poses a challenge regarding standardization of protocols for clinical use. Studies listed in the table below use AM, AM/FM or AM2 stimuli.	   31	      AC ASSR Studies Modulation Multiple (M)/Single (S) Age 500 Hz 1000 Hz 2000 Hz 4000 Hz Mean ± SD Norm MAX  (dB HL) Mean ± SD Norm MAX  (dB HL) Mean ± SD Norm MAX (dB HL) Mean ± SD Norm MAX  (dB HL) Lins et al., (1996) AM M 1-10 mos 45±13 48 29±10 43 26±8 41 29±10 40 Cone-Wesson, Parker, et al., (2002) AM S <4 mos - >71 - >72 - 50 - 54 John, Brown et al., (2004) MM, AM, AM2 M 3-15 wks - >46 - >50 - >50 - 40 Rance et al., (2005) MM S 1-3 mos 32±8 52 33±7 47 24±6 40 28±8 43 Swanepoel & Steyn (2005) MM M 3-8 wks 37±8 50 34±10 >50 34±11 >50 30±11 40 Luts et al., (2006) MM M <3 mos 42±10 >44 35±10 >50 32±10 42 36±9 44 Rance & Tomlin (2006) MM S 6 wks 40±7 50     33±8 40 Van Maanen & Stapells (2009) AM M <6 mos 39±7 49 33±5 45 29±7 36 24±10 32 Van Maanen & Stapells (2009) AM M 6.1-66 mos 41±7 49 37±11 45 31±8 36 22±10 32 Ribeiro et al., (2010) MM M 1-8 days 39 - 28 - 24 - 27 - Alaerts et al., (2010) MM M < 3 mos 40 - 38 - 30 - 33 - Hatton & Stapells (2011) AM S/M 1-9.5 mos 48 - - - - - - - Casey & Small (2014) AM2 M  6.5-19 mos 27.9±8.0 40 22.3±8.3 40 20.0±7.8 30 22.3±9.0 38 BC ASSR Studies            Swanepoel et al. (2008) MM M 0.5-11 yrs 18(7) - 16(11) - 24(7) - 26(8) - Small & Stapells (2008) MM M  0-11 mos  12-24 mos 14  22 30  40 5  13 20  20 26  26 40  40 14  13 30  30 Small & Stapells (2008) MM M 2-11 mos 12.5±12.9 30 5.0±5.2 20 20.0±12.8 40 9.2±7.9 30 Casey & Small (2014) AM2 M 6.5-19 mos 11.1±11 30 9.0±9.1 30 20.5±10.5 40 14.5±11.4 33 Table adapted with permission from "Frequency-Specific Threshold Assessment in Young Infants Using the Transient ABR and the Brainstem ASSR" by David R. Stapells in Comprehensive Handbook of Pediatric Audiology ( p. 434) by Richard Seewald and Anne Marie Tharpe. Copyright© 2011 Plural Publishing, Inc. All rights reserved.   Table 1. 2. A summary of mean and maximum ASSR Thresholds for Normal Hearing Infants in dB HL from the literature.	   32	  Data with normal hearing infants is not enough to bring the ASSR into the clinic diagnostically; it is imperative to know how well the ASSR predicts behavioural thresholds and ABR thresholds for hearing-impaired infants with different types and degrees of hearing loss.  ASSR vs. ABR   Reliability, variability and accuracy of the ASSR in assessing thresholds are critical features to evaluate if clinicians intend to use the ASSR diagnostically. To determine the clinical utility of this measure, it is useful to compare ASSR thresholds to thresholds obtained using a “gold-standard” measure. At present, the “gold-standard” for measuring electrophysiological response thresholds in infants is the tone-ABR (BCEHP, 2012; Joint Committee on Infant Hearing, 2007). ASSR studies have made this comparison using frequency-specific (e.g., tone) and non-frequency-specific (e.g., click) stimuli, with the majority using the latter. Some of these studies will be reviewed below.   The majority of adult data compares ASSR thresholds with the adult gold-standard threshold measure, behavioural thresholds measured by pure-tone audiometry. There are few studies directly comparing ABR and ASSR thresholds in adults. Of the adult data that exists, it has been shown that tone-burst AC ABR and ASSR thresholds were similar (Cone-Wesson et al., 2002)   . Jewett and Williston (1971) showed that ABR and ASSR thresholds were highly correlated (r=0.91) and that there is variation in the prediction of behavioural thresholds between ABR and ASSR. In addition,  Swanepoel et al., (2004) compared ASSRs at 500, 1000, 2000, and 4000 Hz with a click-evoked ABR and 500 Hz tone ABR. They found ABR thresholds to be approximately 14-18 dB better than ASSR thresholds and that there was no significant difference between ABR and ASSR techniques in estimating normal hearing. D'Haenens et al. (2009) studied adults with normal hearing, conductive hearing loss and sensorineural hearing loss. Their AC ASSR threshold data using AM2/FM stimuli showed that mild sensorineural hearing loss was difficult to differentiate from normal hearing and moderate sensorineural hearing loss. For conductive losses, ASSR thresholds accurately predicted behavioural thresholds but suggest that these findings need to be confirmed for infants. 	   33	  The repertoire of infant studies including a direct comparison of ABR to ASSR thresholds is certainly larger than that of adults, but is nonetheless quite small. In infants who are old enough (over six months of age), comparisons tend to be made between ASSR and behavioural thresholds, and these comparisons have yielded encouraging results. It has been shown by several studies that ASSR and behavioural thresholds in normal hearing infants and infants with hearing loss correlate highly (Jerger & Hall, 1980; M¯ller, 1995; Moller & Jannetta, 1983; Parkkonen, 2009). It is important to note that behavioural thresholds are not equivalent to electrophysiological thresholds but can be estimated from electrophysiological thresholds (Lins et al., 1995; Nousak, 1992). Casey and Small (2014)’s findings showed that bone-conduction thresholds were, on average, 7 to 16 dB poorer for ASSR compared with behavioural thresholds, but varied widely across infants. This is similar to the findings of Foxe (1993) described next. There are also studies that compare ABR to ASSR thresholds and behavioural thresholds. For example, Luts et al., (2006) compared ASSR thresholds at 500, 1000, 2000 and 4000 Hz to click-ABR thresholds, and followed-up with behavioural testing approximately 7 months later. They attempted to use ASSR thresholds to predict behavioural thresholds and found that 83% of prediction errors are within 15 dB, 61% within 10 dB and 35% within 5 dB with the mean being 10±12.6dB. Values were similar across frequencies. This suggests that ASSR thresholds tend to correspond to behavioural thresholds as expected with reasonable accuracy.   Studies comparing ASSR thresholds directly to the gold-standard tone-ABR are even fewer. However, among the studies that do exist, a trend seems to exist that shows ASSR thresholds are consistently higher in infants compared to ABR thresholds (Rance et al., 2006; Klein, 1984. Sininger et al., 1997). Also, both AC ASSR and ABR thresholds tend to be higher at lower frequencies compared to higher frequencies (Rance et al., 2006). In addition, the response amplitudes vary significantly with the ASSR being much smaller than the ABR. Luts et al., (2006) and John et al., (2004) measured mean infant ASSR amplitudes from 10-20 nV at 50 dB HL which are substantially lower than ABR amplitudes of approximately 250 nV at the same level (Singer et al., 1997). One infant study, Moller and Jannetta (1983), categorized young infants’ hearing status based on the results of the tone-bust ABR and showed that the correlation values comparing the ASSR measured at <6 months of age and the behavioral thresholds measured >6 months are highly correlated and accurate. Interestingly, infants with greater 	   34	  degrees of sensorineural hearing loss showed better accuracy in the ASSR prediction of behavioural threshold.  The literature comparing frequency-specific tone-ABR thresholds with ASSR thresholds in infants is significantly lacking, and is the focus on the present study. This comparison is relevant as electrophysiological testing is often the only way to test the hearing status of young infants under the age of 6 months. The studies in the previous paragraph all demonstrated very strong correlations between ASSR and behavioural thresholds in newborns and children. We know that maturation has an effect on ASSR thresholds (discussed more in the maturation section), and so it is especially important to establish a comprehensive body of literature for infants who are too young to respond behaviourally. This age group is particularly important as most universal hearing programs (i.e., BCEHP, 2012) require diagnosis by 3 months and intervention by 6 months of age, all before behavioural testing is possible. As mentioned earlier, adult studies have generally shown that ASSR thresholds are usually poorer than tone-ABR thresholds (Cone-Wesson, Dowell et al., 2002; Johnson & Brown, 2005; Swanepoel et al., 2004), and to date, normal hearing infant studies have tended to support this trend (Rance et al., 2002, Van Maanen & Stapells, 2010, Rance et al., 2006). The question is, do the same rules apply for infants with hearing loss?   S. Kramer (1992) and Van Maanen and Stapells (2010) compared ASSR thresholds to tone-ABR thresholds in infants (and children) with sensorineural hearing loss. Van Maanen and Stapells (2010) was the first study to compare tone-ABR and multiple ASSR thresholds in infants with normal hearing and sensorineural hearing loss in the same test session. They used used cosine3-windowed sinusoids, which are nearly equivalent to AM2 stimuli. This is an useful study as it controlled for any changes in hearing between test sessions by completing all testing during one session. They were also able to show that ASSR thresholds to multiple stimuli were strongly correlated (r =.97) to tone-ABR thresholds for 500, 1000, 2000, and 4000 Hz stimuli and that their proposed normal levels of 50, 45, 40, 40 for 500, 1000, 2000 and 4000 Hz respectively correctly differentiate normal hearing from elevated thresholds. Rodrigues et al., (2010) also showed strong correlations between ABR and ASSR thresholds in infants with sensorineural hearing loss. Correlations were 0.91, 0.76, 0.81 and 0.89 for 500, 1000, 2000 and 4000 Hz, 	   35	  respectively. They also compared these thresholds to behavioural thresholds and found r values >0.94 at each frequency. More studies with hearing-impaired infants are needed, but the existing data look promising.   To date, there is a significant gap in the literature. The studies mentioned above studied air-conduction thresholds only. Presently, no data comparing bone-conduction ASSR to bone-conduction tone-ABR thresholds in infants with hearing loss exist. Swanepoel et al., (2008) did provide a small amount of bone-conduction ASSR data for infants with hearing loss, but did not compare to the gold-standard tone-ABR and tested a broad age range (0.25-11.5 years).     	   36	  MATURATION  In addition to the discussed differences between the ABR and ASSR, the two responses do not appear to develop in the same way. The ABR becomes adult-like in the beginning years of life, while the ASSR is completely mature at approximately 12 years of age. One practical reason that this topic is interesting for the purposes of this study is that the presence of infant-adult differences in normal ASSR thresholds presents a need for different normal levels to be established according to age. It is also important to understand, however, that the change in ASSR threshold that occurs in the first six months of life makes it difficult and perhaps inappropriate to determine normal levels using age ranges such as 0-6 months of age (Small & Stapells, 2003). Eventually, it may be necessary to develop normal levels for a more narrow age range. An additional challenge is the variability of thresholds, which appears to be different with age. Maturation of the ABR and ASSR will be discussed separately for air- and bone-conduction thresholds as their development differs.  Measuring thresholds using ABR is slightly different than using the ABR for other purposes, like to detect disruptions in the auditory system. In infants, wave V is the focus as it is the most resilient to changes in intensity and can be tracked down to threshold. The ABR does not undergo a large degree of change following the 18th month of life and is considered to be fully mature around 3 years of age (Picton, John, Purcell, et al., 2003). The beginning months of life, however, show maturational changes, particularly in Wave V latency and response amplitudes. Finitzo-Hieber (1982) and Gorga et al., (1987) showed a steady decease in wave V latency in the beginning weeks of life in pre-term infant. They recommend that an ABR should not be recorded in infants younger than 37 weeks gestational age as this may result in the incorrect interpretation of an impairment that is purely a result of an immature auditory system.   Air-Conduction  Thresholds undergo maturational changes as infant thresholds differ from those of adults. This difference is frequency-dependent; at low frequencies (500 and 1000 Hz), response thresholds have been shown to be higher than those at higher frequencies (2000 and 4000 Hz). With 	   37	  maturation, thresholds improve and become adult-like by approximately 18 months of age (Rance et al., 2006; Stapells et al., 1995; Stapells, 2000b, Vander Werff et al., 2009). Studies have investigated ABR thresholds measured based on dB SPL measured at the eardrum in infants to eliminate biases caused by individual differences in the outer ear. Stapells, Gravel, and Martin (1995) found that “in situ” ABR thresholds for infants decreased significantly during the first 6 months of life. When their thresholds were compared to those of normal-hearing, they showed that the ABR response reached maturity in infants between 4 and 6 months of age for 500 and 2000 Hz, but remained immature, or elevated, at 6 months of age for 4000 Hz. These findings suggest that infant-adult maturational differences should be considered when determining “normal levels” for ABR thresholds. These are taken into account when dB nHL to dB eHL correction factors were established (BCEHP, 2012).  There are also some maturational trends observed for the ASSR. It has been shown, in general, that air-conduction ASSR thresholds tend to improve with age in the order of about 10 dB between infancy and adulthood1 (Savio et al., 2001, Rance & Tomlin, 2006;). High frequency responses tend to mature faster than low frequency responses, so that 500 Hz tends to be the last frequency to reach adult-like levels (John et al., 2004, Savio et al., 2001, Rance & Tomlin, 2006). From a maturational standpoint, the variance in thresholds is higher for infants than for adults. Studies have demonstrated that thresholds for normal hearing neonates can vary by more than 30 dB, which they suggest is due to low response amplitude, and improves with age (Luts et al., 2006; Rance & Tomlin, 2006; Savio et al., 2001).  Bone-Conduction    The maturation of the ABR wave V latency is such that latencies tend to increase with age for 500 Hz stimuli and decrease for 2000 Hz stimuli. This is likely due to changes in skull properties 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  1 ASSR thresholds are measured in dB HL. dB HL refers to a normal hearing level in decibels that was established using young adults with normal hearing. dB HL represents a conversion from dB SPL (decibel sound pressure level) to a normalized scale based on the intensity required for normal hearing adults to respond at threshold. It is important to acknowledge that this unit of measurement was developed based on adult hearing levels, but is also used in the assessment of infant hearing. A separate unit of measurement for infants has not been developed. 	   