You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cortical auditory evoked potentials to gaps in broadband noise in infants Jordan, Rachel 2016

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2016_november_jordan_rachel.pdf [ 2.21MB ]
JSON: 24-1.0319258.json
JSON-LD: 24-1.0319258-ld.json
RDF/XML (Pretty): 24-1.0319258-rdf.xml
RDF/JSON: 24-1.0319258-rdf.json
Turtle: 24-1.0319258-turtle.txt
N-Triples: 24-1.0319258-rdf-ntriples.txt
Original Record: 24-1.0319258-source.json
Full Text

Full Text

CORTICAL AUDITORY EVOKED POTENTIALS TO GAPS IN BROADBAND NOISE IN INFANTS by  Rachel Jordan  B.Mus., McGill University, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Audiology And Speech Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2016  © Rachel Jordan, 2016   ii Abstract Purpose: There are currently no objective measures to evaluate hearing function in young infants with impaired temporal processing (e.g., Auditory Neuropathy Spectrum Disorder (ANSD)). The present study investigated cortical auditory evoked potentials (CAEPs) elicited to gaps in broadband noise to assess temporal processing ability in infants. This method has potential as a clinical tool in these populations. This study was intended as a first step to determine feasibility.  Method: Participants were 10 adults and 22 infants with normal hearing. Stimuli were continuous broadband noise with 20, 50 and 100 ms gaps inserted once per second. CAEPs were recorded at Cz referenced to M1. Two replications (minimum of 75 trials each) were included for analysis and judged by three raters for response presence or absence. Results: CAEPs were interpreted as present for the majority of infants and adults to 20-, 50- and 100-ms gaps. The adults had responses in almost all cases, consistent with previous results. In most cases, the morphology of the infant response was consistent with previous results, that is, a single peak at approximately 200 ms. In some cases, however, the infant response was a plateau-shaped peak. Conclusions: These results suggest that it is feasible to record CAEPs in infants to gaps and that infants may have better gap detection abilities than previously thought (20 ms or better). Further research is needed to refine this technique and to extend it to clinical populations for clinical use.    iii Preface The research described in this thesis was carried out in collaboration with my supervisor, Dr. Susan Small and committee member Dr. Anthony Herdman. However, the study was originally conceived by Dr. Small and Dr. Herdman. I recruited participants, set up and executed the experiment, and analysed the results. Dr. Small assisted with all aspects of the project, including study design, participant recruitment, experimental setup and writing. Dr. Herdman assisted with technical aspects of experimental setup and analysis. Some of the code used to set up the experiment and analyse the results was modified from existing code. Some of the code to create the stimuli was modified from code in the public domain found at Matlab code for analysis was modified from code provided by Dr. Herdman. This study was previously published as an abstract for the World Congress of Audiology (Poster Presentation), Vancouver, BC (September 18-21, 2016).  This research was approved by the UBC Behavioural Research Ethics Board Certificate H07-01218.   iv Table of Contents  Abstract .......................................................................................................................................... ii	Preface ........................................................................................................................................... iii	Table of Contents ......................................................................................................................... iv	List of Tables ................................................................................................................................ vi	List of Figures ............................................................................................................................. viii	List of Abbreviations ................................................................................................................... ix	Acknowledgements ........................................................................................................................x	Dedication ..................................................................................................................................... xi	Chapter 1: Introduction ................................................................................................................1	1.1	 Auditory Neuropathy Spectrum Disorder (ANSD) ........................................................ 2	1.2	 ANSD, Temporal Processing and Speech Perception .................................................... 5	1.3	 Psychophysical Gap Detection ....................................................................................... 7	1.4	 Electrophysiological Gap Detection ............................................................................... 9	1.5	 Cortical Auditory Evoked Potentials (CAEPs) in Infants ............................................. 11	1.6	 Summary ....................................................................................................................... 12	Chapter 2: Methods .....................................................................................................................14	2.1	 Participants .................................................................................................................... 14	2.2	 Stimuli ........................................................................................................................... 14	2.3	 EEG Recording ............................................................................................................. 15	2.4	 Procedure ...................................................................................................................... 16	2.5	 Analysis......................................................................................................................... 16	  v Chapter 3: Results ........................................................................................................................20	3.1	 Presence and Absence of Responses ............................................................................. 20	3.2	 Waveform Morphology ................................................................................................ 20	3.3	 Threshold Estimation .................................................................................................... 25	3.4	 Inter-Rater Reliability ................................................................................................... 26	Chapter 4: Discussion ..................................................................................................................29	Bibliography .................................................................................................................................36	Appendix A ...................................................................................................................................45	   vi List of Tables Table 3-1. Total number of infants included in each control and experimental condition, and number of infants with “present,” “absent” or “ambiguous” waveform interpretations. The brackets show the ratings from raters 1, 2 and 3 respectively. ..................................................... 20	Table 3-2. Mean (1SD) amplitudes and latencies of infant and adult CAEPs for each experimental condition. ................................................................................................................. 24	Table 3-3. Mean (1SD) SDR values for each control and experimental condition by waveform interpretation. Blank cells indicate that no data fit that category. ................................................ 24	Table 3-4. Summary of waveform interpretations for each individual infant across conditions and estimated gap detection threshold (GDT). “P” refers to “present,” “A” refers to “absent,” and “?” refers to “ambiguous.” Table is in order of youngest to oldest infant. Blank cells indicate that no data was collected or analyzed for that infant/condition. ............................................................. 26	Table A.1. Breakdown of participant information and SDR values, p-values, number of trials, and ratings for the 0 ms control condition. Table is sorted in age order as in Table 3-4. ............. 45	Table A.2. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 10 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” ..................................... 46	Table A.3. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 20 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” ..................................... 47	Table A.4. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 50 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” ..................................... 48	  vii Table A.5. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 100 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” ..................................... 49	   viii List of Figures Figure 2.1. Sample rating image from infant I05 in the 50 ms condition. .................................... 17	Figure 3.1. Grand mean average (thick) and individual (thin) waveforms for infants and adults with “present” interpretations across all conditions. The vertical line represents gap onset. One infant waveform was excluded from the grand mean for the 0 ms condition, because the file was corrupted. ...................................................................................................................................... 22	Figure 3.2. Cumulative ratings of waveforms for each condition for infants (top) and adults (bottom). Each rating figure received three ratings, one from each rater as described in Section 2.5. Ratings of 1, 2 or 3 (the purple, blue, and turquoise columns) represent judgments of “absent”; ratings of 5, 6 or 7 (the yellow, orange, and red columns) represent judgments of “present”, and ratings of 4 (the green column in the middle) represent “could not evaluate” judgments. ..................................................................................................................................... 