Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cochlear microphonics from auditory brainstem responses of infants with auditory neuropathy spectrum… Smith, Kyle 2018

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

Item Metadata

Download

Media
24-ubc_2018_november_smith_kyle.pdf [ 1.13MB ]
Metadata
JSON: 24-1.0372052.json
JSON-LD: 24-1.0372052-ld.json
RDF/XML (Pretty): 24-1.0372052-rdf.xml
RDF/JSON: 24-1.0372052-rdf.json
Turtle: 24-1.0372052-turtle.txt
N-Triples: 24-1.0372052-rdf-ntriples.txt
Original Record: 24-1.0372052-source.json
Full Text
24-1.0372052-fulltext.txt
Citation
24-1.0372052.ris

Full Text

Cochlear Microphonics from Auditory Brainstem Responses of Infants with Auditory Neuropathy Spectrum Disorder by  Kyle Smith  B.Sc., The University of Victoria, 2015  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)  September 2018  © Kyle Smith, 2018   ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  Cochlear Microphonics from Auditory Brainstem Responses of Infants with Auditory Neuropathy Spectrum Disorder  submitted by Kyle Smith in partial fulfillment of the requirements for the degree of Master of Science in Audiology and Speech Sciences  Examining Committee: Dr. Anthony Herdman, Associate Professor  Supervisor  Dr. Navid Shahnaz, Associate Professor and Clinical Audiologist Supervisory Committee Member  Rae Riddler, Clinical Pediatric Audiologist Supervisory Committee Member   iii  Abstract  Objectives: The cochlear microphonic (CM) is a bioelectric potential detectable through early latency auditory brainstem responses (ABR). It has been described as large in amplitude and long in duration in auditory neuropathy spectrum disorder (ANSD), a complex form of hearing loss. ANSD is identified through a combination of diagnostics, including but not limited to present otoacoustic emissions (OAEs) and/or a robust CM, and absent/abnormal ABR. Based on previous literature, this study hypothesized that CMs would differ significantly between infants with ANSD and normal-hearing infant data from well-baby, and neonatal intensive care unit (NICU) populations; it further hypothesized that CM/ABR wave V (CM/V) amplitude ratios would not differ significantly across ANSD groups presenting with and without OAEs.  Design: Retrospective analysis was performed on click-ABR recordings from 16 infants with ANSD (24 ears; mean 3.5 months, corrected) for comparison to published normative data. ANSD was identified by the presence of OAEs (OAE+) and absent/abnormal ABR, and by CM presence with abnormal/absent ABR and support from later behavioral results for infants with absent OAEs (OAE-). Waves were analyzed by comparing condensation and rarefaction polarities to highlight the CM. Proposed CM/V ratio values for ANSD identification were also explored.  Results:  Mean CM durations were significantly longer in combined ANSD ears (4.197 ± 1.154ms) than normally-hearing well-baby (0.73 ± 0.3ms) and NICU (0.82 ± 0.51ms) infants. CM amplitudes were significantly larger in ANSD (0.322 ± 0.173µV) than well-baby (0.24 ± 0.09µV), but not NICU (0.26 ± 0.13µV) infants. OAE+ and OAE- ANSD ears did not differ significantly in CM duration or amplitude but did differ significantly in mean CM/V ratios iv  (6.602 ± 2.987, 2.040 ± 1.112, respectively). Ratios correctly identified ANSD in 16 of 19 ANSD ears with a wave V.   Conclusions: Significant group differences in CM duration suggest that this measure could be useful for identification of ANSD in infants. CM amplitude was less definitive, with confounds between datasets. The CM/V ratios failed to correctly categorize all OAE- infants for which the measure would be most applicable. Results should be viewed with caution given the retrospective nature and limited sample size of the analysis.  v  Lay Summary  Early identification of hearing loss in infancy can allow for better child speech and language development. Quantifying infant hearing loss relies on predictably timed bioelectric responses to sound from the cochlea and brainstem. Typically, absent brainstem responses at increasing loudness indicates the degree of hearing loss; however, widespread testing has identified a type of loss with unexpectedly absent or abnormal brainstem responses, despite present cochlear responses, called ANSD. The mechanisms of ANSD are not clear, but may include poor neural activity within the cochlea, at the synaptic juncture to the auditory nerve, or along the auditory nerve itself. Poor neural activity may produce systematically different cochlear responses in ANSD than other types of hearing loss due to breakdown of feedback systems within the brainstem. This study compares data from infants with normal hearing to infants with ANSD to assess if cochlear bioelectric measures can reliably discriminate between groups.   vi  Preface  This thesis was carried out in collaboration with my supervisor Dr. Anthony Herdman, and my research colleague Alicia Parfett. Dr. Anthony Herdman conceptualized the study and advised on the project’s development, including study design and statistical measurement. He also coded the SmartEP Reader program in MatLabTM for waveform analysis.   Deidentified subject data was provided by the British Columbia Early Hearing Program (BCEHP) in association with the BC Children’s Hospital (BCCH). Jenny Hatton gathered patient data from the British Columbia Early Hearing Surveillance Tool (BEST) and created a master excel data sheet at BCCH. Paper charts were scanned in to an electronic version by Marlina Anderson at BCCH; deidentification of these electronic charts and retrieval of ABR waveform data was performed by Rae Riddler, also at BCCH. Data relevant to the current study, and to Alicia Parfett’s thesis was identified and categorized by Alicia Parfett and myself at the University of British Columbia (UBC). I envisioned the current study design, researched the presented causal theory of CM distortions, conducted all waveform measures and analysis of the results, and wrote the manuscript.   Ethics approval was initially provided by the UBC Children's & Women’s Research Ethics Board, certificate number H14-01969, and continued under H18-01605.  Startup funding was provided to Dr. Anthony Herdman through a UBC Faculty of Medicine internal funds grant. No conflicts of interest were declared. vii  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ vii List of Figures .................................................................................................................................x List of Abbreviations .................................................................................................................. xii Acknowledgements .................................................................................................................... xiv Chapter 1: Auditory neuropathy spectrum disorder (ANSD) ...................................................1 1.1 Mechanisms and auditory deficits of ANSD .................................................................. 1 1.2 Clinical presentation of ANSD ....................................................................................... 4 1.2.1 ANSD Diagnosis ......................................................................................................... 6 1.2.2 Suspected ANSD ........................................................................................................ 9 1.3 The cochlear microphonic ............................................................................................. 15 1.3.1 The cochlear microphonic in normally hearing infants ............................................ 19 1.3.2 The cochlear microphonic in ANSD ......................................................................... 20 1.3.3 Reported values of abnormal CM waveforms in ANSD .......................................... 23 1.4 Summary and justification ............................................................................................ 25 1.5 Hypotheses of the current investigation ........................................................................ 26 Chapter 2: Methods .....................................................................................................................27 2.1 Participants .................................................................................................................... 27 2.1.1 Exclusion criteria ...................................................................................................... 28 viii  2.1.2 Inclusion criteria ....................................................................................................... 28 2.2 Instrumentation and equipment..................................................................................... 31 2.3 Waveform analysis........................................................................................................ 33 2.3.1 CM duration measurement ........................................................................................ 35 2.3.2 Amplitude measurements and CM/V amplitude ratio calculation ............................ 36 2.4 Statistical analysis ......................................................................................................... 38 Chapter 3: Results........................................................................................................................41 3.1 CM amplitude, duration, and CM/V amplitude ratio findings...................................... 41 3.1.1 NICU and ANSD comparisons ................................................................................. 41 3.1.2 Well-baby and ANSD comparisons .......................................................................... 42 3.1.3 OAE+ and OAE- comparisons and CM/V amplitude ratio results ........................... 43 Chapter 4: Discussion ..................................................................................................................46 4.1 CM Duration in ANSD compared to normally-hearing infants.................................... 46 4.2 CM amplitude in ANSD compared to normally-hearing infants .................................. 50 4.3 CM duration, amplitude, and CM/V amplitude ratios in OAE+ and OAE- infants ..... 53 4.4 Strengths and limitations of the current study .............................................................. 58 4.5 Potential applications of the current findings ............................................................... 62 4.6 Future research directions ............................................................................................. 64 Bibliography .................................................................................................................................67 Appendices ....................................................................................................................................71 Appendix A ............................................................................................................................... 