38	  (Foxe & Stapells, 1989; Nousak & Stapells, 1992). Wave V amplitudes, and therefore thresholds also mature in a frequency-dependent manner. The maturation of bone-conduction thresholds shows a trend opposite to air-conduction; bone-conduction thresholds tend to be better for infants than adults in the low and high frequencies and worse in the mid frequencies. The largest infant-adult difference is seen at 500 Hz; during the first 6 months of age, one can expect to see an improvement in 500 Hz thresholds of approximately 16-27 dB, and between 6 months of age and adulthood, an improvement of 2-7 dB (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989). This pattern, coupled with the maturation of air-conduction thresholds  results in a small air-bone gap being present in the low frequencies that is not clinically significant (Vander Werff et al., 2009; Foxe & Stapells, 1993). Perhaps the most notable aspect of the maturation of bone-conduction ABR responses is the ipsilateral/contralateral asymmetries discussed in the “ABR” section. This asymmetry allows clinicians to determine which cochlea is responding to a stimulus and is typically reliable until approximately seven months of age.  ASSR thresholds to bone-conducted stimuli tend to improve at mid frequencies (i.e., 2000 and 4000 Hz), and worsen in the low frequencies (500 and 1000 Hz) (Small & Stapells, 2006, 2008). Small and Stapells (2008a) compared BC ASSR thresholds for young (mean age:16 weeks) and older infants (mean age: 18 months) and found that thresholds were significantly higher for young infants, so maturational effects are apparent even at early stages. Bone-conduction ASSRs appear to continue to develop until at least two years of age (Small & Stapells, 2008). Similarly to the ABR, threshold maturation results in a small air-bone gap being present in the low frequencies that is not clinically significant. This will be discussed further below. ASSR thresholds also show asymmetries between the ipsilateral and contralateral recording channels. To date, Small and Stapells (2008b) have shown that infant ASSR response amplitudes were larger in the ipsilateral channel compared to the contralateral channel at 500, 1000, 2000 and 4000 Hz. Small and Love (2014) showed that this finding was not consistent for 1000 and 2000 Hz, and clinical masking would still be needed for these frequencies. Although these findings are positive for 500 and 4000 Hz, it is not yet clear if ipsilateral/contralateral asymmetries can be used clinically until research is conducted using infants with hearing loss.       	   39	  Air-Bone Gap  For a population in which conductive hearing loss is prevalent, the estimation of the air-bone gap (ABG) can be useful to estimate the amount of hearing loss caused by the outer and/or middle ear. Recall that in adults, the air-bone gap is the difference between air-conduction and bone-conduction thresholds and typically represents the conductive component of a hearing loss (Gorga, Kaminski, Beauchaine, & Bergman, 1993; Norton et al., 2000). Jeng et al., (2004) studied adults and determined that the ASSR air-bone gap was reasonably accurate in estimating the behavioural air-bone gap. As mentioned earlier, there are maturational differences between the development of air-conduction and bone-conduction ASSR thresholds that result in an air-bone gap. For this reason, it is important to be cautious when assessing the significance of an ABG in infants, especially for the low frequencies. It is not appropriate to use the same criterion (ABG >10 dB) that is used for adults to determine an ABG is clinically significant in infants. Maturational trends are opposite at the low frequencies for AC versus BC ASSRs such that air-conduction thresholds tend to improve with age and BC thresholds tend to worsen at the low frequencies. At early stages, this may appear as a significantly large air-bone gap (between 20-25 dB) that is clinically insignificant and likely due to maturation (Swanepoel et al., 2008; Small & Stapells, 2006; John et al., 2004, Savio et al., 2001, Rance & Tomlin, 2006, Small & Stapells, 2008; Van Maanen & Stapells, 2009, Casey & Small, 2014; Hulecki & Small, 2011). This has been coined the “maturational air-bone gap” by Hulecki and Small (2011). Lee, Hsieh, Pan, and Hsu (2007) published a relatively small study that included children with otitis media (n=4, participant ages at time of testing ranged from 2 months- 2;7 years) examining the air-bone gap using multiple ASSRs in children with otitis media pre- and post-treatment. They showed that previous to placement of ventilation tubes, air- and bone-conduction ASSR thresholds showed air-bone gaps of 32.5 dB at 500 Hz, 22.5 dB at 1000 Hz, 7.5 dB at 2000 Hz and 22.5 dB at 4000 Hz. The size of the ABGs decreased post-treatment so that air-conduction thresholds were closer to the bone-conduction thresholds. They suggest that bone-conduction ASSR can accurately assess the hearing status in children with conductive hearing loss. As mentioned above, due to the effects of maturation, it remains unclear whether the same trend is generalizable in the infant population under six months of age.  	   40	  Let us now discuss some of the physical changes of the ear that may contribute to the above-mentioned maturational trends found for the ABR and ASSR. This is not meant to be a particularly exhaustive list of the changes seen in the system as the child grows, but rather a general summary of some key points. There are anatomical changes that occur to the outer and middle ear that affect the transmission of sound through the system. The outer ear and middle ear are not fully mature at birth. It is well known that the infant ear canal changes in length to become longer with age. The ear canal length and diameter, for example, is suggested to be approximately 50 % shorter in the newborn compared to the adult (Lee, Jaw, Pan, Hsieh, & Hsu, 2008; Ribeiro & Carvallo, 2008). For the same input, as the ear canal lengthens, there is a decrease in dB SPL that reaches the eardrum compared to when the ear canal was shorter. The middle ear ossicles’ developmental status at birth is debated. Some researchers claim that they are fully developed and adult-like at birth, while others suggest that the ossicles are smaller in weight and size than adults (Lasky, 1997; Starr & Hamilton, 1976; Vander Werff et al., 2009; Wolfe, Skinner, & Burns, 1978). In addition, the middle ear system of the infant is known to be significantly more mass dominated than the adult system that is stiffness dominated (Campbell & Bartoli, 1986; Janssen, Usher, & Stapells, 2010). Due especially to the change in ear canal size during infancy, we might expect infant air-conduction thresholds to be better than adults. There is skepticism in the literature about whether the changes in the middle ear would be the main cause of differences in infant’s and adult’s thresholds. Additionally, regardless of whether these changes significantly affect air-conduction thresholds, we also see differences in bone-conduction thresholds for which these changes in the outer and middle ear systems cannot account for (Rance et al., 2006; Small et al., 2007; Small & Hu, 2011; Stenfelt & Goode, 2005).  The cochlea is mature at birth and is unlikely to be responsible for any maturational changes in threshold (e.g., Abdala et al., 1996). The immaturity of the skull and surrounding skin and tissue has also been suggested to contribute to infant-adult differences observed in bone-conduction testing. Mackey, Hodgetts and Small (submitted) showed differences in mechanical impedances between young infants and older infants and children. What appears to be most likely is that the development of the brainstem and cortex have the most effect on the electrophysiological responses and result in these infant-adult differences. 	   41	  Given these maturational changes, it is reasonable to ask, why is there a difference between the ABR and ASSR which are both electrophysiological responses? Rance (2008) discusses some of the possible answers to this question. He suggests that this difference may lie in the generators of the responses. As discussed in the Generators section, the 80-Hz ASSR is primarily a brainstem response but still does contain cortical responses. The ABR, on the other hand, is a true brainstem response. Another of his hypotheses is that the processing ability for different stimuli rates may be the underlying mechanism. The 80-Hz ASSR stimuli are presented at a very fast rate, and perhaps infants’ ability to process such fast rates is impaired at early stages due to immature auditory systems and high stimulus rates that exceed the refractory periods of the nerve. ABR stimulus rates are do not affect wave V up to approximately 50 Hz, but above 50 Hz there is a decrease in amplitude and latency. The suggested rate for acquiring ABRs in infants is 39.1/s (BCEHP, 2012). Finally, the ABR and ASSR are calibrated differently, which may also contribute to the difference. The ABR is measured in dB nHL and the ASSR in dB HL. Peak-to-peak equivalent dB SPL values are between 5-15 dB higher for tone bursts at 0 dB nHL compared to modulated tones at 0 dB HL.   Stimulus Conduction Considerations Artifact  A caution in ASSR testing at high intensities is the issue of artifactual responses. The risk is that the stimulus artifact will result in energy being aliased to frequencies within the range that is measured for a response. If this energy passes the filter and is amplified, it may appear from the FFT that there is a response, when in fact, there was no response from the participant. This is referred to as a “spurious ASSR” (Gutschalk et al., 1999; Jeng et al., 2004; Jiang, 1991; Kuwada et al., 2002; Linden et al., 1985). Adult data suggest that when using high intensity stimuli (bone-conduction >50 dB HL and air-conduction >110 dB HL), spurious ASSRs are produced, with 500 and 1000 Hz being the most vulnerable. The authors concluded that with appropriate EEG filters (low-pass) and appropriate analog-digital conversion rates (at least twice the maximum frequency that is present in the EEG), spurious responses due to aliasing can be avoided. They 	   42	  caution, however, that non-auditory physiologic responses may contribute to artifactual responses at approximately 50 dB HL using bone-conducted stimuli (Small & Stapells, 2004).   Similarly, in an infant/child study, Swanepoel et al., (2008) showed that the minimum levels for spurious BC ASSR occurred at 40, 60, 60, 60 dB HL for 500, 1000, 2000 and 4000 Hz respectively. They concluded that for this group, sensorineural hearing loss that is moderate or greater from 1000-4000 Hz and mild or greater at 500 Hz cannot be quantified using bone-conduction ASSR.  	   43	  EARLY HEARING  Earlier sections of this thesis have illustrated the diagnostic power that the electrophysiological tests afford us. Reasonable questions at this point are “Why are we so interested in testing infants at such an early age? Why not wait until they are able to respond behaviourally?”. These questions have been researched for decades and the answers to which form the basis of early hearing programs like the BCEHP. Levitt and colleagues’ research pointed out that when language ability is assessed in hearing impaired children, even across different degrees of hearing loss (mild to profound), the greatest predictor of language delay was age of diagnosis rather than degree of loss; Those that were identified with hearing loss earlier had less severe delays than those identified later (Picton et al., 2007). Almost any document one reads on the subject of early hearing will cite the seminal research papers by Yoshinaga-Itano et al., (1998a,b) which showed that early identification and intervention for hearing loss before six months of age resulted in better language outcomes in early-identified compared to later-identified children. In essence, one answer to the questions “Why are we so interested in testing infants at such an early age? Why not wait until they are able to respond behaviourally?” is there exists a critical period in which intervention for infants with hearing loss is required to optimize language development.   Although speech and language development is often the most cited justification for early screenings by clinicians and position statements, it is not the only concern for hearing impaired children. In another study by Yoshinaga-Itano, it was found that early-identified children had higher scores on the personal-social quotients than later identified children (Yoshinaga-Itano, 1998). Social development can be negatively affected by hearing loss and can interfere with later success in school.   Major task forces have endorsed early identification of hearing loss. As early as 1993, position statements from major groups have been published, like the National Institude on Deafness and Other Communication Disorders (NIDCD) statement that emphasized the importance of early hearing screenings (NIDCD, 1993). The Joint Committee on Infant Hearing supports early hearing identification. Their most recent position statement in 2007 addresses the cognitive, social and emotional effects of hearing loss due to their dependency on language skills. They 	   44	  reference the Yoshinaga-Itano (2014) paper that describes that delays that are a consequence of hearing impairment may result in lower educational and employment levels in adulthood.   Early Hearing Screening Programs  The evidence discussed above has supported the need for Universal hearing screening programs to facilitate the early identification of hearing impaired infants. Such programs began to be implemented in several areas of the world (e.g., parts of the United Kingdom, United States of America and Canada). As Yoshinaga-Itano describes, studies investigating the effect of efficacy of these programs in detecting hearing loss and providing intervention earlier than if programs did not exist vary significantly. These variations were typically due to the inconsistent criteria used to include a hearing screening program in the evaluation (Eisenberg et al., 2007). In general, however, they did support that early hearing programs were resulting in earlier identification and intervention for children with hearing loss. After the Widen et al., (2000) paper became available and showed that it was very feasible to behaviourally test four test frequencies in each ear, it provided ammunition to early hearing programs to insist upon ear-specific evaluations. Ear-specific measures are insisted upon for both behavioural and electrophysiological testing (BCEHP, 2012).  The British Columbia Early Hearing Program is an exemplary evidence-based program that has gained international recognition (BCEHP, 2012). Goals of the British Columbia Early Hearing Program can be found in their training manuals and are listed below:  Goal of the Screening Component: The goal of the screening component of the BCEHP is to identify infants with hearing loss or with late onset indicators, in a timely and family-sensitive manner.  Purpose of Screening: To identify infants who may have a target disorder and require further diagnostic assessment to confirm hearing status, and infants who may develop target disorder (delayed onset or progressive) and who require periodic reassessment to monitor for target disorder. 	   