27	Figure 3.3. Cumulative ratings for rating figures for each individual rater. Each bar represents the number of times a rater gave a rating figure a given rating. Colour scheme is the same as in . ... 28	   ix List of Abbreviations ABR – Auditory Brainstem Response ACC – Acoustic Change Complex AEP – Auditory Evoked Potential ANSD – Auditory Neuropathy Spectrum Disorder CAEP – Cortical Auditory Evoked Potential CM – Cochlear Microphonic dB SPL – decibels Sound Pressure Level EHDI – Early Hearing Detection and Intervention ISI – InterStimulus Interval GDT – Gap Detection Threshold MMN – Mismatch Negativity ms - milliseconds NICU – Neonatal Intensive Care Unit OAE – Otoacoustic Emissions PBK – Phonetically Balanced Kindergarten PCHL – Permanent Childhood Hearing Loss SD – Standard Deviation SDR – Standard Deviation Ratio   x Acknowledgements The work presented here was made possible with the support of a multitude of people and organizations. First of all, I thank my supervisor, Dr. Susan Small, whose expertise, patience and support pointed me in the right direction over and over again, and gave me the confidence to keep moving forward.  Secondly, I thank the many families who came to participate in this research, and the various organizations that allowed me to tell them about my project (Vancouver Coastal Health and many midwifery clinics). Their willingness to make the trek to our lab and persist in spite of whatever challenges the babies threw our way made this project possible.  Third, I thank the volunteers and classmates who assisted with data collection, Annie Wang, Geneva Gamble, Rui Lui and Myron Huen. I would also like to thank my committee members, Dr. Anthony Herdman, and Jenny Hatton, for their time and feedback. This project was also facilitated by financial support from the Natural Sciences and Engineering Research Council (NSERC) via a Canada Graduate Scholarship, and a stipend supported by Dr. Small’s Discovery Grant.  Finally, no graduate student exists in a vacuum. Behind the research presented here is a mountain of love and support that kept me going on the most challenging of days. Thank you to my husband Wendelin who tolerated me in the fowlest of moods and took an interest in topics that otherwise may not have interested him, to Myron Huen for commiserating with me, but also celebrating milestones, to Nanuq for giving me a reason to go outside, and cuddling with me when I was sad, and to my parents, siblings and in-laws who appear to have always believed that I would someday complete this journey!   xi Dedication   To Nanuq, my dog, who cheered me on through the whole process, and even served as my first audience for the defense! And to Echo, my hearing-impaired dog, who inspires me to think creatively about hearing loss.    1 Chapter 1: Introduction  Over the past decade, early hearing detection and intervention (EHDI) programs to identify infants with permanent childhood hearing loss (PCHL) have become commonplace. The predominant impetus for these programs is that beginning intervention in children with hearing loss before the age of six months has been shown to lead to significant improvements in language development outcomes (Yoshinaga-Itano, Sedey, Coulter, & Mehl, 1998). Prior to the introduction of EHDI programs, children with hearing loss were commonly identified as late as 3 years of age (Coplan, 1987; Van Naarden, Decoufle, & Caldwell, 1999). The Joint Committee on Infant Hearing (JCIH) provides recommendations for the implementation of EHDI programs (Joint Committee on Infant Hearing, 2007). Specifically, they recommend that all infants be screened by one month of age, that referring infants are diagnosed by three months of age, and that infants with PCHL begin receiving intervention by six months of age. Screening is done using otoacoustic emissions (OAEs) or automated Auditory Brainstem Responses (ABR), often in the hospital before the family goes home. Infants who refer from screening then receive a full diagnostic ABR. A diagnostic ABR is able to distinguish between sensorineural and conductive hearing loss, as well as estimate air- and bone-conduction pure-tone thresholds with sufficient accuracy to fit amplification and begin other interventions. Auditory Neuropathy Spectrum Disorder (ANSD) presents us with challenges when trying to meet the JCIH goals. While infants can be identified with ANSD within the recommended timeframe, it is not possible to begin management of the hearing loss with confidence because the severity of the hearing loss and the degree of neural involvement are difficult to assess in a young infant. Essentially, we cannot acquire sufficient information about the child’s case to best support speech and language development.   2 Hopefully, with future research, we will be able to improve on this situation. Given what we know about the perceptual deficits of patients with ANSD, it may be possible to develop an objective assessment tool, such as cortical auditory evoked potentials (CAEPs) to give us more information about a child’s hearing, especially within the first year of life. This project was intended as a first step towards this goal. 1.1 Auditory Neuropathy Spectrum Disorder (ANSD)  ANSD is a hearing disorder characterized by an absent or abnormal ABR inconsistent with behavioural audiometry and present otoacoustic emissions (OAEs) (Starr, Picton, Sininger, Hood, & Berlin, 1996). These patients also have poorer speech perception abilities than expected based on pure-tone thresholds, as evidenced by poor word recognition scores and reported difficulties understanding speech in everyday life. Because clinical findings typically show evidence of an intact cochlea, or at least intact outer hair cells, the disorder is believed to result from a neuropathy in the auditory nerve (Starr et al., 1996; reviewed in Starr & Rance, 2015).  When the disorder was first described by Starr and colleagues (1996), they coined the term “Auditory Neuropathy.” The description, “neuropathy” was based on the observation that eight of their ten patients with the disorder had additional peripheral neuropathies. Since this original description of the disorder, hundreds of cases have been identified, many of which did not involve these additional neuropathies, which has led to much debate regarding the appropriate term to use (Berlin, Hood, & Rose, 2001, 2002; Marsh, 2002; Rapin & Gravel, 2006). Some researchers argued for the term auditory dys-synchrony, citing ambiguity of the site of lesion and the finding that cochlear implants help these patients in a way that would not be expected in cases of simple damage to the auditory nerve (Berlin et al., 2001, 2002). Others argued in favour of the more general term “neural hearing loss,” suggesting that this is the most   3 accurate term (Marsh, 2002; Rapin & Gravel, 2006). The conflict over the name for the disorder was resolved by consensus in Lake Como, Italy in 2008; since then, experts have allegedly agreed to the term Auditory Neuropathy Spectrum Disorder (ANSD) (Hayes & Sininger, 2008).  Given the clinical findings of present OAEs and absent or abnormal ABR, possible sites of lesion include the inner hair cells, the synapse between the inner hair cells and the auditory nerve, the auditory nerve itself (pre- or post-ganglionic, or the cell body itself), or even the brainstem. If the damage is in the auditory nerve, it can be demyelinating or axonal, and this distinction may impact the associated sensory deficits and the progression of the disorder (Rapin & Gravel, 2006). In some patients, the disorder is a symptom of another disease or syndrome, such as Charcot-Marie-Tooth disease, and in these cases inferences can be made about the underlying pathology (Starr et al., 1996). However, in most cases, it is not possible to distinguish between these possible pathologies. In infants in particular, where speech perception deficits cannot be evaluated, it is impossible to develop a complete clinical picture.   The majority of infants with ANSD indicators, that is, present OAEs and abnormal ABRs, are premature or have suffered hypoxia (Berlin et al., 2010; Harrison, Gordon, Papsin, Negandhi, & James, 2015; Rea & Gibson, 2003). Other known risk factors for ANSD include hyperbilirubinemia, genetic conditions and peripheral neuropathies (Berlin et al., 2010; Rance et al., 1999). Regardless of the etiology of the disorder, screening techniques must be adapted to identify ANSD in infants. Given that OAEs are by definition present in patients with ANSD, EHDIs that use only OAEs are likely to miss most or all newborns that have the disorder. Rather than recommending an alternate screening for all infants, the JCIH recommends that the automated ABR be used only for babies in the neonatal intensive care unit (NICU), given the NICU related risk factors described above (Joint Committee on Infant Hearing, 2007). While   4 some estimates suggest that as many as 35% of infants with ANSD present with no known risk factors, and these infants will be missed in OAE screening (Berlin et al., 2010), the disorder is either considered sufficiently rare that this does not amount to a substantial number of cases, or this estimate is considered unreasonably high. The highest estimates suggest that 10% of cases of hearing loss are ANSD (Rance et al., 1999). Given a prevalence of 1 in 900 for infants with hearing loss (A. Davis & Wood, 1992), this amounts to a prevalence of 1 in 9,000 for children with ANSD. If 35% of cases were missed, the prevalence of missed cases would be approximately 3.8 in 100,000.  