71   ix  List of Tables  Table 1-1. ANSD category determination. In cells labeled “see ratio” the CM/V ratio is calculated. Ratios exceeding 1.5 suggest “probable ANSD” (Hyde et al., 2016). ......................... 9 Table 2-1. Demographic data and BCEHP risk factors for hearing loss in infants with ANSD (Bremner et al., 2012). Multiple risk factors may present in an infant. Note: BCEHP protocols also accept any hospital-standard high bilirubin criteria as indicative of hyperbilirubinemia as a risk factor for hearing loss. ........................................................................................................... 31 Table 3-1. Combined ANSD ears (n=24) compared to normally-hearing infants from the neonatal intensive care unit (NICU; n=58). Statistically significant (p<0.05) results are indicated with (*) symbols. Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again................................................................................................................................ 42 Table 3-2. Combined ANSD ears (n=24) compared to normally-hearing well-baby (WB; n=53) ears. Statistically significant (p<0.05) results are indicated with (*) symbols.  Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again. ............................. 43 Table 3-3. Independent t-test results of OAE+ infant ears (n=17) compared to OAE- ears (n=7). Statistically significant (p<0.05) results are indicated with (*) symbols.  Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again. ............................. 44 Table 4-1. Summary of referenced research studies of the cochlear microphonic ....................... 50  x  List of Figures  Figure 1-1. ABR responses from the normal-hearing ear of a five-year-old child with unilateral ANSD, elicited by 80 dB nHL click stimuli (data from this thesis). From top: overlapping R & C waves showing the antiphasic CM, (R - C)/2 CM wave isolating the CM with trough-to-peak amplitude indicated between 0.5 and 1.5ms, and (R + C)/2 alternating wave (alt) isolating the non-inverting neural aspects of the ABR, including waves I, III, and V. A peak-and-trough latency shift is evident between R and C polarities starting at approximately 2.5ms due to neural asymmetries produced during electrical-mechanical transduction in many individuals. The y-axis presents amplitude in microvolts (uV). No clamped-tube trials were recorded for this ear. ........ 17 Figure 1-2. Schematic of MOC connections and the proposed OHC brake model (Elgoyhen & Katz, 2012; Moser & Starr, 2016). Sound input provokes basilar membrane motion, activating the OHC gain mechanism. Red X’s indicate systematic breakdown due to dys-synchronous neural activity at points 1, 2, 3, or 4, corresponding to potential loci of lesion in ANSD. In the absence of MOC efferent feedback through 5, OHCs act on the basilar membrane, which in turn reacts on the OHCs, creating a feed-forward, ringing pattern possibly leading to long duration CMs. .............................................................................................................................................. 21 Figure 2-1. Waveform averaging for a representative 2.25-month-old (corrected) OAE+ infant from 95 dB nHL click stimuli. From top: rarefaction replications (Rare 1, 2), rarefaction average (Rare Ave.), condensation average (Cond Ave.), condensation replications (Cond 1, 2), butterfly plots (R and C) presenting antiphasic averages (Rare Ave., Cond Ave.) with the clamped artifact subtracted, alternating waveform [(R + C)/2 Alt] showing the average of clamp-subtracted R and C averages, CM wave in isolation [(R – C)/2 CM] showing the CM response alone from clamp-xi  subtracted R and C averages, individual clamped tube tracings [R(c), C(c) Clamp]. Amplitude is presented on the y-axis at a 1µV scale. CM-bars above (R and C), and overlaying [(R – C)/2 CM] tracings indicate the measured CM duration and amplitude, respectively. Wave V measurement is shown on the (R + C)/2 tracing as wave V, V’. .................................................. 35 Figure 2-2. Representative 95 dB nHL click-ABR results from a 2.25-month-old (corrected) OAE+ infant with ANSD. From top: overlapping (R & C) butterfly plots used for duration measures, (R – C)/2 CM wave in isolation for amplitude measures, and (R + C)/2 alternating waveform for wave V measurement. Amplitude is presented on the y-axis at a scale of 0.5 µV. CM-bars indicate CM duration and amplitude measures. Wave V indicates the onset of neural wave V, with V’ indicating the opposite phase maxima for peak-to-trough amplitude determination. The clamped-tube waves (not shown) have been subtracted from these averages........................................................................................................................................................ 38  xii  List of Abbreviations  ABR – Auditory Brainstem Response ALT – Alternating polarity ANSD – Auditory Neuropathy Spectrum Disorder BCCH – British Columbia Women and Children’s Hospital BCEHP – British Columbia Early Hearing Program BEST – British Columbia Early Hearing Surveillance Tool C – Condensation Polarity (c) – Clamped Tube Trial CI – Confidence Interval CM – Cochlear Microphonic CNE – Could Not Evaluate DNT – Did Not Test EcochG – Electrocochleography FDR – False Discovery Rate Hz – Hertz IHC – Inner Hair Cell IHS – Intelligent Hearing SystemsTM MOC – Medial Olivocochlear Complex ms – Milliseconds nV - Nanovolt NICU – Neonatal Intensive Care Unit xiii  OAE – Otoacoustic Emission OCB – Olivocochlear Bundle OHC – Outer Hair Cell R – Rarefaction Polarity SHL – Sensory Hearing Loss TM - Trademark VRA – Visual Reinforcement Audiometry WB – Well-Baby μM – Micromoles  µV – Microvolt ± - Plus, or Minus xiv  Acknowledgements  My gratitude to the everyone at the UBC School of Audiology and Speech Pathology for making the school the excellent learning environment it is. Thanks also to all my fellow AUDI students, without whom I may not have survived the evenings and weekends.   Thanks to Dr. Anthony Herdman for his knowledge, humor, and skill, and for pushing me to challenge everything, including myself. Thanks also to my supervisory committee Dr. Navid Shahnaz and Rae Riddler for their valued input on this manuscript, and to the good people with the BC Early Hearing Program, working to leave no child behind.    Enduring thanks to Helena Burrows, my partner, for hearing my stress and doubts, and for making life better. 1  Chapter 1: Auditory neuropathy spectrum disorder (ANSD) In 1996, Starr and colleagues reported on 10 unusual subjects with mild-to-moderate hearing loss, who exhibited robust otoacoustic emissions (OAEs) and cochlear microphonic (CM) potentials. These measures suggested normal cochlear activity despite the subjects’ functional hearing impairments. Subjects’ auditory brainstem responses (ABR) were absent, or grossly abnormal with a notably absent wave I (Starr, Picton, Sininger, Hood, & Berlin, 1996); further, these patients’ behavioral pure-tone hearing thresholds were disproportionately good compared to their poor speech perception and absent middle-ear reflexes to loud sounds. Since 8 of the 10 presented with other peripheral neuropathies, and testing suggested faulty auditory-neural conductance, the authors coined this affliction “auditory neuropathy” (Starr et al., 1996). This constellation of results has since come to be known as auditory dys-synchrony, or auditory synaptopathy, with a current consensus of auditory neuropathy spectrum disorder (ANSD) (Moser & Starr, 2016). For clarity, hearing loss due to outer hair cell impairments, which are typically referred to as sensorineural hearing loss, will be referred to here as sensory hearing loss (SHL).  1.1 Mechanisms and auditory deficits of ANSD Aptly named, ANSD includes a spectrum of auditory neural-conductive deficits, each appearing to cause dys-synchronized communication between the cochlea and auditory brainstem (Rance & Starr, 2015). According to Moser & Starr’s 2016 review of the neural and synaptic mechanisms of ANSD, site-of-lesion may include: 1) Presynaptic disorders of neurotransmitter release, 2) faulty postsynaptic spike generation at unmyelinated dendrites, 3) conduction disorders of myelinated ganglion cells, and 4) demyelinating disorders of the 2  auditory brainstem itself. These lesions disrupt auditory nerve activity by two basic mechanisms, including: Reduction of the number of activated auditory nerve fibres, and/or reduction in the degree of neural synchrony in firing (Rance & Starr, 2015). In the first mechanism, neural spike amplitudes have been shown to decrease over time with no corresponding change in neural conduction speed, as evidenced by consistent ABR peak latencies but decreasing peak amplitudes. In the second, ABR interpeak latencies increase over time, with peak amplitudes decreasing due to dys-synchronization of neural firing and/or loss of functioning nerve fibres. Due to dys-synchrony, excitatory temporal summation in nerve fibers may be less dependable, raising neural spike thresholds (Moser & Starr, 2016; Rance & Starr, 2015), and reducing auditory acuity. Perceptually, this irregular neural conduction can impair gap detection and temporal discrimination of sound, which is critical for low-frequency sound localization, complex sound and speech comprehension, and figure-ground separation in background noise (Elgoyhen & Katz, 2012; Michalewski, Starr, Nguyen, Kong, & Zeng, 2005). Some research has found that psychophysical measures of gap detection can discriminate between ANSD and SHL (Rance & Starr, 2015), with ANSD participants requiring significantly longer gap durations for detection. This impairment of gap detection in cases of ANSD is problematic for young language learners, as phonetic gaps are linguistically informative. For stop-closures of plosive consonants, voice-onset time is useful, if not crucial to phoneme identification in hard of hearing populations (Elangovan & Stuart, 2008). In noisy environments such gaps in phonetic structure are effectively masked by noise (Starr et al., 1996; (Berlin, Hood, Morlet, & Wilensky, 2010), further interfering with phoneme discrimination, and inhibiting young language learner’s ability to recognize and learn individual speech sounds. 