45	  Target Disorder: 35 dB HL2 between 500 and 4000 Hz (frequencies important for speech) unilateral or bilateral sensorineural hearing loss, congenital conductive hearing loss, or conductive loss secondary to craniofacial anomalies (such as Atresia, Cleft Palate, Down’s Syndrome, Congenital Ossicular fixation).  Screening Targets: Minimum 95 percent of babies screened, screening completed (both stages if needed) by age 1 month or within 1 month of hospital discharge. Overall referral rate to diagnostics less than 4 percent well baby population, less than 10 percent Neonatal Intensive Care (NICU) baby population (greater than 48 hours in NICU) miss rate in hospital of less than 5 percent.  The Goals of the overall BC Early Hearing Program  • 95% of babies screened by 1 month of age • Diagnostics completed by 3 months of age • Intervention started by 6 months of age  Outcomes: • A child with hearing loss will have age-appropriate language skills by school • entry. • The child and his/her family will have meaningful and accessible communication. • Parents will have knowledge, skills and confidence to parent/advocate for their child with hearing loss.  • The program is provided to families based on fully informed parent/guardian choice and consent, and will comply with confidentiality requirements.  • The program was developed using the principles of evidence-based practice.  • The program is monitored and evaluated on an ongoing basis.  • The quality of the program is to be continuously improved, based on evaluation and new evidence.  It should be noted that goals and protocols vary among programs. As can be seen from the goals of the BCEHP program, transient conductive hearing losses, like those encountered in infants with otitis media, are not necessarily target disorders. Conductive losses secondary to otitis media do, however, impinge on the clinician’s ability to detect a target disorder using typical screening protocols such as otoacoustic emissions and tympanometry. Patients with otitis media will have absent OAEs and abnormal tympanograms. This screening profile is ambiguous as it 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  2 This level was selected to identify infants with mild or greater degrees of hearing loss. It is recognized that this level does not identify infants with minimal hearing loss and some infants with mild hearing loss; the rationale being that management in terms of hearing devices would not be pursued for a minimal/mild range of hearing loss. For the purposes of this study, "normal" refers to hearing thresholds 25 dB eHL or better, which means that infants with minimal and some mild degrees of loss would be included with infants with normal hearing . 	   46	  may be an indication of a temporary conductive hearing loss, no hearing loss or a permanent (sensorineural or conductive) hearing loss (Moeller, Tomblin, Yoshinaga-Itano, Connor, & Jerger, 2007; Yoshinaga-Itano, 1999, 2003, 2004). For this reason, it is often necessary to escalate the assessment for infants with otitis media to a diagnostic assessment using electrophysiologic measures. As discussed in other sections, the gold-standard diagnostic measure for infants is the tone-ABR, which can be time-consuming. The audiology community is hopeful that more efficient methods of measuring hearing sensitivity, like the ASSR, will be able to provide the target information in a more efficient way. At this point in time, ASSR research looks promising for infants with normal hearing and sensorineural hearing loss (Van Maanen & Stapells, 2010; Casey & Small, 2014; Rance et al., 2002; Rance et al., 2006; Luts et al., 2006; Swanepoel et al., 2008), but there is a need for studies including infants with conductive hearing loss.  Although transient conductive hearing loss secondary to otitis media is not a target disorder according to the BCEHP, it has a particularly high prevalence in childhood (with an especially high prevalence between 3 and 46 percent among aboriginal children) that has consequences for development (Downs & Yoshinaga-Itano, 1999). Yoshinaga-Itano and Apuzzo (1998) showed children with long-standing conductive losses had generally poorer language performance than their normal hearing peers. In addition, Needleman (1977) investigated comprehension and production aspects of phonology and found that the otitis media group had poorer production of phonemes and words.  Teele et al., conducted a prospective cohort study of 2500 children and found a significant correlation between amount of time with otitis media and low scores on vocabulary, comprehension and verbal ability (Bowd, 2005; Yoshinaga-Itano, Sedey, Coulter, & Mehl, 1998). Friel-Patti and Finitzo (1990) and Winskel (2006) also showed that the presence of otitis media had a negative effect on language abilities that persisted into later childhood, but a more recent study that on the effects of chronic otitis media in the first 2 years of life showed that the negative effects on language were resolved by 7 years of age (Zumach et al., 2010). While the evidence is mixed, there is still a strong body of literature suggesting the negative effects of otitis media and the associated hearing loss, and thus, we must take this into account when assessing and planning interventions for young children with transient conductive hearing loss. 	   47	  STUDY RATIONALE  Given the potential of the ASSR (using both air- and bone-conduction) to identify hearing loss and differentiate between types of hearing loss, it is an ideal diagnostic test for infants. This test fits well within the framework of early hearing programs at is it possible to test infants electrophysiologically who are too immature to perform behaviourally; the mandate of the BCEHP is to have all infants screened for hearing loss by one month of age, to have an ear and frequency specific diagnostic assessment completed by three months of age and an intervention implemented by six months of age (BCEHP, 2012). Within the structure of publically funded hearing programs in Canada, it is always desirable to have more efficient methods of obtaining reliable results. As stated earlier, to employ this method with confidence in the infant population, more studies are needed. Although data for normal hearing infants and infants with sensorineural hearing loss look promising, no data exist for infants with conductive loss within the BCEHP essential timeframe of 0-6 months and no published ASSR data exist using bone-conduction ASSR in infants with hearing loss. This study aims to act as a starting point for filling in this existing gap, and based on previous infant data (normal hearing studies: Van Maanen & Stapells 2009, 2010; Casey & Small, 2014), it is expected that the conductive hearing loss data will follow the same trends of accurately predicting hearing thresholds in infants.      	   48	  METHODS AND MATERIALS  Participants  Infants were recruited through the newborn hearing screening program at the Royal University Hospital by the pediatric audiologist and newborn hearing screener and participation was entirely voluntary. A total of 64 infants between the ages of 0 and 6 months participated in the study (normal hearing mean age = 7.36 weeks, CHL mean age = 6.71 weeks, ranged from 0.6-22 weeks); 61 from the well-baby nursery and 3 graduates from the NICU. 14 infants were excluded because they did not sleep and did not complete any conditions of the testing session. Partial or complete results for AC and BC ABR and ASSR were obtained from 51 infants.  Of the 50 who completed at least two conditions of the test, 15 completed testing for one ear for both 2000 and 500 Hz (air- and bone-conduction), 8 completed only 2000 Hz and 27 only 500 Hz. All test results were obtained during one session. A summary of the number of thresholds obtained for each condition is shown below in table 2.1.        Table 2.1. Sample size for 500 and 2000 Hz  air- and bone-conduction stimuli using ASSR and ABR.   Infants were included in the study if they referred from the newborn hearing screening or if they were unable to be screened at birth. To verify the status of the middle ear and hearing at the time of testing, tympanometry and transient-evoked otoacoustic emissions (TEOAEs) were performed using a Madsen Accuscreen and the Otoflex. TEOAE stimulus levels ranged from 70-84 dB SPL and used noise-weighted averaging. The response detection method involved the counting of significant signal peaks with self-calibration depending on ear canal volume. To pass the TEOAE test, a total of eight valid peaks in alternating directions (counted both above and below Mode  500 Hz 2000 Hz   ASSR ABR ASSR ABR Air-conduction Normal 21 23 22 23 CHL 15 15 0 0 Bone-conduction Normal 23 24 23 23 CHL 15 15 0 0 	   49	  the median line) must be present.  Of the 51 participants who completed the testing, 31 did not pass tympanometry and 31 did not pass OAEs. A tympanogram was considered to be a “refer” if there was no identifiable peak, or maximum admittance, and thus, the tympanogram was flat. TEOAEs were considered to be a “refer” if there was a response in fewer than three bands. The primary purpose of the screening measures was to corroborate the presence of middle ear pathology for participants with conductive hearing loss shown by abnormal tympanograms and absent TEOAEs and to determine the follow-up protocol as per the Royal University Hospital. Screening results by participant are shown in tables 2.2 and 2.3 below.  Subject Number  Age (weeks) Tympanometry OAE 6 3.3 Refer Refer 7 3.3 Refer CNT 8 12.9 Refer Pass 18 8.1 Pass Pass 21 6 Refer Refer 24 6 Pass Pass 27 22.1 Pass Refer 29 7.4 Pass Pass 35 3.3 Pass Pass 36 3.7 Refer Refer 37 5.3 Pass Pass 38 16.6 Refer Refer 39 11.6 Pass Pass 40 4.7 Pass Pass 43 11.6 Pass Pass 44 7.1 Pass Pass 46 13.3 Refer Refer 49 0.6 Refer Refer 50 10.4 Pass Pass 52 5.9 Refer Refer 53 3.4 Pass Pass 54 2.7 Refer Refer 56 5.1 Pass Pass 57 7.1 Pass Pass 60 2.4 Pass Pass  Table 2.2. 500 Hz Normal Hearing Group: Immittance (tympanometry) and TEOAE screening results for individual participants.    	   50	  Subject Number Age (weeks) Tympanometry OAE 10 6 Refer Refer 11 6.4 Refer Refer 14 4.9 Refer Refer 16 5.7 Refer Refer 17 4.1 Refer Refer 25 6.3 Refer Refer 28 20.6 Refer Refer 30 5.6 Refer Refer 32 5.3 Refer Refer 33 2.9 Refer Refer 34 3 Refer Refer 41 4.1 Refer Refer 42 4.1 Refer Refer 58 11.3 Refer Refer 62 10.4 Refer Refer  Table 2.3. 500 Hz Conductive Hearing Loss Group: Immittance (tympanometry) and TEOAE screening results for individual participants.    Participants were considered to be “normal” if ABR thresholds for air- and bone-conducted stimuli were within the normal levels provided by BCEHP, and were considered to have “conductive hearing loss” if bone-conduction levels were within normal levels, while air-conduction thresholds were above BCEHP normal levels. The BCEHP normal levels are 35 dB nHL at 500 Hz; 35 dB nHL at 1000 Hz; 30 dB nHL at 2000; or 25 dB nHL at 4000 Hz (BCEHP, 2012).  Stimuli  Air-conduction stimuli were presented to participants using ER-3A insert earphones in one ear (the same ear that was used to establish bone-conduction thresholds). Bone-conduction stimuli were presented to participants using the B-71 bone oscillator placed on the mastoid, slightly posterior to the upper portion of the pinna for both ABR and ASSR testing.  Small, Hatton and Stapells (2007) showed no difference between lower mastoid and upper portion, so the upper portion was chosen to avoid interfering with the nearby electrode. This was coupled to the head 	   51	  using approximately 4.0 Newtons of force using the hand-held method held by the experimenter. This method of coupling was used as it was the least disruptive method to the infants’ sleep and was found to have no significant differences to thresholds obtained by the elastic headband coupling method (Small et al. 2007). For the ABR portion of testing, the Intelligent Hearing Systems (IHS) SmartEP was used to generate and present stimuli. System default stimuli were used for both 500 and 2000 Hz. These stimuli were Blackman-windowed tones (five-cycle tonal duration) presented at 39.1/sec to one ear. For the ASSR portion of testing, the two-channel Master II Clinical System was used to generate and present ASSR stimuli with carrier frequencies 500, 1000, 2000 and 4000 Hz. These stimuli were AM2 at modulation frequencies 78, 85, 93 and 101 Hz for carrier frequencies 500, 1000, 2000 and 4000 Hz, respectively, and were presented simultaneously to one ear (monotic multiple ASSR).  Calibration  Air-conduction ASSR stimuli were calibrated in dB HL and ABR stimuli in dB nHL using a Quest 177 sound level meter and G.R.A.S. DB 0138 2-CC coupler with 1-inch microphone. The acoustic calibration for 0 dB nHL using insert earphones was 22 and 20 dB ppe SPL for air-conduction (Stapells, 2011), and using the bone-conductor, 67 and 49 re: 1µN for bone-conduction at 500 and 2000 Hz, respectively (Small & Stapells, 2003). For ASSR stimuli, each of the four frequencies were calibrated separately in dB HL and then combined. Bone-conduction stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re: 1 µN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using the Quest 177 sound level meter and B & K Mastoid 4930 artificial mastoid. Calibrations for 0 dB HL using insert earphones was 5.5 and 3 dB ppe SPL for air-conduction, and using the bone-conductor, 58 and 31 re: 1µN for bone-conduction at 500 and 2000 Hz, respectively (ANSI, 1996).      	   52	  Recording  All participants were tested at the Royal University Hospital, Saskatoon in a double-walled sound-attenuating booth. All recordings were obtained using the Intelligent Hearing Systems SmartEP ABR system and the Master II Clinical ASSR System. Four disposable electrodes were placed on the infant’s scalp using the typical electrode montage for infant ABR testing: one electrode on the vertex (non-inverting), an inverting electrode on each mastoid; and the ground electrode off-center on the forehead. Impedances between all electrodes were less than 3 kOhms.  For ABR testing, standard BCEHP parameters were used. Gain was set to 100,000 and band-pass filtering from 30 to 1500 Hz with an artifact rejection of ±25 µV. A minimum of two replications of 2000 trials was obtained at threshold levels and one step (of 10 dB) below threshold. The presence and/or absence of a response were determined visually by the experimenter. A response being “present” was determined by a visually identifiable wave V in the averaged waveform. In accordance with BCEHP guidelines, “no response” was determined only when no visually identifiable wave V was present and residual noise was less than 0.08 µV (BCEHP, 2012). An example of this procedure is shown in figure 2.1 in the “Procedure” section.  