Secondly, diagnostic ABR results from infants with ANSD could be confused with a profound hearing loss of cochlear origin if the ABR response is absent. Therefore, a thorough diagnostic protocol also includes additional testing to distinguish ANSD from profound hearing loss. One typical finding in patients with ANSD is the cochlear microphonic (CM), a response that can be detected in the ABR that originates in the outer hair cells (Rance et al., 1999). The CM is distinct from the ABR in that it reverses polarity when the stimulus reverses polarity. Therefore, it can be detected by conducting a click ABR at high intensity using rarefaction and condensation clicks separately. The presence of a CM is a clear indication of ANSD. A click ABR response will typically not be present in infants with profound hearing loss, and therefore this test can distinguish between these two types of hearing loss. Unfortunately, once ANSD has been identified in young infants, there are no measures currently available to determine its severity, expected outcomes or recommended intervention. The absent or abnormal ABR results do not predict the pure-tone thresholds, and regardless, the pure-tone thresholds would not be indicative of the severity of speech perception deficits (Starr et al., 1996). At this time, pure-tone thresholds must be tested behaviourally when the child is old   5 enough to participate in visual reinforcement audiometry (about 6-9 months), and speech perception can only be measured as the child begins to develop sufficient language skills around two years of age.  Given the lack of predictive information available following identification of ANSD, the most common approach is to wait and see how the child develops (Harrison et al., 2015). Depending on the child’s development and results from subsequent behavioural testing, the child may receive a trial with hearing aids. If the trial with hearing aids is unsuccessful, and language development is not progressing, the family may consider cochlear implants. The severity of the disorder can range from mild enough to require no intervention, to severe enough that even with a cochlear implant, the patient is unable to develop spoken language (Berlin et al., 2010). According to Harrison et al. (2015), the average age of implantation for a young child with ANSD receiving a cochlear implant is 3 years, much later than for children with profound cochlear hearing loss.  Regardless of the end result, the JCIH recommendation of intervention by 6 months is not met in the same way for children with ANSD as it is for children with more common forms of permanent childhood hearing loss (PCHL). Hence, there is a need for improved assessment protocols for these children that can enable parents and clinicians to make more informed decisions about intervention for these children.  1.2 ANSD, Temporal Processing and Speech Perception  The speech perception deficits in ANSD are widely believed to be a result of deficits in temporal processing. Zeng, Kong, Michalewski, & Starr (2005) measured a variety of auditory psychophysical abilities in 21 adults and children (aged 6-53 years) with ANSD and found a pattern of deficits in temporal processing, including gap detection, temporal modulation   6 detection, forward and backward masking and sound localization involving inter-aural time differences. Specifically, gap detection thresholds for ANSD patients were 15-20 milliseconds (ms) longer than normal controls at suprathreshold intensities. Other tasks not involving temporal processing, such as sound localization involving inter-aural level differences and intensity discrimination, were not impaired. Furthermore, Rance et al. (2008) argued that the specific speech perception deficits in three patients with ANSD were related to temporal processing issues. Compared with patients with a cochlear hearing loss, they showed more difficulty distinguishing voiced from voiceless consonants (e.g. /b/ vs. /p/). Acoustically, the difference between these consonants relates to the voice onset time, which is a temporal feature.  Other researchers have demonstrated a relationship between psychophysical temporal processing and speech perception ability in patients with ANSD. Zeng, Oba, Garde, Sininger, & Starr (1999) found a significant correlation between word recognition scores and both gap detection and amplitude modulation detection in 8 patients with ANSD. Rance, McKay, & Grayden (2004) also reported a strong correlation between amplitude modulation thresholds and monosyllabic word recognition scores in 14 children with ANSD.  Objective measures, such as CAEPs have also been used to investigate temporal processing in ANSD patients. CAEPs are cortically generated evoked potentials with a latency of approximately 50 ms or longer (P. A. Davis, 1939). The adult P1-N1-P2 response is typically measured around 50-250 ms after the onset of a stimulus or change in acoustic parameters of a stimulus. In the latter case, this response is sometimes referred to as the acoustic change complex (ACC) (Martin & Boothroyd, 1999). The P1-N1-P2 is primarily generated in the auditory cortex, although there appear to be multiple generators throughout the cortex (Picton et al., 1999). In spite of the absent ABR, CAEPs tend to be present in patients with ANSD (Rance, Cone-  7 Wesson, Wunderlich, & Dowell, 2002). He et al. (2015) found a negative correlation between objective gap detection using the ACC and aided Phonetically Balanced Kindergarten (PBK) word recognition scores in 19 children between 2 and 15 years old with ANSD. That is, children with more elevated gap detection thresholds had poorer word recognition scores. Gap detection thresholds in these children ranged from 10 to 100 ms. Michalewski, Starr, Nguyen, Kong, & Zeng, (2005) compared behavioural and electrophysiological gap detection thresholds in adults with and without ANSD, and found a relationship between those measures in both groups. Gap detection thresholds in normal hearing adults were around 5 ms, but were 10 – 50 ms in ANSD patients. No published study to date has examined these abilities in infants with ANSD, either behaviourally or electrophysiologically. Electrophysiological measures would, however, be particularly valuable for infants, as this form of testing would not require the infant to respond or use potentially delayed language skills.  In summary, the literature demonstrates a strong and repeatable relationship between deficits in temporal processing and speech perception ability in children and adults with ANSD, both when these abilities are measured behaviourally and when they are measured objectively. Therefore, it is plausible that objective measures of temporal processing in infancy would relate to later speech perception performance. 1.3 Psychophysical Gap Detection Gap detection has been thoroughly examined in normal-hearing adults (for review, see Phillips, 1999). Among other things, gap detection thresholds are affected by stimulus level (Plomp, 1964), marker frequency (Fitzgibbons, 1983), and marker duration (Schneider & Hamstra, 1999). Typically, these experiments use a two-alternative-forced-choice paradigm. Each alternative will contain a leading marker and a trailing marker, but only one will contain a   8 gap. This methodology allows for control over other stimulus characteristics that might alert the listener to the presence of a gap, such as spectral splatter in the case of tonal stimuli. Most studies find gap detection thresholds of a few milliseconds using broadband noise, when stimulus parameters are optimized for best possible detection ability (Fitzgibbons, 1983; Plomp, 1964; Schneider & Hamstra, 1999). The existing literature measuring gap detection in infants behaviourally is somewhat limited. One study investigated gap detection in broadband noise, and found that thresholds for infants aged 3, 6 and 12 months were significantly higher than adults (Werner, Marean, Halpin, Spetner, & Gillenwater, 1992). Gap detection thresholds in broadband noise in this study were approximately 50 ms for infants 12 months and younger, as compared to approximately 5 ms in adults. In contrast, Trehub, Schneider, & Henderson (1995) found that while gap detection thresholds to 500-Hz tone bursts in infants aged 6.5 and 12 months were elevated as compared to adults and 5-year-old children, the difference was less than 5 ms. While Trehub and colleagues did not find elevated thresholds in 5-year-old children when compared to adults, other studies have found elevated thresholds in school-aged children (Irwin, Ball, Kay, Stillman, & Rosser, 1985; Lister, Roberts, & Lister, 2011). Lister and colleagues also found that the thresholds in 7- and 8-year-old children were highly variable. This suggests that the development of temporal processing is complex. Stimulus parameters often imply multiple neural mechanisms which may develop separately (Lister et al., 2011). Neither of the studies involving infants reported on the range of performance across children, but Trehub & Henderson (1996) found that the infants from the earlier study who had better than average gap detection abilities had better language abilities in the toddler years according to parent-report language measures. This suggests that the variability of temporal   9 processing abilities in children and infants is highly relevant to language development, and warrants further investigation. 1.4 Electrophysiological Gap Detection Gap detection ability has also been investigated using auditory evoked potentials (AEPs). As discussed above, temporal processing has been investigated in subjects with ANSD using primarily cortically-generated AEPs. Werner, Folsom, Mancl, & Syapin (2001) also measured gap detection ability, but at the level of the brainstem using ABRs. Using broadband noise, they found that behavioural and ABR gap detection thresholds were similar in adults. On average, the adults had behavioural thresholds that were 2.6 ms and ABR thresholds of 2.4 ms. In contrast, Werner and colleagues also measured both behavioural and ABR thresholds in infants, and found large differences. In 3- and 6-month-old infants, they found behavioural thresholds between 50 and 60 ms, but ABR thresholds in a separate group of 3-month-old infants were not significantly different from adults. The fact that these differences exist behaviourally but not in the ABR suggests that the behavioural differences between infants and adults originate at a level of the auditory system more central than the brainstem. To my knowledge, no study has independently investigated electrophysiological gap detection abilities in older children. Mismatch Negativity (MMN) has also been used to measure gap detection in infants. In an MMN paradigm, responses to an oddball stimulus are compared to responses to a standard stimulus (Näätänen, Gaillard, & Mäntysalo, 1978). Using this paradigm, Trainor, Samuel, Desjardins, & Sonnadara (2001) found evidence of physiological gap detection to 4 ms gaps in 2000-Hz Gaussian tone pips in 6-month-old infants. However, Trainor et al. (2003) found no gap detection response in 2- to 4-month-old infants for 16 ms gaps. For the 6-month-old infants, the responses were similar to adults, but the responses from the younger infants were more   10 consistent with the behavioural literature discussed above. Overall, however, the literature describing gap detection ability in infants both behaviourally and electrophysiologically is very limited with small sample sizes, and the results cannot be interpreted as more than a guideline for future research.  