3  This deficient afferent neural synchrony may also preclude efferent feedback to the outer hair cells (OHCs) within the cochlea. Feedback systems are thought to enable brainstem modulation and fine-tuning of the peripheral cochlear amplifier in response to sound input (Elgoyhen & Katz, 2012; Santarelli, Scimemi, Dal Monte, & Arslan, 2006). This fine-tuning of sound input is thought to modulate speech in noise perception, temporal discrimination, selective attention, and the dynamic range of hearing. Interestingly, ANSD listeners typically show normal cochlear processing of frequency and intensity parameters of sound input, supporting the findings that the fault does not specifically involve the cochlear amplifier generated by the outer hair cells (Rance & Starr, 2015). Taken together, these deficits in auditory processing may create an unpredictable cacophony of sound, that interferes with a child’s acquisition of speech and language (Sharma & Cardon, 2015).  Early hearing programs continue to improve screening and diagnostic methods to identify infants with ANSD as early as possible; however, infants with ANSD may still undergo multiple tests before ANSD is confidently diagnosed and this may delay intervention with amplification or cochlear implantation (Teagle, Roush, Woodard, Hatch, Zdanski, Buss, & Buchman, 2010). Lacking a confident diagnosis, clinicians may opt for a conservative approach and delay intervention until reliable behavioural results can be obtained. Clinicians may be concerned about prescribing hearing aids that might over-amplify sounds to infants with ANSD because behavioural thresholds might be lower than physiologically-measured ABR thresholds (Berlin et al., 2010). The counter-argument to this is that delaying intervention is also a form of under-amplification thereby limiting sound input needed for typical speech and language development (Yoshinaga-Itano, Sedey, Coulter, & Mehl, 1998). Unfortunately, there is currently limited evidence to indicate if over- or under-amplification occurs in infants with ANSD.  4  Due to prolonged under-amplification, poor neural pathway activation may lead to neuroplastic deafferentation and reorganization of cortical auditory centers, with more stable sensory modalities invading neural space typically reserved for auditory language centers. This cortical reorganization might be mitigated by cochlear implantation in some cases (Sharma & Cardon, 2015; Teagle, et al., 2010). It’s possible that differences in responses to intervention for children with ANSD may be due to the different sites-of-lesion included in the spectrum of ANSD. Research to identify diagnostics for localizing the source of ANSD is ongoing (Santarelli, Del Castillo, Rodríguez-Ballesteros, Scimemi, Cama, Arslan, & Starr, 2009) and may help predict speech and language outcomes in the future, with and without cochlear implantation.   1.2 Clinical presentation of ANSD ANSD is rare, presenting in one-to-three per 10,000 births, or 8.4 to 11 percent of children with permanent hearing loss (Soares, Menezes, Carnaúba, de Andrade, & Lins, 2016), with reports indicating roughly equal incidence in males and females (Berlin et al., 2010). Some evidence suggests that 30 to 50 percent of new cases are missed by newborn hearing screening programs, which primarily rely on the assessment of OAEs as an indicator of auditory function in well-baby (WB) populations presenting with no known risk factors for hearing loss (Berlin et al., 2010; Hyde, Bagatto, Martin, Pigeon, Scollie, & Witte, 2016; Rodríguez Domínguez, Cubillana Herrero, Cañizares Gallardo, & Pérez Aguilera, 2007). Problematically, ANSD may present in infants with no known risk factors for hearing loss (Berlin et al., 2010), resulting in many delayed diagnoses in these infants. Congenitally, gene mutations affecting the expression of certain proteins, such as otoferlin (DFNB9 associated) and pejvakin (DFNB59 associated) appear to impair vesicle 5  release at the inner hair cell (IHC) to spiral ganglion synapse, resulting in the diagnostic profile of ANSD (Moser & Starr, 2016). In the auditory nerve, defective MPZ genes that code for myelin proteins may also cause dys-synchronous neural activity due to breakdown in conduction along the nerve fibers. Similarly, demyelinating disorders like Charcot-Marie-Tooth, and polyneuropathy disorders such as Friedrich’s Ataxia have also been implicated, along with other genetic and syndromic targets that are beyond the scope of this report (Berlin et al., 2010; Moser & Starr, 2016; Vlastarakos, Nikolopoulos, Tavoulari, Papacharalambous, & Korres, 2008).  ANSD may also be acquired. For instance, hyperbilirubinemia has been shown to selectively damage auditory brainstem nuclei and cell bodies (Vlastarakos et al., 2008); natal hypoxia may selectively damage sound-transducing IHCs, while sparing the cochlear gain mechanism of the OHCs and consequent OAE presentation, consistent with ANSD. Low birth weight, prematurity, cytomegalovirus, and meningitis are all risk factors for hearing loss that have been linked to ANSD (Berlin et al., 2010; Soares et al., 2016; Starr et al., 1996). Unfortunately, risk factors such as these often co-occur in infants admitted to the NICU, complicating etiological reports on the disorder. Further, the site of impairment along the auditory pathway and the specific cause of disorder often cannot be distinguished, except in cases of severe cochlear nerve deficiency where diagnostic imaging may reveal agenesis or inadequate growth of the auditory nerve for normal functioning (Sharma & Cardon, 2015). There is no obvious pattern why one infant with a similar neo-natal status to another may present with ANSD and the other not (Soares et al., 2016; Vlastarakos et al., 2008). Adding to the complexity, SHL and ANSD share risk factors, and the two forms of hearing loss may co-occur in the same ear, along with complicating disorders of the middle-ear (Hyde et al., 2016). Interestingly, research on noise exposure and auditory-synaptic degradation, also known as hidden hearing 6  loss, suggests that similar pathology could be acquired in older patients (Kujawa & Liberman, 2009), and that ANSD-like pathology may be more widespread than conventionally described.   1.2.1 ANSD Diagnosis ABR measures of neural integrity, OAE assessments, and determinations of CM presence or absence are all essential diagnostics for identification of ANSD in infants. As noted by Starr and colleagues in 1996, poor neural summation may eliminate ABR waves I and II, sourced at the spiral ganglion and auditory nerve, and delay, reduce, or otherwise eliminate waves III, IV and V, which arise from activation of ascending auditory brainstem structures (Katz, 2015). Particularly of interest in ABR testing is wave V, the clinical indicator of an auditory “response present”, or “response absent” for determination of threshold and frequency-specific hearing sensitivity in infants (Bremner, Davies, Hatton, Hyde, Ishida, Janssen, & Van Maanen, 2012). This wave V potential is thought to originate from the lateral lemniscus and fibers of the superior olivary complex (Hyde et al., 2016; Katz, 2015). In 2001, Starr and colleagues found that a low amplitude wave V could be elicited with high amplitude clicks in 20 percent of subjects with ANSD. This indicated that the severity of neural dys-synchrony in ANSD may be graded and thus resulted in different physiological responses (Starr, Sininger, Nguyen, Michalewski, Oba, & Abdala, 2001). Different levels of severity could arise from the different sites-of-lesion in ANSD, or from different degrees of dys-synchrony at each site-of-lesion (Santarelli et al., 2006). Importantly, the click stimuli used for ANSD diagnostics are not frequency specific, and produce broader cochlear excitation than frequency-specific stimuli. The use of click ABRs are generally reserved to explore the possibility of neural dys-synchrony disorders, such as ANSD, when frequency-specific ABR is absent or abnormal (Bremner et al., 2012). According to the most 7  recent Canadian clinical protocols out of Ontario, collaboratively established with the BC health authority, ANSD may fall under three diagnostic categories: definite, probable, and not suspected (Hyde et al., 2016). A definite ANSD diagnosis requires repeatable frequency-specific OAEs at any one frequency from 2-4 kHz, and broadband click-ABR results showing an absent, small (<0.1µV), or delayed (>10ms) wave V at 75 dB nHL or greater. Present OAEs with slightly larger wave V amplitudes from 0.1-0.2µV suggests a less definitive, but probable ANSD component, with further assessment required (Hyde et al., 2016). Absent OAEs with small CM and wave V amplitudes (<0.1µV) would indicate little to no OHC function and thus ANSD would be unsuspected/unlikely.  Infants with such a profile of results would typically be diagnosed with SHL based on these findings.   Concerning OAEs, these cochlear responses are low amplitude acoustic emissions that validate the mechanical integrity of the OHCs in response to sound input. Frequency-specific, distortion product OAEs (DPOAEs) are elicited by the combined presentation of two pure-tone frequencies, which are identified as F1 and F2. Due to the non-linear enhancement of incoming sound by the OHCs, interactions of these tones on the basilar membrane within the cochlea produce recordable emissions at a third frequency band that was not present in the original signal. A F2/F1 ratio of 1.2 is typically used (Bremner et al., 2012), as this ratio appears to produce the most stable distortion products at the 2F1/ F2 frequency band that is monitored for the OAE response from within the cochlea. A microphone placed in the external ear canal is used to detect these faint emissions that are generated from the OHC within the cochlea. For OAEs to be generated, several chain of events must occur: First, sound energy imparted on the tympanic membrane must be sufficiently conducted through the middle-ear system (forward transmission) to impart mechanical energy into the cochlea. Once activated, OHCs contract and amplify the 8  incoming mechanical energy. This OHC amplification process generates mechanical energy that is propagated out of the cochlea through the middle-ear system (reverse transmission) into the ear canal where the microphone sits. Even minor middle-ear pathology may impede the forward and reverse transmission of the mechanical energy. Due to the low amplitude of the reverse-transmission, OAE measurement is more susceptible to middle-ear pathology than CM measurement, which relies only on forward-transmission through the middle-ear system; researchers have reported that OAEs may be abolished with as little as 5 dB of conductive hearing loss (Hyde et al., 2016; Bremner et al., 2012). Similarly, mild OHC dysfunction may reduce the OAE response to clinically absent levels due to reduced OHCs activation and contraction (Katz, 2015; Hyde et al., 2016).  