For ASSR testing, two-channels were recorded but only the ipsilateral channel was considered when determining if a response was present or absent and was the only channel analyzed in this study. The EEG was filtered using a 30-150 Hz filter and amplified 10 000 times with artifact rejection set to 1-125 µV. The analog-to-digital conversion rate was 1200 Hz. Each sweep consisted of 16 epochs of 1024 data points and took 13.107 seconds of recording time. The ASSRs were averaged in the time domain and analyzed online in the frequency domain using a fast Fourier transform (FFT) with a resolution of 0.076 Hz over a range of 0-625 Hz. Amplitudes were measured baseline-to-peak and expressed in nV. Recording continued until there was a response present with a minimum of 10 sweeps, or the residual noise levels were at least <15 nV and there was a minimum of 10 sweeps completed; whichever came first.  A F-ratio was calculated by the MASTER II and a response was considered present if a significant response value (p<0.05), was obtained from the F-ratio compared to critical values for F(2, 240) for at least three consecutive sweeps. The F-ratio estimated the probability that the amplitude of the 	   53	  ASSR at the modulation frequency (80-Hz) was significantly different from the average amplitude of the noise at adjacent frequencies . This was calculated within 120 bins, or +/-60 bins from the modulation frequency (John and Picton, 2000a). A response was considered absent if no significant response value was obtained (p>0.05) and the noise value was appropriately low (<15 nV).   Procedure  This study involved only one recording session lasting between one and three hours. Before testing began, caregivers consented to participation in the study and were provided an honorarium. All infants completed a hearing screening3 (tympanometry and TEOAEs), ABR and ASSR testing on the same day in the same session. One ear was chosen to be tested using electrophysiologic methods. The selection of test ear was made based on the outcome of the hearing screening. If only one ear referred, that ear was tested using ABR and ASSR; if both ears failed, or both ears passed, the ear was chosen based on the most comfortable position for the infant and the caregiver. Testing was completed with the examiner inside the booth, next to the infant and caregiver. The examiner held the oscillator and continually monitored the placement of the earphone in the infant’s ear.  Hearing screening was conducted and electrodes were applied while the infant was awake. The infant remained asleep during the test in the caregiver’s arms inside the double-walled sound attenuated booth. TEOAEs and tympanometry was completed to identify impaired middle-ear function for participants with conductive hearing losses measured by the ABR.   Electrophysiological testing always began with ABR testing as it was important to provide parents with information from a “gold-standard” test before proceeding with ASSR. Air-conduction ABR was followed by bone-conduction.   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  3 Although stapedial reflex screening is a regular part of the hearing screening protocol at the Royal University Hospital, reflexes were not included in the screening procedures in the present study due to technical difficulties with the equipment. 	   54	  AC ABR testing began at 35 dB nHL for 500 Hz and 30 dB nHL for 2000 Hz. BC ABR testing began at 30 dB HL for 500 and 2000 Hz. A 10 dB bracketing method was used, as described above. Threshold levels were defined as the lowest level at which a response is present and the level below resulted in an absent response. The lowest level tested was 0 dB nHL for 2000 Hz AC and BC and 500 Hz BC, and 5 dB nHL for 500 Hz AC. An example of a representative normal response to air- and bone-conducted stimuli and representative elevated response to air-conducted stimuli at 500 Hz is shown below                         	   55	  Normal Air-Conduction Threshold Elevated Air-Conduction Threshold                Normal Bone-Conduction Threshold     Figure 2.1. Representative ABR response series to 500 Hz air- and bone-conduction stimuli. A representative normal air-conduction response series is shown on the upper left and a representative elevated air-conduction response series on the upper right. A representative normal bone-conduction response series is shown on the lower left. Waveform averages are directly below the replications at each intensity for air-conduction. Ipsilateral waveform averages are shown directly below the replications and contralateral directly above the replications at each intensity. Wave V is indicated by “V” on the waveform. *NR indicates “no response”, RN<0.08 nV.  35 dB nHL 25 dB nHL 10 dB nHL 45 dB nHL 35 dB nHL 30 dB nHL 20 dB nHL 10 dB nHL ipsi ipsi 0 dB nHL ipsi ipsi contra contra ipsi ipsi contra contra contra contra 	   56	  ASSR testing began at the highest “normal level” obtained by Casey and Small (2014) (30 dB HL). Similar to the ABR testing procedure, a 10 dB bracketing method was used, and threshold was defined as the lowest level at which a response is present and the level below resulted in an absent response. The lowest level tested was 0 dB HL for ASSR. As our study was more interested in bone-conduction comparisons, bone-conduction ASSR was prioritized over air-conduction. Thresholds were found using a 10 dB bracketing procedure. If a response was present, intensity was decreased by 10 dB. If no response was present, intensity was increased by 20 dB. Testing continued down to threshold, regardless of whether or not the infant’s threshold was below the BCEHP “normal levels”.    	   57	  DATA ANALYSIS  Individual and mean AC and BC ABR and ASSR thresholds were determined for normal-hearing infants and infants with conductive hearing loss. “Normal levels” were determined by calculating the cumulative percent of responses present at each intensity for each testing method  and mode of presentation. The level at which greater than 90 percent of infants had a response was considered the “normal level”. Results are shown in table and graph formats.   500 and 2000 Hz thresholds were averaged separately for air- and bone-conduction and 500 Hz thresholds were compared between normal and conductive hearing loss groups. Analyses were performed using (1) a parametric test: independent samples t-test and (2) and a non-parametric test: independent samples Mann-Whitney test. Differences in thresholds were considered significant at the p<0.05 level and a trend at the p<0.1 level. The Wilcoxon signed-rank test was used to compare ABR and ASSR thresholds within participant. Differences in thresholds were considered significant at the p<0.05 level and a trend at the p<0.1 level.  Air-bone gaps were calculated for ABR and ASSR for air- and bone-conduction separately by subtracting the bone-conduction threshold from the air-conduction threshold. An independent samples t-test was then conducted to determine if means between groups were significantly different. Differences in thresholds were considered significant at the p<0.05 level and a trend at the p<0.1 level. Difference scores were computed by subtracting ABR thresholds from ASSR thresholds for air- and bone-conduction separately.   	   58	  RESULTS  As shown in Table 2.4, air- and bone-conduction ABR thresholds at 2000 Hz for normal-hearing infants were, on average, 18 and 15 dB nHL, respectively. Air- and bone-conduction ABR thresholds at 500 Hz for normal-hearing infants were, on average, 25 and 10 dB nHL, respectively. For infants with confirmed conductive hearing loss, air-conduction ABR thresholds increased to approximately 48 dB nHL, while bone-conduction ABR thresholds were approximately 12 dB nHL. For the ASSR, air- and bone-conduction thresholds for normal-hearing infants were, on average, 20 and 26 dB nHL at 2000 Hz and 30 dB HL and 17 dB HL at 500 Hz, respectively. Air-conduction ASSR thresholds at 500 Hz for infants with conductive hearing loss increased to approximately 37 dB HL and bone-conduction thresholds were approximately 15 dB HL. Standard deviations were greater for ASSR data compared to ABR data for both air- and bone-conduction. Overall, mean bone-conduction thresholds appeared similar between groups, while mean air-conduction thresholds were higher for infants with conductive hearing loss than infants with normal-hearing for both ABR and ASSR testing methods. As mean bone-conduction thresholds were similar between groups, bone-conduction data for normal hearing and CHL groups were combined (see Table 2.5 below).       Normal Hearing Group Conductive Hearing Loss Group   2000 Hz 500 Hz 500 Hz   ASSR  (dB HL) ABR  (dB nHL) ASSR  (dB HL) ABR  (dB nHL) ASSR  (dB HL) ABR  (dB nHL) Air-conduction Mean 20.47 17.72 29.52 25.43 36.67 48.33 Median 20.00 20.00 30.00 25.00 30.00 45.00 SD 12.03 9.22 9.20 7.05 11.12 4.88 n 21.00 22.00 21.00 23.00 15.00 15.00 min 0.00 0.00 10.00 5.00 20.00 45.00 max 40.00 30.00 40.00 35.00 50.00 55.00 Bone-conduction Mean 21.00 15.00 17.39 10.41 15.33 12.00 Median 26.10 15.00 20.00 10.00 10.00 10.00 SD 13.96 10.12 9.63 7.50 12.63 7.93 n 21.00 22.00 23.00 24.00 15.00 15.00 min 0.00 0.00 0.00 0.00 0.00 0.00 max 50.00 30.00 30.00 20.00 40.00 20.00  Table 2.4. Normal Hearing and CHL: Air- and bone-conduction 2000 and 500 Hz ASSR and ABR threshold measures of central tendency and dispersion.   	   59	     All Participants 500 Hz  ASSR (dB HL) ABR (dB nHL) Bone-conduction Mean 16.58 Mean 11.03 Median 15.00 Median 10.00 SD 10.97 SD 7.18 n 38.00 n 39.00 min 0.00 min 0.00 max 40.00 max 20.00  Table 2.5. All participants: bone-conduction 500 Hz ASSR and ABR threshold measures of central tendency and dispersion.   Normal Levels  Table 2.6 shows the cumulative percentage for AC and BC ABR and ASSR thresholds for the normal hearing group with “normal levels” for 2000 and 500 Hz in bolded text. For 2000 Hz ABR, normal levels were 30 dB and 30 nHL for air- and bone-conduction, respectively and for ASSR, 40 and 40 dB HL for air- and bone-conduction, respectively.  For 500 Hz ABR, normal levels were 35 dB and 20 nHL for air- and bone-conduction, respectively and for ASSR, 40 and 30 dB HL for air- and bone-conduction, respectively. Normal levels are higher for air-conduction compared to bone-conduction and higher for ASSR compared to ABR. The 90th percentile was the intended percentile to define the “normal levels”, but due to the distribution of threshold data and the step size used, the normal hearing level represented the level at which 100 percent of infants showed responses. The true 90th percentile is likely 5 dB below the 100th percentile, and thus, normal levels are slightly overestimated.              	   60	                            ABR ASSR  2000 Hz 500 Hz  2000 Hz 500 Hz  dB HL Cumulative Percent dB nHL Cumulative Percent dB HL Cumulative Percent dB HL Cumulative Percent Air- 0.0 4.5 5.0 4.3 0.0 9.5 10.0 4.8 Conduction 10.0 45.5 15.0 13.0 10.0 33.3 20.0 33.3  20.0 72.7 25.0 78.3 20.0 66.7 30.0 66.7  30.0 100.0 35.0 100.0 30.0 85.7 40.0 100.0      40.0 100.0   Bone- 0.0 18.2 .0 25.0 0.0 9.5 .0 8.7 Conduction 10.0 50.0 10.0 70.8 10.0 19.0 10.0 43.5  20.0 81.8 20.0 100.0 20.0 47.6 20.0 73.9  30.0 100.0   30.0 66.7 30.0 100.0      40.0 95.2        50.0 100.0    Table 2.6. The cumulative percentage of responses present for AC and BC ABR and ASSR at 500 Hz and 2000 Hz. Normal levels for each condition are shown in bold and represent the intensity at which more than 90% of subjects demonstrated a response.  Figure 2.2 compares 500 and 2000 Hz air- and bone-conduction ABR and ASSR thresholds by group (with the normal hearing levels denoted by the dashed line). The number of participants at each data point is shown by the colour scale. As mentioned in the Methods section, participants were categorized as having normal hearing4 (normal AC and BC ABR thresholds) or conductive hearing loss (elevated AC ABR thresholds with normal BC ABR thresholds). One will note that there is no overlap in AC ABR thresholds for normal and CHL groups, but significant overlap for ASSR thresholds.    	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  4 8 of 23 infants in the normal hearing group referred on the screening tests (OAEs and tympanometry), but had both AC and BC ABR thresholds at normal levels (Table 2.2). 	   61	                               Figure 2.2. A comparison of 500 and 2000 Hz AC and BC ABR and ASSR thresholds by group. Normal levels for normal-hearing participants for air- and bone-conduction represents the intensity at which all subjects demonstrated a response and is shown by the dotted lines.  	   62	   The Air-Bone Gap  The air-bone gaps at 500 Hz for the groups with normal hearing and conductive losses are shown in Figure 2.3. On average, the mean ABGs for the two groups were as follows: (i) ABR normal hearing: 15(9.25), (ii) ASSR normal hearing: 12(10.56) (iii) ABR CHL: 36(6.40), (iv) ASSR CHL: 21(13.56).  The independent samples t-test comparing mean ABR ABG for the normal hearing and CHL group showed significantly larger ABGs for the CHL (p<0.0001). Similarly, for the ASSR, the ABGs were also significantly larger in CHL group (p=0.028). Levene’s Test for equality of variances was not significant, so equal variances were assumed (see Table 2.7).  The non-parametric independent-samples Mann-Whitney U test (that does not violate the normality assumption like the t-test) also showed that the mean ABG was significantly larger for the conductive group than the normal hearing group for ABR (p<0.0001). For the ASSR, a trend was found showing a tendency for the ABG to be larger in the CHL group, but significance was not reached (p=0.07) [Table 2.7 and Figure 2.3].  The average air-bone gaps at 2000 Hz for normal hearing infants were as follows: (1) ABR: 2.72(12.41) and (2) ASSR: 4.0(16.03). As there were no data points collected at 2000 Hz with conductive hearing loss, no comparison can be made between ABGs between groups.          	   63	  500 Hz                  Figure 2.3. A comparison of the 500 Hz air-bone gap for normal and CHL groups. The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses.	   64	   Independent Samples Test Comparing the 500 Hz ABG for Normal and CHL Groups  t-test for Equality of Means t df Sig. (2-tailed) Mean Diff Std. Er Diff 95% Confidence Interval Lower Upper ABR ABG Equal variances assumed -7.74 35 <0.0001* -21.33 2.76 -26.93 -15.73 ASSR ABG Equal variances assumed -2.29 33 0.03* -9.33 4.07 -17.62 -1.04 * = Significant using a criterion of p<0.05  Table 2.7. Independent samples t-test comparing the 500 Hz air-bone gap for normal and CHL groups.   500 Hz AC and BC ASSR Thresholds for Conductive vs. Normal-Hearing Groups  As mentioned earlier, air-conduction thresholds for the conductive hearing loss group were higher than thresholds for the normal-hearing group for both ABR and ASSR. Bone-conduction thresholds did not differ (Figure 2.4 and 2.5). Due to the non-normal distribution of data, using the Mann-Whitney independent samples non-parametric test may be considered more appropriate than a parametric t-test. Results of this test show that AC ASSR thresholds did not differ significantly (p = 0.089) between normal (median = 30) and CHL (median = 30) groups using a p<0.05 criterion, but did show a trend for thresholds to be higher for the CHL group. Additionally, no difference existed between BC ASSR thresholds for normal (median = 20) and CHL (median = 10) groups, p= 0.425. A parametric test, like the t-test, however, has more power to detect a smaller difference and shows slightly different results (see Table 2.8). The t-test shows that air-conduction ASSR thresholds were significantly higher for the conductive hearing loss group (mean = 36.67) compared with the normal-hearing group (mean = 29.52) p=0.043. There was no significant difference for BC ASSR thresholds between normal (mean = 17.39) and CHL (mean = 15.62) groups, p=0.58, (Figure 2.5). Levene’s Test for equality of variances was not significant, so equal variances were assumed. 	   