In adults, much of the research on AEPs to gaps has focused on some unique morphological features. Michalewski et al. (2005) first measured CAEPs to gaps in noise and identified a double N1 peak, instead of the usual single peak. Although this appeared to be overlapping N1 peaks to noise offset and noise onset, they ruled out this explanation when they measured the same double peak in response to gaps large enough that the noise onset response was entirely separate. Clearly, the electrophysiological response to stimulus offset, as in the beginning of a gap, is different from the response to stimulus onset. Pratt and colleagues further explored this unusual morphology, and hypothesized that this second peak is a separate process with separate generators, dubbed “N(egation)-process” (Pratt, Bleich, & Mittelman, 2005). Interestingly, they observed a shift in hemispheric lateralization between the two N1 peaks. Both of these papers reported a gap detection threshold of 5 ms to gaps in broadband noise in normal hearing adults (Michalewski et al., 2005; Pratt et al., 2005). Finally, Atcherson and colleagues investigated this N-complex in a much larger sample size, using narrowband noise instead of broadband noise, and found that only 64 out of 144 subjects had this morphology to 50 ms gaps (Atcherson, Gould, Mendel, & Ethington, 2009). Clearly, CAEPs to noise offset are unique, with significant individual variability.  Therefore, while the existing data on children and adults suggest that electrophysiological measures of gap detection may eventually prove to be useful in predicting ANSD severity in infants, there is currently too little research on infants, both behaviourally and   11 electrophysiologically to implement this idea.1 The current study will provide some of the much-needed data on typical responses elicited in normal hearing infants and explore the question of feasibility. 1.5 Cortical Auditory Evoked Potentials (CAEPs) in Infants The morphology of the infant CAEP to stimulus onset is distinct from the adult version, consisting in most cases of one large positive peak at around 100 to 200 ms (for review, see Wunderlich & Cone-Wesson, 2006). Mapping the developmental trajectory is confounded by differences in interstimulus interval (ISI) and type of stimulus, making it difficult to compare results across studies. For example, Gilley and colleagues found that the more complex form of the CAEP involving two positive peaks with a negative peak between them, which was always present in adults, was present more often in young children for slower stimulation rates (Gilley, Sharma, Dorman, & Martin, 2005). According to Schafer and colleagues, the additional peaks begin to emerge between 6- and 30-months (Shafer, Yu, & Wagner, 2015). Ponton and colleagues demonstrated the importance of scalp location on the developmental trajectories, and also observed abrupt changes in the amplitude and latency of the peaks in 10-year-old children (Ponton, Eggermont, Kwong, & Don, 2000). Furthermore, they found that the P1 and N1 peaks and their various generators followed a maturational path distinct from the P2 peak, and suggested that these peaks represent different pathways in the central auditory system with unique developmental trajectories. As further support for this hypothesis, Barnet and colleagues reported that P2 becomes stable in amplitude and latency around 12 months of age (Barnet,                                                 1 CAEP data were collected in a small group of infants using gaps inserted in a vowel stimulus (K. Gardner-Berry, personal communication, January 2014), but to my knowledge have not been published.   12 Ohlrich, Weiss, & Shanks, 1975), whereas Ponton and colleagues found major changes in P1 and N1 in children as old as 10 years. Given that the human infant brain is not fully developed at birth, changes in the brain over the course of development must be considered when looking at AEP development. Starting in the last few weeks of gestation, the infant brain begins developing synapses, and this continues until the infant is 2- to 4-years-old (Huttenlocher & Dabholkar, 1997). Interestingly, this process happens at different rates in different areas of the infant cortex, and within a given area, this process happens at different rates in different layers of the cortex. Furthermore, development is not complete when the synapse count reaches its peak; rather, synaptic pruning continues into adulthood. Changes in the number of synapses could easily impact the amplitude of an evoked potential. Also, as AEP morphology is dependent on the orientation of dipoles, changes in the organization of the auditory cortex would lead to changes in the morphology and topography of the scalp-recorded response. Other ways that the human brain develops over childhood include an increase in myelination, leading to faster conduction rates, and improved synchronicity, leading to sharper peaks in evoked potentials. 1.6 Summary The present study was designed to investigate potential clinical measures for the assessment of ANSD severity in infants. Given the relationship between gap detection ability and speech perception ability in older children and adults with ANSD, this study explored gap detection in normal-hearing infants using CAEPs. This response can be expected to be present in patients with ANSD, and this testing method requires less cooperation from patients than behavioural methods.   13 Gap durations for this study were chosen based on the existing behavioural literature. Specifically, a sub-threshold gap duration of 20 ms, a threshold gap duration of 50 ms and a supra-threshold gap duration of 100 ms were used in addition to a control condition with 0 ms gaps, or no gaps. The 100 ms gap was predicted to elicit CAEPs from all infants, while the 50 ms gap was predicted to elicit smaller CAEPs from some infants, and the 0 ms and 20 ms gaps were predicted to elicit no CAEPs from infants.     14 Chapter 2: Methods 2.1 Participants  Study participants were 22 infants and 10 adults all with normal hearing. Participants were recruited from the community via social media, drop-in groups and personal networking. All infants were screened for hearing loss in both ears with transient-evoked OAEs using a Madsen Accuscreen. Testing was conducted if the infant passed screening in at least one ear. All but three infants passed screening in at least one ear. Two infants passed screening in one ear only, and that ear was used in the experiment. Data (partial or complete) was collected from 19 infants (age range: 9-50 weeks, mean age: 21 weeks, standard deviation: 12 weeks, 11 female). A group of 10 adults who reported normal hearing (age range: 23 – 64 years, mean age: 37.3 years, standard deviation: 15.8 years) was also tested for comparison.  2.2 Stimuli  Stimuli were gaps in broadband noise produced using code written in Matlab and C software languages, and presented using the software Presentation. Continuous broadband noise was presented with gaps occurring every 1000 ms. Gap durations were 0, 10, 20, 50 and 100 ms. ISI varied as a function of gap duration, and stimulus onset asynchrony was constant. Sound files lasting 30 seconds (i.e. containing 30 gaps) were generated, and repeated in a continuous loop. The experimenter controlled when the next condition could begin by button press. The last gap in each loop was omitted from analysis, as it could not be guaranteed that the next loop would begin after the correct gap duration. The final 30 ms of the sound file for the 0 ms condition was faded out, so that if the next loop began early, the sum of the two sound files would not cause a fatal error in Presentation. Cue markers to trigger the recording computer marked the timing of both gap onset and offset.    15  Stimuli were presented at 60 dB SPL using an EAR 3A (50 Ohms) insert earphone in the right ear, with two exceptions. One infant passed screening in the left ear only, and another infant was tested in the left ear for the 50 and 100 ms conditions due to experimenter error. The right ear was chosen due to the unexplored possibility of a right ear advantage (Berlin, Hughes, Lowe-Bell, & Berlin, 1973). Stimulus intensity was calibrated in dB SPL using a Larsen Davis Systems 824 sound level meter and a GRAS RA0113 2 cc coupler; gap duration and trigger timing was verified in milliseconds using a Bitscope BS120 oscilloscope. 2.3 EEG Recording 	 Electrodes were applied at CZ, C3, C4, M2, and forehead. The left mastoid (M1) was the reference electrode, and the forehead was the ground electrode. Electrode application was adjusted if impedance was measured greater than 3 kOhms. Continuous EEG was recorded using Scan 4.5 at a sampling rate of 1000 Hz, and online filtered between 0.1 and 30 Hz. For each condition, two replications were obtained, and stimuli were presented continuously until 150 trials were completed in each replication.  The continuous data were offline filtered through a bandpass filter between 1 and 15 Hz with a slope of 24 dB/octave. Each trial was epoched from a time period 200 ms prior to gap onset to 600 ms following gap onset. Trials were baseline corrected using the prestimulus interval (200 ms before gap onset) before averaging was conducted. Recordings from trials containing large artefacts (+/- 75 microVolts (µV) within the interval from 200 ms before gap onset and 400 ms after gap onset) were excluded from analysis. A condition was only included for analysis if a minimum of 150 trials total were accepted between the two replications. An averaged waveform was generated from a minimum of two replications. Each replication contained at least 75 accepted trials.   