Related to the DPOAE, transient-evoked OAEs (TEOAEs) are elicited by broadband (500–5000Hz) click stimuli (Katz, 2015), presented at approximately 82 dB SPL (peak equivalent). These emissions are predictably time-delayed due to the tonotopic organization of the cochlea, with measurement occurring from 4-to-10ms or 12.5ms delay following click stimulus presentation. Statistical algorithms are applied to the responses of both elicitation techniques to identify the cochlear emission (signal) from the ambient environmental noise (Katz, 2015). Normative emission amplitudes and signal-to-noise values are used to compare recorded responses to healthy ears. Clinically, the frequency-specificity of DPOAEs makes them the more informative and more typically used in a clinical setting (Bremner et al., 2012; Hyde et al., 2016). Regardless of recording method, the presence of OAEs implies that ABR responses should be detectable in a patient, therefore a finding of abnormal or absent ABRs with present OAEs indicates a neural source of dysfunction.  9  Measurement of the CM could also be a valuable diagnostic method for quantifying cochlear integrity in complex cases. Some evidence has shown that large CMs with small wave V amplitudes are present in children and adults with ANSD (Hyde et al., 2016); however, a robust CM on its own currently does not differentiate between SHL and ANSD. To capitalize on present or otherwise robust CM amplitudes and diminished wave V in ANSD, a CM/wave V amplitude ratio has been proposed as a possible diagnostic tool (Hyde et al., 2016). This ratio is discussed more in the following section 1.2.2. A summary of CM, wave V and CM/wave V (CM/V) ratios is presented in table 1-1, below (Hyde et al., 2016). CM amplitude (µV) Wave V amplitude (µV)  < 0.1 0.1 – 0.2 > 0.2 < 0.1 Not Suspected Not Suspected Not Suspected 0.1 – 0.2 Probable See Ratio Not Suspected > 0.2 Definite Probable See Ratio Table 1-1. ANSD category determination. In cells labeled “see ratio” the CM/V ratio is calculated. Ratios exceeding 1.5 suggest “probable ANSD” (Hyde et al., 2016).    1.2.2 Suspected ANSD  Infants with suspected ANSD and absent OAEs present a particular challenge to clinicians, because physiological assessments (objective test measures) are the only developmentally appropriate option available until the infant can reliably participate in behavioral audiometry after about six months of age. Problematically, OAEs may be abolished by middle-ear pathologies, such as middle-ear effusion, abnormal middle-ear pressure, or ossicular malfunction, which may impede the transmission of the low-amplitude emission from within the cochlea, as described above. OAEs may also be reduced in amplitude or considered absent by a standalone or co-occurring SHL. For these cases the above-mentioned CM/V amplitude ratio has been proposed as a possible diagnostic method that takes advantage of the 10  typically reliable presentation of the CM when OAEs are absent, via the relative strength of only requiring a forward-transmission of the signal. Categorically, CM/V ratios ≥2.0 indicates definite, 1.5-2.0 indicates probable, and ≤1.5 indicates no suspicion for ANSD (Hyde et al., 2016). Essentially, when the CM is small, or an identifiable wave V is generated within 10ms and above 0.2uV, suspicions of ANSD are reduced. In the absence of OAEs, a small CM amplitude further indicates cochlear malfunction typical of SHL, and a recordable wave V diminishes the likelihood of retrocochlear disorder. These ABR waves are recorded using a split-alternating method, presenting high-amplitude click stimuli in both rarefaction (R) and condensation (C) polarities to show the antiphasic response of the CM. A clamped tube recording of each polarity is used to rule out stimulus artifact (Hunter, Blankenship, Gunter, Keefe, Feeney, Brown, & Baroch, 2018; Soares et al., 2016). CM amplitude is measured at the maximum peak-to-trough/trough-to-peak from (R – C)/2 averaged waveforms, typically within the first millisecond after stimulation (Shi, Ji, Lan, Liang, Ding, Wang, & Wang, 2012), or between 0.5-1.5 ms latancy (Hyde et al., 2016). The resulting waveform must be divided by two (or multiplied by 0.5) to mathematically correct for the subtraction of out-of-phase components (i.e. 1- (-1)), while in-phase components are effectively removed from the waveforms (i.e. 1-1). Wave V is measured at the largest appropriate negative deflection within 4-10ms post stimulus, if at all identifiable in the (R + C)/2 averaged waveforms, also known as the alternating waveform, which effectively remove the phase-reversing aspects of the waveform like the CM (Hyde et al., 2016). Marking these points on the ABR waveform simultaneously measures latency and amplitude. The wave V symbol denotes wave V on the ABR, with latency measured between the marked point (V) and the initial waveform node (i.e. 0ms latency). This 0ms point is time-compensated to allow for sound to physically travel from the transducer to the ear. Peak-to-11  peak wave V amplitude is measured as a comparison between the V marker placed at the maximum peak of one polarity and a corresponding V’ marker placed at the maximum peak of the opposite polarity within same cycle of the waveform (i.e. peak-to-trough). Crucially, abnormal or absent ANSD neural waves present unpredictably, with little correlation to normally-hearing and SHL neural amplitudes and latency values for waves I-V in click or tone-evoked ABR, leading to the result of a so-called abnormal ABR (Starr et al., 2001; Hyde et al., 2016). It is for this reason clinicians might delay interventions with amplification because the ABR thresholds might not represent the actual auditory ability in adults and children (Berlin et al., 2010), but we have no evidence for this in infants (<12 months of age).   Another cochlear receptor potential, the summating potential (SP), has been explored as a potential identifier for ANSD. As opposed to the alternating polarity of the CM, the SP is a direct current potential which appears to reflect the non-linearities of cochlear sound transduction due to the flow of ions within the cochlea and the synapse to the auditory nerve (Katz, 2015). The SP has been found to be difficult to distinguish from a prolonged, low amplitude wave I in early auditory-evoked recordings in ANSD, which may also occur in normally-hearing ears (Santarelli et al., 2009; Starr et al., 2001). Santarelli and colleagues (2009) found that cochlear potentials such as the SP/wave I complex may present as low as 50 to 90 dB SPL below the behavioral hearing thresholds of children with confirmed presynaptic disorders from OTOF mutations resulting in ANSD. In normally-hearing child ears, behavioral and cochlear receptor thresholds were found to be closely correlated (Santarelli et al., 2009). Exploration of the SP in cases of ANSD has been limited and the potential is not described in current Ontario or BCEHP clinical practice guidelines for the identification of ANSD. In addition, the SP is challenging to measure using ABR tracing recorded from the mastoid. Electrocochleographic recordings to slow (10-12  20/s) and fast presentation rate (>60/s) clicks are typically needed to disentangle the SP from the compound action potential (wave I). Given that this study was a retrospective study of ABR recorded from mastoid electrodes, the SP component was explored in this thesis. Other supporting diagnostics for ANSD include ipsilateral and contralateral acoustic reflexes of the middle-ear and the olivocochlear bundle, which are typically absent in ANSD (Soares et al., 2016). The absence of these responses supports the diagnosis of a neural site-of-lesion and may contribute to the test battery as a cross-check to the ABR results described above. Problematically, acoustic reflexes will also be absent in severe-to-profound SHL and thus only provide limited clinical value with differentiating between ANSD and SHL. Middle-ear reflexes may also be abolished by middle-ear pathologies that cause significant conductive hearing impairments, where reduced forward transmission of the acoustic stimulus to the cochlea does not activate the cochlea sufficiently enough to trigger a measurable reflex from the stapedius muscle. The application of bone-conducted stimuli in the ABR would be the definitive sign of such a conductive disorder, provided that ABR responses are present. Interestingly, reports have described infants with ANSD who present with poorly replicating middle-ear reflex responses at elevated stimulus levels (Berlin et al., 2010; Hunter et al., 2018). Reflexes in response to broadband noise (BBN) stimuli typically present at lower stimulation intensities in normal-hearing ears, which suggests broadband stimuli may be more effective than tonal for reflex assessment in cases of ANSD (Hyde et al., 2016). If a reflex can be identified in patients with ANSD, this could help inform the clinician regarding the levels of amplification to prescribe in an amplification device. For instance, where a tone-evoked ABR shows absent responses at testing limits, suggesting a profound hearing loss, the presence of BBN reflexes may indicate that amplification targets directly reflecting this degree of hearing loss may be uncomfortably 13  high. More research is required to determine the clinical validity of using BBN reflexes as an indicator for true sensation levels that are lower than those estimated from tone ABR thresholds. This is beyond the scope of the current thesis.  Similar to absent reflexes, research has shown that contralateral, ipsilateral and binaural suppression of OAEs via the brainstem may also be absent in adult cases with ANSD (Berlin, 1998). The mechanisms of OAE suppression follow the same afferent neural pathway as those for middle-ear reflex responses. As with absent reflexes, absent OAE suppression further supports the retrocochlear involvement indicated by an absent ABR. Unfortunately, these measures offer no extra information in the presence of middle-ear dysfunction or SHL due to the necessary reverse-transmission through the middle-ear system, and the reliance on the OHCs for cochlear amplification and OAE generation, respectively. In the 2016 Ontario clinical practice guidelines, acoustic reflex assessment is not considered a mandatory part of the ANSD test battery (Hyde et al., 2016); the assessment is considered a cross-check that is secondary to click-ABR and OAE measures of primary interest. Methods for the application of measuring OAE suppression are not in either Ontario or BCEHP current clinical guidelines, but may be a useful diagnostic with further research (Bremner et al., 2012; Hyde et al., 2016).  