65	  Air- Conduction    500 Hz     Bone-Conduction      Normal Hearing Group                    Normal Hearing Group                             Figure 2.4. A Comparison of AC and BC ABR (dB nHL) and ASSR (dB HL) thresholds at 500 Hz by participant. Participant numbers are shown on the x-axis.0 10 20 30 40 50 60 6 7 8 18 21 24 27 29 35 36 37 38 39 40 43 44 46 49 50 52 53 54 56 57 60 0 10 20 30 40 50 60 6 7 8 18 21 24 27 29 35 36 37 38 39 40 43 44 46 49 50 52 53 54 56 57 60 	   66	                                         500 Hz Air-Conduction                                                                                       Bone-Conduction                  Figure 2.5. A comparison of 500 Hz air- conduction and bone-conduction ASSR thresholds by group. The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses.	   67	   Independent Samples T-test :  500 Hz Normal vs. CHL groups  t-test for Equality of Means T df Sig. (2-tailed) Mean Diff Std. Er Diff 95% Confidence Interval Lower Upper AC ASSR Equal variances assumed      -2.10 34 0.043* -7.14 3.39 -14.04 -0.24 BC ASSR Equal variances assumed 0.56 36 0.57* 2.06 3.67 -5.39 9.51 * = significant using a p<0.05 criterion  Table 2.8. Independent Samples T-tests: comparing 500 Hz AC ASSR thresholds between groups and BC ASSR thresholds between groups.  Difference Scores   ASSR minus ABR threshold 500 Hz difference scores were calculated for air- and bone-conduction stimuli and are shown in Table 2.9. Overall, mean bone-conduction ASSR thresholds were higher (i.e., poorer) than mean bone-conduction ABR thresholds (Figure 2.5). This mean ASSR-ABR difference score was significant for bone-conduction using the Wilcoxon signed-rank test (p=0.006); although significance was reached for bone-conduction, this was not consistent across subjects and showed large variability (Figure 2.4). On averages, air-conduction showed negligible differences (p=0.486).   ASSR-ABR Difference Scores  N Range (dB) -20 - 30 -35 - 25   Mean (dB) Std. Deviation BC ASSR-BC ABR 37 6.21 12.32 AC ASSR-AC ABR 36 -2.50 15.00 Valid N (listwise) 35    Table 2.9. Air- and bone-conduction ASSR-ABR mean difference scores and measures of dispersion.  	   68	  DISCUSSION  Normal Hearing ASSR Thresholds at 500 Hz  Previous studies using the ASSR have shown maximum normal levels for 500 Hz measured at approximately 50 dB HL for air-conduction and between 20-30 dB HL for bone-conduction stimuli (see Table 2.10 adapted from Table 1.2 shown earlier). The present study showed a maximum of 40 dB for air-conduction and 30 dB HL for bone-conduction. These values are not dramatically different from other studies using different stimuli, but do differ by approximately 10 dB. These differences may be attributed to differences in sample size, stimuli, whether presentation was multiple or single, the age range tested, stopping criteria, EEG noise, recording system (and the system’s detection algorithm).   (Casey & Small, 2014; Foxe, 1993; Holden-Pitt & Albertorio, 1998; Kirikae, 1959; Small & Hu, 2011; Small Sa, 2008; Small & Stapells, 2003, 2008; Stapells & Ruben, 1989; Van Maanen A, 2010; Van Maanen & Stapells, 2009)  Modulation Multiple (M) /Single (S) Age 500 Hz Norm MAX (dB HL) Air Conduction ASSR studies          Lins et al., (1996)** AM M 1-10 mos 48 Cone-Wesson, Parker, et al., (2002) AM S <4 mos >71 John, Brown et al., (2004) MM, AM, AM2 M 3-15 wks >46 Rance et al., (2005) MM S 1-3 mos 52 Swanepoel & Steyn (2005) MM M 3-8 wks 50 Luts et al., (2006)** MM M <3 mos >44 Rance & Tomlin (2006) MM S 6 wks 50 Van Maanen & Stapells (2009) AM (cos3 sinusoids*) M <6 mos 49 Casey & Small (2014) AM2 M  6.5-19 mos 30 Present AM2 M 0-6 mos 40  Bone Conduction ASSR studies      Small & Stapells (2008) MM M  0-11 mos  12-24 Mos 30  40 Small & Stapells (2008) MM M 2-11 mos 30 Casey & Small (2014) AM2 M 6.5-19 mos 20 Present AM2 M 0-6 mos 30 *Cosine3-windowed sinusoids are nearly equivalent to AM2 ** Thresholds were converted from dB SPL to dB HL using ANSI-1996 adjustment values. (Casey & Small, 2014; Foxe, 1993; Holden-Pitt & Albertorio, 1998; Kirikae, 1959; Small & Hu, 2011; Small Sa, 2008; Small & Stapells, 2003, 2008; Stapells & Ruben, 1989; Van Maanen A, 2010; Van Maanen & Stapells, 2009) Table 2.10. A summary of 500 Hz air- and bone-conduction ASSR “normal levels” from the literature.  	   69	  The most appropriate comparisons are to Casey and Small (2014), who used very similar procedures and stimuli, as well as Van Maanen and Stapells (2009) and (2010) who used stimuli that is very similar to AM2. Data for the four studies were collected using different systems. Van Maanen and Stapells used the Intelligent Hearing Systems SmartEP – ASSR, Casey & Small (2014) used the research version of the MASTER, and the present study, the MASTER II clinical software. This is the first study to estimate AC and BC ASSR thresholds using AM2 stimuli for infants using this system. Data from the present study appears to be very similar to Casey and Small (2014) with maximum thresholds differing by only 10 dB and group means by 1.5 dB and 6 dB for air-conduction and bone-conduction, respectively. Both studies show significantly lower thresholds than other studies in the literature (Table 2.10).  We hypothesized that the lower thresholds found in this study could be due to the fact that AM2 stimuli may elicit larger amplitudes that are detectable at lower stimulus intensities. The present study and Casey and Small (2014) are the only published studies to date that used AM2 stimuli for bone-conduction, and both studies showed lower than average thresholds. However, amplitudes are not found to be significantly different when compared between these studies and a study using AM/FM stimuli (Small & Stapells, 2008) (discussed in more detail later in this discussion), so the mechanism underlying this difference in thresholds remains an open question.    500 Hz ASSR Thresholds in Infants with Conductive Hearing Loss  In general, the present study follows the expected trend of air-conduction thresholds from the CHL group being higher than their bone-conduction thresholds, and higher than the AC thresholds from the normal hearing group. Given the small sample size, more data are needed to determine if this relationship will generalize to a larger sample and broader range of hearing loss. With the data that are presented here, it is important to recognize that although a trend is present showing a difference between groups, there is also a significant overlap of thresholds between these groups (Figure 2.5). When several proposed normal hearing levels were compared, the significant overlap between the groups made it extremely difficult to accurately differentiate 	   70	  between normal hearing infants and infants with conductive hearing loss regardless of the normal level chosen (Table 2.11 and 2.12).  500 Hz Accuracy of Classification and Time to Test  Table 2.11 and 2.12 show how accurately the 500 Hz ASSR classifies normal hearing versus conductive hearing loss in terms of percentage of error compared to the gold-standard method, the ABR. This is based on different normal levels proposed by the present study and previously published studies.    Study Criterion  (dB HL)  # with CHL # without CHL Total Van Maanen & Stapells (2009) 50 AC ASSR Positive for HL 0 0 0  50 AC ASSR Negative for HL 15 21 36   Total Error % 100% 0%        Present study 40 AC ASSR Positive for HL 11 0 11  40 AC ASSR Negative for HL 4 21 25   Total Error % 27% 0%        Casey & Small (2014) and Small & Stapells (2008) 30 AC ASSR Positive for HL 6 7 13  30 AC ASSR Negative for HL 9 14 23   Total Error % 60% 33%   Table 2.11. Accuracy of categorization for 500 Hz AC ASSR using different “cut-off criteria”. Percentage of error is calculated for each group.     Study Criterion  (dB HL)  # with CHL # without CHL Total Casey & Small (2014) 30 BC ASSR Positive for HL 0 2 39  30 BC ASSR Negative for HL 0 36 36   Total Error % N/A 5%        Present study 20 BC ASSR Positive for HL 0 9 9  20 BC ASSR Negative for HL 0 29 29   Total Error % N/A 23%   Table 2.12. Accuracy of categorization for 500 Hz AC ASSR using different “cut-off criteria”. Percentage of error is calculated for each group.   	   71	  Choosing a normal-hearing cut-off criteria is still a difficult task. As shown in Tables 2.11 and 2.12, the accuracy of classification varies based on which normal level was used. Using the data from the current study, a 50 dB HL criterion for air-conduction is likely too generous when using AM2 stimuli, as 100% of the infants with hearing loss were considered normal using the ASSR. If a more conservative criterion of 30 dB HL is used, many infants would be placed in the incorrect category so that some infants with hearing loss would be missed and some infants with normal hearing would be detected as having hearing loss. Choosing a criterion in the middle, 40 dB HL, seems to result in the least amount of error. The percentage of error for bone-conduction levels were also calculated, but because the sample for the current study did not include infants with sensorineural hearing loss, it was not possible to determine the number of false negatives for bone-conduction ASSR. Using normal hearing infants, the normal level with the least amount of error was 30 dB HL. More data are needed to make a more confident assessment of the amount of error for a larger population using these levels5.   Overlap of 500 Hz Thresholds Between Groups  Why is there such a large degree of overlap between normal hearing and conductive hearing loss groups when a clearer separation is expected given the ABR results? We have several possible explanations: (1) Perhaps this is the trend for milder degrees of hearing loss and the ASSR is not sensitive enough to reliably detect mild losses, (2) given the study’s dependency on the “gold-standard ABR” for categorization, these results may reflect the intended target population for the BCEHP, which may miss some mild hearing losses, (3) the sample was primarily “at risk” infants who failed or missed the initial screening, and thus may not be entirely representative of true normals, (4) maturation of the ASSR response makes “cutoff” points less clear for 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  5 For a cost-benefit analysis of a new clinical tool, testing time and accuracy are considered. Testing time was not assessed in the present study for the following reasons: (i) testing did not follow the standard BCEHP protocol for either method (i.e., starting levels were not always the “normal” levels and thresholds were estimated rather than stopping when responses were obtained at normal levels) and (ii) testing order (ABR versus ASSR) was not randomized which introduced test bias (i.e., infants were usually more restless for ASSR testing because this method was always tested after the ABR).   	   72	  categorizing hearing loss and the use of an age range of 0-6 months may be too large. These points are explored in more detail below.  (1) Differentiating Between Normal Hearing and Mild Hearing Loss  Although Van Maanen and Stapells (2010) reported success in differentiating between normal and hearing loss groups, this study is not the first to suggest difficulty in differentiating between normal hearing and mild hearing loss. D’haenens et al., (2009), an adult study using the ASSR technique claimed to clearly distinguish moderate SNHL from normal hearing, but the “mild SNHL and NH” and “mild SNHL and moderate SNHL” differentiation was particularly difficult at 0.5 and 2.0 kHz, respectively. They suggested the most appropriate ASSR cutoff point for normality seems to be the 30-dB HL-or-lower criterion. Rance et al., (2006) also commented on having some difficulty capturing normal levels. They discussed that ASSR thresholds showed large variability (large standard deviations) compared to TB-ABR thresholds. The large variations in ASSR thresholds in this study are thus not a new finding (Rance et al., 2006; Luts et al., 2006; Rickards et al., 1994; Savio et al., 2001) and may explain the difficulty in differentiating between degrees of hearing loss. This resulted in significant errors in classifying infants into normal or hearing loss groups using the ASSR with normal levels from Van Maanen and Stapells (2009), Small and Stapells (2008), Casey and Small (2014) and the present study (Table 2.11 and 2.12). A much larger sample size is needed to measure the sensitivity and specificity of the ASSR as a diagnostic tool.   (2) The BCEHP has established its “normal levels” based on the target population for the program and correction factors from dB nHL to dB eHL, or estimated behavioral hearing level. When normal levels were developed for the ABR, these levels were developed using two very different pieces of information; (a) the actual electrophysiological data collected for normal hearing infants compared to behavioural data, and (b) the levels at which clinicians believed an intervention should be planned and, (c) the infant should not be discharged from the program. What this means in a practical sense is that based on these three types of information together, the program will be discharging some number of infants that have a slight hearing loss or a mild 	   73	  hearing loss on the lower end of mild. This was then coupled with differences in protocols for well-babies versus at risk babies in order to identify those at risk for hearing loss under a more strict testing protocol. Next, correction factors were determined to estimate the behavioural audiogram from the ABR. These values differed for normal hearing infants compared to infants with more severe degrees of hearing loss as the severe hearing losses more closely approximated behavioural thresholds than did normal thresholds. This limitation is then carried through to the present study, and so even when using the normal ABR levels provided by the program to determine presence or absence of hearing loss, by definition, we still expect to miss some infants with slight and mild hearing losses. It follows that some degree of overlap will exist between the normal hearing group and hearing loss group simply based upon the normal levels that are used to categorize. For this reason, it may not be possible to detect all mild hearing losses with the ASSR if we continue to use the same ABR normal hearing levels.   (3) The present study showed that 15/38 infants in the study had conductive hearing loss, which is significantly higher than the incidence of conductive hearing loss predicted in the general infant population6. It is likely due to recruitment procedures being biased toward at-risk infants (failed or missed screening at birth). Given this bias, it is possible that this sample represented infants with hearing sensitivity at the top of the normal range, and the bottom of the mild conductive loss range. This would be more susceptible to an overlap in thresholds than would low-risk infants with thresholds on the better end of normal. Using the sample of healthy, low-risk infants from Casey and Small (2014) as the normal hearing group, this overlap between normal and CHL groups is not as large (see Figure 2.6 below).     	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  6The incidence of conductive hearing loss and middle ear pathology in infancy reported in an Australian study is 2.97 per 1,000 while prevalence of middle ear pathology (with or without conductive hearing loss) was 4.36 per 1,000 (Aithal, 2012). In British Columbia, births for 2013 totaled approximately 40,000 (according to BC statistics, 2014). If the prevalence of conductive hearing loss is similar to that shown by the Australian study, this supposes that approximately 120 infants in BC will be diagnosed with conductive hearing loss each year. 	   74	               Figure 2.6. A comparison of AC ASSR thresholds at 500 Hz using normal data from Casey and Small (2014). The median is denoted by the bolded line.  The box surrounding the median indicates the middle 50 percent of the distribution, and the top and bottom 25 percent are denoted by the whiskers. Outliers are shown by the crosses.    The reality is, in clinical practice, we see patients that come from both types of samples; patients like those in Casey and Small (2014) and patients like those in the present study. It is important that that the diagnostic tools we use clearly define hearing loss and normal hearing. So, while the separation between the groups appears to be larger in Casey and Small (2014), it is still important to keep in mind the large overlap that exists using other populations.  (4) Maturation  As discussed in previous sections (see “Maturation”), although the inner ear is completely developed at birth, the auditory neural pathway is not. To remind the reader, for the ASSR, air-conduction thresholds tend to become better with age, while bone-conduction thresholds tend to worsen. Rance et al., (2006) showed this time course of maturation up to 6 weeks of age for air-conduction thresholds, and Small and Stapells (2006, 2008a) show this trend for bone-conduction thresholds. Rance et al., (2006) suggested that at earlier ages, variation in thresholds is quite large. It is possible that the age range of participants somewhat contributed to the overlap shown between conductive and normal hearing groups’ AC ASSR thresholds.   	   75	  The ASSR air-bone gap shown in the present study showed a trend for ABG in the conductive hearing loss group to be larger than the normal hearing group. This was very much expected. The maturational air-bone gap has been estimated to be approximately at least 20 dB (Small & Stapells, 2008; Van Maanen & Stapells, 2009), therefore with a mean ABG of 20 dB for the conductive hearing loss group, this would not be considered clinically significant. Again, given the large amount of variation in ASSR thresholds in normal infants, it is still unclear what an appropriate “normal level” for thresholds and “normal size” for the ABG should be.  500 Hz ABR and ASSR Threshold Differences  Another reasonable question to ask is “why is there a difference between ASSR and ABR thresholds for the same infant?” Previous studies have shown strong correlations between ABR and ASSR thresholds, and they are thought to measure responses from approximately the same part of the auditory system. For these reasons, it is a reasonable expectation for ABR and ASSR thresholds to be similar. Studies like Rance et al., (2006), however, have shown that ASSR thresholds are consistently higher than TB-ABR thresholds for 500 and 4000 Hz stimuli that they used in testing. For 500 Hz, ASSR thresholds were, on average, 4.8 dB higher. They suggested that this difference might be due to maturational effects following different time courses between the two tests and calibration differences. Being that the ASSR involves more than just a brainstem response, it is expected that the ABR and ASSR may have different developmental time courses. Peak-to-peak equivalent dB SPL values are between 5-15 dB higher for tone bursts at 0 dB nHL compared to modulated tones at 0 dB HL. When Rance et al., (2006) calibrated both stimuli using the same units (peak equivalent dB SPL), the results for ABR and ASSR thresholds were similar. Studying the maturation of the responses in the 6-week period, their study found that over time, ABR thresholds remained stable, whereas the ASSR thresholds changed and ultimately reduced the difference between ABR and ASSR thresholds over time. On average, the mean bone-conduction ASSR thresholds in the present study were approximately 7 dB higher than BC ABR thresholds, and air-conduction ASSR thresholds were 4.1 dB higher than ABR thresholds. When thresholds are converted to dB ppe SPL for air-conduction and dB re:1µN for bone-conduction (ANSI 1996; Small & Stapells, 2003), differences still exist (see Table 2.13). 	   76	  For example, mean ABR thresholds were still approximately 12 dB SPL and 2 dB re:1µN  higher than ASSR thresholds for air- and bone-conduction, respectively. The ASSR-ABR differences for the present study are in part due to a calibration artifact, but this is clearly not the only reason for the differences. Other contributors to the ASSR-ABR differences are likely an insufficiently large sample size, difference in stimuli, different starting levels and the use of a 10 dB step-size. For example, with a starting level that differed by 5 dB between ABR (35 dB nHL) and ASSR (30 dB HL) and by using a 10 dB step size, we automatically introduced an ASSR-ABR difference score of -5 dB. If smaller step sizes, or similar starting levels were used, this difference may have followed the trend of Rance et al., when converted to like units.    Normal Hearing Group  RETSPL* RETFL* Mean AC dB ppe SPL Mean BC dB re:1µN ABR ASSR 22 67 25. 43 dB nHL 47.43 10.41 dB nHL 77.41 5.5 58 29.52 dB HL 35.02 17.39 dB HL 75.39 *ANSI (1996); Small & Stapells(2003)   Table 2.13. Normal Hearing Group: 500 Hz ASSR (dB HL) and ABR (dB nHL) mean thresholds converted to dB ppe SPL and dB re:1µN.  Differences in response amplitudes are also important to take into account when discussing differences. ASSR response amplitudes are known to be significantly smaller than ABR amplitudes, so that ABR thresholds can typically be tracked down and detected at much lower intensities than ASSR thresholds (John et al., 2004; Luts et al., 2006; Sininger et al., 1997). Although response amplitudes at each intensity were not analyzed here, compared to other studies, mean AC ASSR amplitudes at threshold in the present study were quite similar to amplitudes reported at 10 and 20 dB in other studies. Van Maanen and Stapells (2009) showed responses with mean amplitudes between 15-20 nV, and Casey and Small (2014) with mean amplitudes between at 12-18 nV 10 and 20 dB HL. The present study showed mean amplitudes of 21.92 nV at threshold (ranging from 10-40 dB HL) for normal hearing infants and 24.5 for CHL infants (ranging from 20-50 dB HL).   BC ASSR amplitudes were also comparable to Small and Stapells (2008a), but significantly larger than Casey and Small (2014). Small and Stapells (2008a) found that BC amplitudes at 500 Hz were between 20-40 nV for stimuli between 10-30  dB HL. Casey and Small (2014) showed 	   77	  mean amplitudes between 12-18 nV at 10 and 20 dB HL. The present study showed amplitudes 31.5 and 28.2 nV for normal hearing (range 0-30 dB HL) and CHL (range 0-40 dB HL) groups, respectively. These studies used a slightly smaller no-response noise criterion (<11 nV), which may have allowed for responses of smaller amplitude to be detected.   It has been proposed that reduced ability to process responses at high presentation rates may be the cause for the reduced ASSR response amplitude when compared to the ABR as effects of increasing stimulus rate have been shown on the ABR in the form of latency increases and amplitude decreases (Burkhard, Shi, & Hecox, 1990; Sininger & Don, 1989; Rance, 2008).      	   78	  CONCLUSION AND CLINICAL APPLICATIONS   In conclusion, although the present study was conducted using a relatively small group of infants, it is clear that there is significant overlap between 500 Hz AC ASSR thresholds for groups with normal hearing and mild conductive hearing loss. This issue poses a significant challenge for clinicians hoping to be able to identify a mild hearing loss. Clinicians must decide what degree of hearing loss (and at which frequencies) is clinically relevant for intervention before determining if the ASSR is an appropriate as a screening tool. For example, perhaps identifying a mild conductive hearing loss only at 500 Hz is not significant enough to warrant intervention or rigorous follow-up. It also appears that a trend is emerging that shows a difference between air-conduction thresholds for normal hearing and conductive hearing losses. Although this relationship was not shown to be statistically significant using a conservative criterion, a more liberal criterion shows a significant difference between the groups. The clinical community can remain hopeful that with the addition of more studies to the literature, we may be able to find optimal criteria and stimuli to use the ASSR diagnostically with normal hearing infants and infants with hearing loss. Until then, this study has suggested that we must be very cautious when using the ASSR, even as a screening tool, as it is difficult to determine a cut-off criterion for normal hearing may be missing even mild conductive hearing losses in infants. By comparing the present study to other studies in the literature using AM2 stimuli, it is suggested that 40 dB HL and 30 dB HL be used as a normal hearing maximum response for air- and bone-conduction respectively at 500 Hz.   In addition, as the air-bone gap is a clinically useful tool for confirming the presence of conductive hearing loss, it is desirable to have a clear guideline for the clinically significant size of the gap. To do this, it is important to understand the time course of the maturational air-bone gap. The present study showed a difference between the mean size of the ABG between normal hearing and conductive hearing loss groups, but again, there was a significant overlap in the size of the ABG for both groups. For this reason, until more data is added to the literature, it is not possible to determine the size of a clinically significant ABG.(Edwards, Durieux-Smith, & Picton, 1985; J. L. Hatton, Janssen, & Stapells, 2012; J. L. Hatton & Stapells, 2013; Mehl & Thomson, 1998; Probst, Lonsbury-Martin, Martin, & Coats, 1987; Stapells & Mosseri, 1991; Winskel, 2006; Zumach, Gerrits, Chenault, & Anteunis, 2010)  	   79	  This study has provided a starting point for comparing ASSR thresholds to AM2 to an electrophysiological gold-standard test in infants with conductive hearing loss, but much more research is needed before we can confidently use the ASSR as a diagnostic tool in practice.  (Linden et al., 1985; Plourde et al., 1991; Rodriguez et al., 1986)   	   80	  FUTURE RESEARCH AND LIMITATIONS  The methodology for like studies in the literature is extremely inconsistent. It makes it very challenging to compare data between studies and to create a large repertoire of a like data to increase the power of analyses. The recruitment process for infants with hearing loss and infant sleep time required can be extremely challenging. For this reason, sufficient sample sizes are very difficult to obtain in reasonable amounts of time. It seems to be unclear what an attainable and yet large enough sample size is to be able to have strong statistical power, but is also be feasible to collect in the clinical environment. The number of participants in the present study certainly does not allow strong statistical inferences to be made. The collaboration of several research teams to compile similar data is certainly recommended.  More studies comparing ASSR thresholds to the gold-standard tone-burst ABR thresholds with infants with hearing loss are clearly warranted. Increased sample sizes, narrow age ranges, strict stopping criteria, decreased noise levels are recommended to investigate which parameters are optimal for maximal sensitivity and specificity of the ASSR. In addition, studies using infants with varying degrees of hearing loss (CHL and SNHL) using AM2 are to determine if different trends exist for hearing losses worse than mild. It is also necessary for other frequencies (especially 1000, 2000, and 4000 Hz) to be tested to determine if the same trends apply at other frequencies. With these pieces of information, it may be possible to use the ASSR, at least as a screening tool, to determine if follow-up and/or intervention is needed. For example, it is likely that the ASSR will be better able to differentiate between normal hearing and mild conductive hearing loss at other frequencies, but the research needs to be conducted. It may be acceptable to miss a mild conductive hearing loss at 500 Hz if the other frequencies have thresholds within the range for normal hearing.  There also seems to be a need to investigate what the best type of analysis is for the sample sizes and type of data obtained. Being that the end-goal of this type of research is to bring a more efficient diagnostic tool into the clinic, there may be more appropriate analyses for the type of data collected that has a more clinical decision-making focus for a field in healthcare.    	   81	  REFERENCES 	  Anson, B., Bast, H., & Richany, S. (1955). The fetal and early postnatal development of the tympanic ring and related structures in man. Ann Otol Rhinol Laryngol(64), 802-823.   Aoyagi, M., Kiren, T., Furuse, H., Fuse, T., Suzuki, Y., Yokota, M., & Koike, Y. (1994). Pure-tone threshold prediction by 80-Hz amplitude-modulation following response. Acta Otolaryngol Suppl, 511, 7-14.   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. Hear Res, 83(1-2), 1-8.   Barany, E. (1938). A contribution to the physiology of bone conduction. Acta Otolaryngol, 22, 1-233.   Basar, E., Rosen, B., Basar-Eroglu, C., & Greitschus, F. (1987). The associations between 40 Hz-EEG and the middle latency response of the auditory evoked potential. Int J Neurosci, 33(1-2), 103-117.   BCEHP. (2012). Audiology Assessment Protocol, retrieved from http://www.phsa.ca/NR/rdonlyres/2CEAFB48-­‐66C2-­‐4D7D-­‐954A-­‐4285C7771C36/0/BCEHPAudiologyAssessmentProtocol.pdf on September 1, 2015.  Bowd, A. D. (2005). Otitis media: health and social consequences for aboriginal youth in Canada's north. Int J Circumpolar Health, 64(1), 5-15.   Brownell, W. E. (1990). Outer hair cell electromotility and otoacoustic emissions. Ear Hear, 11(2), 82-92.   Campbell, K. B., & Bartoli, E. A. (1986). Human auditory evoked potentials during natural sleep: the early components. Electroencephalogr Clin Neurophysiol, 65(2), 142-149.  	   82	   Carhart, R. (1950). The clinical application of bone conduction audiometry. Trans Am Acad Ophthalmol Otolaryngol, 54, 699-707.   