16 2.4 Procedure  Testing was conducted in a sound attenuated booth. Each gap duration condition was run twice, with 150 trials per condition each time, for a total of 300 trials in each of the four conditions. Order of conditions was randomized, however the 10 ms condition was only included after all other conditions were complete, if the infant was not too fussy and the parent/guardian agreed to continue. During testing, the infant was seated on the caregiver’s lap, and entertained by a video and a research assistant who distracted the child with toys to maintain the child’s attention. The infant’s behaviour was monitored throughout the experiment, and the experiment was paused if the infant became fussy or vocalized excessively. The experiment continued after a break or attending to the infant’s needs. The adult participants sat in a comfortable chair and watched a silent video.  A complete data set (0 ms control condition; 20 ms, 50 ms and 100 ms experimental conditions) was collected for 10 adults and 16 infants. One infant was tested only for the 0 and 20 ms conditions; two infants contributed data only for the 0, 50 and 100 ms conditions, because the 20 ms condition was excluded due to either insufficient accepted trials or excessive movement during testing. Of the 16 infants who provided a complete data set, five infants also provided data for the 10 ms condition. 2.5 Analysis  For the purposes of judging the presence or absence of responses, figures for each individual infant were created and contained the following: the averaged waveform for each replication, an averaged waveform of both, the standard deviation ratio (SDR), p-value and t-test data. These figures will be referred to as rating images. A sample is shown in Figure	2.1. The SDR, that is, the ratio of the standard deviation of the average to the standard deviation of a plus-  17 minus reference, was calculated for the period between 0 and 400 ms after gap onset and provided for each infant/condition. A p-value was obtained using the F-distribution; the degrees of freedom were calculated as follows: 2 ∗ 14	&'	 ()*+,-+.ℎ	01	1-2.34 ∗ 	0.4	7	 +84).-0* = 	11.2 A one sample t-test was also performed on each sample in the average, and samples that were significant (p<0.05) were marked with a black bar above the waveforms. Figure 2.1. Sample rating image from infant I05 in the 50 ms condition.   Three researchers experienced in interpreting auditory evoked potentials judged the figures for each condition for response presence/absence based on a visual inspection of the waveform, SDRs and t-tests provided. The raters included the experimenter (Rater 1) and two experienced researchers (Raters 2 and 3) at the same institution. They were instructed to make a Baby	541		4					SDR:	1.273	P-value:	0.341				  18 judgment on whether or not CAEPs to a gap were present, absent or could not be evaluated, as well as whether replicability was good, medium or poor. The result was a seven-point scale as follows: 1 - Absent, with good replicability; 2 – Absent, with medium replicability; 3 - Absent, with poor replicability; 4 - Could Not Evaluate; 5 - Present, with poor replicability; 6 - Present, with medium replicability; 7 - Present, with good replicability. The raters were given all rating images for a given infant in a single bundle, but blinded to condition, and not given any other information about the infant. Because rater 1 was directly involved in data collection and processing, the rating images were coded to ensure the raters were blind to which infant’s data was being rated.  Ratings were completed in two rounds. In the first round, all raters judged all waveforms from the Cz electrode. By experimenter error, the data for infant I01 contained the data from C3, instead of Cz. In the second round, waveforms from Cz, C3 and C4 were randomly distributed in bundles of one infant, one channel to each of the raters, so that each rater judged one third of the waveforms. The second round of ratings were done to validate the results from the first round, and to explore any hemispheric differences in the responses.  For analysis purposes, a “present” interpretation was assigned when at least two raters selected “present” with “medium” or “good” replicability. Similarly, an “absent” rating was assigned when at least two raters selected “absent” with “medium” or “good” reliability. All other waveforms were considered ambiguous. The 10 ms condition was excluded from most of the analyses due to the small sample size; only five infants participated in this condition.  Latencies and amplitudes were calculated for all “present” interpretations. For the infants, the positive peak (P) described was defined as the maximum point unless there was a broad plateau, in which case the amplitude and latency of the beginning of the plateau were noted.   19 Raters also participated informally in peak selection. For the adults, N1 was the lowest point between 50 and 150 ms after gap onset, and P2 was the highest point after this before approximately 250 ms.   20 Chapter 3: Results 3.1 Presence and Absence of Responses All adults had “present” interpretations to 20 and 50 ms gaps, and 9/10 adults had “present” interpretations to 100 ms gaps. Nine adult waveforms were interpreted as “absent” in the 0 ms condition. One adult in the 100 ms condition and one adult in the 0 ms condition had interpretations of “ambiguous.” As shown in Table 3-1, 35/39 of the unambiguous waveforms for the infants in the experimental conditions (20, 50 and 100 ms gaps) were interpreted as “present.” In the control condition (0 ms gaps), 1/19 was interpreted as “present” (i.e.; one false positive judgment). Waveforms were interpreted as “ambiguous” for 21/72 of the infant averaged waveforms. A more detailed breakdown of individual data is shown in Appendix A. Table 3-1. Total number of infants included in each control and experimental condition, and number of infants with “present,” “absent” or “ambiguous” waveform interpretations. The brackets show the ratings from raters 1, 2 and 3 respectively.  Condition [ms] N “Present” “Absent” “Ambiguous” 0 19 1 (1, 1, 2) 11 (12, 10, 9) 7 (6, 8, 6) 20 17 11 (8, 12, 11) 2 (2, 2, 2) 4 (7, 3, 4) 50 18 11 (10, 11, 11) 1 (1, 2, 1) 6 (7, 5, 6) 100 18 13 (9, 15, 13) 1 (2, 2, 0) 4 (7, 1, 5) Note: For the 10 ms condition, two infants had “present” interpretations and three had “ambiguous” interpretations.  3.2 Waveform Morphology Figure 3.1 shows the grand mean average and individual waveforms for infants and adults with “present” interpretations across conditions, and all responses in the control condition. For both infants and adults, there were no discernable responses for the control conditions in the   21 grand mean. The infant responses in the experimental conditions consisted of a single peak at approximately 200 ms. The peak was more plateau-shaped in the grand mean waveform for the 100 ms condition. Of the infant responses defined as “present,” two individual responses to 50 ms gaps and five individual responses to 100 ms gaps were plateau-shaped.     22 Figure 3.1. Grand mean average (thick) and individual (thin) waveforms for infants and adults with “present” interpretations across all conditions. The vertical line represents gap onset. One infant waveform was excluded from the grand mean for the 0 ms condition, because the file was corrupted. 100 ms2 μV05010020GapDuration[ms]Infants Adults  23 The adult responses were more complex, consisting in most cases of N1 at approximately 100 ms and P2 at approximately 190 ms. Three adults had a double-peaked N1 to 50 ms gaps, and two adults had a double-peaked N1 to 100 ms gaps. One adult had a plateau-shaped P2 to 50 ms gaps.  As shown in Table 3-2, there were no noticeable differences in amplitude or latency between conditions in either adults or infants. As shown in Table 3-3, the SDR values were highest for present responses, and lowest for absent responses, and this effect was magnified in the adult results. The t-test bars did not seem to assist the raters in making judgments. They were frequently absent in responses judged present, and present outside of the area of interest.   24 Table 3-2. Mean (1SD) amplitudes and latencies of infant and adult CAEPs for each experimental condition. Condition [ms] Infants Adults P N1 P2 N1-P2 Peak-to-peak Amplitude [µV] Latency [ms] Amplitude [µV] Latency [ms] Amplitude [µV] Latency [ms] Amplitude [µV] 20 3.091 (0.906) 200 (63) -3.400 (1.402) 109 (13) 2.991 (1.871) 192 (25) 6.392 (2.414) 50 3.841 (1.600) 174 (25) -2.937 (1.323) 91 (9) 3.045 (1.827) 199 (26) 5.982 (2.582) 100 3.524 (1.721) 200 (48) -2.897 (1.805) 102 (33) 1.739 (1.559) 169 (22) 4.172 (2.398)  Table 3-3. Mean (1SD) SDR values for each control and experimental condition by waveform interpretation. Blank cells indicate that no data fit that category. Condition [ms] Infants Adults Present Absent Ambiguous Present Absent Ambiguous 0 2.959 0.824 (0.410) 1.222 (0.251)  1.299 (0.538) 1.660 20 1.512 (0.380) 1.115 (1.092) 1.203 (0.673) 3.423 (2.088)   50 1.813 (0.504) 1.508 1.503 (0.601) 3.314 (1.565)   100 1.816 (0.856) 0.755 1.582 (0.667) 2.415 (0.914)  1.360    25 3.3 Threshold Estimation All adults had responses to 20 ms gaps, and could therefore be said to have a gap detection threshold better than 20 ms. As shown in Table 3-4, the majority of infants (15 infants) had waveform interpretations of “present” at the smallest gap size for which unambiguous interpretations could be obtained. Interpretations of absent to 20 ms gaps in two infants suggested a threshold between 20 and 50 ms (I11) or between 20 and 100 ms (I08). One infant (I16) had no present interpretations, and one infant (I09) had an inconsistent pattern of interpretations including an interpretation of present in the 0 ms condition.   26 Table 3-4. Summary of waveform interpretations for each individual infant across conditions and estimated gap detection threshold (GDT). “P” refers to “present,” “A” refers to “absent,” and “?” refers to “ambiguous.” Table is in order of youngest to oldest infant. Blank cells indicate that no data was collected or analyzed for that infant/condition. Infant Age [weeks] 0 ms 10 ms 20 ms 50 ms 100 ms GDT I01 9 A   P P <50 I11 10 A  A P ? 