Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electrophysiological investigation of auditory temporal processing MacDonald, Katie Mary 2012

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

Item Metadata

Download

Media
24-ubc_2013_spring_macdonald_katie.pdf [ 1.78MB ]
Metadata
JSON: 24-1.0073418.json
JSON-LD: 24-1.0073418-ld.json
RDF/XML (Pretty): 24-1.0073418-rdf.xml
RDF/JSON: 24-1.0073418-rdf.json
Turtle: 24-1.0073418-turtle.txt
N-Triples: 24-1.0073418-rdf-ntriples.txt
Original Record: 24-1.0073418-source.json
Full Text
24-1.0073418-fulltext.txt
Citation
24-1.0073418.ris

Full Text

ELECTROPHYSIOLOGICAL INVESTIGATION OF AUDITORY TEMPORAL PROCESSING  by  KATIE MARY MACDONALD BSc, The University of Victoria, 2008  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Audiology and Speech Sciences) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2012 © Katie Mary MacDonald, 2012  Abstract Infant hearing-health programs have a goal of identifying infants with a permanent hearing loss by the age of three months and treating these infants by the age of six months. However, deficits in hearing thresholds are not the only deficits that exist in the auditory system. The ability of an infant's auditory system to resolve rapid changes in acoustic signals (i.e., temporal resolution) and integrate acoustic information over time (i.e., temporal integration) is important for typical language development. Because behavioural responses are unreliable for diagnostic purposes before the age of six months, electrophysiological measures of temporal resolution and integration could be beneficial. The main objective of my thesis was to validate in adults if 80-Hz auditory steady-state responses (ASSRs, an objective electrophysiological measure) can be used to assess temporal resolution and integration. Physiological temporal resolutions of adults were estimated from cortical auditory evoked potentials (CAEPs) and ASSR resets evoked by gaps (1.5625, 3.125, 6.25, 12.5, and 25 ms) within 40-Hz and 80-Hz amplitude-modulated white-noise bursts. Physiological gapdetection thresholds for CAEPs (8 ± 6 ms, averaged across conditions), 40-Hz ASSR resets (6 ± 5 ms), and 80-Hz ASSR resets (5 ± 4 ms) were comparable to behavioural gap-detection thresholds (5 ± 2 ms). However, 40- and 80-Hz ASSRs maximally reset to half-cycle gap durations (i.e. 12.5 and 6.25 ms respectively), thus ASSR resets might not be truly measuring gap-detection thresholds. Conflicting results from CAEPs and ASSR resets to gaps provides evidence that CAEPs respond preferentially to all gaps down to threshold; whereas, ASSRs preferentially reset to gaps that violate their modulation periodicity. Physiological integration times (117 ± 48 ms, averaged across conditions), as measured from the rise time of the ASSR resets, were comparable to behavioural measurements of temporal integration (132 ± 83 ms). ii  However, more research is required to determine if physiological and behavioural integration times are correlated or are coincidentally similar. These results indicate that CAEPs are accurate measures of temporal resolution. However, further research is required to determine the utility of ASSR resets in assessing temporal resolution and integration.  iii  Preface This study was reviewed and approved by the Behavioural Research Ethics Board of the University of British Columbia. The certificate number of the ethics certificate obtained is H1200525.  iv  Table of Contents Abstract ....................................................................................................................................................... ii Preface ........................................................................................................................................................ iv Table of Contents .........................................................................................................................................v List of Tables............................................................................................................................................. vii List of Figures .......................................................................................................................................... viii List of Abbreviations.................................................................................................................................. ix Acknowledgements ......................................................................................................................................x Dedication .................................................................................................................................................. xi Chapter 1: Introduction ................................................................................................................................1 1.1 Auditory Temporal Processing..........................................................................................................3 1.1.1 Auditory temporal resolution & integration ...............................................................................3 1.1.2 Hypotheses of temporal resolution & integration. .....................................................................6 1.1.3 Behavioural measures ................................................................................................................8 1.2 Specific Language Impairments ........................................................................................................9 1.3 Early Identification of Hearing Function ........................................................................................11 1.4 Behavioural Responses ...................................................................................................................12 1.5 Human Electrophysiological Recordings ........................................................................................12 1.5.1 Cortical auditory evoked potentials..........................................................................................13 1.5.2 Mismatch negativity .................................................................................................................14 1.5.3 Middle-latency responses .........................................................................................................14 1.5.4 Auditory steady-state responses ...............................................................................................15 1.6 Aims of Study and Hypotheses .......................................................................................................17 Chapter 2: Methods ....................................................................................................................................18 2.1 Participants ......................................................................................................................................18 2.2 General Study Procedures ...............................................................................................................18 2.3 Behavioural Measurements of Temporal Processing ......................................................................19 2.3.1 Temporal resolution: gap-detection task. .................................................................................20 2.3.2 Temporal integration time: threshold-duration task. ................................................................21 2.4 Electrophysiological Measurements of Temporal Processing. .......................................................23 Chapter 3: Results ......................................................................................................................................35 3.1 Behavioural Measurements of Temporal Processing ......................................................................35 3.1.1 Temporal resolution: gap-detection task. .................................................................................35 v  3.1.2 Temporal integration time: threshold-duration task. ................................................................35 3.2 Electrophysiological Measurements of Temporal Processing ........................................................36 3.2.1 Physiological temporal-resolution estimated from CAEPs and ASSR results. ........................36 Chapter 4: Discussion ................................................................................................................................53 4.1 Temporal Resolution .......................................................................................................................53 4.1.1 CAEPs. ......................................................................................................................................53 4.1.2 ASSR reset. ..............................................................................................................................55 4.2 Temporal Integration .......................................................................................................................57 4.3 Theoretical Implications ..................................................................................................................58 4.4 Clinical Implications .......................................................................................................................60 4.5 Caveats ............................................................................................................................................60 4.6 Conclusions .....................................................................................................................................62 References ..................................................................................................................................................63 Appendices .................................................................................................................................................71 Appendix A: Distribution of waveforms across all electrodes...................................................................71 Appendix B: Rating results from CAEPs and ASSR resets. ......................................................................75  vi  List of Tables Table 1: Behavioural Gap-detection Results………………………………………………. …..35 Table 2: Estimated Behavioural Integration Time Measures (tau)………………………… …..36 Table 3: CAEP Gap-detection Results……………………………………………………...…..46 Table 4: ASSR Reset Gap-detection Results………………………………………………. …..47 Table 5: Number of Response Present per Gap Duration………………………………….. …..49 Table 6: Estimated Physiological Integration Time Measures (tau)……………………….. …..49 Table 7: Physiological Integration Time Measures Min-to-max peak…………………….. …..50  vii  List of Figures Figure 1.1: Threshold by duration …………………………………………………………...…..4 Figure 2.1: Visual representation of behavioural stimuli …………………………………..…..20 Figure 2.2: Visual representation of electrophysiological stimuli ……………………………...24 Figure 2.3: Envelope of the ASSR from the Hilbert-transform …………………………… …..26 Figure 2.4: Rating for CAEP thresholds …………………………………………………... …..30 Figure 2.5: Rating for ASSR reset thresholds Measures (tau)……………………………... …..33 Figure 3.1: Grand-averaged waveforms for CAEPs 40- and 80-Hz ASSRs …………………...38 Figure 3.2: CAEP onset responses for 40- and 80-Hz conditions ……………………………...39 Figure 3.3: Topographies for 40-Hz CAEPs …………………...................................................41 Figure 3.4: Topographies for 80-Hz CAEPs …………………………………………………...42 Figure 3.5: Topographies for 40-Hz ASSR resets …………………...........................................44 Figure 3.6: Topographies for 80-Hz ASSR resets …………………...........................................45 Figure 3.7: Grand-averaged behavioural and physiological integration times……..…………...52  viii  List of Abbreviations ABR AEP AM ASSR BCEHP CAEP CANS EEG ERP dB LI MMN RAP SLI SOA VRA  Auditory Brainstem Response Auditory Evoked Potential Amplitude Modulated Auditory Steady-State Response British Columbia Early Hearing Program Cortical Auditory Evoked Potential Central Auditory Nervous System Electroencephalography Event Related Potential Decibel Language Impairment Mismatch Negativity Rapid Auditory Processing Specific Language Impairment Stimulus Onset Asynchrony Visual Reinforcement Audiometry  ix  Acknowledgements First and foremost, I must thank my supervisor Dr. Anthony Herdman for without him this thesis would never have happened. His confidence that I was capable of conquering the task, the endless hours spent on the project and his enthusiasm towards research, learning and teaching made this thesis possible. Thank you also to my committee members: Dr. Anna Van Maanen and Dr. Navid Shahnaz for their input and support during the research process. I also thank my colleagues and friends, Heather Holliday, Osamu Takai and Allison Mackey for their constant encouragement and support during this time. This research was supported by a Natural Sciences and Engineering Research Council (NSERC) of Canada Discovery Grant awarded to Dr. Anthony Herdman.  x  Dedication To my family: All I have and will accomplish would not be possible without your endless love, support and encouragement.  xi  Chapter 1: Introduction The ability of an infant's auditory system to resolve rapid changes in acoustic signals (i.e., temporal resolution) and to integrate acoustic information over time (i.e., temporal integration) is important for typical language development (Tallal, Miller, & Fitch, 1993). Speech is a very complex acoustic signal; it is made up of spectral and temporal details that are essential for understanding the message of the speaker. The capacity to process and categorize auditory stimuli occurring within tens of milliseconds is critical to language acquisition (Benasich & Tallal, 2002). Periodicity, formant structure, frequency transitions, acoustic onsets, and the speech envelope are acoustic features of speech that are vital for speech understanding (Katz, 2009). The central auditory nervous system (CANS) is constantly being bombarded with simultaneous sounds of different frequencies and intensities. The CANS translates features of the speech signal into a meaningful message for the individual to process and comprehend (Katz, 2009). If all the pathways and connections are functioning normally, the CANS has an amazing ability to decode, organize, and package the acoustic signals it receives with exquisite temporal resolution. However, if there are any deficits in processing along the pathway or in the connections between neural populations, meaningful linguistic events could be improperly constructed. Two such deficits that can affect speech perception are the abilities to segregate and to integrate acoustic information across time (Tallal et al., 1993). The speech signal is changing over time, both constantly and quickly. Deficiencies in the ability of the CANS to resolve and integrate temporal changes in an acoustic signal will have significant impacts on how speech sounds are received by the auditory system and ultimately how they are represented in long-term memory. An example of a commonly occurring class of English speech sounds that changes acoustic properties very quickly is the stop consonants: /p/, 1  /b/, /t/, /d/, /k/, and /g/ (Kent & Read, 2002). These phonemes are distinguished from one another in terms of place and manner of articulation. Deficits in temporal resolution and integration abilities can have significant impacts on the identification of these phonemes because the acoustic cues necessary to identify the place and manner of articulation occur very quickly. When an initial stop consonant is produced, there is a constriction at a particular point