Casey, K. A., & Small, S. A. (2014). Comparisons of auditory steady state response and behavioral air conduction and bone conduction thresholds for infants and adults with normal hearing. Ear Hear, 35(4), 423-439. doi: 10.1097/aud.0000000000000021  Chou, Y. F., Chen, P. R., Yu, S. H., Wen, Y. H., & Wu, H. P. (2012). Using multi-stimulus auditory steady state response to predict hearing thresholds in high-risk infants. Eur Arch Otorhinolaryngol, 269(1), 73-79. doi: 10.1007/s00405-011-1604-0  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. J Acoust Soc Am, 90(5), 2467-2479.   Cone-Wesson, B., Dowell, R. C., Tomlin, D., Rance, G., & Ming, W. J. (2002). The auditory steady-state response: comparisons with the auditory brainstem response. J Am Acad Audiol, 13(4), 173-187; 225-176.   Cone-Wesson, B., Parker, J., Swiderski, N., & Rickards, F. (2002). The auditory steady-state response: full-term and premature neonates. J Am Acad Audiol, 13(5), 260-269.   D'Haenens, W., Dhooge, I., Maes, L., Bockstael, A., Keppler, H., Philips, B., . . . Vinck, B. M. (2009). The clinical value of the multiple-frequency 80-Hz auditory steady-state response in adults with normal hearing and hearing loss. Arch Otolaryngol Head Neck Surg, 135(5), 496-506.   Dauman, R. (2013). Bone conduction: An explanation for this phenomenon comprising complex mechanisms. European Annals of Otorhinolaryngology, Head and Neck Diseases European Annals of Otorhinolaryngology, Head and Neck Diseases, 130(4), 209-213.  	   83	   Despland, P. A., & Galambos, R. (1980). Use of the auditory brainstem responses by prematures and newborns infants. Neuropadiatrie, 11(2), 99-107.   Dimitrijevic, A., John, M. S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., . . . Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. J Am Acad Audiol, 13(4), 205-224.   Dirks, D., & Swindeman, J. G. (1967). The variability of occluded and unoccluded bone-conduction thresholds. J Speech Hear Res(10), 232-249.   Dobie, R. A., & Wilson, M. J. (1996). A comparison of t test, F test, and coherence methods of detecting steady-state auditory-evoked potentials, distortion-product otoacoustic emissions, or other sinusoids. J Acoust Soc Am, 100(4 Pt 1), 2236-2246.   Edwards, C. G., Durieux-Smith, A., & Picton, T. W. (1985). Neonatal auditory brain stem responses from ipsilateral and contralateral recording montages. Ear Hear, 6(4), 175-178.  Elpern, B., & Naunton, R. (1963a). The stability of the occlusion effect. Arch Otolaryngol(77), 376-384.   Elpern, B., & Naunton, R. (1963b). The stability of the occlusion effect. Arch Otolaryngol(77), 44-52.   Finitzo-Hieber, T., Freeman, F. J., Gerling, I. J., Dobson, L., & Schaefer, S. D. (1982). Auditory brainstem response abnormalities in adductor spasmodic dysphonia. Am J Otolaryngol, 3(1), 26-30.   Galambos, R., Makeig, S., & Talmachoff, P. J. (1981). A 40-Hz auditory potential recorded from the human scalp. Proc Natl Acad Sci U S A, 78(4), 2643-2647.   	   84	  Geisler, C. D. (1960). Average response to clicks in man recorded by scalp electrodes. M.I.T. Technical Report(380), 1-158.   Gorga, M. P., Kaminski, J. R., Beauchaine, K. L., Jesteadt, W., & Neely, S. T. (1989). Auditory brainstem responses from children three months to three years of age: normal patterns of response. II. J Speech Hear Res, 32(2), 281-288.   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 Hear, 25(3), 302-307.   Gorga, M. P., Reiland, J. K., Beauchaine, K. A., Worthington, D. W., & Jesteadt, W. (1987). Auditory brainstem responses from graduates of an intensive care nursery: normal patterns of response. J Speech Hear Res, 30(3), 311-318.   Gutschalk, A., Mase, R., Roth, R., Ille, N., Rupp, A., Hahnel, S., . . . Scherg, M. (1999). Deconvolution of 40 Hz steady-state fields reveals two overlapping source activities of the human auditory cortex. Clin Neurophysiol, 110(5), 856-868.   Han, D., Mo, L., Liu, H., Chen, J., & Huang, L. (2006). Threshold estimation in children using auditory steady-state responses to multiple simultaneous stimuli. ORL J Otorhinolaryngol Relat Spec, 68(2), 64-68.  Hashimoto, I., Ishiyama, Y., Yoshimoto, T., & Nemoto, S. (1981). Brain-stem auditory-evoked potentials recorded directly from human brain-stem and thalamus. Brain, 104(Pt 4), 841-859.   Hatton, J., & Stapells, D. R. (2011). The efficiency of the single- versus multiple-stimulus auditory steady state responses in infants. Ear Hear, 32(3), 349-357.  	   85	  Herdman, A. T., Picton, T. W., & Stapells, D. R. (2002). Place specificity of multiple auditory steady-state responses. J Acoust Soc Am, 112(4), 1569-1582.   Herdman, A. T., & Stapells, D. K. (2003). Auditory steady-state response thresholds of adults with sensorineural hearing impairments. Int J Audiol, 42(5), 237-248.   Hodgson, W., & Tillman, T. (1966). Reliability of bone conduction occlusion effects in normals. J Auditory Res, 6, 141-151.   Holden-Pitt, L., & Albertorio, J. (1998). Thirty years of the Annual Survey of Deaf and Hard-of-Hearing Children & Youth: a glance over the decades. Am Ann Deaf, 143(2), 72-76.   Holm, V. A., & Kunze, L. H. (1969). Effect of chronic otitis media on language and speech development. Pediatrics, 43(5), 833-839.   Holte, L., Margolis, R. H., & Cavanaugh, R. M., Jr. (1991). Developmental changes in multifrequency tympanograms. Audiology, 30(1), 1-24.   Homma, K., Shimizu, Y., Kim, N., Du, Y., & Puria, S. (2010). Effects of ear-canal pressurization on middle-ear bone- and air-conduction responses. Hear Res, 263(1-2), 204-215. doi: 10.1016/j.heares.2009.11.013  Hooks, R. G., & Weber, B. A. (1984). Auditory brain stem responses of premature infants to bone-conducted stimuli: a feasibility study. Ear Hear, 5(1), 42-46.   Huizing, E. (1960). Bone conduction: the influence of the middle ear. Acta Otolaryngol(155), 1-99.   Ishida, I. M., Cuthbert, B. P., & Stapells, D. R. (2011). Multiple auditory steady state response thresholds to bone conduction stimuli in adults with normal and elevated thresholds. Ear Hear, 32(3), 373-381. 	   86	   Janssen, R. M., Usher, L., & Stapells, D. R. (2010). The British Columbia's Children's Hospital tone-evoked auditory brainstem response protocol: how long do infants sleep and how much information can be obtained in one appointment? Ear Hear, 31(5), 722-724.   Jeng, F. C., Brownt, C. J., Johnson, T. A., & Vander Werff, K. R. (2004). Estimating air-bone gaps using auditory steady-state responses. J Am Acad Audiol, 15(1), 67-78.   Jewett, D. L. (1970). Volume-conducted potentials in response to auditory stimuli as detected by averaging in the cat. Electroencephalogr Clin Neurophysiol, 28(6), 609-618.   Jewett, D. L., & Williston, J. S. (1971). Auditory-evoked far fields averaged from the scalp of humans. Brain, 94(4), 681-696.   Jiang, Z. D., Brosi, D. M., & Wilkinson, A. R. (1998). Immaturity of electrophysiological response of the neonatal auditory brainstem to high repetition rates of click stimulation. Early Hum Dev, 52(2), 133-143.   John, M. S. (2003). Stimulus Setup Instructions, retrieved from http://www.mastersystem.ca/index.php?section=700	  -­‐	  mod_4174 on September 1, 2015.  John, M. S., Brown, D. K., Muir, P. J., & Picton, T. W. (2004). Recording auditory steady-state responses in young infants. Ear Hear, 25(6), 539-553.   John, M. S., Dimitrijevic, A., & Picton, T. W. (2001). Weighted averaging of steady-state responses. Clin Neurophysiol, 112(3), 555-562.   John, M. S., Dimitrijevic, A., & Picton, T. W. (2003). Efficient stimuli for evoking auditory steady-state responses. Ear Hear, 24(5), 406-423.   	   87	  John, M. S., Lins, O. G., Boucher, B. L., & Picton, T. (1998). Multiple auditory steady-state responses (MASTER): Stimulus and recording parameters. Audiology(37), 59-82.   John, M. S., Purcell, D. W., Dimitrijevic, A., & Picton, T. W. (2002). Advantages and caveats when recording steady-state responses to multiple simultaneous stimuli. J Am Acad Audiol, 13(5), 246-259.   Johnson, T. A., & Brown, C. J. (2005). Threshold Prediction Using the Auditory Steady-State Response and the Tone Burst Auditory Brain Stem Response: A Within-Subject Comparison. Ear and hearing, 26(6), 559-576.   Keefe, D. H., Bulen, J. C., Arehart, K. H., & Burns, E. M. (1993). Ear-canal impedance and reflection coefficient in human infants and adults. J Acoust Soc Am, 94(5), 2617-2638.   Kiren, T., Aoyagi, M., Furuse, H., & Koike, Y. (1994). An experimental study on the generator of amplitude-modulation following response. Acta Otolaryngol Suppl, 511, 28-33.   Kirikae, I. (1959). An experimental study on the fundamental mechanism of bone conduction. Acta Otolaryngol Suppl, 145, 1-111.   Korczak, P., Smart, J., Delgado, R., Strobel, T. M., & Bradford, C. (2012). Auditory steady-state responses. J Am Acad Audiol, 23(3), 146-170. doi: 10.3766/jaaa.23.3.3  Kramer, S. J. (1992). Frequency-specific auditory brainstem responses to bone-conducted stimuli. Audiology, 31(2), 61-71.   Kuwada, S., Anderson, J. S., Batra, R., Fitzpatrick, D. C., Teissier, N., & D'Angelo, W. R. (2002). Sources of the scalp-recorded amplitude-modulation following response. J Am Acad Audiol, 13(4), 188-204.   	   88	  Lasky, R. E. (1984). A developmental study on the effect of stimulus rate on the auditory evoked brain-stem response. Electroencephalogr Clin Neurophysiol, 59(5), 411-419.   Lasky, R. E. (1997). Rate and adaptation effects on the auditory evoked brainstem response in human newborns and adults. Hear Res, 111(1-2), 165-176.   Lasky, R. E., & Rupert, A. L. (1982). Temporal masking of auditory evoked brainstem responses in human newborns and adults. Hear Res, 6(3), 315-334.   Lee, C. Y., Hsieh, T. H., Pan, S. L., & Hsu, C. J. (2007). Thresholds of tone burst auditory brainstem responses for infants and young children with normal hearing in Taiwan. J Formos Med Assoc, 106(10), 847-853.  Lev, A., & Sohmer, H. (1972). Sources of averaged neural responses recorded in animal and human subjects during cochlear audiometry (Electro-Cochleogram). Arch. klin. exp., 201(2), 79-90.   Levi, E. C., Folsom, R. C., & Dobie, R. A. (1993). Amplitude-modulation following response (AMFR): effects of modulation rate, carrier frequency, age, and state. Hear Res, 68(1), 42-52.   Levitt, H., McGarr, N., & Geffner, D. (1987). Development of language and communication skills in hearing-impaired children. Introduction. ASHA Monogr(26), 1-8.   Lierle, D. M., & Reger, S. N. (1946). Correlations between bone and air conduction acuity measurements over wide frequency ranges in different types of hearing impairments. Laryngoscope, 56, 187-224.   Lins, O. G., Picton, T. W., Boucher, B. L., Durieux-Smith, A., Champagne, S. C., Moran, L. M., . . . Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear Hear, 17(2), 81-96.  	   89	   Luts, H., Desloovere, C., & Wouters, J. (2006). Clinical application of dichotic multiple-stimulus auditory steady-state responses in high-risk newborns and young children. Audiol Neurootol, 11(1), 24-37.   Luts, H., Van Dun, B., Alaerts, J., & Wouters, J. (2008). The influence of the detection paradigm in recording auditory steady-state responses. Ear Hear, 29(4), 638-650.   Mackersie, C. L., & Stapells, D. R. (1994). Auditory brainstem wave I response prediction of conductive component in infants and young children. American Journal of audiology(3), 52-58.   Marcoux, A. M. (2011). Maturation of auditory function related to hearing threshold estimations using the auditory brainstem response during infancy. Int J Pediatr Otorhinolaryngol, 75(2), 163-170.   Maurizi, M., Almadori, G., Paludetti, G., Ottaviani, F., Rosignoli, M., & Luciano, R. (1990). 40-Hz steady-state responses in newborns and in children. Audiology, 29(6), 322-328.   Mehl, A. L., & Thomson, V. (1998). Newborn hearing screening: the great omission. Pediatrics, 101(1), E4.   Mo, L., & Stapells, D. R. (2008). The effect of brief-tone stimulus duration on the brain stem auditory steady-state response. Ear Hear, 29(1), 121-133.  Moller, A. R., Jannetta, P., Bennett, M., & Moller, M. B. (1981). Intracranially recorded responses from the human auditory nerve: new insights into the origin of brain stem evoked potentials (BSEPs). Electroencephalogr Clin Neurophysiol, 52(1), 18-27.   Moller, A. R., & Jannetta, P. J. (1982a). Auditory evoked potentials recorded intracranially from the brain stem in man. Exp Neurol, 78(1), 144-157. 	   90	    Moller, A. R., & Jannetta, P. J. (1982b). Comparison between intracranially recorded potentials from the human auditory nerve and scalp recorded auditory brainstem responses (ABR). Scand Audiol, 11(1), 33-40.   Moller, A. R., & Jannetta, P. J. (1983). Auditory evoked potentials recorded from the cochlear nucleus and its vicinity in man. J Neurosurg, 59(6), 1013-1018.  Moller, A. R., Jannetta, P. J., & Moller, M. B. (1981). Neural generators of brainstem evoked potentials. Results from human intracranial recordings. Ann Otol Rhinol Laryngol, 90(6 Pt 1), 591-596.   Moller, A. R., Jho, H. D., Yokota, M., & Jannetta, P. J. (1995). Contribution from crossed and uncrossed brainstem structures to the brainstem auditory evoked potentials: a study in humans. Laryngoscope, 105(6), 596-605.   Nagashima, H., Udaka, J., Chida, I., Shimada, A., Kondo, E., & Takeda, N. (2013). Air-bone gap estimated with multiple auditory steady-state response in young children with otitis media with effusion. Auris Nasus Larynx, 40(6), 534-538.   Nousak, J. M., & Stapells, D. R. (1992). Frequency specificity of the auditory brain stem response to bone-conducted tones in infants and adults. Ear Hear, 13(2), 87-95.   OIHP. (2008). Ontario Infant Hearing Program Audiologic assessment protocol, retrieved from https://http://www.mountsinai.on.ca/care/infant-­‐hearing-­‐program/documents/IHPAudiologicAssessmentProtocol3.1FinalJan2008.pdf on September 1, 2015.  Olszewski, J. (1990). The morphometry of the ear ossicles in humans during development. anatomischer Anzeiger, 171(3), 187-191.   	   91	  Picton, T. W., Dimitrijevic, A., John, M. S., & Van Roon, P. (2001). The use of phase in the detection of auditory steady-state responses. Clin Neurophysiol, 112(9), 1698-1711.   Picton, T. W., Dimitrijevic, A., Perez-Abalo, M. C., & Van Roon, P. (2005). Estimating audiometric thresholds using auditory steady-state responses. J Am Acad Audiol, 16(3), 140-156.   Picton, T. W., & John, M. S. (2004). Avoiding electromagnetic artifacts when recording auditory steady-state responses. J Am Acad Audiol, 15(8), 541-554.   Picton, T. W., John, M. S., Purcell, D. W., & Plourde, G. (2003). Human auditory steady-state responses: the effects of recording technique and state of arousal. Anesth Analg, 97(5), 1396-1402.   Picton, T. W., & Smith, A. D. (1978). The practice of evoked potential audiometry. Otolaryngol Clin North Am, 11(2), 263-282.   Picton, T. W., van Roon, P., & John, M. S. (2007). Human auditory steady-state responses during sweeps of intensity. Ear Hear, 28(4), 542-557.   Probst, R., Lonsbury-Martin, B. L., Martin, G. K., & Coats, A. C. (1987). Otoacoustic emissions in ears with hearing loss. Am J Otolaryngol, 8(2), 73-81.   Qi, L., Funnell, W. R., & Daniel, S. J. (2008). A nonlinear finite-element model of the newborn middle ear. J Acoust Soc Am, 124(1), 337-347.   