20<T<50 I19 10 A P P P P <10 I05 11 A  P P ? <20 I13 11 A  P P P <20 I06 14 ?  P ? P <20 I02 15 ?  P ? ? <20 I14 16 ? P ? P P <10 I20 17 A  P P ? <20 I22 17 A  P P P <20 I18 20 A  ? ? P <100 I07 22 A  P P P <20 I04 24 ?  P P P <20 I10 26 ? ? ? ? P <100 I16 26 A ? ? ? A ? I21 28 A  P   <20 I08 29 ? ? A ? P 20<T<100 I03 47 ?   P P <50 I09 50 P  P A P ? 3.4 Inter-Rater Reliability As shown in Figure 3.2, the ratings for the infants were more variable than for the adults, but there was still a general trend to rate the control condition as absent and the experimental conditions as present. As shown in Figure 3.3, there were no systematic differences between the raters, however, some interrater variability in interpretations was noted. Compared to raters 2 and 3, rater 1 judged fewer waveforms as “present” (29 vs. 39-41) and more waveforms as   27 “ambiguous” (30 vs. 20 or 26) and rater 3 judged the fewest waveforms as “absent” compared to raters 1 and 2 (12 vs. 16-18). Figure 3.2. Cumulative ratings of waveforms for each condition for infants (top) and adults (bottom). Each rating figure received three ratings, one from each rater as described in Section 2.5. Ratings of 1, 2 or 3 (the purple, blue, and turquoise columns) represent judgments of “absent”; ratings of 5, 6 or 7 (the yellow, orange, and red columns) represent judgments of “present”, and ratings of 4 (the green column in the middle) represent “could not evaluate” judgments. 114 2 2202 2 21263 4836 4259539 817122 24 200510152025Cumulative RatingInfants1 2 3 4 5 6 7260 0 12 0 0 01 0 0 01 0 0 10 0 0 00 0 03030 30250510152025300 20 50 100Gap Condition  [ms]Adults  28 Figure 3.3. Cumulative ratings for rating figures for each individual rater. Each bar represents the number of times a rater gave a rating figure a given rating. Colour scheme is the same as in . Figure 3.2.  Rater	1 Rater	2 Rater	3Absent	-	good	replicability13 4 2Absent	-	medium	replicability5 12 10Absent	-	poor	replicability13 6 8Could	not	evaluate 9 13 4Present	-	poor	replicability 8 1 14Present	-	medium	replicability9 11 20Present	-	good	replicability20 30 19134 2512 10136 8913481149 1120203019051015202530Rater	1 Rater	2 Rater	3Cumulative	Rating	by	Rater  29 Chapter 4: Discussion  The present study sought to record CAEPs to gaps in infants to determine whether this would be a feasible method of determining ANSD severity in pre-lingual infants. CAEPs to gaps in broadband noise in infants were successfully measured in normal-hearing infants. Moreover, responses were elicited to 20 ms gaps (and even 10 ms gaps in two infants), even though it was expected that a minimum gap of 50 ms would be needed to elicit a response. These data suggest that infants can detect shorter gaps in noise at the cortical level of processing than previous behavioural results would suggest (Werner et al., 1992).  There are a few possible explanations for the difference between the present CAEP results and the previous behavioural results. First of all, it is possible that the infants did not actually perceive the smaller gaps. As reported by Werner and colleagues, electrophysiological thresholds do not always match psychophysical results (Werner et al., 2001). Werner and colleagues suggested that the infant brain was able to detect smaller gaps (less than 5 ms) at the brainstem level, but that cortical immaturity did not support actual perception of these small gaps. Thus, while the infant brain appears to be able to detect smaller gaps than expected at the cortical level, more processing may be needed to produce a behavioural response to these smaller gaps. As gap detection thresholds do not match behaviourally and for CAEPs, it will likely be necessary for future research to investigate the relationship between the two threshold types. It may also be that the behavioural methods used in previous studies are not sensitive to what infants can perceive, and other methods are necessary to approach this problem.  Smaller gap durations were predicted to elicit smaller CAEPs, but this was not the case. This prediction was based on other threshold-based findings, where responses closer to threshold become smaller in amplitude (Cone & Whitaker, 2013; Purdy, Sharma, Munro, & Morgan,   30 2013). These studies found lower amplitudes for CAEPs to stimuli with lower intensities. Purdy and colleagues found that the amplitudes of the response increase with increasing intensity until they reached a plateau at a supra-threshold level. It may be that the current study did not find an effect of gap duration on amplitude because the gap durations used were all suprathreshold, and smaller gap durations would find smaller amplitudes. It may also be that the amplitude of the response is not sensitive to gap duration as it is to intensity. In contrast to the relationship of amplitude to gap duration, there were changes in morphology for at least some infants for the longer gap durations. This finding may be analogous to the findings in adults, which have shown that a double-peaked N1 is sometimes present (Atcherson et al., 2009). In the current study, 6/20 adult responses for the two larger gap conditions (50 and 100 ms) had a varied morphology. It is not clear why adults show a double-peaked N1 component, or why it is elicited in some but not other adults. In infants, this morphological difference warrants further study to see whether it really is analogous to the N-process described by Pratt and colleagues (Pratt et al., 2005). This morphological difference was detected in a similar proportion of infants to adults (7/26 unambiguous responses in the 50 and 200 ms conditions), but was more common in the 100 ms condition.  Aside from these exceptions, the morphology of the responses detected was consistent with previous CAEPs measured in infants, i.e. a single positive peak (e.g. Cone & Whitaker, 2013; Golding, Purdy, Sharma, & Dillon, 2006; Purdy et al., 2013; Wunderlich, Cone-Wesson, & Shepherd, 2006). Mean latencies of the responses in previous results vary enormously from less than 100 ms (Wunderlich et al., 2006) to over 300 ms (Cone & Whitaker, 2013). The current results fall somewhere in the middle of this range at around 200 ms. Amplitudes in the current results were generally smaller than previous results (Golding et al., 2006; Purdy et al., 2013). In   31 the present study, the period between stimulus presentations was not silent as it is in most studies, but rather, there was continuous noise, and the stimulus itself was silence. It is possible that the amplitude of the response measured here was reduced because the neurons were constantly responding to the noise, and therefore less able to respond to gap onset. Amplitudes of the adult responses in the present results were also smaller than previous results (Michalewski et al., 2005), which suggests that the stimulus and recording parameters such as ISI or stimulus intensity could be optimized to elicit larger amplitudes. Larger amplitudes would also increase the probability of detecting responses in noisy data, thus decreasing the number of ambiguous results.  Nearly one third of the results were judged ambiguous. To minimize the number of ambiguous results, either the amplitude of the response must be increased, or the noise must be reduced. Some simple changes in the way the data is presented, such as the use of jitter instead of a regular stimulus presentation or an increase in the number of trials presented, may lead to larger amplitudes. Furthermore, changes to the way the data is processed, such as the filter settings, the dimensions of the image, or the information available to the rater may improve the rater’s ability to judge the presence or absence of the response.  One significant way to increase the amplitude of the response is increasing the ISI. In the current study, a relatively short ISI of 900 – 1000 ms was used in order to minimize study duration. Because infants have a limited attention span, testing time must be as brief as possible. Longer ISIs require more time in the sound booth to collect the same number of epochs. However, longer ISIs may increase the amplitude of the response to a degree that outweighs the convenience of shorter booth time. Golding and colleagues compared responses to ISIs of 750, 1125 and 1500 ms, and found effects for one stimulus but not the other (Golding et al., 2006).   32 Specifically, they found an increase in amplitude with an increase in ISI for a /t/ stimulus, but not for a /m/ stimulus. That is, the effect of ISI on the response appears to be stimulus-specific and it is worth investigating the effects of ISI on infant CAEPs to gaps specifically. Furthermore, in older children, an increase in ISI is known to change the morphology of the response to a more adult-like morphology (Gilley et al., 2005). This suggests that the generators of the additional peaks have longer refractory periods in young children than adults. Longer ISIs may enable the recording of more complex, adult-like responses and/or larger amplitudes in infants compared to the current study.  It may also be possible to reduce the noise in future data. Measuring CAEPs in awake infants is fundamentally challenging due to the high myogenic noise associated with movement. Vocalization during gaps likely also contributes to some ambiguous responses, as it produces noise, and also potentially masks the perception of the stimulus. However, it may be possible to measure CAEPs in sleeping infants. Early studies of infant CAEPs reported responses to tone bursts in healthy sleeping newborns (Taguchi, Picton, Orpin, & Goodman, 1969), so the assumption that cortical AEPs should only be measured in awake infants may be an assumption from adult data that has been unquestioningly passed on to infant research. If so, this assumption is certainly worth questioning. Future studies should investigate differences between responses in awake and asleep infants using the same stimulus, and if possible, within the same infant.  