along the vocal tract that causes a brief closure. When the vocal tract is constricted, a small amount of acoustic energy, if any, is produced. When no acoustic energy is produced, a silent gap is created. The time between the release of the constriction of the vocal tract and the onset of voicing (when the vocal folds start vibrating for the accompanying vowel) is called the voice onset time (VOT), which is an acoustic cue for the manner of articulation. If the auditory system cannot resolve the gap or the VOT, the manner of articulation may be misinterpreted and the wrong phoneme could be perceived. When a stop is followed by a vowel, the vocal tract changes shape in order to transition from the stop consonant to the vowel. The shape of the changing vocal tract causes the formants to change, which changes the speech sound that is produced. Formant transitions are very important speech cues needed for the perception of stop consonants, and occur within an interval of 50 ms (Kent & Read, 2002). If an individual’s auditory system does not contain the temporal resolution needed to decipher the stop gap and VOT or cannot integrate the important acoustic information from the formant transitions during these short intervals, the phoneme could be improperly perceived. For example, the words “pat” and “bat” only differ in their initial phoneme. If an individual is not able to resolve this fine temporal detail of the stop gap between /pa/ and /ba/ or integrate the information over the specified time interval of the changing formant transition, then the consonant being produced could be improperly perceived. This could change the meaning of the intended message. If a 2  child’s CANS receives improper temporal information about the speech signal during language development, then his/her brain could improperly construct their phonemic templates for speech. The necessity of the CANS to resolve and integrate these acoustic changes further increases when background noise is present. Background noise can affect the ability to segregate speech from noise by masking key acoustic changes, including gaps. As mentioned previously, resolving and integrating these fine changes is necessary for the CANS to construct the intended message of the speaker. In everyday listening situations, speech is rarely present without competing noise; therefore, the importance of resolving and integrating acoustic changes is further increased. In addition, binaural temporal cues are important for segregating a speaker’s voice from background noise (Bregman, 1990). Thus, temporal processing deficits could further impair the CANS ability to segregate speech from noise by impairing temporal cues needed for binaural localization. 1.1 Auditory Temporal Processing Auditory temporal processing is how the brain sorts through acoustic information over time. I focused on two aspects of auditory temporal processing: temporal resolution and temporal integration. A detailed discussion of both of these concepts is found below. 1.1.1 Auditory temporal resolution & integration Temporal resolution and integration are greatly interconnected. Temporal resolution can be defined as how small of a change the auditory system can detect (Viemeister & Plack, 1993). The difference between perceiving two speech sounds can be as small as tens of milliseconds. If the auditory system does not have the capacity to resolve these fine changes, the individual will not be able to perceive the difference between two consonants. Broadly described, temporal 3  integration is the combination of auditory information over time (Yost, 2007). In order to perceive complex sounds, such as speech, the auditory system constantly combines acoustic information across time. Speech and other complex auditory signals also require that the auditory system resolve very fine temporal changes. Listening to speech is not an instantaneous task; parts of the speech signal are being received at the ear when the auditory system is trying to decipher the previous acoustic information. This acoustic information is stored in short-term memory and later accessed to be combined with the next string of acoustic information. If the auditory system cannot store the information and integrate what it has heard over a certain interval of time, the meaning of the spoken message could be constructed with errors or lost altogether. Even though the broad description of temporal integration is simply “the combination of information over time,” this combination is complex. An example of temporal integration is how the auditory system benefits from longer stimulus durations. The auditory system requires that a certain amount of energy be present in order to detect an acoustic event; once an acoustic event has reached this amount of required energy, it can be detected. The formula P=E/T is useful in explaining this example, where P=power, E=energy, and T= time (Yost, 2007). If the equation is rearranged in terms of energy, then E=PT. The amount of energy will stay constant (because the auditory system only requires a certain amount to reach detection) therefore; when one variable increases the other will decrease. When duration (T) is increased, then there is less power required for the acoustic signal to be detected, i.e., the threshold will be lower. As stimulus duration continues to increase, the participant’s threshold will continue to decrease until increasing the duration does not change the threshold measured. The integration time of the auditory system can be described as the time required for the auditory system to reach maximum 4  performance (Yost, 2007). Because threshold change by stimulus duration is an exponential function, it is difficult to determine the exact point on the curve to define as maximum performance, an individual’s integration time (Figure 1.1). The time constant (i.e. tau) of an exponential function could be used to define an individual’s integration time.  Figure 1.1. Threshold by duration.  Figure 1.1. As signal duration increases, measured threshold decreases. The asterisk denotes tau. Because the threshold gain by stimulus duration function grows exponentially, it is difficult to determine the exact time in which the auditory system reaches maximum performance for integrating temporal information. Tau is commonly used in science to represent the timing constant of an exponential function. Tau is the time it takes to rise 63.2% (1-1/e) of the magnitude of threshold benefit a participant gains for increasing stimulus duration. The length of time that is required for the combination of acoustic information to occur has been under debate for a number of years. Different hypotheses and models have been proposed throughout the literature to help explain and understand temporal integration (Moore, 2003; Näätänen 1990, 1992; Viemeister & Wakefield, 1991). These models of temporal integration can generally be broken down into two categories: models that suggest a long period of integration time (in the order of hundreds of milliseconds, see Näätänen, 1990) and models 5  that suggest a very short integration time (in the order of tens milliseconds, see Moore, 2003). The auditory system benefits (by having a lower threshold) from a longer stimulus (see Figure 1.1 above) in the order of hundreds of milliseconds, which is an argument for the auditory system having a longer integration period. However, the auditory system is capable of resolving very fine changes in an acoustic signal as demonstrated in gap-detection experiments (Viemeister & Plack, 1993). The resolution of the auditory system (found from gap-detection experiments) is finer (approximately 2-10 ms) than what the long integration period would predict (approximately 100-200 ms). These two conflicting facts of the auditory system led to the “integration-resolution” paradox by de Boer (1985). How does one reconcile the fact that the auditory system benefits from having a long integration period, with the fact that the auditory system is able to resolve very fine acoustic changes? An explanation for this discrepancy is that the auditory system has two different integration systems with different time constants that can be used depending on the task. 1.1.2 Hypotheses of temporal resolution & integration. Temporal window of integration. Näätänen (1990, 1992) proposed the temporal window of integration hypothesis to explain temporal integration. According to this hypothesis, there is a sliding window of approximately 200 ms over which temporal integration occurs. The onset of the initial stimulus signals the window to begin and the window of integration continues for approximately 200 ms. Any auditory event occurring within approximately 200 ms of the stimulus onset will be combined with the first stimulus and treated as one acoustic event. Behavioural evidence to support this hypothesis is that an individual’s threshold is shown to improve with stimulus durations up to 200 ms (Pedersen & Salomon, 1977). Yabe and colleagues (1997) investigated this hypothesis using mismatch negativity (MMN; an 6  electrophysiological measure of signal change detection, see section 1.4.2). Acoustic stimuli with different stimulus onset asynchronies (SOAs) were used and the smallest SOA that elicited a MMN was determined to be the temporal window of integration. They found that the 100 and 125 ms SOAs elicited the MMN because they occurred less than 150 ms which is hypothesized to be the temporal window of integration. Because the 150 ms SOA falls outside of the 150 ms integration window, when a stimulus is omitted, the CANS did not notice the change. Multiple looks. Viemeister and Wakefield (1991) proposed the “multiple looks hypothesis” to help explain and understand temporal resolution and integration. In their experiment, a participant’s threshold was found for one pulse and was then compared to their threshold for two pulses while the duration was varied between the pulses. They found that when the two pulses were separated by 1ms, the participant’s threshold was 3-5dB better (lower) as compared to their threshold for one pulse. This decrease in threshold was observed when the separation between the two pulses ranged from 1-5 ms, and not for longer separations of the pulses. Their conclusion was that the auditory system takes multiple “looks” at a stimulus in very short durations and holds these “looks” in short-term memory. If the two pulses fall within the same “look” their energy is summed and treated as one event. If there is more energy (from combining the two pulses) the threshold measured will be lower (better). This opposes the temporal window of integration hypothesis which predicts that there should not be a difference in the thresholds for the small duration separations of 1-5ms; the threshold should have only decreased when separations were in the order of hundreds of milliseconds. The multiple looks hypothesis explains both the resolution of rapid changes in acoustic signals, and the long integration window of the auditory system. The long-integration window can be explained within the multiple-looks hypothesis by considering the sampling probability of the acoustic 7  signal. The more times the auditory system “looks” at a stimulus (a longer stimulus), the more chances it has to detect the energy change in the acoustic signal (i.e., overcoming threshold). With the multiple looks hypothesis, the auditory system is “looking” at the acoustic event every few milliseconds, instead of once every 100 to 200ms, which is suggested by the temporal window of integration hypothesis (Viemeister & Wakefield, 1991). The “look” can also be thought of as the resolution of the auditory system. 1.1.3 Behavioural measures Temporal resolution. Behavioural measurements of auditory temporal resolution are often referred to as a measure of temporal acuity. Temporal acuity is commonly measured from behavioural gap-detection experiments. Participants are asked to listen to a signal and report if they heard one or two sounds. The duration between the two sounds is varied and the smallest gap that an individual can detect in a supra-threshold continuous sound (when the individual reports 2 sounds instead of 1) is considered to be their behavioural gap-detection threshold. Gap-detection thresholds for gaps placed within broadband noise were found to be between 2 and 3 ms (Forrest & Green, 1987). Temporal integration. A gold standard for measuring auditory temporal integration has yet to be defined in this field. A few different methods have been proposed to obtain an estimate of temporal integration, but the threshold by duration test appears to be the most popular (Plomp & Bouman, 1959). In this test, an individual’s threshold is found for acoustic signals of various durations. These thresholds are then plotted threshold against duration of the stimulus. There will be a point at which the auditory system does not benefit (i.e., the threshold does not decrease) from a longer stimulus. This is said to be temporal integration. The estimated integration time of the auditory system depends on the stimulus being measured. For example, 8  300 ms is required for the detection of a tone compared to a few milliseconds for the detection of a click (Yost, 2007). Auditory temporal integration depends on what kind of signal is being measured and can range from a few milliseconds to 500 ms and beyond (Yost, 2007). 1.2 Specific Language Impairments Temporal processing impairments have been linked to Specific Language Impairment (SLI; Muluk, Yalcinkaya, & Keith, 2011; Tallal et al. 1993). Tallal et al. (1993) defined SLI as “children who were developing normally in every respect, but failed to develop language at the expected rate” (Tallal, Miller & Fitch, 1993, p. 28). It is estimated that 7.4% of kindergarten children have a SLI (Tomblin, Records, Buckwalter, Zhang, Smith, & O'Brien, 1997). Tallal and colleagues (1993) studied two groups of children: children who had been diagnosed with language impairment (LI) and a group of age-matched controls. When two tones were presented at rapid rates, the LI group had more difficulty when asked to discriminate and sequence the two tones than their age-matched controls; the former performed close to chance on the task. However, if the LI children were given enough time to integrate the auditory stimuli (the length of time between the two tones was increased), they performed the same as their age-matched controls. In the same study, the LI group was compared to their age-matched controls in another task that required the children to discriminate both vowels and syllables. The vowel stimuli did not change in their acoustic properties for 250 ms, but the syllables only differed in the first 40 ms of their formant transitions. The LI group performed very well when asked to discriminate the vowels. However, these children could not report a difference between the two syllables when asked to discriminate the two syllables from one another. The authors also found that children who were significantly impaired in the temporal processing tasks were also impaired in their receptive language. The authors concluded that when a child is not able to resolve acoustic 9  information within a very short time interval (tens of milliseconds), it can lead to a disturbance of the development of phonological processes and subsequently a delay of receptive language. This provides evidence for the need to identify these children with impaired temporal resolution early in life (Tallal et al., 1993). Benasich & Tallal (2002) presented a longitudinal study of infants tested at 7.5 months with a behavioural rapid auditory processing test (RAP). Infants who performed with RAP thresholds of greater than 150 ms had significantly poorer language outcomes at twenty four months of age as compared to those who obtained lower RAP thresholds. Tallal et al. (1993) hypothesized that children diagnosed with a SLI have a basic temporal processing impairment that gives rise to difficulty in integrating sensory information that occurs in a fast sequence. Impairment in temporal processing could have a cascade of negative effects in a child’s language development caused by difficulty in identifying and separating auditory stimuli that occur within milliseconds (Tallal et al., 1993). Following the same logic for early identification (i.e. <6 months) of a permanent hearing loss, early identification of impairment in temporal integration could help clinicians intervene promptly and potentially reduce the severity of SLI. Currently, children with SLIs are not identified until they fail to develop language or show signs of language delay, which is typically around three and a half years of age (Paul, 2006). Improving the clinician’s ability to identify such children early could help improve the child’s outcome for language development. However, in order to do so effectively there is a need to understand the effectiveness of different forms of intervention in treating individuals with temporal processing deficits. Controversy remains in the literature regarding the best interventions for children with temporal processing problems, but comparing these interventions is beyond the scope of this thesis. 