It is not entirely clear what impact middle-ear dysfunction may have on measures of the CM. It has been reported by Hyde and colleagues (2016) that CM amplitudes may be reduced below detectable levels in the presence of a mid-frequency conductive component of 20 dB eHL when elicited by a 90 dB nHL click-stimulus. It is similarly unclear what impact conductive impairments may have on CM duration if CM response amplitudes are attenuated. For instance, if CM amplitudes are variably reduced by a conductive hearing impairment, phase-reversing components of the CM may become indistinguishable from background electrical noise more 14  rapidly, leading to a shorter identifiable CM duration in ears with middle-ear dysfunction. This is purely speculation, as no reviewed literature has directly explored CM measures in ears with known middle-ear dysfunction. Since the CM relies only on forward-transmission through the middle-ear and is generated by both the inner and outer hair cells, it may be that the CM is measurable even with significant conductive or outer hair cell impairments. Like the CM, neural wave V response amplitudes may be similarly reduced below the background electrical noise for detection of ANSD in the presence of such a conductive component (Katz, 2015). It is possible that CM/V amplitude ratios would remain unaffected, as both CM and wave V response amplitudes may be equally attenuated; however, this is also purely speculation. At approximately six months of age (in normally developing infants) further diagnostic tests may include visually-reinforced pure-tone audiometry (VRA). This behavioral test may show flat, reverse sloping, high frequency, or fluctuating hearing losses, as well as a complete absence of response to sound (Narne, Prabhu, Chandan, & Deepthi, 2014; Soares et al., 2016). When reliable, behavioral hearing thresholds in cases of ANSD do not appear to correlate with the frequency-specific thresholds obtained with tone-ABR assessment or indicated by click-ABR results (Berlin, 1998). For normally-hearing infants and infants with SHL, ABR thresholds are typically within 10 to 20 dB HL of thresholds obtained with behavioral testing (Stapells, 2000). Cortical auditory-evoked electrical potentials (CAEPs), which benefit from longer neural integration intervals, may be better predictors of functional hearing thresholds than ABR when behavioral audiometry is not a reliable test option; however, these CAEP responses still rely on variable neural synchrony, and are generally considered an upper limit of actual hearing ability (Hyde et al., 2016; Teagle et al., 2010). Research on CAEP efficacy for diagnosing infants with ANSD is ongoing.  15  Beyond infancy, suspected ANSD may be supported through pure-tone/speech perception mismatches, with speech results disproportionately poor for a given pure-tone loss (Soares et al., 2016; Starr et al., 1996); however, this is also not a reliable measure, as children may range from zero awareness of speech, to performing at near normal levels as noted above (Soares et al., 2016; Vlastarakos et al., 2008), further suggesting that ANSD severity may be graded (Starr et al., 2001). Critically, this testing would occur beyond the six-month target for hearing loss interventions to be in place to maximize language learning (Bremner et al., 2012; Rodríguez Domínguez et al., 2007). Late identified children may be at risk of neuroplastic reorganization of the auditory cortex as described above (Sharma & Cardon, 2015; Teagle et al., 2010), and delayed intervention has been linked to lifelong social, academic and vocational frustration (Yoshinaga-Itano, Sedey, Coulter, & Mehl, 1998). Further, this extensive diagnostic scenario, which often occurs over a very long period of time, comes at a cost to both families and public health systems. A developmentally appropriate, objective, and reliable indicator of ANSD is needed, especially for the diagnostically complex infants presenting with absent OAEs.   1.3 The cochlear microphonic The CM reflects the electrical depolarization and repolarization of both IHCs and OHCs induced by motion of the basilar membrane in the cochlea (Hyde et al., 2016; Shi et al., 2012), with greater contributions from the OHCs due to their greater number in healthy cochlea. The morphology of the CM follows the phase of the stimulus (Rance & Starr, 2015). As such, it is described as an alternating current, believed to be the vector sum of extracellular components of mechanical-electrical transduction by all the cochlear hair cells, with a significant portion of the response arising from the 2-4 kHz region of the cochlea in ABR measures (Prabhu, Narne, & 16  Barman, 2014; Riazi & Ferraro, 2008). The response is detectable even when OAEs are absent due to the contributions of the IHCs, and in the presence of middle-ear dysfunction, as long as the stimulus intensity exceeds the degree of conductive hearing loss caused by this dysfunction (Starr et al., 2001; Vlastarakos et al., 2008). Response amplitude of the CM in these situations would be reduced compared to healthy ears (Katz, 2015; Shi et al., 2012). Extensive cochlear lesions due to SHL may reduce CM amplitude to undetectable levels; however, the CM amplitude might indicate some amount of cochlear integrity in patients with suspected ANSD when OAEs are absent. As noted, CMs are not as effected by middle-ear pathology, but there is a caveat to strictly relying on CM presence as a clinical indicator of OHC integrity in that IHCs also generate CMs. Thus, the CM only indicates that hair cells are being activated, regardless of CMs being of IHC or OHC origin, which may complicate any direct comparisons of CM amplitude to true behavioral hearing thresholds.  CM response amplitude has been shown to increase with greater stimulation amplitudes in healthy cochlea (Shi et al., 2012), with the latency of CM onset remaining constant regardless of stimulus intensity in healthy ears (Katz, 2015). CMs are a frequency-following response that reliably track the frequency of the stimulus. Although this appears to be a good measure of cochlear function, electrical artifacts can confound the CM recording. Specifically, stimulus transducers and preamplifiers may ring, producing electrical artifacts that mimic the CM (Hyde et al., 2016; Riazi & Ferraro, 2008). This is especially problematic in far-field recordings like the ABR where distance reduces CM amplitudes compared to the stimulus artifact (Riazi & Ferraro, 2008; Santarelli et al., 2006). It is for this reason that clamped-tube recordings to isolate the artifact from the physiological response are a mandatory part of neurologic ABRs. Clinically, the non-invasive nature of the ABR is ideal for infant assessment, and advances in equipment and 17  techniques have reduced the issue of artifact interference in testing, including: electrical shielding, optical cabling, regular calibration and replacement of equipment, and recording clamped-tube trials to quantify the level of artifact. CM duration can also be deceptive, because ABR neural components may shift by up to 0.4ms due to asymmetries in receptor activation between R and C polarities, mimicking a long duration CM (see figure 1-1, below).  Figure 1-1. ABR responses from the normal-hearing ear of a five-year-old child with unilateral ANSD, elicited by 80 dB nHL click stimuli (data from this thesis). From top: overlapping R & C waves showing the antiphasic CM, (R - C)/2 CM wave isolating the CM with trough-to-peak amplitude indicated between 0.5 and 1.5ms, and (R + C)/2 alternating wave (alt) isolating the non-inverting neural aspects of the ABR, including waves I, III, and V. A peak-and-trough latency shift is evident between R and C polarities starting at approximately 2.5ms due to neural asymmetries produced during electrical-mechanical transduction in many individuals. The y-axis presents amplitude in microvolts (uV). No clamped-tube trials were recorded for this ear.   18  The CM is subject to top-down, central nervous system (CNS) control through acetylcholinergic (ACh) and GABAergic efferent fibers, originating primarily in the contralateral medial olivocochlear complex (MOC), with ipsilateral contributions (Elgoyhen & Katz, 2012; Lichtenhan, Wilson, Hancock, & Guinan, 2016); these efferent fibers synapse directly onto the base of OHCs. In a healthy cochlea, OHCs perform work on the basilar membrane, delivering non-linear, positive feedback energy for enhanced hearing sensitivity (Katz, 2015; Lichtenhan et al., 2016). That is, low intensity stimuli are enhanced to a greater degree than high intensity stimuli in a non-linear manner, via contraction of the OHCs increasing the deflection of the basilar membrane in the cochlea. ACh released via the MOC is believed to cause calcium (Ca2+) influx into the OHCs via these ACh receptors on the OHC. These Ca2+ ions activate nearby potassium (K+) channels through a second-messenger system, resulting in outflow of K+ from within the cell (Elgoyhen & Katz, 2012). The net effect is hyperpolarization of OHCs due to a greater efflux of K+ than the influx of Ca2+. Overall, this reduces the likelihood of depolarization from influx of K+ when stereo cilia tilt in response to basilar membrane deflection, thus limiting the enhancement of incoming sound stimuli. CM amplitudes are believed to be increased by this hyperpolarization as well, with obligatory attenuation of neural activity beyond the cochlea (Starr et al., 2001). In effect, MOC activity is proposed to apply a braking mechanism to the nonlinear cochlear amplifier through hyperpolarization of the OHCs by limiting their action on the basilar membrane (Elgoyhen & Katz, 2012). As noted above, fine-tuning of sound input within the cochlea is thought to modulate speech in noise perception, selective attention, and the dynamic range of hearing (Lichtenhan et al., 2016), all of which have been reported to be deficient in ANSD (Berlin, 2010; Vlastarakos et al., 2008). Breakdown of this MOC to OHC 19  efferent system has been suggested as a potential source of distorted CMs in ANSD, which is discussed in detail in section 1.3.2, below.   1.3.1 The cochlear microphonic in normally hearing infants The CM has historically received little attention in the literature, and meaningful amounts of normative data have not been available, especially for infants (Hunter et al., 2018; Riazi & Ferraro, 2008). A recent report by Hunter and colleagues (2018) found no significant differences between normally hearing infants (n=32 infants, 64 ears) with known risk factors for hearing loss (e.g. low birth weight, prematurity, hyperbilirubinemia) from the neonatal intensive care unit (NICU), and normally-hearing well-baby infants (n=30 infants, 60 ears) with no known risk factors for hearing loss. Infants from both groups were approximately 1.25 months of age, corrected. Hunter and colleagues (2018) explored several electrophysiological measures using high-amplitude clicks (70 dB nHL), a rate of 37.1 per second, split-alternating recordings, and a desirably non-invasive, extratympanic recording method (ABR). At 70 dB nHL, the intensity of the stimuli in this normative study are lower than those typically used for ANSD assessment, probably to limit the risk of noise damage in healthy ears. Despite this lower intensity, the study presents a detailed bank of normally-hearing infant data that includes CM duration (well-baby: 0.73 ± 0.3ms, NICU: 0.82 ± 0.51ms) and CM amplitude (well-baby: 0.24 ± 0.09μV, NICU: 0.26 ± 0.13μV), as well as wave V amplitudes (well-baby: 0.28 ± 0.09μV, NICU: 0.29 ± 0.10μV). Normative data from this Hunter and colleagues (2018) study will be used for comparison to the ANSD data in the current manuscript, as described in the methods section. The researchers also include a case study of an infant identified with ANSD (Hunter et al., 2018), allowing for comparisons to infants with ANSD from the present dataset. 