Qi, L., Liu, H., Lutfy, J., Funnell, W. R., & Daniel, S. J. (2006). A nonlinear finite-element model of the newborn ear canal. J Acoust Soc Am, 120(6), 3789-3798.   Rance, G., & Rickards, F. (2002). Prediction of hearing threshold in infants using auditory steady-state evoked potentials. J Am Acad Audiol, 13(5), 236-245.  	   92	   Rance, G., Rickards, F. W., Cohen, L. T., De Vidi, S., & Clark, G. M. (1995). The automated prediction of hearing thresholds in sleeping subjects using auditory steady-state evoked potentials. Ear Hear, 16(5), 499-507.   Rance, G., & Tomlin, D. (2006). Maturation of auditory steady-state responses in normal babies. Ear Hear, 27(1), 20-29.   Rance, G., Tomlin, D., & Rickards, F. W. (2006). Comparison of auditory steady-state responses and tone-burst auditory brainstem responses in normal babies. Ear Hear, 27(6), 751-762.  Regan, D. (1966). Some characteristics of average steady-state and transient responses evoked by modulated light. Electroencephalogr Clin Neurophysiol, 20(3), 238-248.   Rickards, F. W., Tan, L. E., Cohen, L. T., Wilson, O. J., Drew, J. H., & Clark, G. M. (1994). Auditory steady-state evoked potential in newborns. Br J Audiol, 28(6), 327-337.   Rodrigues, G. R., Lewis, D. R., & Fichino, S. N. (2010). Steady-state auditory evoked responses in audiological diagnosis in children: a comparison with brainstem evoked auditory responses. Braz J Otorhinolaryngol, 76(1), 96-101.   Rodriguez, R., Picton, T., Linden, D., Hamel, G., & Laframboise, G. (1986). Human auditory steady state responses: effects of intensity and frequency. Ear Hear, 7(5), 300-313.   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. Hear Res, 83(1-2), 9-18.   Saunders, J., Kaltenbach, J., & Relkin, E. (1983). The structural and functional development of the outer and middle ear Development of auditory and vestibular systems. New York, NY: Academic. 	   93	   Savio, G., Cardenas, J., Perez Abalo, M., Gonzalez, A., & Valdes, J. (2001). The low and high frequency auditory steady state responses mature at different rates. Audiol Neurootol, 6(5), 279-287.   Shehata, W. E., Brownell, W. E., & Dieler, R. (1991). Effects of salicylate on shape, electromotility and membrane characteristics of isolated outer hair cells from guinea pig cochlea. Acta Otolaryngol, 111(4), 707-718.   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 Hear, 28(1), 83-98.   Small, S. A., & Hu, N. (2011). Maturation of the occlusion effect: a bone conduction auditory steady state response study in infants and adults with normal hearing. Ear Hear, 32(6), 708-719.   Small, S. A., & Love, A. (2014). An investigation into the clinical utility of ipsilateral/contralateral asymmetries in bone-conduction auditory steady-state responses. Int J Audiol, 53(9), 604-612.  Small Sa, S. D. R. (2008). Maturation of bone conduction multiple auditory steady-state responses. International journal of audiology, 47(8), 476-488.   Small, S. A., & Stapells, D. R. (2003). Normal brief-tone bone-conduction behavioral thresholds using the B-71 transducer: three occlusion conditions. J Am Acad Audiol, 14(10), 556-562.   Small, S. A., & Stapells, D. R. (2004). Artifactual responses when recording auditory steady-state responses. Ear Hear, 25(6), 611-623.   	   94	  Small, S. A., & Stapells, D. R. (2008). Normal ipsilateral/contralateral asymmetries in infant multiple auditory steady-state responses to air- and bone-conduction stimuli. Ear Hear, 29(2), 185-198.   Stapells, D. R. (2000). Threshold estmation 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. Electroencephalogr Clin Neurophysiol, 71(4), 289-295.   Stapells, D. R., Gravel, J. S., & Martin, B. A. (1995). Thresholds for auditory brain stem responses to tones in notched noise from infants and young children with normal hearing or sensorineural hearing loss. Ear Hear, 16(4), 361-371.   Stapells, D. R., & Mosseri, M. (1991). Maturation of the contralaterally recorded auditory brain stem response. Ear Hear, 12(3), 167-173.   Stapells, D. R., & Ruben, R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. Ann Otol Rhinol Laryngol, 98(12 Pt 1), 941-949.   Starr, A., Amlie, R. N., Martin, W. H., & Sanders, S. (1977). Development of auditory function in newborn infants revealed by auditory brainstem potentials. Pediatrics, 60(6), 831-839.   Starr, A., & Hamilton, A. E. (1976). Correlation between confirmed sites of neurological lesions and abnormalities of far-field auditory brainstem responses. Electroencephalogr Clin Neurophysiol, 41(6), 595-608.   Stenfelt, S. (2011). Acoustic and physiologic aspects of bone conduction hearing. Adv Otorhinolaryngol, 71, 10-21.  	   95	   Stenfelt, S., & Goode, R. L. (2005a). Bone-conducted sound: physiological and clinical aspects. Otol Neurotol, 26(6), 1245-1261.   Stenfelt, S., & Goode, R. L. (2005b). Transmission properties of bone conducted sound: measurements in cadaver heads. J Acoust Soc Am, 118(4), 2373-2391.   Stenfelt, S., Hato, N., & Goode, R. L. (2002). Factors contributing to bone conduction: the middle ear. J Acoust Soc Am, 111(2), 947-959.   Stenfelt, S., & Reinfeldt, S. (2007). A model of the occlusion effect with bone-conducted stimulation. ija International Journal of Audiology, 46(10), 595-608.   Stenfelt, S., Wild, T., Hato, N., & Goode, R. L. (2003). Factors contributing to bone conduction: the outer ear. J Acoust Soc Am, 113(2), 902-913.  Stuart, A., Yang, E. Y., & Stenstrom, R. (1990). Effect of temporal area bone vibrator placement on auditory brain stem response in newborn infants. Ear Hear, 11(5), 363-369.   Swanepoel, D., Schmulian, D., & Hugo, R. (2004). Establishing normal hearing with the dichotic multiple-frequency auditory steady-state response compared to an auditory brainstem response protocol. Acta Otolaryngol, 124(1), 62-68.   Swanepoel de, W., & Steyn, K. (2005). Short report: establishing normal hearing for infants with the auditory steady-state response. S Afr J Commun Disord, 52, 36-39.   Swanepoel, D. W. E. S. F. P. S. A. P. L. (2008). Auditory steady-state responses to bone conduction stimuli in children with hearing loss. International Journal of Pediatric Otorhinolaryngology, 72(12), 1861-1871.   	   96	  Swanepoel, W., Ebrahim, S., Friedland, P., Swanepoel, A., & Pottas, L. (2008). Auditory steady-state responses to bone conduction stimuli in children with hearing loss. Int J Pediatr Otorhinolaryngol, 72(12), 1861-1871.  Teele, D. W., Klein, J. O., & Rosner, B. (1989). Epidemiology of otitis media during the first seven years of life in children in greater Boston: a prospective, cohort study. J Infect Dis, 160(1), 83-94.   Teele, D. W., Klein, J. O., & Rosner, B. A. (1984). Otitis media with effusion during the first three years of life and development of speech and language. Pediatrics, 74(2), 282-287.   Tonndorf, J. (1966). Bone Conduction: studies in experimental animals. Acta Otolaryngol(213), 1-132.   Tonndorf, J., Campbell, R. A., Bernstein, L., & Reneau, J. P. (1966). Quantitative Evaluation of Bone Conduction Components in Cats. oto Acta Oto-Laryngologica, 61(S213), 10-38.   Valdes, J. L., Perez-Abalo, M. C., Martin, V., Savio, G., Sierra, C., Rodriguez, E., & Lins, O. (1997). Comparison of statistical indicators for the automatic detection of 80 Hz auditory steady state responses. Ear Hear, 18(5), 420-429.   van der Reijden, C. S., Mens, L. H., & Snik, A. F. (2005). EEG derivations providing auditory steady-state responses with high signal-to-noise ratios in infants. Ear Hear, 26(3), 299-309.   Van Maanen A, S. D. R. (2010). Multiple-ASSR thresholds in infants and young children with hearing loss. Journal of the American Academy of Audiology, 21(8), 535-545.   Van Maanen, A., & Stapells, D. R. (2009). Normal multiple auditory steady-state response thresholds to air-conducted stimuli in infants. J Am Acad Audiol, 20(3), 196-207.   	   97	  Vander Werff, K. R., Prieve, B. A., & Georgantas, L. M. (2009). Infant air and bone conduction tone burst auditory brain stem responses for classification of hearing loss and the relationship to behavioral thresholds. Ear Hear, 30(3), 350-368.   Von Bekesy, G. (1932). Zur theorie des horens bei der schalluafnahme durch knochenleirung. Annals Physiologie, 111-136.   Winskel, H. (2006). The effects of an early history of otitis media on children's language and literacy skill development. Br J Educ Psychol, 76(Pt 4), 727-744.  Wolfe, J. A., Skinner, P., & Burns, J. (1978). Relation between sound intensity and the latency and amplitude of the brainstem auditory evoked response. J Speech Hear Res, 21(2), 401-407.   Wood, L. (2009). Multiple brainstem auditory steady-state response interactions for different stimuli. (Master of Science), The University of British Columbia.     Yoshida, M., & Uemura, T. (1991). Transmission of cerebrospinal fluid pressure changes to the inner ear and its effect on cochlear microphonics. Eur Arch Otorhinolaryngol European Archives of Oto-Rhino-Laryngology, 248(3), 139-143.   Yoshinaga-Itano, C. (2004). Levels of evidence: universal newborn hearing screening (UNHS) and early hearing detection and intervention systems (EHDI). J Commun Disord, 37(5), 451-465.   Yoshinaga-Itano, C., Sedey, A. L., Coulter, D. K., & Mehl, A. L. (1998). Language of early- and later-identified children with hearing loss. Pediatrics, 102(5), 1161-1171.   Zumach, A., Gerrits, E., Chenault, M., & Anteunis, L. (2010). Long-term effects of early-life otitis media on language development. J Speech Lang Hear Res, 53(1), 34-43.  	   98	  APPENDIX  ASSR Data Normal Hearing Group 500 Hz     S Age AC ASSR BC ASSR AC ASSR-ABR BC ASSR-ABR ASSR ABG AC ASSR Amp AC ASSR Noise AC ASSR Sweeps BC ASSR Amp BC ASSR Noise BC ASSR Sweeps 6 3.3 30 20 -5 10 10 14.16 6.49 17 16.85 6.86 15 7 3.3   10   0         14.99 9.48 10 8 12.9   20   0         24.62 7.63 20 18 8.1 30   -5 -10   12.78 4.8 10       21 6       -20   11.83 5.39 24       24 6 20 0 -15   20 19.09 7.09 10 26.01 6.66 12 27 22.1   20   20         35.18 10.64 16 29 7.4 20 10 -5 10 10 12.28 5.38 15 29.32 13.34 10 35 3.3 40 20 15 10 20 11.19 5.69 19 16.95 6.56 15 36 3.7 40 30 15 10 10 29.92 7.55 10 61.17 12.86 10 37 5.3 40 10 15 0 30 30.66 13.33 10 21.83 10.12 10 38 16.6 40 30 5 10 10 29.24 7.68 10 12.09 5.62 15 39 11.6 20 10 5 0 10 15.46 4.96 10 32.52 6.29 10 40 4.7 30 10 5 10 20 19.86 9.09 14 75.17 20.1 10 43 11.6 30 10 5 0 20 52.87 8.29 10 16.2 7.88 11 44 7.1 40 20 15 0 20 12.77 5.36 17 39.98 8.19 10 46 13.3 20 10 -5 0 10 30.51 9.22 10 22.33 7.7 10 49 0.6 30 10 15 0 20 23.44 9.61 10 26.52 8.75 11 50 10.4 40 30 15 20 10 22.62 7.53 16 39.41 16.98 24 52 5.9 10 30 -25 30 -20 9.56 4.96 11 32.2 12.75 18 53 3.4 40 30 15 30 10 24.86 9.69 13 38.75 18.09 11 54 2.7 20 0 -5 -10 20 13.18 5.17 15 15.12 6.13 10 56 5.1 30 30 5 10 0 15.04 5.88 12 75.81 33.62 21 57 7.1 20 20 -5 0 0 14.11 5.28 10 14.25 5.53 11 60 2.4 30 20 25 20 10 56.83 9.57 10 37.23 11.01 10 Mean  29.52 17.39          SD  9.20 9.63          n  21 23          	   99	  ABR Data Normal Hearing Group 500 Hz    S Age AC ABR BC ABR AC ASSR-ABR BC ASSR-ABR ABR ABG 6 3.3 35 10 -5 10 25 7 3.3   10   0   8 12.9 25 20   0 5 18 8.1 35 10 -5 -10 25 21 6   20   -20   24 6 35   -15     27 22.1 25 0   20 25 29 7.4 25 0 -5 10 25 35 3.3 25 10 15 10 15 36 3.7 25 20 15 10 5 37 5.3 25 10 15 0 15 38 16.6 35 20 5 10 15 39 11.6 15 10 5 0 5 40 4.7 25 0 5 10 25 43 11.6 25 10 5 0 15 44 7.1 25 20 15 0 5 46 13.3 25 10 -5 0 15 49 0.6 15 10 15 0 5 50 10.4 25 10 15 20 15 52 5.9 35 0 -25 30 35 53 3.4 25 0 15 30 25 54 2.7 25 10 -5 -10 15 56 5.1 25 20 5 10 5 57 7.1 25 20 -5 0 5 60 2.4 5 0 25 20 5 Mean  25.43 10.41    SD  7.05 7.5    n  23 24    	   100	  ASSR Data Conductive Hearing Loss Group  500 Hz    S Age AC ASSR  BC ASSR AC ASSR-ABR BC ASSR-ABR ASSR ABG AC ASSR Amp AC ASSR Noise AC ASSR Sweeps BC ASSR Amp BC ASSR Noise BC ASSR Sweeps 10 6 20 10 -35 -10 10 19.89 8.3 10 19.05 8.44 10 11 6.4 30 20 -15 10 10 35.79 12.07 10 27.84 9.36 14 14 4.9 30 30 -25 10 0 32.05 15.28 10 114.64 22.47 10 16 5.7 50 20 5 10 30 26.94 10.25 15 24.51 12.42 11 17 4.1 30 0 -15 -20 30 16.97 7.94 10 18.47 6.29 12 25 6.3 40 10 -5 0 30 32.21 9.24 10 18.31 7.63 21 28 20.6 30 20 -15 10 10 25.55 11.65 14 37.12 12.37 16 30 5.6 30 10 -15 -10 20 24.79 10.06 15 21.01 9.11 11 32 5.3 50 0 5 0 50 14.09 5.44 15 20.43 8.15 10 33 2.9 30 10 -25 -10 20 26.86 8.53 10 36.55 9.45 11 34 3 20 0 -35 -10 20 13.41 5.83 10 15.75 5.51 10 41 4.1 50 40 -5 30 10 24.2 9.9 10 16.82 6.27 15 42 4.1 50 40 5 30 10 41.25 16.84 10 29.41 10.4 11 47 7.1 30 20 -15 -10 10 16.72 6.84 10 10.76 5.31 15 58 11.3 50 10 5 0 40 10.19 3.86 23 10.98 4.54 14 62 10.4 40 10 -5 10 30 31.45 10.15 10 29.79 10.82 13 Mean  36.67 15.33          SD  11.13 13.02          n  15.0 15.0          	   101	  ABR Data Conductive Hearing Loss Group 500 Hz                         S Age AC ABR BC ABR AC ASSR-ABR BC ASSR-ABR ABR ABG 10 6 55 20 -35 -10 35 11 6.4 45 10 -15 10 35 14 4.9 55 20 -25 10 35 16 5.7 45 10 5 10 35 17 4.1 45 20 -15 -20 25 25 6.3 45 10 -5 0 35 28 20.6 45 10 -15 10 35 30 5.6 45 20 -15 -10 25 32 5.3 45 0 5 0 45 33 2.9 55 20 -25 -10 35 34 3 55 10 -35 -10 45 41 4.1 55 10 -5 30 45 42 4.1 45 10 5 30 35 58 11.3 45 10 5 0 35 62 10.4 45 0 -5 10 45 Mean  48.33 12.00    SD  4.88 6.76    n  15 15    	   102	  ASSR Data Normal Hearing Group 2000 Hz                                 S Age AC ASSR BC ASSR AC ASSR-ABR BC ASSR-ABR ASSR ABG AC ASSR Amp AC ASSR Noise AC ASSR Sweeps BC ASSR Amp BC ASSR Noise BC ASSR Sweeps 2 1 40 40 -10 -10 0 42.91 11.32 10 66.58 29.05 10 3 9.9 30 40 -20 -30 10 42.95 14.4 10 23.17 7.12 15 4 2.9 0 0 10 0 0 22.4 6.41 10 13.31 5.65 15 5 17.9   50 30 -20 50       49 10.18 11 6 3.3 20 20 -10 0 0 18.25 5.84 23 11.71 5.53 15 7 3.3 20 20 -10 0 0 23.7 11.79 14 19.84 7.5 10 8 12.9 10 30 20 -20 20 14.38 5.77 10 27.08 11.19 15 10 6 30 30 -10 -10 0 19.29 8.94 14 35.07 10.72 10 11 6.4 20 10 0 20 -10 33.25 11.45 10 15.51 7.73 18 13 10.3 10 40 0 -40 30 20.57 7.62 11 56.37 26.15 16 14 4.9 30 0 -20 10 -30 29.22 13.71 10 24.98 12.35 14 15 10.3 20 20 -10 -10 0 20.57 7.62 11 56.37 26.15 16 16 5.7 40 10 -10 -10 -30 40.64 12.71 10 15.52 6.86 27 17 4.1 0 30 20 15 30 8.62 3.74 14 23.37 10.53 13 18 8.1 10   10 20 -10 31.45 11.82 10 21.95 6.55 11 20 3.1 30 30 -10 -10 0 13.39 4.68 15 20.65 8.74 19 21 6 20 40 0 -30 20 10.55 4.59 24 21.42 12.39 17 22 7.9 40 40 -20 -20 0 18.12 4.95 10 27.42 12.39 17 24 6 10 20 -10 -10 10 29.03 13.8 16 27.3 13.59 24 25 6.3 20 20 10 0 0 24.84 5.53 11 25.27 8.2 20 33 2.9 10 20 0 -10 10 24.26 8.54 10 23.88 4.49 10 34 3 20 40 -10 -10 20 20.73 5.13 10 16.12 6.12 11 Mean  20.47 21.0          SD  12.03 13.96          n  21 21          	   103	  ABR Data Normal Hearing Group 2000 Hz  S Age AC ABR BC ABR AC ASSR-ABR BC ASSR-ABR ABR ABG 2 1 30 30 -10 -10 0 3 9.9 10 10 -20 -30 0 4 2.9 10 0 10 0 -10 5 17.9 30 30 30 -20 0 6 3.3 10 20 -10 0 10 7 3.3 10 20 -10 0 10 8 12.9 30 10 20 -20 -20 10 6 20 20 -10 -10 0 11 6.4 20 30 0 20 10 13 10.3 10 0 0 -40 -10 14 4.9 10 10 -20 10 0 15 10.3 10 10 -10 -10 0 16 5.7 30 0 -10 -10 -30 17 4.1 30 0 20 15 25 18 8.1 20 20 10 20 0 20 3.1 20 20 -10 -10 0 21 6 20 10 0 -30 -10 22 7.9 20 20 -20 -20 0 24 6 0 10 -10 -10 10 25 6.3 30 20 10 0 -10 33 2.9 10 10 0 -10 0 34 3 10 30 -10 -10 20 Mean  17.72 15.0    SD  9.22 10.12    n  22 22    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0167692/manifest

Comment

Related Items