10  1.3 Early Identification of Hearing Function The importance of identifying hearing loss early in an infant’s life and subsequently treating it promptly has been well established in the literature (Kennedy et al., 2006; YoshinagaItano, Sedey, Coulter, & Mehl, 1998). Yoshinago-Itano and colleague’s (1998) showed that infants who were identified with a hearing loss and treated before the age of 6 months (with amplification) had better speech and language outcomes as compared to infants who were treated (with amplification) after the age of 6 months (Yoshinaga-Itano et al., 1998). Kennedy and colleagues (2006) also found similar results that support the need to identify and treat hearing loss early in life. Children who were identified by the age of nine months with a permanent hearing loss by a universal newborn hearing screening program and treated with amplification had better language skills as compared to children who did not have access to a screening program and were identified after the age of 9 months. Many hearing-health programs for children therefore have a goal of identifying children with a permanent hearing loss by the age of three months and treating these children (e.g., amplification) by the age of six months (BCEHP, 2012). However, deficits in hearing thresholds are not the only deficits that exist in the auditory system. Deficits in auditory temporal integration are also known to occur in children and adults (Tallal et al., 1993). Trehub and Henderson (1996) showed that six- and twelvemonth old infants with poorer temporal processing abilities (i.e., higher gap-detection thresholds) have poorer language and speech functions as reported by parents at 16 to 30 months of age (Trehub & Henderson, 1996). Little is known about such deficits in infants under the age of six months because behavioural measures for measuring temporal integration in infants and young children are unreliable. An objective measure of temporal resolution and integration in  11  infants would be a first step toward understanding the prevalence and possible need for early intervention for temporal processing deficits. 1.4 Behavioural Responses Behavioural responses are the gold standard for determining auditory thresholds for a hearing assessment. However, before the age of 6 months behavioural responses such as those observed using behavioural observation audiometry are unreliable for diagnostic purposes. Visual reinforcement audiometry is clinically possible and reliable only after the age of 6 months (Moore, Wilson, & Thompson, 1977; Muir, Clifton, & Clarkson, 1989; Thompson & Wilson, 1984). Because of the inconsistency of responses that are able to be observed with behavioural measures up to the age of 6 months, electrophysiological responses are an alternative technique that can be used to assess auditory function in this population. The British Columbia Early Hearing Program (BCEHP) uses electrophysiological responses, like the Auditory Brainstem Response (ABR) and Auditory Steady-State Response (ASSR) to assess the auditory integrity of these children (BCEHP, 2012). Electrophysiological responses do not depend on a behavioural response from the individual which is why they have become useful in audiology for assessing hearing in difficult to test populations, such as infants. Similarly, measuring temporal resolution and integration is currently challenging in infants younger than six months because behavioural responses are unreliable for diagnostic purposes. Electrophysiological measures might be a more feasible means to objectively assess infant hearing with respect to temporal integration and resolution before six months of age. 1.5 Human Electrophysiological Recordings Human electrophysiological recordings to auditory stimuli contain many different responses that reflect different structural and functional processing stages of the auditory 12  system. Auditory evoked potentials can be classified into three different categories: transient, sustained, and steady-state responses (Stapells, Linden, Suffield, Hamel, & Picton, 1984). These different response types have been extensively used to assess human auditory systems. The following is a discussion of transient and steady-state responses that show promise in assessing auditory temporal processing in humans. 1.5.1 Cortical auditory evoked potentials Cortical auditory evoked potentials (CAEPs) are transient evoked responses to the onset of stimuli with a typical morphology of P1-N1-P2 waves (Hillyard & Picton, 1978). CAEPs can also be evoked by a change in an auditory stimulus feature, such as frequency, intensity, or duration change (Martin & Boothroyd, 2000). A gap that is placed within a tone or noise alters the intensity and temporal characteristics of the stimulus. If the CANS has sufficient resolution to detect such changes, a cortical response will be elicited and can be used to assess the physiological gap-detection thresholds (Michalewski, Starr, Nguyen, Kong, & Zeng, 2005; Ross & Pantev, 2004). However, CAEPs can be challenging to record from infants; their responses differ significantly compared to adults in morphology, latency, and amplitude (Wunderlich & Cone-Wesson, 2006). A recent study by Small & Werker (2012) showed that a phonemic change (a change in stimulus feature) evoked adult-like N1-P2 complexes in four to seven month old infants. More research is required to determine if CAEPs to stimulus changes can be recorded in infants. For adults, Ross & Pantev (2004) showed that gaps within AM tones can evoke CAEP responses (Ross & Pantev, 2004). The smallest gap duration that elicited a CAEP response was suggested to indicate the physiological auditory temporal resolution.  13  1.5.2 Mismatch negativity Mismatch Negativity (MMN) is an evoked potential that can be observed when a ‘deviant’ stimulus is presented in the middle of regularly occurring ‘standard’ stimuli (Näätänen, 1979). Näätänen (1985) proposed there are two hypotheses for MMN elicitation. The first is to assume that new afferent elements are responding and that they generate the MMN (Yabe, Tervaniemi, Reinikainen, & Näätänen, 1997). Or, the MMN could be produced by the stimulus change found when the deviant stimulus is compared to the train of the repetitive stimuli (Yabe, Tervaniemi, Reinikainen, & Näätänen, 1997). Desjardins, Trainor, Hevenor, and Polak (1999) first measured young adults MMN to gaps inserted within 2000 Hz tone pips and found that gaps of 4 ms elicited MMN. Trainor, Samuel, Desjardins, and Sonnadara (2001) then measured infants’ (6-to-7 month old) MMN to gaps inserted within pure tones. Their findings showed that MMN could be evoked by a gap as short as 4ms. However, mismatch negativity is not widely adopted as a clinical tool for assessing infant hearing function because it is inconsistently recorded from infants (Morr, Shafer, Kreuzer, & Kurtzberg, 2002). 1.5.3 Middle-latency responses Middle-latency responses (MLRs) are auditory evoked potentials with typical morphology of Na and Pa waves (Picton, 2011). Rupp, Gutschalk, Hack, and Scherg (2002) and Rupp, Gutschalk, Uppenkamp, and Scherg (2004) recorded MLRs to gaps placed within broadband noise bursts. They found present MLRs for a gap inserted into a broadband noise burst as low as 3 ms and 6 ms. However, MLRs are not reliably recorded in infants and are not reproducible in certain stages of sleep, which is typically the best state to record infants for quiet (i.e., minimal noise) EEG (Kraus, McGee, & Comperatore, 1989; Okitsu, 1984).  14  1.5.4 Auditory steady-state responses Auditory steady-state responses (ASSRs) are evoked when a steady-state auditory stimulus is either amplitude or frequency modulated and regularly repeated (Galambos, Makeig, & Talmachoff, 1981; Picton, 2011). When high stimulus rates are used, a response can be observed that is a sinusoidal waveform that has the same fundamental frequency as the stimulation rate (Stapells, Herdman, Small, Dimitrijevic & Hatton, 2005). When the external auditory stimulus remains constant in its amplitude and frequency content, the steady-state response is then able to become stable in phase and amplitude in relation to the external auditory stimulus. ASSRs follow the frequencies of external sounds (Picton, 2011). When an auditory stimulus is amplitude modulated (AM) (e.g., tone or white-noise) at either 40- or 80-Hz, the response observed after averaging is either a 40- or 80-Hz response. When a fast-Fourier transform is done with ASSRs a peak occurs at the rate of the modulation, even though the stimulus that was presented to the individual contained no energy at that frequency. The ASSR was first recorded in adults by Geisler (1960). A later publication of the “40Hz ASSR” sparked considerable interest and multiple studies and research went into investigating the 40-Hz response (Galambos, Makeig, & Talmachoff, 1981). Audiological interest waned from ASSRs when it was determined that the 40-Hz response was significantly reduced in sleeping infants (Maurizi 1990; Stapells, Galambos, Costello & Makeig, 1988; Suzuki & Kobayashi 1984); however, it was also discovered that it was possible to record ASSRs at higher presentation rates (Lins, Picton, Picton, Champagne, & Durieux-Smith, 1995). Subsequently, significant interest has focused on developing techniques and equipment to record ASSRs for clients of all ages. It is becoming common practice to record auditory thresholds with ASSRs. Tones that are AM at 40- or 80-Hz can be used to evoke 40- or 80-Hz ASSRs, which 15  are EEG measures commonly used for testing hearing thresholds (Lins et al.,1995; Stapells et al., 2005; Stapells, Makeig & Galambos, 1987; Van Maanen & Stapells, 2005). The largest responses observed in the auditory domain are seen when a sound is AM by 40- (Galambos, 1981) or 80-Hz (Cohen, Rickards, & Clark, 1991; Lins et al., 1995). In 2004, Ross & Pantev completed a study that measured physiological temporal resolution and integration time using 40-Hz ASSRs and CAEPs in adults (Ross & Pantev, 2004). Gaps within AM tones evoked a sharp decline in the 40-Hz ASSR amplitude followed by a rise back to the pre-gap steady-state amplitude (referred to as an ASSR reset; Ross & Pantev, 2004). The rise time for the ASSR reset of the 40-Hz ASSR was suggested to indicate the physiological auditory temporal integration time (Ross & Pantev, 2004). They found that gaps of 3 to 6ms placed within the 40-Hz AM tones were the smallest gap durations that elicited a CAEP and a 40-Hz ASSR reset (Ross & Pantev, 2004). Their study indicates that the 40-Hz ASSR could be used to measure temporal resolution and integration time of the auditory system. Because the 40-Hz ASSR is not easily recorded in young infants (Maurizi 1990; Stapells, Galambos, Costello & Makeig, 1988; Suzuki & Kobayashi 1984), using 40-Hz AM tones to measure temporal processing would only be appropriate for adults. This is likely because infants 40-Hz ASSRs are significantly smaller in amplitude than adults 40-Hz ASSRs (Suzuki & Kobayashi, 1984). In addition, the 40-Hz ASSR is decreased in sleeping infants which is typically the most reliable state for recording quiet (i.e., minimal noise) EEG in this population (Cohen, Rickards & Clark 1991; Stapells et al., 2005). In contrast, 80-Hz ASSRs are easily recorded in sleeping infants (Stapells et al., 2005), thus the 80-Hz ASSRs have been recommended to assess hearing thresholds in infants (Stapells et al., 2005).  16  1.6 Aims of Study and Hypotheses A question that lies unresolved is whether the 80-Hz ASSR could be used to identify the auditory integration times and temporal resolution of an infant less than 6-months old. Before this question can be answered, we need to determine whether the 80-Hz ASSRs show similar resets and CAEPs to gaps within AM stimuli that evoke 40-Hz ASSR resets in adults. The main objective of this thesis was to investigate if the 80-Hz response showed the same physiological gap-detection threshold and integration time as the 40-Hz response when a gap is placed within AM white-noise bursts. In the current study, five main hypotheses were tested: 1) ASSRs have similar reset functions to gaps inserted in 40 and 80-Hz AM noise. 2) Physiological gap-detection thresholds are similar among CAEP, 40-Hz ASSR, and 80-Hz ASSRs. 3) Physiological gap-detection thresholds for CAEP, 40-Hz ASSR, and 80-Hz ASSRs correlate with behavioural thresholds of gap detection. 4) Temporal integration times (i.e., ASSR reset duration) are similar for 40-Hz and 80Hz ASSRs. 5) Physiological integration time for 40-Hz and 80-Hz ASSRs correlate with behavioural integration times.  17  Chapter 2: Methods 2.1 Participants Nine female and six male volunteer participants with an average age of 27.2 (range 22 to 39) from the University of British Columbia (UBC) and surrounding areas participated in this study. The study was reviewed by the Behavioural Research Ethics Board at the University of British Columbia and all participants provided informed consent after receiving information of what the study entailed. Participants received an honorarium for their time. 2.2 General Study Procedures After the participants received information regarding the study and provided informed consent, they received a hearing assessment to determine normal audiological status to be included as a participant in the study. Normal audiological status was defined as air-conducted pure-tone thresholds less than 20 dB HL at 500, 1000, 2000, and 4000Hz and normal immittance (Jerger type A tympanograms and ipsilateral reflexes present at 85 dB HL using broad band noise). If a participant met all the inclusion criteria and provided informed consent, he/she was enrolled in the full study. Test sessions began with the electrophysiological tasks and the participants received a short break (approximately 10 minutes) before the behavioural tasks were performed. The behavioural measurements were completed after the electrophysiological measurements in order to not prime the participants during the electrophysiological recordings to pay attention for gaps in the stimuli because this could have affected their recorded physiological responses. Testing took place in the BRANE lab at UBC. Total session time was approximately 2.5 hours and sessions were scheduled at the participant’s convenience. The stimuli for the electrophysiological and behavioural sessions were white-noise bursts that were AM by 40-Hz 18  for condition A and 80-Hz for condition B. During pilot testing to replicate findings from Ross and Pantev’s 2004 study, inconsistent results were found when a 40-Hz AM tone within 1000Hz carrier frequency was used (MacDonald & Herdman, personal communication). Contrary to what was found by Ross and Pantev (2004), ASSRs did not reset to a 25 ms gap placed within a 40-Hz AM tone. Investigating this further, gap durations that were the same as half of the cycle of the modulation rate (e.g., 12.5 ms for 40-Hz or 6.25 ms for 80-Hz) elicited the largest ASSR reset, whereas gap durations equal to a full modulation cycle had little or no effect on ASSR amplitudes. Inserting a gap within an AM tone could alter the phase of the carrier frequency, unless explicitly controlled. This change in phase of the carrier frequency might be detected by the auditory system and cause the ASSR reset instead of the duration of the gap. Thus, it was important to alleviate this potential confound. Modulating the amplitude of white-noise will not be confounded by changes in phase of the carrier frequency because by definition white-noise has random phases. Thus, the gap duration is the main cue available to the auditory system for AM noise for detecting gaps. This is why I chose to use AM white-noise bursts instead of AM tones. 