It is also possible that as we become more familiar with infant CAEPs to gaps, including morphological differences for different gaps, expertise in identifying responses will improve. While two of the raters in this project were very experienced at identifying CAEPs, they had not recorded an infant CAEP to a gap prior to this study. Experience has been shown to improve reliability of judgments in infant ABRs (Gans, Zotto, & Gans, 1992; Zaitoun, Cumming, &   33 Purcell, 2014; Zaitoun, Cumming, Purcell, & O’Brien, 2016), and this likely extends to corticals (Carter, Golding, Dillon, & Seymour, 2010). As more data are collected, better guidelines can be developed to identify CAEPs specifically to gaps.  In the present study, raters were provided with some statistical information to aid in the decision making process. However, this information appeared to provide limited assistance in the interpretation of the waveforms. While the SDR values were higher for present responses, there was significant overlap in the distribution of values (i.e. large standard deviations). Furthermore, ratings of “present” were frequently found in the absence of significant t-test results. Other researchers have attempted to address this problem as well. For example, Cone & Whitaker (2013) evaluated the potential of Schimmel’s plus-minus average (an estimate of signal-to-noise ratio) to judge response presence, but found, not too surprisingly, that this method missed a substantial number of responses that were deemed present with the replication approach. Some automatic detection algorithms have been used to identify CAEP responses in infants, and preliminary data suggest that the Hotelling’s T2 statistic may be as accurate as an experienced examiner (Carter et al., 2010). However, these algorithms cannot be applied to a new stimulus without detailed stimulus-specific investigation. As these methods potentially improve the clinical utility of CAEPs, this investigation will be worthwhile.  The wide age range of infants tested, nearly ten months, was another limitation of the present study with respect to understanding the effect of gap duration on the presence, amplitude, latency and morphology of the infant CAEP. The latency, amplitude and even morphology of the response is known to develop over this time (Barnet et al., 1975; Kurtzberg, Hilpert, Kreuzer, & Vaughan, 2008; Ohlrich & Barnet, 1972; Wunderlich et al., 2006). There did not appear to be any age-related trends in response presence, morphology, amplitude or latency, however, there   34 were not enough infants in any smaller age range to investigate this systematically. Therefore, future investigations should either focus on a smaller age range, particularly those likely to be tested in an ANSD protocol (i.e. three to six months), or more thoroughly investigate the development of the response for a range of gap durations over the first year. Ideally, future testing of age-related changes would investigate both shorter gaps than those tested here (i.e. less than 20 ms) and repeat the same gap durations here for a larger group of infants in systematic age ranges. Furthermore, future testing involving infants with ANSD could use age-matched controls.  Finally, the present study investigated the infant CAEP to a gap presented to the right ear only in most infants. This means that the results may not be generalizable to stimuli presented to the left ear or bilaterally, especially given the potential effects of a right ear advantage (Berlin et al., 1973). Future studies must investigate the potential differences so that clinical results can be reliably interpreted in cases where the stimuli are presented in other configurations, such as bilaterally via soundfield or headphones, or to the left ear.  While the present study established that it is possible to measure CAEPs to gaps in infants, one major question that must first be answered before the method can be used clinically is whether CAEPs to gaps in infancy are predictive of speech perception ability in early childhood for children with ANSD. A longitudinal study which compares CAEP gap thresholds when the infant is diagnosed to behavioural gap thresholds at different time points and early childhood language outcomes would be ideal. Intervention decisions, such as amplification or cochlear implantation, made between diagnosis and the end of the study would be an important consideration in interpreting the results.   35  In conclusion, this study has demonstrated that it is feasible to measure CAEPs to gaps in broadband noise in infants and adults with normal hearing. Responses were measured to shorter gap durations than expected, suggesting that infant temporal resolution may be better than previously thought. Further research will need to explore ways to develop this approach into a useful clinical tool for infants with ANSD. Hopefully, this body of research will improve our ability to make intervention decisions earlier on for these children, and optimize their opportunity for speech and language development.    36 Bibliography Atcherson, S. R., Gould, H. J., Mendel, M. I., & Ethington, C. A. (2009). Auditory N1 Component to Gaps in Continuous Narrowband Noises: Ear and Hearing, 30(6), 687–695. Barnet, A. B., Ohlrich, E. S., Weiss, I. P., & Shanks, B. (1975). Auditory evoked potentials during sleep in normal children from ten days to three years of age. Electroencephalography and Clinical Neurophysiology, 39(1), 29–41. Berlin, C. I., Hood, L. J., Morlet, T., Wilensky, D., Li, L., Mattingly, K. R., … Frisch, S. A. (2010). Multi-site diagnosis and management of 260 patients with Auditory Neuropathy/Dys-synchrony (Auditory Neuropathy Spectrum Disorder*). International Journal of Audiology, 49(1), 30–43. Berlin, C. I., Hood, L. J., & Rose, K. (2001, December). On renaming Auditory Neuropathy as Auditory Dys-synchrony. Audiology Today, 13(6), 15–17. Berlin, C. I., Hood, L. J., & Rose, K. (2002, June). Auditory Neuropathy or Auditory Dys-synchrony. Audiology Today, 14(3), 38–39. Berlin, C. I., Hughes, L. F., Lowe-Bell, S. S., & Berlin, H. L. (1973). Dichotic right ear advantage in children 5 to 13. Cortex; a Journal Devoted to the Study of the Nervous System and Behavior, 9(4), 394–402. Carter, L., Golding, M., Dillon, H., & Seymour, J. (2010). The Detection of Infant Cortical Auditory Evoked Potentials (CAEPs) Using Statistical and Visual Detection Techniques. Journal of the American Academy of Audiology, 21(5), 347–356.   37 Cone, B., & Whitaker, R. (2013). Dynamics of infant cortical auditory evoked potentials (CAEPs) for tone and speech tokens. International Journal of Pediatric Otorhinolaryngology, 77(7), 1162–1173. Coplan, J. (1987). Deafness: ever heard of it? Delayed recognition of permanent hearing loss. Pediatrics, 79(2), 206–213. Davis, A., & Wood, S. (1992). The epidemiology of childhood hearing impairment: Factors relevant to planning of services. British Journal of Audiology, 26(2), 77–90. Davis, P. A. (1939). EFFECTS OF ACOUSTIC STIMULI ON THE WAKING HUMAN BRAIN. Journal of Neurophysiology, 2(6), 494. Fitzgibbons, P. J. (1983). Temporal gap detection in noise as a function of frequency, bandwidth, and level. The Journal of the Acoustical Society of America, 74(1), 67. Gans, D., Zotto, D. D., & Gans, K. D. (1992). Bias in scoring auditory brainstem responses. British Journal of Audiology, 26(6), 363–368. Gilley, P. M., Sharma, A., Dorman, M., & Martin, K. (2005). Developmental changes in refractoriness of the cortical auditory evoked potential. Clinical Neurophysiology, 116(3), 648–657. Golding, M., Purdy, S., Sharma, M., & Dillon, H. (2006). The Effect of Stimulus Duration and Inter-Stimulus Interval on Cortical Responses in Infants. Australian and New Zealand Journal of Audiology, 28(2), 122–136.   38 Harrison, R. V., Gordon, K. A., Papsin, B. C., Negandhi, J., & James, A. L. (2015). Auditory neuropathy spectrum disorder (ANSD) and cochlear implantation. International Journal of Pediatric Otorhinolaryngology, 79(12), 1980–1987. Hayes, D., & Sininger, Y. (Eds.). (2008). Guidelines for Identification and Management of Infants and Young Children with Auditory Neuropathy Spectrum Disorder. Aurora, CO: The Children’s Hospital. Retrieved from He, S., Grose, J. H., Teagle, H. F. B., Woodard, J., Park, L. R., Hatch, D. R., … Buchman, C. A. (2015). Acoustically Evoked Auditory Change Complex in Children With Auditory Neuropathy Spectrum Disorder: A Potential Objective Tool for Identifying Cochlear Implant Candidates. Ear and Hearing, 36(3), 289–301. Huttenlocher, P. R., & Dabholkar, A. S. (1997). Regional differences in synaptogenesis in human cerebral cortex. The Journal of Comparative Neurology, 387(2), 167–178.<167::AID-CNE1>3.0.CO;2-Z Irwin, R. J., Ball, A. K. R., Kay, N., Stillman, J. A., & Rosser, J. (1985). The Development of Auditory Temporal Acuity in Children. Child Development, 56(3), 614. Joint Committee on Infant Hearing. (2007). Year 2007 Position Statement: Principles and Guidelines for Early Hearing Detection and Intervention Programs. PEDIATRICS, 120(4), 898–921.   39 Kurtzberg, D., Hilpert, P. L., Kreuzer, J. A., & Vaughan, H. G. (2008). DIFFERENTIAL MATURATION OF CORTICAL AUDITORY EVOKED POTENTIALS TO SPEECH SOUNDS IN NORMAL FULLTERM AND VERY LOW-BIRTHWEIGHT INFANTS. Developmental Medicine & Child Neurology, 26(4), 466–475. Lister, J. J., Roberts, R. A., & Lister, F. L. (2011). An adaptive clinical test of temporal resolution: Age effects. International Journal of Audiology, 50(6), 367–374. Marsh, R. R. (2002, June). IS IT AUDITORY DYS-SYNCHRONY? Comment on “On renaming Auditory Neuropathy as Auditory Dys-synchrony.” Audiology Today, 14(3), 36–37. Martin, B. A., & Boothroyd, A. (1999). Cortical, auditory, event-related potentials in response to periodic and aperiodic stimuli with the same spectral envelope. Ear and Hearing, 20(1), 33–44. 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, 116(3), 669–680. Näätänen, R., Gaillard, A. W. K., & Mäntysalo, S. (1978). Early selective-attention effect on evoked potential reinterpreted. Acta Psychologica, 42(4), 313–329.   