20  1.3.2 The cochlear microphonic in ANSD Some research investigating the CM in ANSD has found the response to be larger in amplitude (Starr et al., 2001), and longer in duration compared to normally-hearing ears (Santarelli et al., 2006; Soares et al., 2016). Other research has found no significant differences between groups on these measures (Shi et al., 2012; Prabhu et al. 2014). It has been suggested that poor IHC afferent synchrony may cause large and/or long CMs through diminished olivocochlear excitation and insufficient efferent feedback to the OHCs (Santarelli et al., 2006; Soares et al., 2016). This afferent dys-synchrony may reduce the efficacy of the MOC’s physiological brake on the OHC cochlear gain mechanism. Without this brake, sound input may mechanically activate the basilar membrane and the OHCs in turn, with the OHCs adding electro-chemical energy to, and amplifying the mechanical action of the basilar membrane, which then reactivates the OHCs, producing a feed-forward, ringing system proposed to produce long-duration CMs. This theory is supported by the absence of OAE suppression in patients with ANSD noted by Berlin in 1998, and it parallels the neural deficits underlying absent acoustic reflexes in ANSD, albeit through a different efferent nerve pathway (Berlin, 1998; Rance & Starr, 2015). If this hypothesis is correct, CM distortions should present regardless of the site-of-lesion or source of the dys-synchrony, so long as the lesion lies lower in the ascending system than the MOC. Overactivity of the efferent system has been suggested to produce persistent hyperpolarization of the OHCs, resulting in large amplitude CMs with depolarization (Starr et al., 2001). A schematic representation of this mechanism and breakdown are presented below in figure 1-2.  21   Figure 1-2. Schematic of MOC connections and the proposed OHC brake model (Elgoyhen & Katz, 2012; Moser & Starr, 2016). Sound input provokes basilar membrane motion, activating the OHC gain mechanism. Red X’s indicate systematic breakdown due to dys-synchronous neural activity at points 1, 2, 3, or 4, corresponding to potential loci of lesion in ANSD. In the absence of MOC efferent feedback through 5, OHCs act on the basilar membrane, which in turn reacts on the OHCs, creating a feed-forward, ringing pattern possibly leading to long duration CMs.  22  This proposed efferent-brake model for the source of abnormal CM duration and amplitude characteristics is by no means alone. A related, but somewhat opposite theory posits that overactivity of the MOC may lead to large CMs (Starr et al., 2001). Citing evidence that CMs may double in amplitude with electrical stimulation of the olivocochlear bundle in animal models, Starr and colleagues (2001) suggested that OHC hyperpolarization due to MOC hyperactivity could increase cochlear receptor potentials, with an accompanying decrease in auditory nerve activity. It is unclear what mechanism would produce this hyperactivity, as the cited literature applied electrical stimulation from an external source to produce the hyperpolarization (Starr et al., 2001). In 2012 Shi and colleagues proposed that long-ringing CMs are merely dys-synchronized and poorly averaged residual neural components of the ABR that only appear to represent the stimulus (Shi et al., 2012). The authors argued that given the brief duration of click stimuli, only phase reversing components before 1ms can be defined as CMs. This theory does not explain large amplitude CMs found in click-ABRs with absent neural waveforms; however, support is shown in a study by Santarelli and colleagues (2006), who described significantly high oscillation frequencies in the CMs of ANSD subjects compared to normally-hearing individuals with and without CNS disorders (Santarelli et al., 2006). This finding was unexpected, as CMs by definition should be faithful to the stimulus, as it is characterized as a phase-following response (Shi et al., 2012). In their own take on the source of abnormal CM duration and amplitude characteristics, these authors have suggested that different CM durations and amplitudes may arise from different sites-of-lesion and sources of dys-synchrony (Santarelli et al., 2006), though under the umbrella of the MOC efferent theory described above.  23  Of import to the current study, some research has suggested that age may impact measures of the CM. In 2001, Starr and colleagues found that CM amplitudes decreased as a function of age in 21 normally-hearing subjects aged one-week to 43 years. In 2006, Santarelli and colleagues reported that CM amplitudes increased up to five years of age in 43 ears of subjects with a CNS disorder. While declining responses in aged ears is not surprising, it is unclear what role maturational neural synchrony and normal myelination of nerve fibers in infants and young children may play in the presentation of CMs. Similarly, it is unclear what role these factors play in the OHC brake model. Given the poor controls for age in previous studies comparing ANSD and normally-hearing participants, it remains possible that age differences between groups may have produced the reported differences in CM amplitude and duration. Finally, the efferent pathways remain a source of uncertainty and have been little explored in the ANSD literature, aside from brief explorations of the possible causes of abnormal CM characteristics reported in these studies (Starr et al. 2001; Santarelli et al. 2006; Soares et al., 2016). These concerns will not be addressed directly here but could be an avenue for future research.  1.3.3 Reported values of abnormal CM waveforms in ANSD In their 2016 systematic review of the CM in ANSD, Soares and colleagues reported that phase reversing components of the ABR may persist for a mean of 4ms after presentation of a transient stimulus. Across studies that included a wide range of participant ages, this duration was found to be significantly longer in ANSD ears than normally-hearing controls (Soares et al., 2016). Contrary to this review, long duration CMs have been recorded from normally-hearing children and adults with CNS disorders (Santarelli et al., 2006), and were reported to be visibly 24  long in duration in normally-developing children with healthy ears (Prabhu et al., 2014). Large CM amplitudes have also been reported in ANSD, with no correlation between CM amplitude and degree of hearing loss measured with behavioral methods, or with wave V amplitudes when detectable in the ABR (Starr et al., 2001). Other research has found no statistical CM amplitude differences between children and adults with ANSD and normal-hearing controls when OAEs were present in the ANSD group (Santarelli et al., 2006; Shi et al., 2012). Furthermore, statistically smaller CM amplitudes were found for ANSD than normal-hearing controls when OAEs were absent in the ANSD group (Shi et al., 2012). Across studies, the systematic review noted above did not support the theory that large amplitude CMs are a distinctive sign of ANSD (Soares et al., 2016). Even given this evidence, the presence of a large CM has been used as an indicator of cochlear integrity when ABR wave V responses are absent or abnormal  and OAEs are absent (Bremner et al., 2012; Hyde et al., 2016). Conflicting CM duration and amplitude findings such as those described above have led to the development of the CM/V ratio as described in section 1.2.1. The CM/V amplitude ratio may better discriminate between patients with normal hearing, SHL, and ANSD (Hyde et al., 2016); however, evidence is limited to support this recommendation. Part of this thesis investigates whether or not CM amplitude and CM/V amplitude measures are valid for discriminating between normal and ANSD ears.  In a novel approach to CM analysis, Santarelli and colleagues (2006) have found the mean frequency of CM oscillation to be significantly higher in patients with ANSD than in normally-hearing patients with a CNS disorder; both groups were reported to have long duration CMs compared to normal-hearing controls with no CNS disorder (Santarelli et al., 2006). This oscillation analysis was performed using transtympanic electrocochleography and Fourier transformations, which provided a quantitative ruler for CM duration. Through this method, CM 25  onset and offset could be delineated from electrical noise by rapid and sustained increases in oscillation frequency. Unfortunately, this measure proved impossible in the current study due to the relatively small signal-to-noise ratio in far-field ABR recordings as compared to transtympanic methods used by Santarelli and colleagues (2006).   1.4 Summary and justification There are gaps in our understanding of the physiological presentation of ANSD and limitations in our current diagnostic ability, as described above. Diagnosis is especially complicated in young infants (<12 months of age) who do not fit the definite diagnostic profile with present OAEs and absent/abnormal click-evoked ABRs. CM measures of amplitude and duration may be useful for identifying ANSD beyond the current use of the CM for supporting cochlear integrity. Previous research has suffered from poor matching between test and control groups and included limited or no infant-age (<12 months) data. Thanks to the rigorous work of clinicians with the British Columbia Early Hearing Program (BCEHP), deidentified infant physiological data has been made available, with later behavioral results and expert review to support diagnosis of ANSD in patients with absent OAEs. With this information, the present study evaluated CMs from infants exclusively. This study compared measures of CM duration and amplitude to normally-hearing infant data presented by Hunter and colleagues (2018), with further comparisons across infants with ANSD presenting with OAEs (OAE+) and without (OAE-). A CM/V amplitude ratio comparison is also performed across these groups.   26  1.5 Hypotheses of the current investigation H1: CM duration is significantly longer in infants with ANSD than normally-hearing infants from the NICU and well-baby nurseries.  