2.3 Behavioural Measurements of Temporal Processing I behaviourally measured temporal processing of participants using two tasks: 1) a gapdetection task and 2) a threshold-duration task. For four participants, the threshold-duration task was conducted in a subsequent session because of time constraints for recording the electrophysiological data. No apparent differences were evident for these four participants as compared to others. The behavioural tasks are explained in detail below.  19  2.3.1 Temporal resolution: gap-detection task. Stimuli. The stimuli for the behavioural gap-detection task were two 100% AM whitenoise bursts separated by a 1000 ms interval (Figure 2.1). One of the AM white-noise bursts was randomly assigned (50% probability) to have a gap embedded at a null point in the modulation envelope. The gap was randomly assigned to occur between 20 to 40 cycles after stimulus onset so that gap onset could not be used as a cue. The gap durations included 0.3906, 0.78125, 1.5625, 3.125, 6.25, 12.5, and 25 ms. The AM white-noise burst after the gap continued for a random duration of 20 to 40 cycles, therefore stimulus duration could not be used as a cue for detecting a gap. A 40-Hz AM white-noise burst was used in Condition A and an 80-Hz AM white-noise burst was used in condition B. The gap durations were inserted at the place of 100% amplitude modulation (zero stimulus energy) in order to minimize spectral splatter that could act as a cue for identifying the gap. Figure 2.1. Visual representation of behavioural stimuli. AM AM white-noise gap white-noise  1000 ms  AM white-noise  20 + 0-20 cycles 500 ms – 1000 ms  Figure 2.1. Visual representation of behavioural stimuli. The gap was embedded at a null point in the modulation envelope randomly assigned between 20 to 40 cycles after stimulus onset.  Stimuli were calibrated using a SoundPro sound level meter (SoundPro®) and were attenuated to 80 dB SPL for the recording session. The stimuli were presented to the participants left ear through an ER-3A insert earphone while the right ear was occluded with a foam plug during the task to minimize the ambient stimulus energy from crossing over to the non-test ear.  20  Task. The gap-detection task was completed in a sound-treated booth while a participant remained awake, alert, and sitting up. This task used an adaptive staircase method to find the smallest gap that a participant could detect between two AM white-noise bursts. It was a 2alternate forced choice paradigm where a participant was required to indicate, by button press, which of the two AM tones contained the gap. Participants started at detecting a 25 ms gap. If a participant correctly chose the stimulus that contained the gap for three successive presentations, the gap duration decreased one level (e.g. 25 ms to 12.5 ms). This continued until the participant made an incorrect response of indicating the stimulus with no gap as having a gap. After an incorrect identification, the gap duration increased one level (e.g. 3.125 ms to 6.25 ms). A reversal was defined as when a participant made three successive correct responses and then an incorrect response (or an incorrect response and three successive correct responses). The testing stopped after ten reversals. The behavioural gap-detection task took approximately ten minutes to complete. Analyses. Participant’s behavioural gap-detection thresholds were calculated for the 40and 80-Hz conditions by averaging the gap durations that occurred during the last seven reversals. Seven reversals were counted backwards from the end of the task and gap durations that occurred during the last seven reversals were averaged to find the participant’s gapdetection threshold. Successive gap durations for three correct responses were considered as one value when averaging. 2.3.2 Temporal integration time: threshold-duration task. Stimuli. The stimuli were 100% AM white-noise bursts of twelve durations: 25, 50, 100, 150, 200, 250, 300, 350, 400, 450, 500, and 1000 ms. These time durations were chosen in order to obtain a more specific time estimate of behavioural temporal integration. A 40-Hz AM white21  noise burst was used in Condition A and an 80-Hz AM white-noise burst was used in Condition B. The integration time task was completed in a double-walled sound-attenuated room while the participant was awake, alert, and sitting up. The stimuli were presented to the participants left ear through an ER-3A insert earphone. Task. The objective of the task was to estimate a participant’s integration time from finding his/her auditory threshold for the AM white-noise bursts of the different durations (please see above). Standard audiometric threshold bracketing techniques were performed with a starting presentation level of 60 dB SPL. Presentation of the AM white-noise burst durations were randomized for obtaining a participant’s thresholds. A participant was asked to raise his/her hand when he/she heard a sound. The presentation level was then decreased in 10 dB steps until the participant failed to respond. The stimuli were then increased in 1 dB steps until the participant responded. Auditory threshold was estimated at the intensity level that was responded to twice in two consecutive ascending steps (increasing by 1 dB). The integration task took approximately ten minutes to complete. Analyses. To estimate participant’s behavioural integration times for the 40- and 80-Hz conditions, I fitted an exponential curve to the adjusted threshold data as a function of time using Matlab’s fitting software package (Curve Fitting Toolbox, Matlab Mathworks). The adjusted threshold data were obtained by subtracting the threshold at 1000 ms from all other thresholds. This was done to adjust for variability in hearing sensitivity and to normalize the threshold decrement across participants. The threshold obtained for the 1000 ms stimulus duration was considered the reference point because the participant would gain little benefit from a stimulus longer than 1000 ms. An estimate of the time constant of the threshold by duration function, referred to here as tau, was estimated from the fitted exponential function. 22  Tau is commonly used in science to represent the timing related to exponential functions, such as the amount of time needed to charge neuronal membranes or the amount of time needed for radioactive materials to decay. For my studies, tau is the time it takes to rise 63.2% (1-1/e) of the magnitude of threshold benefit a participant gains for increasing stimulus duration. Because the threshold gain by stimulus duration function grows exponentially, it is difficult to determine the exact time in which the auditory system reaches maximum performance for integrating temporal information. Thus, I considered tau as an estimate of the integration time of each participant. Participant NG011 was removed from the integration time analysis because her 40Hz physiological response had no reset to model. 2.4 Electrophysiological Measurements of Temporal Processing. Stimuli. White-noise bursts of 800 ms in duration were 100% AM at 40 Hz for condition A and 80 Hz for condition B. Gaps of 0 (no-gap control condition), 1.5625, 3.125, 6.25, 12.5, and 25 ms were randomly inserted after 400 ms of the AM white-noise burst (Figure 2.2). The gaps were inserted at the null point (zero stimulus energy) of the modulation envelope in order to minimize spectral spread, leaving the gap duration the main cue available to the auditory system. Inter-stimulus intervals were 600ms. The stimuli were calibrated using a SoundPro sound-level meter (SoundPro®) and were presented at 80 dB SPL to the participant’s left ear through an ER-3A insert earphone. The right ear was occluded with a foam plug to minimize the ambient stimulus energy from crossing over to the non-test ear. The 40-Hz condition included two blocks, each containing 100 trials of each gap yielding a total of 1200 trials. The 80-Hz condition included two blocks, each containing 200 trials of each gap yielding a total of 2400 trials. All stimuli were randomly presented at equal probability of occurrence (i.e., 16.7%) across blocks. 23  Figure 2.2. Visual representation of electrophysiological stimuli.  Figure 2.2. AM white-noise bursts for the electrophysiological recordings. 40-Hz stimuli are depicted on the left and 80-Hz stimuli are depicted on the right. The stimuli are 800 ms in duration with a 600 ms ISI (200 ms lead and 400 ms follow). The waveform is zoomed in between 350 and 475 ms to show where the gaps occurred. The gaps were inserted at 400 ms (dashed line) during the null point of the stimulus. Stimuli are zoomed in to visualize the timing of the gap durations (25, 12.5, 6.25, 3.125, 1.5625 ms). Task. Participants sat in a comfortable chair in a double-walled, sound-attenuated booth. The 40-Hz condition was recorded while the participants were awake, alert, and watching a publicly available DVD of their choice with closed captioning. The 80-Hz condition was recorded while the participants were asleep or very relaxed and reclined in the comfortable chair. Participants were encouraged to remain in these states in order to maximize the signal-to24  noise ratios of the 40- and 80-Hz ASSRs, respectively. The 40-Hz response provides the largest signal-to-noise response when a participant is awake (Cohen, Rickards & Clark 1991; Stapells et al., 2005), whereas the 80-Hz response provides the best signal-to-noise response while a participant is sleeping or in a very relaxed state (Stapells et al., 2005). A participant’s arousal states were monitored via visual observation of the participant (e.g., eyes open or closed) and ongoing EEG (e.g. changes in EEG alpha rhythms from occipital electrodes). Participants were instructed not to pay attention to the auditory stimuli during the recording session. Set up and EEG recordings took approximately 2.5 hours to complete. The participants were randomly assigned to start with the 40- or 80-Hz condition to reduce possible order effects. Recordings. EEG data were collected using a BIOSEMI system (BIOSEMI, www.biosemi.com) from 64 electrodes mounted in an electrode cap and positioned on the participant’s scalp in a modified 10-10 system. Seven additional electrodes were placed on the head: one on each mastoid (left and right; M1 and M2), one on the nape of the neck (neck), one above and one below the left eye (SO1 and IO1; bipolar configuration for vertical electrooculogram), and one on each outer canthi (LO1 and LO2; bipolar configuration for horizontal electrooculogram). During the online recording, all electrodes were referenced to a common electrode, CMS, placed midway between Pz and PO3. Data were collected continuously at a sampling rate of 1024 Hz. Evoked Potential Analyses. The EEG data were re-referenced to an average reference of the 67 scalp electrodes (excluding electrodes SO1, IO2, LO1, and LO2). Epochs of -500 to 1200 ms relative to the AM noise onset were created. Epochs with EEG exceeding +/- 150 microV were rejected and the remaining epochs were averaged to generate two replications of equal number of epochs. All participants’ EEG data had more than 50 artefact-free epochs for all 25  stimuli. Epoched data were band-pass filtered between 1-20 Hz, 35-45 Hz, and 75-85 Hz to extract the time-domain waveforms for the slow cortical AEPs (CAEPs), 40-Hz ASSRs, and 80Hz ASSRs, respectively. I also applied a Hilbert-transform to the ASSR data to calculate the envelope of the 40- and 80-Hz ASSR amplitudes (see Figure 2.3). The ASSR envelopes were used for all subsequent analyses. Figure 2.3. Envelope of the ASSR from the Hilbert-transform  Figure 2.3. Example of a 40-Hz ASSR (blue thin line) with the envelope of the ASSR amplitude from the Hilbert-transform (blue thick line). The asterisks depict where the minima and maxima were chosen for the calculation of integration times. I calculated the signal-to-noise ratio of the replications to determine whether a participant’s data was too noisy to include for further analyses. Signal-to-noise ratios for each participant were estimated by calculating a standard deviation ratio which is a standard deviation of the averaged CAEP amplitude across 50-300 ms (i.e., signal + noise) divided by the standard deviation of the plus-minus reference of two replications across 50-300 ms (i.e., noise) 26  (Picton, Linden, Hamel & Maru, 1983). I determined whether there were outliers in the data by viewing a box plot of the standard deviation ratios from all 15 participants. Three participants’ (NG004, NG005, and NG008) standard deviation ratios were below the 99% confidence limits of the group’s median ratios. These participants’ data were thus rejected from all analyses (behavioural and electrophysiological). There were no relevant factors other than a poor signalto-noise ratio that differentiated these three participants from the 12 participants whose data were included for further analyses. Physiological temporal resolution estimated from CAEPs and ASSR results. Two measures of physiological temporal resolution were used for the analysis. According to Ross & Pantev (2004), the smallest gap that elicited an ASSR reset and/or a CAEP can be used as a measure of temporal resolution of the auditory system. To be consistent with the literature, I estimated the physiological temporal resolution separately for CAEPs and ASSRs by detecting the presence of CAEPs or ASSR resets after gaps embedded within AM stimuli, respectively. Group-averaged CAEP gap-detection threshold. To determine the group-averaged physiological gap-detection threshold for CAEPs responses, I averaged the CAEP waveforms for each of the six gaps in the 40- and 80-Hz AM noise conditions across the group. CAEPs to gap durations of 1.5625, 3.125, 6.25, 12.5, and 25 ms were statistically compared to CAEPs to the no-gap control condition (a priori planned contrasts) for each sample point between 400 to 800 ms using a Student’s t-test. To minimize type I errors in performing multiple t-tests, I applied a false-discovery rate (FDR) correction to the t-test p-values (Benjamini & Hochberg, 1995). Statistical analyses were performed using the statistics toolbox in Matlab and custom statistics software developed in the BRANE lab (UBC). The smallest gap duration that was found to evoke significantly different CAEP waveforms from the no-gap control condition was 27  determined to be the