40 Ohlrich, E. S., & Barnet, A. B. (1972). Auditory evoked responses during the first year of life. Electroencephalography and Clinical Neurophysiology, 32(2), 161–169. Phillips, D. P. (1999). Auditory gap detection, perceptual channels, and temporal resolution in speech perception. Journal of the American Academy of Audiology, 10(6), 343–354. Picton, T. W., Alain, C., Woods, D. L., John, M. S., Scherg, M., Valdes-Sosa, P., … Trujillo, N. J. (1999). Intracerebral sources of human auditory-evoked potentials. Audiology & Neuro-Otology, 4(2), 64–79. Plomp, R. (1964). Rate of Decay of Auditory Sensation. The Journal of the Acoustical Society of America, 36(2), 277. Ponton, C. W., Eggermont, J. J., Kwong, B., & Don, M. (2000). Maturation of human central auditory system activity: evidence from multi-channel evoked potentials. Clinical Neurophysiology, 111(2), 220–236. Pratt, H., Bleich, N., & Mittelman, N. (2005). The composite N component to gaps in noise. Clinical Neurophysiology, 116(11), 2648–2663. Purdy, S. C., Sharma, M., Munro, K. J., & Morgan, C. L. A. (2013). Stimulus level effects on speech-evoked obligatory cortical auditory evoked potentials in infants with normal hearing. Clinical Neurophysiology, 124(3), 474–480. Rance, G., Beer, D. E., Cone-Wesson, B., Shepherd, R. K., Dowell, R. C., King, A. M., … Clark, G. M. (1999). Clinical Findings for a Group of Infants and Young Children with   41 Auditory Neuropathy: Ear and Hearing, 20(3), 238. Rance, G., Cone-Wesson, B., Wunderlich, J., & Dowell, R. (2002). Speech perception and cortical event related potentials in children with auditory neuropathy. Ear and Hearing, 23(3), 239–253. Rance, G., Fava, R., Baldock, H., Chong, A., Barker, E., Corben, L., & Delatycki, M. B. (2008). Speech perception ability in individuals with Friedreich ataxia. Brain, 131(8), 2002–2012. Rance, G., McKay, C., & Grayden, D. (2004). Perceptual Characterization of Children with Auditory Neuropathy: Ear and Hearing, 25(1), 34–46. Rapin, I., & Gravel, J. S. (2006). Auditory Neuropathy: A Biologically Inappropriate Label Unless Acoustic Nerve Involvement Is Documented. Journal of the American Academy of Audiology, 17(2), 147–150. Rea, P. A., & Gibson, W. P. R. (2003). Evidence for surviving outer hair cell function in congenitally deaf ears. The Laryngoscope, 113(11), 2030–2034. Schneider, B. A., & Hamstra, S. J. (1999). Gap detection thresholds as a function of tonal duration for younger and older listeners. The Journal of the Acoustical Society of America, 106(1), 371. Shafer, V. L., Yu, Y. H., & Wagner, M. (2015). Maturation of cortical auditory evoked potentials (CAEPs) to speech recorded from frontocentral and temporal sites: Three months to eight   42 years of age. International Journal of Psychophysiology, 95(2), 77–93. Starr, A., Picton, T. W., Sininger, Y., Hood, L. J., & Berlin, C. I. (1996). Auditory neuropathy. Brain, 119(3), 741–753. Starr, A., & Rance, G. (2015). Auditory neuropathy. In Handbook of Clinical Neurology (Vol. 129, pp. 495–508). Elsevier. Retrieved from Taguchi, K., Picton, T. W., Orpin, J. A., & Goodman, W. S. (1969). Evoked response audiometry in newborn infants. Acta Oto-Laryngologica. Supplementum, 252, 5–17. Trainor, L. J., McFadden, M., Hodgson, L., Darragh, L., Barlow, J., Matsos, L., & Sonnadara, R. (2003). Changes in auditory cortex and the development of mismatch negativity between 2 and 6 months of age. International Journal of Psychophysiology, 51(1), 5–15. 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 and Hearing Research, 39(6), 1315–1320. Trehub, S. E., Schneider, B. A., & Henderson, J. L. (1995). Gap detection in infants, children, and adults. The Journal of the Acoustical Society of America, 98(5), 2532. Van Naarden, K., Decoufle, P., & Caldwell, K. (1999, March). Prevalence and Characteristics of Children With Serious Hearing Impairment in Metropolitan Atlanta, 1991-1993. Pediatrics, 103(3), 570.   43 Werner, L. A., Folsom, R. C., Mancl, L. R., & Syapin, C. L. (2001). Human Auditory Brainstem Response to Temporal Gaps in Noise. Journal of Speech, Language, and Hearing Research, 44(4), 737–750. Werner, L. A., Marean, G. C., Halpin, C. F., Spetner, N. B., & Gillenwater, J. M. (1992). Infant Auditory Temporal Acuity: Gap Detection. Child Development, 63(2), 260–272. Wunderlich, J. L., & Cone-Wesson, B. K. (2006). Maturation of CAEP in infants and children: A review. Hearing Research, 212(1-2), 212–223. Wunderlich, J. L., Cone-Wesson, B. K., & Shepherd, R. (2006). Maturation of the cortical auditory evoked potential in infants and young children. Hearing Research, 212(1-2), 185–202. Yoshinaga-Itano, C., Sedey, A. L., Coulter, D. K., & Mehl, A. L. (1998). Language of Early- and Later-identified Children With Hearing Loss. Pediatrics, 102(5), 1161–1171. Zaitoun, M., Cumming, S., & Purcell, A. (2014). Review: Inter and intra-reader agreement among audiologists in reading auditory brainstem response waves. Can J Speech Lang Pathol Audiol, 38, 440–9. Zaitoun, M., Cumming, S., Purcell, A., & O’Brien, K. (2016). Inter and intra-reader variability in the threshold estimation of auditory brainstem response (ABR) results. Hearing, Balance and Communication, 14(1), 59–63. Zeng, F.-G., Kong, Y.-Y., Michalewski, H. J., & Starr, A. (2005). Perceptual Consequences of Disrupted Auditory Nerve Activity. Journal of Neurophysiology, 93(6), 3050–3063.   44 Zeng, F.-G., Oba, S., Garde, S., Sininger, Y., & Starr, A. (1999). Temporal and speech processing deficits in auditory neuropathy. NeuroReport, 10(16), 3429–3435.  45 Appendix A Table A.1. Breakdown of participant information and SDR values, p-values, number of trials, and ratings for the 0 ms control condition. Table is sorted in age order as in Table 3-4. Subject Age [weeks] Screen Gender SDR p-value Number of trials in T1 Number of trials in T2 Total number of trials Rater 1 Score Rater 2 Score Rater 3 Score I01 9 B M 0.358 0.956 112 114 226 1 1 1 I11 10 B F 0.299 0.977 136 137 273 1 2 3 I19 10 B F *Corrupt  148 147 295 1 2 3 I05 11 B F 0.987 0.509 122 113 235 1 2 2 I13 11 B F 0.739 0.696 137 135 272 2 2 3 I06 14 B2 M 1.222 0.367 87 129 216 4 4 6 I02 15 B M 1.312 0.323 144 135 279 3 4 2 I14 16 B M 1.53 0.236 131 121 252 3 4 3 I20 17 B F 0.7 0.727 121 121 242 2 2 2 I22 17 B F 1.4 0.285 117 111 228 2 3 2 I18 20 B M 1.319 0.32 150 139 289 1 2 2 I07 22 B F 0.838 0.617 114 85 199 3 2 2 I04 24 B F 1.432 0.272 107 109 216 3 2 3 I10 26 B F 0.915 0.56 148 132 280 1 6 5 I16 26 R M 1.244 0.356 122 163 285 1 4 2 I21 28 B M 0.355 0.957 116 116 232 1 1 2                                                 2 Although both ears passed screening, the left ear was used due to experimenter error in the 50 and 100 ms conditions for this infant.    46 Subject Age [weeks] Screen Gender SDR p-value Number of trials in T1 Number of trials in T2 Total number of trials Rater 1 Score Rater 2 Score Rater 3 Score I08 29 B F 0.856 0.604 120 124 244 2 4 5 I03 47 L M 1.287 0.334 95 97 192 3 4 3 I09 50 B F 2.959 0.036 89 103 192 7 4 6  Table A.2. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 10 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” Subject SDR p-value Number of trials in T1 Number of trials in T2 Total number of trials Rater 1 Score Rater 2 Score Rater 3 Score Amplitude Latency I19 0.804 644 118 131 249 5 6 6 2.931 0.21 I14 2.041 0.115 127 132 259 6 7 7 6.41 0.148 I10 2.731 0.047 137 137 274 4 3 4   I16 0.787 0.657 133 118 251 2 4 5 2.247 0.2 I08 2.122 0.104 92 80 172 4 3 4        47 Table A.3. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 20 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” Subject SDR p-value Number of trials in T1 Number of trials in T2 Total number of trials Rater 1 Score Rater 2 Score Rater 3 Score Amplitude Latency I11 1.887 0.143 140 138 278 1 7 1   I19 1.176 0.392 132 131 263 7 7 7 2.445 0.223 I05 1.472 0.257 135 183 318 6 7 7 2.25 0.168 I13 2.001 0.122 121 136 257 6 7 7 3.592 0.155 I06 1.803 0.16 142 121 263 5 7 7 3.353 0.166 I02 1.209 0.374 147 129 276 3 6 6 2.388 0.244 I14 0.762 0.677 121 130 251 3 3 5   I20 1.356 0.303 125 120 245 7 6 6 2.571 0.179 I22 1.402 0.284 140 118 258 7 7 7 2.729 0.19 I18 0.64 0.774 105 105 210 5 2 3   I07 1.662 0.196 96 136 232 7 7 6 2.439 0.155 I04 2.168 0.097 126 122 248 7 7 7 4.138 0.171 I10 2.12 0.104 145 129 274 3 4 4   I16 1.291 0.332 129 139 268 4 3 5   I21 1.197 0.38 112 121 233 7 7 7 2.922 0.183 I08 0.342 0.962 136 116 252 1 1 2   I09 1.074 0.452 95 85 180 5 6 6 5.171 0.371      48  Table A.4. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 50 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” Subject SDR p-value Number of  trials in T1 Number of  trials in T2 Total number  of trials Rater 1 Score Rater 2 Score Rater 3 Score Amplitude Latency I01 2.065 0.112 142 79 221 6 7 5 4.279 0.143 I11 1.544 0.232 142 133 275 6 6 5 2.144 0.194 I19 1.962 0.129 138 137 275 5 6 6 3.428 0.222 I05 1.273 0.341 147 154 301 7 7 7 5.137 0.177 I13 2.253 0.087 146 145 291 7 7 7 5.553 0.177 I06 2.266 0.085 116 98 214 3 4 3   I02 1.279 0.338 146 150 296 3 4 6   I14 3.018 0.034 146 134 280 7 7 7 2.833 0.175 I20 1.661 0.196 141 121 262 7 7 7 6.305 0.185 I22 1.798 0.162 113 86 199 7 7 7 5.275 0.1605 I18 1.29 0.333 91 101 192 5 5 6   I07 1.361 0.301 106 100 206 7 7 6 1.447 0.127 I04 1.623 0.207 114 131 245 7 7 7 2.152 0.174 I10 0.937 0.544 135 173 308 4 2 5   I16 2.248 0.087 122 119 241 4 4 5   I08 1 0.5 123 141 264 5 4 5   I03 1.385 0.291 112 113 225 7 7 7 3.698 0.18 I09 1.508  126 138 264 1 1 2        49 Table A.5. Breakdown of SDR values, p-values, number of trials, ratings, amplitudes and latencies for the 100 ms condition. Table is sorted in age order as in Table 3-4. Amplitude and latency cells are blank if the response was not interpreted as “present.” Subject SDR p-value Number of trials in T1 Number of trials in T2 Total number of trials Rater 1 Score Rater 2 Score Rater 3 Score Amplitude Latency I01 2.565 0.058 90 78 168 6 7 6 2.008 0.186 I11 1.5 0.247 138 136 274 4 6 4   I19 1.134 0.416 133 125 258 7 7 7 5.364 0.26 I05 2.547 0.059 128 165 293 3 6 5   I13 2.96 0.036 0 189 189 7 7 7 6.898 0.197 I06 0.61 0.798 125 135 260 4 6 6 ? ? I02 1.193 0.382 135 148 283 3 2 5   I14 3.407 0.022 131 139 270 7 7 7 4.842 0.177 I20 1.086 0.444 122 148 270 3 3 5   I22 1.601 0.213 110 130 240 6 6 6 3.588 0.132 I18 1.697 0.186 93 70 163 7 7 6 5.455 0.243 I07 1.508 0.244 108 116 224 7 7 7 3.131 0.142 I04 1.888 0.142 129 155 284 5 7 6 2.444 0.199 I10 0.795 0.652 127 126 253 1 7 6 1.441 0.143 I16 0.755 0.683 122 123 245 1 2 6   I08 1.343 0.309 123 141 264 4 7 6 2.563 0.257 I03 2.732 0.047 105 94 199 6 7 5 2.736 0.201 I01 1.367 0.298 89 92 181 6 7 6 1.829 0.267  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items