H2: CM amplitude is significantly larger in infants with ANSD than normally-hearing infants from the NICU and well-baby nurseries.  H3: CM/V ratios for all infants diagnosed with ANSD fall under “probable”, or “definite” diagnostic categories.  Ho1: CM duration is not significantly different between ANSD infants with OAE+ and OAE-. Ho2: CM amplitude is not significantly different between ANSD infants with OAE+ and OAE-. Ho3: CM/V amplitude ratio is not significantly different between ANSD infants presenting with OAE+ and OAE-. 27  Chapter 2: Methods 2.1 Participants Physiological and behavioral diagnostic results, intervention, and response-to-intervention data for infants meeting the inclusion criteria (ANSD) was first identified by clinicians with the British Columbia Children’s Hospital (BCCH) and BCEHP. This search was performed using the BC Early Hearing Surveillance Tool (BEST). Patient data was obtained for the years of 2008 through 2015. Identified paper charts for patients with ANSD were scanned into electronic versions, and Intelligent Hearing SystemsTM (IHS) ABR waveform files were collected for these infants. Electronic copies of patient records were examined for unique identifiers, and any such information was removed from the record. We received only anonymized and deidentified patient data for analysis, as submitted to the UBC Children's & Women’s Research Ethics Board. All IHS data provided for patients with ANSD was examined for split-alternating, high-amplitude click-ABR assessment, including clamped-tube trials. Individual waveform files were identified from electronic copies of ABR tracking sheets and the recorded ABR printouts themselves. The IHS files from infants identified as having ANSD were extracted for analysis in custom MatLabTM software (SmartEP Reader) created by Dr. Anthony Herdman. All mathematical calculations and waveform manipulation were performed in this custom software (see section 2.3, below). All calculations could also be performed using the IHS software, but the SmartEP Reader program provided ease use and efficiency in the averaging and measurement of waveforms for the purposes of this study.   28  2.1.1 Exclusion criteria Data for 34 patients with ANSD were identified through BEST by a BCEHP staff member. One further infant was identified with ANSD by researchers at UBC through chart review of infants with SHL who underwent high-amplitude click-ABR assessment. It is unclear why this infant was not identified through BEST given the diagnosis of ANSD presented in the chart files. From these 35 patients, data were excluded for the following reasons: 1) Click-ABR assessment was not performed or not present in the data (n=6 infants, 9 ears). 2) Waveform identifiers for stimulus and/or recording parameters were indecipherable in the scanned electronic copies of waveform files and ABR tracking sheets (n=2 infants, 4 ears) 3) Click-ABR was completed after 12 months corrected gestational age (n=7 infants, 8 ears).  4) No clamped-tube trials were recorded for a given ear (n=8 infants, 9 ears).  5) The stimulus was presented at levels less than 90 dB nHL (n=3 infants, 3 ears).  6) Diagnosis changed with additional testing (n=2 infants, 3 ears). For example, if diagnostic imaging revealed abnormal nerve growth as opposed to dys-synchronous pathology. Infants reported to have cochlear nerve deficiency were excluded from analysis as complete agenesis of the nerve could not be ruled out with the available chart data. These exclusion criteria were not mutually exclusive. For example, late age and absent clamped-tube recordings co-occurred in the same infant.  2.1.2  Inclusion criteria Click-ABR data from 24 ears of 16 infants with ABRs recorded at ages ranging from 0.35 to 11.7 months (corrected) of age (mean age 3.5 ± 3.2 months, corrected) were included for final analysis in this study. Groups were defined as OAE+ (10 infants, 17 ears, mean corrected 29  age of 3.4 ± 3.6 months) and OAE- (6 infants, 7 ears, mean corrected age of 3.8 ± 2.8 months). OAE+ and OAE- determinations were based on the presence or absence of repeatable DPOAEs from 2-4 kHz at any point in the patient history, as specified in the BC clinical practice protocols (Bremner et al., 2012). OAE presence or absence was determined in this study by clinical diagnostic reports (n=13 infants, 19 ears) and by inspection of the DPOAE tests results themselves (n=16 infants, 24 ears), when available. In BC, clinically significant OAEs indicating OAE presence is defined as DPOAE amplitudes exceeding -5 dB SPL, and 3 dB above the measured noise floor, with test-retest amplitude differences of 5 dB at most (Bremner et al., 2012). OAE- confirmation of ANSD relied on the ongoing BCEHP clinician diagnosis of ANSD, in conjunction expert input from the continuous quality insurance panel of professionals, and a review of patient history by researchers performing the current study, including Alicia Parfett and myself. All available middle-ear results, diagnostic imaging, repeat ABR, and later behavioral and outcome measures of intervention were considered alongside initial click-ABR assessments and OAE results.  Assumptions were made regarding the middle ear status of the ANSD infant data. Scanned tympanograms corresponding to the date of the neurologic ABR that verified normal middle-ear status at the time of testing were present for 6 of the included 24 ears. Typed clinical reports corresponding to the ABR assessment date indicated that 14 of the included 24 ears were considered to have normal middle-ear function, including the above-mentioned 6 ears with scanned tympanograms. As such, it is possible that up to 10 ears from 7 infants accepted for analysis in the present study showed abnormal middle-ear function at the time of click-ABR assessment. Instances of ossicular abnormalities and permanent conductive losses would likely have been displayed prominently in the patient files; however, this is another assumption made 30  in preparation for this analysis. Middle-ear status was assumed to be normal in patients where middle-ear assessment results were not available. Justification for the inclusion of ears where middle-ear status was unknown or ambiguous is both rarity of the infant ANSD data, and that BCEHP protocol does suggest that click-ABR assessment should not be conducted in the presence of middle-ear dysfunction (Bremner et al., 2012). The key middle-ear abnormality criterion described in the protocol is a peak static admittance of ≤ 0.6mmho, compensated from the negative tail at -400daPa using a 1kHz probe tone for ages of 6 months or younger, and peak static admittance of 0.1mmho, compensated from the positive tail at +200daPa using a 226hz probe tone for ages of 7-12 months (Bremner et al., 2012). Either tympanometry or evaluation of middle-ear status by an otolaryngologist should have been conducted at each click-ABR assessment. It is also unclear what impact middle-ear dysfunction or abnormal pressure would have on the CM responses obtained from high amplitude click-ABRs, or what degree of positive or negative middle-ear pressure would be necessary to impact these measures. The inclusion of these 10 ears where middle-ear status is unknown is certainly a limitation of the current study. A summary of available infant demographics and risk factors for hearing loss present in the included ANSD infant subjects is presented below in table 2-1. See Bremner et al. (2012), or Hyde et al. (2016) for a clinical discussion of these risk factors for hearing loss.     31  Variable Group All ANSD Present OAEs Absent OAEs Gender  Female 7 2 5 Male 7 7 0 Unknown 2 1 1 Mean age (months) Corrected 3.5 3.4 3.8 Uncorrected 5.6 5.8 5.2 Risk factors for hearing loss Prematurity 8 6 2 Low birth weight (<1500 grams) 4 2 2 Family history 1 1 0 Hyperbilirubinemia (≥400μM) 3 3 0 Respiratory distress 3 2 1 Ototoxic antibiotics (gentamycin, cloxacillin) 2 1 1 Neurological disorder (cerebral palsy, intraventricular hemorrhage) 3 2 2 No known risk factors 6 2 4 Table 2-1. Demographic data and BCEHP risk factors for hearing loss in infants with ANSD (Bremner et al., 2012). Multiple risk factors may present in an infant. Note: BCEHP protocols also accept any hospital-standard high bilirubin criteria as indicative of hyperbilirubinemia as a risk factor for hearing loss.   2.2 Instrumentation and equipment Identified IHS data files were imported into custom MatLabTM software (SmartEP Reader) created by Dr. Anthony Herdman in May of 2018 for ABR waveform analysis. This program was adapted from a simulated ABR program (sABR) used for instructional purposes at UBC. It was assumed that BCEHP protocols were met for machine calibration, ambient noise criteria, and related technical aspects for recording the ABR waves (Bremner et al., 2012), which is a further limitation of this retrospective analysis. Protocol states that electrodes should have been placed at the center midline of the high forehead (FPz), and the mastoid ipsilateral to the ear being stimulated (M1), with a ground electrode placed at or near the nasion (Naz); however, this 32  ground electrode is commonly placed to either side of the FPz electrode due to the small size of infant foreheads. Placing the ground electrode farther from the nasion may reduce the recordable response amplitude of electrical transduction by reducing the distance between reference points (Hyde et al., 2016). Overall impedances should have been ≤ 3 kilo-ohms (kohm) with interelectrode impedance differences ≤ 1 kohm (Bremner et al., 2012). Amplifier gain should have been 100,000 dB, with a widely set bandpass filter of 30-3000 Hz to gather as high-fidelity of an analogue-to-digital conversion as possible. On review, it was found that filters were set at 30-1500 Hz for 5 infants, 7 ears, likely depressing high frequency components of the CM (phase reversals of greater than the 1500 Hz cutoff) but expediting data collection in these infants by filtering more electrical noise from the recorded signal (Hyde et al., 2016; Katz, 2015).   Stimuli of interest were 90 dB nHL or greater (mean 94 dB nHL), 0.1ms clicks, presented at rates of 19.1/s (n=4 ears), 19.3/s (n=17 ears), and 19.5/s (n=1 ear) using ER3A insert transducers. One OAE- subject’s responses (n=2 ears) to a click-stimulus rate of 61.1/s were included due to a general lack of data for the OAE- group. This rate is not believed to impact CM amplitudes or durations, but likely reduced presentation of neural components of the ABR by reducing the time available for nerve fibers to recover after firing, resulting in a relatively smaller wave V. The inability to reconcile the differences in recording parameters (band-filter cutoff, presentation rate, stimulus amplitude) or the different stimulus transducers and electrical/environmental noise present during recording across infants is a limitation of this retrospective analysis.   