group-averaged physiological gap-detection threshold. Physiological gapdetection thresholds were separately estimated for the 40- and 80-Hz conditions. In addition, I determined the topographies of the changes in CAEPs to gaps by calculating the CAEP power for all scalp electrodes. I applied a Hilbert transform to the CAEP responses and then averaged the CAEP power across 475 to 700 ms for each gap condition. This yielded a single CAEP power value for each electrode that I used to plot the topographies of CAEP power. I estimated the amount of significant changes in CAEP power due to gaps by comparing CAEP power for each gap duration condition (1.5625, 3.125, 6.25, 12.5, and 25 ms) to CAEP power for the no-gap control condition (a priori planned). Student’s t-test and FDR correction were used to determine significant differences for the comparison of CAEP response power across electrodes. CAEP power differences were considered significant at FDR corrected p<.05. Individual CAEP gap-detection threshold. To determine participant’s individual CAEP gap-detection thresholds, two raters blindly rated the participants’ CAEPs for the gap conditions (randomly ordered) as having replicable responses different from the no-gap control condition. In order to better compare the CAEPs for gap conditions to CAEPs for the no-gap condition, an estimate of the variance in the no-gap condition was calculated. To do this the ERP data were split-half resampled; half of the number of trials were randomly selected and averaged to create one replication and the other half of the trials were averaged to create another replication. The data were then resampled and averaged using another randomly selected half number of trials (Efron & Tibshirani, 1993). This was repeated 100 times with replacement of the sampled trials to create 200 replications of the CAEPs for the no-gap control condition. The average of the 200 replications for the no-gap responses and plus/minus two standard-deviations were plotted and the two replications of the CAEPs to the gaps were overlaid for comparison. The no-gap results 28  were plotted in order to help raters make a decision if a CAEP to the gap was present, absent, or unknown. The two-replications of CAEPs to the gap conditions were determined by separately averaging the odd and even trials. The waveforms that were presented to the raters for judgment were from four electrodes: FC3, Cz, M2 (right mastoid), and P10 because CAEPs were largest at these sites. The raters’ task was a three-forced choice procedure where the rater was required make a decision of: ‘response’, ‘no response’ or ‘maybe’ (could not evaluate). These three conditions were chosen because this is what is done clinically for auditory brainstem response detection (BCEHP, 2012). Each rater rated 180 waveforms (15 participants x 6 gaps x 2 conditions = 180) which were randomly presented. The 40- and 80-Hz responses (conditions) were rated independently. Figure 2.4 provides an example of what the raters saw for making their ratings. An interrater reliability analysis using the Kappa statistic was performed to determine consistency among raters. The rater results were used to determine the individual participant’s physiological gap-detection threshold. Combined rater judgments of CAEP responses to a gap as being ‘present-present’, ‘present-maybe’, or ‘maybe-maybe’ defined the condition as having a CAEP present. The smallest gap that had a rater-defined CAEP present was considered as that participant’s CAEP gap-detection threshold with one exception. The exception was that the threshold was not considered at the smallest gap with one of the above ratings if two gap levels above were deemed as response absent.  29  Figure 2.4. Rating for CAEP thresholds.  Figure 2.4. Rating individual CAEP thresholds for the 80-Hz condition. Red lines represent the odd and even averaged trials for a particular gap. Thick black line is the mean for the no-gap control condition with the thin black lines representing two standard deviations from the mean. Waveforms were shown for FC3, P10, Cz and M1. Group-averaged ASSR gap-detection threshold. To determine the group-averaged ASSR reset threshold for the 12 participants, ASSRs were calculated for each of the six gaps for the 40- and 80-Hz conditions. In order to determine whether there was a reset in the ASSR, a statistically significant power decrease in the ASSR envelope after a gap had to be observed. Because the ASSR to the no-gap control stimuli should only have natural fluctuations across the 400 to 625 ms interval, there should be only minimal ASSR envelope changes. If gaps caused ASSRs to reset then there should be a significant decline in ASSR envelope amplitudes in this interval following a gap. To determine if significant resets occurred, ASSR envelope amplitudes 30  for gap conditions (1.5625, 3.125, 6.25, 12.5, and 25 ms) were compared to ASSR envelope amplitudes for the no-gap control condition at electrode Fz for each sample point between 400 to 625 ms using Student’s t-tests. FDR correction was applied to minimize type I errors that could be caused by performing multiple t-tests. ASSR envelope amplitude differences were considered significant at p<.05 after FDR correction. The smallest gap duration with a significant ASSR reset was considered to be the group-averaged ASSR gap-detection threshold. This was done for the 40-Hz condition and the 80-Hz condition. To evaluate the spatial distribution of the ASSR resets, I generated topographies by averaging the ASSR envelope amplitudes across the interval of 425 to 475 ms. For the topographies I wanted to narrow in on the most likely time of when the reset was occurring to the gap. If a longer interval was used, this could have washed out gap effects on the ASSR amplitudes as the ASSR return to a steady-state by about 150-200 ms after gap offset. The average ASSR envelope amplitude was calculated for each of the six gaps (1.5625, 3.125, 6.25, 12.5, and 25 ms) and compared to the no-gap condition. Statistical differences in the topographies between the gap and no-gap conditions were determined using Student’s t-tests and FDR correction. Individual ASSR gap-detection threshold. To determine participant’s individual physiological gap-detection thresholds for the ASSRs, two raters blindly rated the participants’ ASSRs for the gap conditions (randomly ordered) as having replicable resets different from the no-gap control condition. In order to better compare the ASSR resets between gap conditions and the no-gap control condition, an estimate of the variance across replications in the no-gap condition was calculated. To do this the ERP data were split-half resampled; half of the number of trials were randomly selected and averaged to create one replication and the other half of the 31  trials were averaged to create another replication. The data were then resampled and averaged using another randomly selected half number of trials. This was repeated 100 times with replacement of the sampled trials to create 200 replications of the ASSRs for the no-gap control condition. The average of the 200 replications and plus/minus two standard-deviations were plotted and the ASSRs to the gaps were overlaid for comparison. The no-gap ASSRs were plotted to provide raters with a visual representation of the natural fluctuations in the ASSR. This helped raters make a decision if a reset of the ASSR to the gap was present, absent, or unknown. The two-replications of ASSRs to the gap conditions were determined by separately averaging the odd and even trials. The waveforms that were presented to the raters for judgment of the 40-Hz condition were from three electrodes: Fz, Cz, and Pz because ASSR resets were largest at these sites. For the 80-Hz condition the waveforms that were presented to the raters for judgment were from four electrodes: Fz, Cz, Pz and the neck electrode. The raters’ task was a three-forced choice procedure where the rater was required make a decision of: ‘response’, ‘no response’ or ‘maybe’ (i.e., could not evaluate). Each rater judged the randomly presented 180 waveforms (15 participants x 6 gaps x 2 conditions = 180). The 40- and 80-Hz ASSRs were rated independently. Figure 2.5 provides an example of what the raters saw for making their ratings. An interrater reliability analysis using the Kappa statistic was performed to determine consistency among raters. The rater results were used to determine the individual participant’s physiological gap-detection threshold. Combined rater judgments of ASSR reset responses to a gap as being ‘present-present’, ‘present-maybe’, or ‘maybe-maybe’ defined the condition as having an ASSR reset present. The smallest gap that had a rater-defined ASSR reset present was considered as that participant’s ASSR gap-detection threshold with one exception. The  32  exception was that the threshold was not considered at the smallest gap with one of the above ratings if two gap levels above were deemed as response absent.  Figure 2.5. Rating for ASSR reset thresholds.  Figure 2.5. Rating individual ASSR reset thresholds for the 80-Hz condition. Red lines represent the odd and even averaged trials for a particular gap. Thick black line is the mean for the no-gap control condition with the thin black lines representing two standard deviations from the mean. Physiological integration time estimated from ASSR results. Physiological integration times were estimated from amplitude envelopes of the rise times for the 40- and 80Hz ASSR resets that occurred after an inserted gap. A 12.5 ms gap (half cycle of 40Hz) elicited the most robust 40-Hz ASSR reset at electrode Fz and was thus used to estimate the physiological integration time for 40-Hz condition. A 6.25 ms gap (half cycle of 80 Hz) elicited the most robust 80-Hz ASSR reset at the neck electrode and was thus used to estimate the 80-Hz 33  physiological integration times. I calculated two estimates of the physiological integration times for each 40-Hz and 80-Hz conditions. The first estimate of the integration times was the latency difference between the minima and maxima point in the ASSR envelope following a gap (similar to Ross & Pantev, 2004). Minima and maxima in the ASSR envelope amplitudes were found between 400 to 700 ms by visually inspecting each participant’s data. The minima points were chosen as the first minima of the ASSR envelope following the gap and the maxima points were chosen as the first peak or shoulder of the plateau following the minima (see Figure 2.3). I calculated a second estimate of the physiological integration times (i.e., physiological tau) that followed the similar exponential fitting methods I used to estimate the behavioural integration times. To do this, I extracted the ASSR envelope amplitude values between minima and maxima points determined above for the resets for each participant. The maxima amplitude value was then subtracted from all other values and then an exponential line was fitted to these adjusted data. Physiological tau was calculated from the exponential line as the time it takes to rise 63.2% (1-1/e) of amplitude.  34  Chapter 3: Results 3.1 Behavioural Measurements of Temporal Processing 3.1.1 Temporal resolution: gap-detection task. The group-mean 40- and 80-Hz behavioural gap-detection thresholds were 4 ± 2 ms and 5 ± 2 ms, respectively (Table 1). A paired-samples t-test showed that there was insufficient evidence to support finding significant differences in gap-detection thresholds between 40- and 80-Hz conditions (t=1.152, df=11, p = 0.274). A Pearson product-moment correlation showed an insignificant relationship between 40and 80-Hz behavioural gap-detection thresholds (r = 0.477, n = 12, p = 0.117). Table 1 Behavioural Gap-detection Results Participant NG001 NG002 NG003 NG006 NG007 NG009 NG010 NG011 NG012 NG013 NG014 NG015 Mean SD Note. SD = standard deviation  40-Hz (ms) 3.30 3.34 3.52 1.88 5.26 4.39 7.50 4.91 1.95 3.67 3.32 7.39 4.20 1.81  80-Hz (ms) 2.55 4.40 4.46 1.95 5.90 6.60 6.94 2.26 6.11 2.93 7.10 7.03 4.85 2.01  3.1.2 Temporal integration time: threshold-duration task. Behavioural integration times, as estimated by behavioural tau measures, were on average 129±77 ms and 136±92 ms for 40- and 80-Hz conditions, respectively (Table 2). A paired-samples t-test found insufficient evidence to indicate significant differences of tau between conditions (t=-0.383, df=10, p = 35  0.710). A Pearson product-moment correlation showed a significant moderate relationship between 40- and 80-Hz behavioural integration times (r = 0.708, n = 11, p = 0.015). Table 2 Estimated Behavioural Integration Time Measures (tau) Participant NG001 NG002 NG003 NG006 NG007 NG009 NG010 NG012 NG013 NG014 NG015 Mean SD Note. SD = standard deviation  Tau(40Hz) 83.33 98.09 163.41 48.89 293.37 65.43 187.87 77.00 208.84 131.18 57.64 128.64 77.03  Tau(80Hz) 92.96 106.78 40.30 23.62 238.62 78.01 303.77 181.46 232.31 146.01 55.08 136.27 91.94  3.2 Electrophysiological Measurements of Temporal Processing 3.2.1 Physiological temporal-resolution estimated from CAEPs and ASSR results. Group-average CAEP and ASSR gap-detection threshold. Figure 3.1 shows the groupaverage CAEPs and ASSR amplitude envelopes for the 40- and 80-Hz conditions to the various gap and no-gap stimuli between 400 to 800 ms after gap onset at FCz, P10, and Neck (please see Appendix A for top view distribution across all electrodes for 0 and 25 ms gaps). Significant changes in CAEPs (presence of P1-N1-P2 complex) after a gap occurred to the 40-Hz AM noise stimuli with 12.5 ms (orange waveforms) and 25 ms (red waveforms) gaps as compared to the no-gap control stimulus (black waveforms) at FCz and P10 (p<.05 FDR corrected). CAEPs to the 6.25 ms gap stimuli (magenta waveforms) had a similar P1-N1-P2 response morphology as compared to CAEPs to the 12.5 and 25 ms gap stimuli; however, responses to the 6.25 ms gap 36  stimulus were not significantly different from those to the no-gap control stimulus. CAEPs after gaps of less than 12.5 ms did not reveal significant changes in responses as compared to those for the no-gap control stimulus. Thus, the group-averaged CAEP threshold for the 40-Hz condition was deemed to be 12.5 ms. The second column of Figure 3.1 shows significant CAEPs after gaps of 6.25, 12.5, and 25 ms in 80-Hz AM noise stimuli. No significant CAEP differences between no-gap and gaps less than 6.25 ms were found. The CAEP gap-detection threshold was thus regarded to be 6.25 ms for the 80-Hz AM noise condition. From these results, the 80-Hz condition appeared to have a lower CAEP threshold of 6.25 ms as compared to 12.5 ms for the 40-Hz condition. To determine if there was an effect of subject’s state, onset responses from the two conditions were compared. Figure 3.2 contains the results from this analysis. CAEPs to the onset of the first AM noise-burst (not after the gaps) were larger for the 80- than 40-Hz condition. The third and fourth columns of Figure 3.1 show the ASSR resets at FCz, P10, and neck electrodes. An ASSR reset was seen as a significant downward deflection in the ASSR amplitude envelope after a gap. This occurred at FCz for ASSRs to 40-Hz AM noise bursts with 12.5 (orange waveforms) and 25 ms (red waveforms) gaps as compared to the no-gap AM noise burst (black waveforms) (p<.05 FDR corrected). The 1.5625 and 3.125 ms gaps also showed a slight deflection suggesting a reset; however, these responses were not significantly different from those to the no-gap stimulus. Thus, the group-averaged ASSR reset threshold for the 40-Hz condition was deemed to be 12.5 ms. The fourth column of Figure 3.1 showed a significant 80Hz ASSR reset after gaps of 6.25 and 25 ms at the neck electrode (p<.05 FDR corrected).  37  Figure 3.1. Grand-averaged waveforms for CAEPs, 40- and 80-Hz ASSRs  Figure 3.1. Grand-average waveforms for CAEPs (top columns) and ASSRs (bottom columns) to AM noise burst with gaps at electrodes FCz, P10, and neck. Significant difference intervals (p<.05 FDR corrected) are shown as coloured bars below the time-domain waveforms for comparisons between responses to no-gap stimuli and responses to 25 ms gaps (red bars), 12.5 ms gaps (orange bars) and 6.25 ms gaps (magenta bars). *please note scale differences for viewing the responses.  38  Figure 3.2. CAEP onset responses for 40- and 80-Hz condition.  Figure 3.2. Onset responses from CAEPs for the 40- (black) and 80-Hz (red) condition. Significance bars underneath where they are significantly different from one another p<0.05 FDR corrected.  Unexpectedly, I found no significant ASSR resets differences between no-gap and gaps of 12.5 ms. Because of this discontinuity in ASSR resets to decreasing gap durations, I could not Figure 3.X. Onset responses from CAEPs for the 40- and 80-Hz condition. confidently determine gap-detection threshold using 80-Hz ASSR resets. Moreover, I hypothesized (hypothesis #1) that ASSRs would have similar reset functions to gaps inserted in 40 and 80-Hz AM noise. The results from Figure 3.1 indicate that 40- and 80-Hz ASSRs did not reset to the same gaps. In addition, I hypothesized (hypothesis #2) that the physiological gap-detection thresholds as estimated from the group-averaged CAEPs, 40-Hz ASSR resets, and 80-Hz ASSR resets would be similar. The CAEP and ASSR gap-detection threshold for the 40-Hz condition were both 12.5 ms. Group-averaged CAEP gap-detection thresholds for the 80-Hz condition was lower at 6.25 ms. For the 80-Hz ASSR resets; however, I could not determine gap-detection thresholds because only gaps of 6.25 and 25 ms elicited a reset of the ASSR resets. Therefore, 39  physiological gap-detection thresholds appear to be similar among the physiological measures with the exception that I could not confidently assess gap-detection thresholds for 80-Hz ASSR resets. Another approach to determining physiological gap-detection threshold was to estimate the power changes in the CAEPs and ASSRs following a gap as compared to the no-gap condition. For the CAEP topographies, an increase in power was observed over frontal and inferior temporal-parietal scalp regions for both 40- and 80-Hz AM stimuli (Figure 3.3 and 3.4). The first column shows the topography of the change in CAEP power (left panel) for each gap duration. The second column shows the difference in CAEP power due to the gaps as compared to the no-gap control condition. The third column shows the Student’s t-test results for the statistical differences (p<.01). The bar charts on the top of both figures contain the power values for electrodes FCz and P10. CAEP power increased as the gap duration increased. For the 40-Hz condition, significant power changes were observed for the 12.5 and 25 ms gaps. These results are consistent with the results found in Figure 3.1. For the 80-Hz condition, gaps of 6.25, 12.5, and 25 ms caused statistically significant increase in CAEP power, consistent with the results from Figure 3.1. The scalp topographies of the averaged amplitude envelopes for the ASSR resets are shown in Figure 3.5 and 3.6. The first column shows the six gap durations with their topographies and associated ASSR power levels. The second columns of Figure 3.5 and 3.6 contains the comparison of the average ASSR envelope amplitude for the six gaps to the no-gap control condition. ASSR power decreases (blue) as compared to the no-gap condition reflect ASSR resets. This is most evident for 40-Hz ASSR resets to the 12.5 ms gap and for the 80-Hz ASSR resets to the 6.25 ms gap. The third column shows the topographies of the significant 40  Figure 3.3. Topographies for 40-Hz CAEPs.  Figure 3.3. The first column contains the topographies of the power calculated from a Hilbert transform of the CAEPs to gaps. The CAEP topographies for the no-gap condition are at the top with incrementally increasing CAEP topographies for gap durations (0, 1.5625, 3.125, 6.25, 12.5, and 25 ms). The second column contains the differences which were created by subtracting each gap duration from the no-gap control duration. The third column is the Student’s t-test results for statistical differences p<0.05 FDR corrected. The bar chart contains the power calculated from the Hilbert transform of the CAEPs recorded at FCz and P10.  41  Figure 3.4. Topographies for 80-Hz CAEPs.  Figure 3.4. The first column contains the topographies of the power calculated from a Hilbert transform of the CAEPs to gaps. The CAEP topographies for the no-gap condition are at the top with incrementally increasing CAEP topographies for gap durations (0, 1.5625, 3.125, 6.25, 12.5, and 25 ms). The second column contains the differences which were created by subtracting each gap duration from the no-gap control duration. The third column is the Student’s t-test results for statistical differences p<0.05 FDR corrected. The bar chart contains the power calculated from the Hilbert transform of the CAEPs recorded at FCz and P10.  42  Student’s t-test results for ASSR power differences between gaps and no-gap stimuli (p<.05). For the 40-Hz condition, gaps of 12.5 and 25 ms indicate significant power reductions over frontal scalp locations. Similar results were also observed in Figure 3.1 for the group-averaged CAEP waveforms. For the 80-Hz ASSRs, only the 6.25 ms gap caused a significant power reduction. However, a significant ASSR power difference was not evident to the 25 ms gap as compared to the no-gap for the 80-Hz condition as was seen in Figure 3.1. This might be due to the fact that different time windows were used in these two calculations and that averaging power across an interval might reduce the power to find a significant difference. The bar charts at the top of both figures show the power measured in mV for the six gap durations. The 40-Hz ASSR resets showed power decreases at FCz and P10 to the 12.5 and 25 ms gaps, whereas the 80-Hz ASSR resets showed a large power decrease at Fz and neck for the 6.25 ms gap and a small reduction for the 25 ms gap. Individual participant CAEP gap-detection threshold. Individual participant CAEP gap-detection thresholds, as determined by two blind raters, were on average 7±7 ms and 9±5 ms for 40- and 80-Hz conditions, respectively (Table 3). Insufficient evidence was found to support significant differences in the individual CAEP thresholds found for the 40- and 80-Hz conditions (t=-0.657, df = 11, p =0.525). A Pearson product-moment correlation showed an insignificant relationship between 40- and 80-Hz individual CAEP gap-detection thresholds (r = 0.226, n = 12, p = 0.117).  43  Figure 3.5. Topographies for 40-Hz ASSR resets.  Figure 3.5. The first column contains the topographies of the power calculated from a Hilbert transform of the ASSR resets to gaps. The ASSR topographies for the no-gap condition are at the top with incrementally increasing ASSR reset topographies for gap durations (0, 1.5625, 3.125, 6.25, 12.5, and 25 ms). The second column contains the differences which were created by subtracting each gap duration from the no-gap control duration. The third column is the Student’s t-test results for statistical differences p<0.05 FDR corrected.The bar chart contains the power calculated from the Hilbert transform of the ASSR resets recorded at FCz and P10.  44  Figure 3.6. Topographies for 80-Hz ASSR resets.  Figure 3.6. The first column contains the topographies of the power calculated from a Hilbert transform of the ASSR resets to gaps. The ASSR topographies for the no-gap condition are at the top with incrementally increasing ASSR reset topographies for gap durations (0, 1.5625, 3.125, 6.25, 12.5, and 25 ms). The second column contains the differences which were created by subtracting each gap duration from the no-gap control duration. The third column is the Student’s t-test results for statistical differences p<0.05 FDR corrected. The bar chart contains the power calculated from the Hilbert transform of the ASSR resets recorded at Fz and the neck.  45  Table 3 CAEP Gap-detection Results Participant NG001 NG002 NG003 NG006 NG007 NG009 NG010 NG011 NG012 NG013 NG014 NG015 Mean SD Note. SD = standard deviation  40-Hz (ms) 25 12.5 3.125 12.5 3.125 6.25 3.125 1.5625 12.5 1.5625 6.25 1.5625 7.42 7.05  80-Hz (ms) 12.5 1.5625 12.5 12.5 12.5 12.5 6.25 6.25 12.5 1.5625 3.125 12.5 8.85 4.73  To determine the inter-rater reliability between the two raters, a Cohen’s Kappa coefficient was calculated which measures the inter-rater agreement while taking into consideration agreement between raters that would occur by chance (Cohen, 1968). The interrater reliability for the raters for the 40-Hz condition was 0.356 (fair agreement) (p<0.001), 95% CI (0.193, 0.519) whereas the weighted Kappa was 0.501 (moderate agreement). The inter-rater reliability for the 80-Hz condition was 0.670 (substantial agreement) (p<0.001), 95% CI (0.0.531, 0.809) with a weighted Kappa of 0.768(substantial agreement). When an unweighted kappa coefficient is calculated, it only takes into consideration the data where the raters agreed. Any disagreement is not considered in the calculation. However, for the purposes of our rating, when rater A said ‘response’ and rater B said ‘maybe’, this should be taken into consideration. Therefore, the weighted Kappa was also calculated because conceptually disagreement between one rater stating ‘response present and the other rater stating  46  ‘response absent’ is greater than one rater stating ‘response present’ and the other stating ‘maybe response present’. Individual participant ASSR gap-detection threshold. Individual participant ASSR reset gap-detection thresholds, as determined by two raters blindly reviewing the results, were on average 6±5 ms and 5±4 ms for 40- and 80-Hz conditions, respectively (Table 4). Insufficient evidence was found to support significant differences in the individual ASSR reset thresholds found for the 40- and 80-Hz conditions (t=0.684, df = 8, p =0.513). I only used the participants in the t-test and correlation that had ASSR resets for both conditions (n=9). A Pearson productmoment correlation showed an insignificant relationship between 40- and 80-Hz individual CAEP gap-detection thresholds (r = -0.327, n = 9, p = 0.390). Table 4 ASSR Reset Gap-detection Results Participant 40-Hz (ms) 80-Hz (ms) NG001 12.5 1.5625 NG002 1.5625 6.25 NG003 1.5625 6.25 NG007 12.5 3.125 NG009 6.25 12.5 NG011 1.5625 1.5625 NG012 12.5 3.125 NG014 6.25 3.125 NG015 3.125 6.25 Mean 6.42 4.86 SD 4.90 3.44 Note. SD = standard deviation. Participant NG006, NG010 and NG013 were omitted from the ASSR gap-detection results as they were not rated as having a reset for the 80-Hz condition.  The inter-rater reliability for the 40-Hz ASSR resets was 0.323 (fair agreement) (p<0.001), 95% CI (0.162, 0.484) with a weighted Kappa of 0.40 (fair agreement). The Kappa  47  for the 80-Hz ASSR resets was 0.210 (fair agreement) (p<0.01(0.007)), 95% CI (0.047, 0.373) with a weighted Kappa of 0.271 (fair agreement). A paired-samples t-test was conducted to compare the individual gap thresholds in the CAEPs and ASSR resets (40-Hz). There was not a significant difference in the scores for CAEP (M=7.42, SD=7.05) and ASSR reset (6.12, SD=4.29) conditions; t(11)=0.637, p = 0.537. Because three participants were not rated as having an ASSR reset to any gap duration, the t-test was not completed on this data. I hypothesized (hypothesis #3) that physiological gap-detection thresholds for CAEP would correlate with behavioural thresholds of gap detection. To test this hypothesis, four A priori comparisons were performed. A Pearson product-moment correlation revealed insufficient evidence to support a relation between behavioural and CAEP gap-detection thresholds for the 40-Hz condition (r = -0.555, n = 12, p = 0.061). A Pearson product-moment correlation revealed insufficient evidence to support a relation between behavioural and CAEP gap-detection thresholds for the 80-Hz condition (r = 0.077, n = 12, p = 0.812). I also tallied the number of participants that were rated as having a response for each of the gap durations. For a participant’s response to be deemed as present, the raters had to report: ‘response-response’, ‘response-maybe’, or ‘maybe-maybe’. For the CAEPs, as the gap duration increases, more participants were rated as having a response present (Table 5). However, for the 40- and 80-Hz ASSR resets a different trend occurred. The greatest number of participants being rated as having an ASSR reset occurred for the gap duration that corresponded to half the cycle of the modulation frequency (12.5 ms for the 40-Hz and 6.25 ms for the 80-Hz). This is consistent with the pattern from the group-averaged ASSR reset data (see Figure 3.1).  48  Table 5 Number of Response Present per Gap Duration. Gap (ms) 25 12.5 6.25 3.125 1.5625 0  CAEP 40-Hz 80-Hz 10/12 12/12 8/12 12/12 4/12 4/12 5/12 2/12 3/12 4/12 0/12 0/12  ASSR 40-Hz 80-Hz 6/12 4/12 10/12 2/12 8/12 8/12 2/12 4/12 4/12 3/12 0/12 0/12  3.2.2 Physiological integration time estimated from ASSR results. Because there is no standard way to measure physiological integration times, two measures were calculated in my study. One integration-time measure, called physiological tau, was found from the best fitted exponential function to the ASSR reset data (Table 6) and the other integration time measure was the absolute latency difference between minimum and maximum of the ASSR reset (Table 7). Figure 3.7 shows the grand-averaged behavioural and physiological integration times. Table 6 Estimated Physiological Integration Time Measures (tau) Participant NG001 NG002 NG003 NG006 NG007 NG009 NG010 NG012 NG013 NG014 NG015 Mean: SD Note. SD = standard deviation  Tau(40Hz) 40.56 24.89 27.54 88.39 32.92 22.64 34.10 38.27 68.45 31.08 41.59 40.95 19.99  Tau(80Hz) 29.23 30.59 146.18 46.68 56.85 21.76 37.86 23.73 102.51 46.34 53.45 54.11 37.88  49  Table 7 Physiological Integration Time Measures Min-to-max peak Participant NG001 NG002 NG003 NG006 NG007 NG009 NG010 NG012 NG013 NG014 NG015 Mean: SD Note. SD = standard deviation  Min-max peak latency (40Hz) (80Hz) 92.66 66.41 55.66 106.43 70.31 182.62 197.27 122.00 94.73 162.11 69.34 54.64 86.93 105.47 149.41 64.45 208.01 164.06 79.10 162.11 124.02 149.41 111.59 121.79 52.10 45.68  I hypothesized (hypothesis #4) that temporal integration times would be similar for 40Hz and 80-Hz ASSRs. A paired-samples t-test revealed insufficient evidence to support significant differences in physiological tau between 40- and 80-Hz conditions (t=-1.077, df=10, p = 0.307). A Pearson product-moment correlation revealed insufficient evidence to support a relation between 40- and 80-Hz physiological tau (r = 0.128, n = 11, p = 0.707). In addition, a paired-samples t-test revealed insufficient evidence to support differences in min-to-max ASSR integration times between 40- and 80-Hz ASSRs (t=-0.521, df=10, p = 0.614). A Pearson product-moment correlation insufficient evidence to support a relation between 40- and 80-Hz min-to-max ASSR integration times (r = 0.122, n = 11, p = 0.720). Furthermore, I hypothesized (hypothesis #5) that physiological integration times for the 40-Hz and 80-Hz ASSRs would correlate with behavioural integration times. Four a priori planned correlations were performed to test this hypothesis. A Pearson product-moment correlation revealed an insignificant relationship between the behavioural and physiological tau 50  for the 40- and 80-Hz conditions (respectively: r= -0.132, n = 11, p = 0.698 and r = -0.101, n = 11, p = 0.767). Because two measures of integration time were used, I also performed a correlation on the min-to-max peak ASSR integration times. A Pearson product-moment correlation revealed an insignificant relationship between behavioural and physiological min-tomax ASSR integration time for the 40- and 80-Hz conditions (respectively: r =-0.061, n = 11, p = 0.858 and r =0.046, n = 11, p = 0.894).  51  Figure 3.7. Grand-averaged behavioural and physiological integration times.  Figure 3.7. Grand-averaged behavioural and physiological integration times. Results for 40-Hz are shown on the left whereas and 80-Hz on the right. The top of the figure represents the minimum to maximum ASSR peak (the thick line on the steady-state response). The bottom plots represent the fitted function to the behavioural and physiological data. Blue and red lines are the fitted exponential functions for 40- and 80-Hz behavioural integration time data (red and blue dots), respectively. Green and magenta lines are the fitted exponential functions for the 40- and 80-Hz min-to-max ASSR reset data (green and magenta dots), respectively. The asterisks represent the tau for each exponential fitted function.  52  Chapter 4: Discussion My thesis focused on investigating physiological methods to assess auditory temporal processing. Paradoxical findings of human abilities to detect very short gaps on the order of a couple milliseconds versus improved thresholds for long duration stimuli on the order of 100200 ms (de Boer, 1985) lead Viemeister and Wakefield to suggest the ‘multiple-looks hypothesis’ (Viemeister & Wakefield 1991). They hypothesized that the auditory system rapidly samples short intervals of the acoustic input and combines them into a longer bank of temporal detectors. If acoustic input level or frequency changes occur within one of the sampled looks then the auditory system could detect the change. Thus, the multiple-looks model can explain how the human auditory system can detect small temporal changes in the acoustic signal while still integrating information across several hundred milliseconds. To be consistent with this model, I conceptualize temporal processing as a combination of short- and long-temporal integrators that reflect temporal resolution and temporal integration, respectively. Previous researchers suggested that CAEPs and 40-Hz ASSRs can be used to physiologically measure auditory temporal resolution as well as temporal integration in adults (Ross & Pantev, 2004). However, this might not be appropriate for infants younger than 6 months because their CAEPs and 40-Hz ASSRs are unreliably recorded. This is where 80-Hz ASSRs might be useful. Thus, a major objective of this thesis was to determine if 80-Hz ASSRs could first be used to measure temporal resolution and integration in normal-hearing adults. The main findings are discussed below in the context of physiological measures of temporal resolution and temporal integration. 4.1 Temporal Resolution 4.1.1 CAEPs. The mean of the individual CAEP gap-detection threshold as determined by the two raters (7.42 ms for 40-Hz condition and 8.85 ms for 80-Hz condition) were consistent 53  with the group-averaged CAEP gap-detection thresholds as determined by statistical testing across the time-domain waveforms (12.5 ms for 40-Hz condition and 6.25 ms for 80-Hz condition). These results are clinically promising because when individual data were subjectively rated, they were comparable to statistical results found from the group data. CAEP gap-detection thresholds from my study correspond well with previous literature; however, they were about 2-6 ms higher than previously reported (Michalewski, Starr, Nguyen, Kong, & Zeng, 2005; Ross & Pantev, 2004). Desjardins et al. (1999) found that gaps of 4 ms elicited significant MMNs. Rupp et al. (2004) found that middle-latency auditory evoked fields were generated after gaps greater than or equal to 3 ms. CAEP gap-detection thresholds were also shown to be 3 and 5 ms by Ross and Pantev (2004) and by Michalewski et al. (2005), respectively. One plausible explanation for the discrepancy between my results and the literature is that most previous studies used non-AM stimuli. Gaps placed in non-AM stimuli cause spectral spread of energy that could be used as an additional cue by the auditory system and thus reducing (improving) gap-detection thresholds. However, placing gaps at null points in AM stimuli has minimal impact on spectral content of the stimulus energy. Another explanation for the discrepant results could be a result of variability in sampling the population’s mean gapdetection thresholds by using limited sample sizes (n<20). Furthermore, in my study gap step sizes (2-12 ms) near behavioural threshold were larger than most previous studies (2-3 ms) which could account for the higher averaged CAEP gap-detection thresholds in my study. In my study, average CAEP gap-detection thresholds were also comparable to behavioural gap-detection thresholds of 4.20 (40-Hz) and 4.85 ms (80-Hz). However, I did not find any relations between behavioural and physiological gap-detection thresholds. The lack of correlation could have resulted from the limited variance of the behavioural gap-detection 54  thresholds due to the participants all having normal temporal processing functions (i.e. restriction of range problem). Even though CAEP and behavioural gap-detection thresholds are comparable, CAEP thresholds are notably more variable than behavioural. Thus, when recording CAEPs to determine gap-detection thresholds, one should be concerned of the large variability in CAEP gap-detection thresholds before suggesting using this as a test for screening of temporal resolution deficits. Larger sample sized studies would be required to fully elucidate appropriate screening thresholds. What is interesting from my study’s data is that all normalhearing participants (n=12) showed significant CAEPs to at least 12.5 ms gaps for the 80-Hz condition. Furthermore, all CAEP gap-detection thresholds were below what is considered the normative behavioural gap-detection threshold of 20 ms for adults (Keith, 2000), with the exception of NG001 having a 25 ms threshold for the 40-Hz condition. CAEPs to gaps were larger for the 80- than 40-Hz condition, which likely led to lower CAEP gap-detection thresholds for the 80- Hz condition. Possible factors contributing to larger CAEPs to gaps for the 80-Hz condition could be differences in participants’ arousal state and attention. For instance, during sleep CAEPs can be larger, particularly P2, and can contain additional components such as a N300 (Picton, 2011). The P2 to the 25 ms gap in the 80-Hz condition is larger than that for the 40-Hz condition and there appears to be an additional N300 component (Figure 3.1). Conversely, if the participants were awake during the 80-Hz condition, they might have been paying more attention to the stimuli, which would increase their CAEP responses to stimulus changes such as gaps (Picton & Hillyard, 1974). 4.1.2 ASSR reset. The mean of the individual 40-Hz ASSR gap-detection thresholds as determined by the two raters was 6.42 ms, which was consistent with the group-averaged 40-Hz ASSR gap-detection threshold of 12.5 ms as determined by statistical testing across the time55  domain waveforms. These results are also comparable with the results from Ross & Pantev (2004) that found 40-Hz ASSR resets to gaps of 3, 6, 9, and 12 ms. Even though I only observed a significant ASSR reset to gaps of 12.5 and 25 ms in the group-averaged data, gaps of 3.125 and 6.25 ms also produced a similar ASSR reset morphology (Figure 3.1). The ASSR reset to these smaller gaps in my study likely did not reach significance because they were only an average of 200 trials; whereas, the Ross and Pantev study averaged 1600 trials. Importantly though, the mean 40-Hz ASSR gap-detection threshold in my study was only 2.2 ms above the mean behavioural threshold, which indicates good physiological-to-behavioural correspondence across individual participants. However, this correlation was not significant. This insignificant finding could be due to a restriction of range problem as mentioned previously. Further investigation with inclusion of participants that have a range of temporal resolution deficits could help determine relations between physiological and behavioural thresholds. From my study’s results, 40-Hz ASSR resets to gaps appear to be a valid method for assessing gapdetection thresholds, at least in individuals with normal temporal resolution abilities. Gap-detection thresholds for 80-Hz ASSRs were difficult to estimate because ASSRs did not reset to 12.5 ms gaps in 83% (n=10/12) of the participants but did reset to a 6.25 ms gap in 67% (n=8/12) of the participants. Thus it was hard to determine if the ASSR reset was actually reflecting a detection of all gaps below an individual’s threshold. It appears more likely that the ASSR reset is occurring as a result of a violation in the periodicity of the 80-Hz modulation frequency. This explanation is consistent with the findings that gap-detection for AM stimuli appeared to depend more on the amount of violation from the modulation frequency than the absolute duration of the gap (Sek & Moore, 2002). What is surprising is that the group-averaged 40-Hz ASSRs appeared to reset at all gap durations between 3 to 25 ms and not preferentially to 56  gaps that mainly violate the periodicity of the modulation frequency (present study; Ross & Pantev, 2004). However, raters classified ASSR resets in 83% (n=10/12) of the participants for the 12.5 ms gap but only 50% (n=6/12) of participants for the longer 25 ms gap. This could indicate that even for 40-Hz ASSRs, gaps that violate the 40-Hz modulation periodicity are more likely to elicit a reset. This could mean that the ASSR resets reflect neural oscillators that detect specific gap durations (i.e. characteristic gaps) that violate the frequency at which they like to oscillate best (characteristic oscillatory firing rates they like to fire at specific rates; Hewitt & Meddis, 1993). In other words, separate groups of neurons could code for discrete gap durations. These lower-level oscillators could send information about specific gaps to higher auditory centres that monitor all lower levels. The auditory cortices are likely the region in which performs this monitoring and thus can detect a full range of gaps. This is evidenced by CAEPs showing responses to all gaps, whereas ASSRs predominately reset to gaps that violate their modulation periodicity (e.g. half cycle). Also, there was no significant group 80-Hz ASSR reset observed to a 12.5 ms gap, whereas all participants (n=12) showed individual CAEPs to this gap duration. Because these two responses (ASSR resets and CAEPs) are the product of filtering of the same data, both responses should have been the same result if they were reflecting similar auditory levels of temporal processing. 4.2 Temporal Integration Gap-detection experiments are conventional measures of behavioural temporal resolution, whereas behavioural and physiological measures of integration time have not been well established. For my study, I obtained one measure of behavioural integration time and two measures of physiological integration time in order to assess any correlation between behavioural and physiological measures. Physiological integration times defined by tau (40-Hz 57  = 41 ms and 80-Hz = 54 ms) were significantly smaller when compared to behavioural integration defined by tau (40-Hz = 129 ms and 80-Hz = 136 ms). Because behavioural tau is within typical behavioural integration times of 100-200 ms for the auditory system (Pedersen & Salomon, 1977), I do not consider the physiological tau to be a good measure of integration time. However, the min-to-max peak latencies of ASSR resets (40-Hz = 112 ms and 80-Hz = 123 ms) showed comparable integration times to behaviour. Although min-to-max peak and behavioural integration times were comparable, the correlation did not reach significance. Again, this could be a restriction of range problem. My results, although shorter, do compare to the Yabe et al. 2007 study that found physiological integration time was between 150-170 ms. Previous research has suggested that the rise time of the ASSR is a measure of temporal integration (Ross & Pantev, 2004). However, my data showed no significant correlation between physiological and behavioural estimates of integration time. Therefore, more research is required to determine if the rise time of the ASSR is in fact a measure of temporal integration, or if it is just a coincidence that physiological estimates compared to behavioural estimates of temporal integration times. In addition, ASSR integration times for my study were about 50-75 ms shorter than those found in the Ross & Pantev study. Possibly, my ASSR integration times are shorter because I presented at a higher stimulus intensity but the Ross & Pantev study did not report dB SPL levels. Therefore, I cannot be certain if stimulus intensity differences can explain this discrepancy. 4.3 Theoretical Implications Brain responses to short gaps within a stimulus are indications that the CANS can detect fast temporal changes in acoustic signals (Desjardins et al., 1999; Michalewski, et al., 2005; Ross & Pantev, 2004; Rupp et al., 2004). My results showing behavioural and physiological 58  detection of short gaps between 4-12.5 ms provide further evidence to support the hypothesis that short-temporal integrators exist and thus support the multiple-looks model of auditory temporal processing (Viemeister & Wakefield 1991). Conflicting results from the CAEPs and the ASSRs that mainly reset to half-cycle gap durations provides evidence that different levels of the auditory system could be responding to the temporal aspects of the stimulus. Lower-level processes assessed by 40- and 80-Hz ASSRs appear to reflect violations in periodicity of temporal structure. To be able to assess the lower auditory centres’ abilities to detect gaps from 3 to 25 ms using ASSRs, modulation frequencies ranging between 20 to 160 Hz would be needed for the gaps to violate modulation periodicity by half a cycle. CAEPs, on the other hand, appear to reflect higher-level processes that combine multiple inputs from these lower centres to provide overall temporal information of the acoustic signal. Additional research to confirm such a model is required. Temporal integration results from my study provide further evidence that auditory information is integrated over a time period; behavioural thresholds improved when longer stimulus durations were presented. However, my results suggest that the auditory integrator is not a sliding window of approximately 200 ms that treats acoustic information within this window as the same acoustic event. Evidence for this is that I observed larger responses for longer gap durations. If the temporal window of integration hypothesis was true, I should have seen the same size of responses for gap durations less than 200 ms. Importantly, because we did not observe significant correlations between physiological and behavioural measures of temporal integration using the rise time of the ASSRs, it remains to be determined if the rise time of the steady-state actually represents temporal integration, or if comparable results were of coincidence. 59  4.4 Clinical Implications Based on the fact that my results corroborated previous findings that CAEPs and 40-Hz ASSR resets occurred to gaps near behavioural gap-detection thresholds (Ross and Pantev, 2004), CAEPs and 40-Hz ASSRs appear to be clinically viable for assessing adults with normal temporal processing. However, large-sample studies of individuals with a range of known temporal processing deficits should be conducted to evaluate the sensitivities and specificities of these methods in detecting individuals with temporal processing deficits. Such large-sample studies would help determine the validity of using ASSR resets and CAEPs to clinically evaluate temporal processing in adults. In the infant population, it is premature to recommend further research until the issue of periodicity violation with 80-Hz ASSRs is worked out in adults. Although behavioural tau and min-to-max ASSR integration times were comparable, further research is needed to determine if they are measuring the same construct of temporal integration or if they simply have coincidentally similar time constants. This is because I did not find a significant correlation between behavioural and physiological integration times. 4.5 Caveats A very large ASSR was observed at M1 (left mastoid) and M2 (right mastoid) for a few participants for the 40-Hz condition. Because the response amplitude was much larger than typical ASSRs, the likely cause of this response is due to the post-auricular muscle (PAM) response contracting at the rate of the modulation frequency (Picton, John, Purcell & Plourde, 2003). Electrodes M1 and M2 were placed behind the pinna, on the mastoid near or over the belly of the PAM. Loud sounds (> 75 dB SPL) typically elicit a large contraction of the PAM. I am not concerned with the electrical response due to the PAM contraction contaminating other electrodes because previous research showed no apparent PAM response in electrodes farther 60  than 2 cm (O’Beirne & Patuzzi, 1999). Because of this, AEPs in M1 and M2 were not considered in any analysis. In hindsight, a lower presentation level should have been used. However, I chose to use a presentation of 80 dB SPL to maximize signal-to-noise ratios for AEPs. Another caveat was that twice the number of trials was recorded for the 80-Hz condition as compared to the 40-Hz condition. This was done in order to improve the signal-to-noise ratio of the 80-Hz, while still maintaining practical recording times. Thus the 40-Hz condition had lower signal-to-noise ratios and likely resulted in reduced ability to detect CAEPs and ASSR resets. Previous studies have used a different approach when determining which gap durations to include for the physiological recordings. Instead of choosing set gap durations that are the same for each participant, the researchers found the participant’s behavioural threshold first, and chose the gaps that were used in the physiological recordings based on this threshold. These gaps usually consisted of a gap that corresponded to their behavioural gap-detection threshold, a gap that was clearly below their threshold (sub threshold), a gap that was clearly above their threshold (suprathreshold) and a control gap of either no gap (0 ms) or a 1 ms gap. If I would have used this approach instead of using set gap durations, then I might not have seen the large reset for the half cycle modulation frequency gap across the participants. If I did not happen to have gaps equal to half of the modulation frequency for the group of participants, I would not have seen this result.  61  4.6 Conclusions The current study investigated electrophysiological measures of auditory temporal resolution and integration. My results from CAEPs were consistent with the literature for assessing physiological gap-detection thresholds. Because behavioural and physiological results were similar, this indicated that CAEPs could be used as an objective assessment of gapdetection thresholds in adults. Therefore, this is evidence to support creating a screening procedure to assess physiological gap detection in adults. Lack of consistent finding in steady-state responses indicated that more research is required to determine physiological measures of temporal integration. The ASSR min-to-max peak latency appeared to be a physiological measure of temporal integration because it was similar to behavioural measures of temporal integration. However, this needs to be further examined. At this point the physiological and behavioural relationship is only speculative in nature. The unpredicted finding that ASSRs most robustly reset to violations in their periodicity supports the concept of a multi-tiered system for temporal processing, whereby lower centers might detect specific gap durations and higher centers integrate the lower center information. If this is true, ASSR resets could still be used as a measure of temporal resolution. However, further research is required in order to determine if this is possible.  62  References British Columbia Early Hearing Program [BCEHP]. (2012). Diagnostic audiology protocols. Retrieved October 4th, 2012 at www.phsa.ca/NR/.../0/BCEHPAudiologyAssessmentProtocol.pdf  Benasich, A., & Tallal, P. (2002). Infant discrimination of rapid auditory cues predicts later language impairment. Behavioural Brain Research, 136(1), 31-49. doi: 10.1016/S01664328(02)00098-0  Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society, Series B, 57, 289-300.  Bregman, A.S. (1990). Auditory Scene Analysis: The Perceptual Organization of Sound. Cambridge: MIT Press  Cohen, J. (1968). Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychological Bulletin, 70, 213-220.  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. The Journal of the Acoustical Society of America, 90(5), 2467-2479.  Desjardins, R.N., Trainor, L.J., Hevenor, S.J., & Polak, C.P. (1999). Using mismatch negativity to measure auditory temporal resolution threholds. Neuroreport, 10(10), 2079-2082.  63  Efron, B., & Tibshirani, R.J. (1993). An introduction to the bootstrap. London: Chapman & Hall.  Galambos, R., Makeig, S., & Talmachoff, P. J. (1981). A 40-hz auditory potential recorded from the human scalp. Proceedings of the National Academy of Sciences of the United States of America, 78(4), 2643-2647.  Geisler, C.D. (1960). Average response to clicks in man recorded by scalp electrodes. M.I.T. Technical Report, 380(1-158).  Hewitt, M.J., & Meddis, R. (1993). A computer model of amplitude-modualtion sensitivity of single units in the inferior colliculus. The Journal of the Acoustical Society of America, 95(4), 2145-2159.  Hillyard, S.A., & Picton, T.W. (1979). On and off components in the auditory evoked potential. Perception & Psychophysics, 24(5), 391-398.  Katz, J., Medwetsky, L. Burkard, R., & Hood, L. (Eds). (2009). Handbook of Clinical Audiology (6th ed), Baltimore, MD: Lippincott Williams & Wilkins.  Keith, R.W. (2000). Random Gap Detection Test (RGDT). St. Louis, MO: AUDiTEC of St. Louis Ltd.  Kennedy, C. R., McCann, D. C., Campbell, M. J., Law, C. M., Mullee, M., Petrou, S., . . . Stevenson, J. (2006). Language ability after early detection of permanent childhood hearing impairment. The New England Journal of Medicine, 354(20), 2131-2141. doi: 10.1056/NEJMoa054915 64  Kent, R.D., & Read, C. (2002). Acoustic Analysis of speech: second edition. Canada: Singular Publishing.  Kraus, N., McGee, T., & Comperatore, C. (1989). MLRs in children are consistently present during wakefulness, stage 1, and REM sleep. Ear and Hearing, 10, 339-345  Lins, O. G., Picton, P. E., Picton, T. W., Champagne, S. C., & Durieux-Smith, A. (1995). Auditory steady-state responses to tones amplitude-modulated at 80-110 hz. The Journal of the Acoustical Society of America, 97(5 Pt 1), 3051-3063.  Martin, B., & Boothroyd, A. (2000). Cortical, auditory, evoked potentials in response to changes of spectrum and amplitude. Journal of the Acoustical Society of America, 107(4), 21552161. doi: 10.1121/1.428556.  Maurizi, M., Almadori, G., Paludetti, G., Ottaviani, F., Rosignoli, M., & Luciano, R. (1990). 40hz steady-state responses in newborns and in children. Audiology : Official Organ of the International Society of Audiology, 29(6), 322-328.  Michalewski, H. J., Starr, A., Nguyen, T. T., Kong, Y. Y., & Zeng, F. G. (2005). Auditory temporal processes in normal-hearing individuals and in patients with auditory neuropathy. Clinical Neurophysiology : Official Journal of the International Federation of Clinical Neurophysiology, 116(3), 669-680. doi: 10.1016/j.clinph.2004.09.027.  Moore, B. C. (2003). Temporal integration and context effects in hearing. Journal of Phonetics, 31(3–4), 563-574. doi: 10.1016/S0095-4470(03)00011-1.  65  Moore, J., Wilson, W., & Thompson, G. (1977). Visual reinforcement of head-turn responses in infants under 12 months of age. Journal of Speech and Hearing Disorders, 42(3), 328-334.  Morr, M. L., Shafer, V. L., Kreuzer, J. A., & Kurtzberg, D. (2002). Maturation of mismatch negativity in typically developing infants and preschool children. Ear and Hearing, 23(2), 118-136.  Muir, D., Clifton, R., & Clarkson, M. (1989). The development of a human auditory localization response - a U-shaped function. Canadian Journal of Psychology-Revue Canadienne De Psychologie, 43(2), 199-216. doi:10.1037/h0084220.  Muluk, N. B., Yalcinkaya, F., & Keith, R. W. (2011). Random gap detection test and random gap detection test-expanded: Results in children with previous language delay in early childhood. Auris, Nasus, Larynx, 38(1), 6-13. doi: 10.1016/j.anl.2010.05.007.  Näätänen, R. (1990). The role of attention in auditory information processing as revealed by event-related potentials and other brain measures of cognitive function. Behavioral and Brain Sciences, 13(2), 201–208.  Näätänen, R. (1992). Attention and Brain Function. Hillsdale, NJ: Lawrence Erlbaum. O’Beirne, G.A., & Patuzzi, R.W. (1999). Basic properites of the sound-evoked post-auricular muscle response (PAMR). Hearing Research, 138(1-2), 115-132.  Okitsu, T. (1984). Middle components of the auditory evoked response in young children. Scandinavian Audiology, 13, 83-86.  66  Paul, R. (2006). Language disorders from infancy through adolescence. New York: Mosby.  Pedersen, C. B., & Salomon, G. (1977). Temporal integration of acoustic energy. Acta OtoLaryngologica, 83(5-6), 417-423.  Picton, T.W. (2011). Human auditory evoked potential. San Diego: Plural Publishing.  Picton, T. W., & Hillyard, S. A. (1974). Human auditory evoked potentials. II. effects of attention. Electroencephalography and Clinical Neurophysiology, 36(2), 191-199.  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. Anesthesia and Analgesia, 97(5). 1396-1402. Picton, T. W., Linden, R. D., Hamel, G., & Maru, J. T. (1983). Aspects of averaging. Seminars in Hearing, 4(4), 327-340). Plomp, R., & Bouman, M.A. (1959). Relation between hearing threshold and duration for tone pulses. The Journal of the Acoustical Society of America, 31(6), 749-758.  Ross, B., & Pantev, C. (2004). Auditory steady-state responses reveal amplitude modulation gap detection thresholds. The Journal of the Acoustical Society of America, 115(5 Pt 1), 21932206.  Rupp, A., Gutschalk, A., Hack, S., Scherg, M. (2002). Temporal resolution of the human primary auditory cortex in gap detection. Neuroreport, 13, 2203-2207.  67  Rupp, A., Gutschalk, A., Uppenkamp, S., & Scherg, M. (2004). Middle latency auditory-evoked fields reflect psychoacoustic gap detection thresholds in human listeners. Journal of Neurophysiology, 92(4), 2239-2247. doi: 10.1152/jn.00163.2004.  Sek, A., & Moore, B.C. (2002). Mechanisms of modulation gap detection. The Journal of the Acoustical Society of America, 111(6), 2783-2792. Small, S. A., & Werker, J. F. (2012). Does the acc have potential as an index of early speechdiscrimination ability? a preliminary study in 4-month-old infants with normal hearing. Ear and Hearing, 33(6), e59-69. doi: 10.1097/AUD.0b013e31825f29be; 10.1097/AUD.0b013e31825f29be Stapells, D. R., Galambos, R., Costello, J. A., and Makeig, S. (1988). Inconsistency of auditory middle latency and steady-state responses in infants. Electroencephalography and Clinical Neurophysiology 71:289–295. Stapells, D.R., Herdman, A., Small, S.A., Dimitrijevic, A. & Hatton, J. (2005). Current status of the auditory steady-state responses for estimating an infant’s audiogram. In: In R. C. Seewald & J.M. Bamford (Eds.) A Sound Foundation Through Early Amplification 2004. Basel: Phonak AG. Pp. 43-59.  Stapells, D. R., Linden, D., Suffield, J. B., Hamel, G., & Picton, T. W. (1984). Human auditory steady state potentials. Ear and Hearing, 5(2), 105-113.  68  Stapells, D., Makeig, S., & Galambos, R. (1987). Auditory steady-state responses - threshold prediction using phase coherence. Electroencephalography and Clinical Neurophysiology, 67(3), 260-270. doi:10.1016/0013-4694(87)90024-1.  Suzuki, T., and Kobayashi, K. (1984). An evaluation of 40-Hz event-related potentials in young children. Audiology 23:599–604.  Tallal, P., Miller, S., & Fitch, R. H. (1993). Neurobiological basis of speech: A case for the preeminence of temporal processing. Annals of the New York Academy of Sciences, 682, 27-47.  Thompson, G., & Wilson, W.R. (1984). Clinical application of visual reinforcement audiometry. Seminars in Hearing, 5, 85-99.  Tomblin, J. B., Records, N. L., Buckwalter, P., Zhang, X., Smith, E., & O'Brien, M. (1997). Prevalence of specific language impairment in kindergarten children. Journal of Speech, Language, and Hearing Research : JSLHR, 40(6), 1245-1260.  Trainor, L. J., Samuel, S. S., Desjardins, R. N., & Sonnadara, R. R. (2001). Measuring temporal resolution in infants using mismatch negativity. Neuroreport, 12(11), 2443-2448.  Trehub, S. E., & Henderson, J. L. (1996). Temporal resolution in infancy and subsequent language development. Journal of Speech & Hearing Research, 39(6), 1315-1320.  69  Van Maanen, A., & Stapells, D. (2005). Comparison of multiple auditory steady-state responses (80 versus 40 hz) and slow cortical potentials for threshold estimation in hearing-impaired adults. International Journal of Audiology, 44(11), 613-624. doi:10.1080/14992020500258628  Viemeister, N. F., & Plack, C. (1993). Temporal processing. In Human Psychoacoustics. W.A. Yost, A.N. Popper, and R.F. Fay, eds. New York: Springer-Verlag.  Viemeister, N.F., & Wakefield, G.H. (1991). Temporal integration and multiple looks. The Journal of the Acoustical Society of America, 90(2), 858-865.  Wunderlich, J. L., & Cone-Wesson, B. K. (2006). Maturation of CAEP in infants and children: A review. Hearing Research, 212(1-2), 212-223. doi: 10.1016/j.heares.2005.11.008  Yabe, H., Tervaniemi, M., Reinikainen, K., & Näätänen, R. (1997). Temporal window of integration revealed by MMN to sound omission. Neuroreport, 8(8), 1971-1974.  Yoshinaga-Itano, C., Sedey, A. L., Coulter, D. K., & Mehl, A. L. (1998). Language of earlyand later-identified children with hearing loss. Pediatrics, 102(5), 1161-1171.  Yost, W.A. (2007). Fundamentals of hearing. San Diego: Elsevier.  70  Appendices Appendix A: Distribution of waveforms across all electrodes.  Appendix A1: CAEP 40 Hz  71  Appendix A2: CAEP 80 Hz  72  Appendix A3: ASSR 40 Hz  Note. Responses shown are the steady-state ASSRs, not the envelope calculated from the Hilberttransform.  73  Appendix A4: ASSR 80 Hz  Note. Responses shown are the steady-state ASSRs, not the envelope calculated from the Hilbert-transform.  74  Appendix B: Rating results from CAEPs and ASSR resets.  The confidence of raters was also determined by determining how the raters responded to the six gap durations. For each gap duration for the CAEPs and ASSR resets, the number of ‘present present’ (YY), ‘present maybe’ (YM), ‘maybe maybe’ (MM), ‘present no-response’ (YN), ‘maybe no-response’ (MN), and ‘noresponse no-response’ (NN) were calculated. Figure 3.4 contains the results of this calculation. Raters were overall in more agreement when rating the CAEPs as compared to the ASSR resets. For the CAEPs, raters were in high agreement for rating the no-gap control condition where they both reported ‘NN’ and for the 12.5 and 25 ms gap where the majority of the responses were ‘YY’.  75  

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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073418/manifest

Comment

Related Items