33  2.3 Waveform analysis A series of waveform averaging was carried out within the SmartEP Reader program for analysis of the CM. All mathematical operations were carried out within the custom MatLabTM software. The R and C waveforms were examined to ensure the presence of at least two trials of 2000 stimulus repetitions each. These R and C waveforms were digitally averaged, and then digitally subtracted from their respective average clamped tube (c) tracings to reduce stimulus artifact within the waveforms. Subtracting the average artifact at the time of the recording further served to normalize this confound across subjects, which was necessary in this context as the data was collected over several years using different transducers in different recording conditions. If clamped trials were only performed for one polarity (e.g. R), the wave was inverted (R x -1) as an imperfect substitution for the opposite polarity (e.g. C). Infrequently, transducers presented with asymmetries in this subtraction, resulting in incomplete artifact removal in the first microseconds of the ABR waveform. The resulting clamp-tube subtracted R and C averaged waveforms were presented in what is commonly referred to as butterfly plots, which is the overlapping of the two averaged waveforms at the position of the first datapoint; these butterfly plots were arranged with as much overlap in the pre-stimulus tracings as visually achievable. Waves were analyzed with a -12 to 12ms window. While this wide time window was not used in the original recordings, the clinically used IHS software does record this timeframe and the full visualization of the baseline was helpful to fully visualize the pre-stimulus baseline and artifact for overlapping of the waveforms R and C. The first datapoint was defined here as the initial antiphasic crossing of the CM at the point where the CM amplitude exceeded the visualized artifact amplitude. This definition is similar to those presented by Hunter and colleagues (2018), Shi and colleagues (2012), and Starr and colleagues (2001) who used similar 34  visual methods for CM analysis in ABRs. Next, non-phase inverting portions of the waveform were highlighted by combining the R and C average waveforms and dividing by two to mathematically correct for addition of in-phase portions of the waveform (R + C)/2, creating what is commonly referred to as the alternating waveform. Extracting the non-phase inverting portions of the waveform revealed the neural contributions to the ABR waves, such as wave V if present. To provide the best estimate of the CM, R and C polarity average waveforms were subtracted and multiplied by 0.5 to mathematically correct for subtraction of out of phase portions of the waveform: This resulting waveform showing the CM response is the (R – C)/2 waveform. This process of subtracting the R from the C average waveform removed wave V and other non-inverting neural potentials, showing the best estimates of the CM alone. As noted, all waveform averaging, and manipulation was performed in the custom MatLabTM software (SmartEP Reader) created by Dr. Herdman. Amplitude scales were adjusted as needed to reveal an ideal representation of the waveforms for measurement. Waves are presented in appendix A, figures A-1a through A-5, at scales suitable for presentation. A representative example of waveform averaging is presented below in figure 2-1.  35   Figure 2-1. Waveform averaging for a representative 2.25-month-old (corrected) OAE+ infant from 95 dB nHL click stimuli. From top: rarefaction replications (Rare 1, 2), rarefaction average (Rare Ave.), condensation average (Cond Ave.), condensation replications (Cond 1, 2), butterfly plots (R and C) presenting antiphasic averages (Rare Ave., Cond Ave.) with the clamped artifact subtracted, alternating waveform [(R + C)/2 Alt] showing the average of clamp-subtracted R and C averages, CM wave in isolation [(R – C)/2 CM] showing the CM response alone from clamp-subtracted R and C averages, individual clamped tube tracings [R(c), C(c) Clamp]. Amplitude is presented on the y-axis at a 1µV scale. CM-bars above (R and C), and overlaying [(R – C)/2 CM] tracings indicate the measured CM duration and amplitude, respectively. Wave V measurement is shown on the (R + C)/2 tracing as wave V, V’.  2.3.1 CM duration measurement The duration of the CM was defined as the interval from the onset time of the initial phase reversal after the stimulus artifact, to the time at which the phase reversals were no longer 36  visually apparent (Hunter et al., 2018; Shi et al, 2012). Duration of the CM was measured from R and C waves overlaid on the first data point, with as much overlap in the pre-stimulus tracings as visually achievable (Hunter et al., 2018; Hyde et al., 2016; Shi et al., 2012). CM onset was defined where R and C amplitudes were greater than their respective clamped-tube tracings. Subtracting clamped tracings and identifying the onset where waveform amplitudes exceeded the artifact in clamped recordings should have reduced the risk of including artifact in the CM duration measurement. CM endpoint was defined where waveform morphology was no longer clearly antiphasic, and excluded spontaneous amplitude growth late in the waveform, especially where similar artifact was visible in the clamped recording. That is, the CM was not expected to spontaneously grow at a later time-point and was assumed to be artifact or incomplete subtraction of neural phase-delayed waves III-VI of the ABR in such cases. This visual method of analysis may also have overlooked low amplitude oscillations late in the waveform (Santarelli et al., 2006; Shi et al., 2012); however, this method is clinically typical for high-amplitude click ABR assessment (Hyde et al., 2016), and utilizes infant-friendly, non-invasive techniques.   2.3.2 Amplitude measurements and CM/V amplitude ratio calculation Amplitudes of CMs and wave V neural components were calculated within the SmartEP Reader MatLabTM software from (R – C)/2 waveforms created with the clamped-tube subtracted R and C averages. Amplitude was identified as the maximum peak of a waveform’s polarity subtracted from the maximum peak of the following, opposite polarity within the measured duration of the phase-following CM (i.e. peak-to-peak amplitude). Wave V amplitudes were measured from (R + C)/2, clamp-tube subtracted tracings as the most prominent negative deflection (i.e. peak-to-trough amplitude) between 5 and 10ms post stimulus onset (Hyde et al., 37  2016), unless absent from visibly flat waveforms in this latency region; measurement was not attempted in these cases (n=5 ears). Amplitude values presented in nanovolts within the program were converted in Excel 2016TM to microvolts for direct comparison to related literature. Amplitude ratios were calculated as CM amplitude/wave V amplitude in a spreadsheet of compiled CM and wave V amplitude results in Excel 2016TM (described below). If these categorical values for indicating ANSD are accurate, the sensitivity of these measures should be 100 percent in this ANSD dataset: Ratios should be >1.5-2, with CMs of >0.2µV, and wave Vs of <0.1µV (Hyde et al., 2016). Duration and amplitude measurement are presented below in figure 2-2.     38   Figure 2-2. Representative 95 dB nHL click-ABR results from a 2.25-month-old (corrected) OAE+ infant with ANSD. From top: overlapping (R & C) butterfly plots used for duration measures, (R – C)/2 CM wave in isolation for amplitude measures, and (R + C)/2 alternating waveform for wave V measurement. Amplitude is presented on the y-axis at a scale of 0.5 µV. CM-bars indicate CM duration and amplitude measures. Wave V indicates the onset of neural wave V, with V’ indicating the opposite phase maxima for peak-to-trough amplitude determination. The clamped-tube waves (not shown) have been subtracted from these averages.   2.4 Statistical analysis An excel spreadsheet of values was compiled with CM measures (latency onset, latency offset, and amplitude), wave V measures (amplitudes and latencies) and the calculated CM/V ratios (see table A-1, appendix A). Left and right ear data was combined for all groups, as was also done for data presented by Hunter and colleagues (2018). Kurtosis, skew, and mean to median differences calculated in Excel (2016) TM suggested that pooled ANSD data was sufficiently normally distributed for analysis (see table A-2, appendix A). One-way independent-39  samples t-tests were conducted to determine the size of CM duration and amplitude mean differences between infants with ANSD and the normative results presented by Hunter and colleagues (2018). NICU and well-baby infants were compared separately. In their report, Hunter and colleagues (2018) explain that their normative data failed the Shapiro–Wilk test for normality (with alpha level set to p<0.05), indicating that their data did not fit the standard bell-curve distribution; it is not clear what degree of abnormality was present in their data, as no p-value was stated for the results of the Shapiro–Wilk test. The lack of normal distribution in the data, with the application of t-statistics that assume a normal distribution is an unfortunate limitation of the current study; analysis was performed with t-tests because ANOVA and non-parametric median-difference statistical methods require original data points, which were not feasibly accessible for the current project. CM amplitudes and duration measures and CM/V ratios from OAE+ and OAE- infants with ANSD were also explored for the size of their mean differences via two-way independent t-tests.  A Benjamini-Hochberg false discovery rate (FDR) was applied to the results of all t-tests to limit the number false-positives (type-1 errors) reported as significant due to repeated measures of the same group data. Test results with p<0.05 (after FDR correction) were considered significant. Ratio results from the current study were compared to the diagnostic categories described in table 1-1 for all groups with a specific focus on ANSD subgroups of OAE+ and OAE- (Hyde et al., 2016). The CM/V results were not tested for significant mean differences between ANSD and normally-hearing infants. Because the ratio is a categorical tool calculated with both CM amplitude and wave V amplitude, any mean differences would not be indicative of the ratio’s performance. Rather, CM/V assessment in the current study focused on whether the ratio could categorize appropriately across ANSD and normally-hearing infants, 40  without sacrificing statistical power through superfluous t-testing and FDR. All statistical analysis was performed using Microsoft ExcelTM (2016). Probability values and confidence intervals for t-test calculations were retrieved online from: http://www.socscistatistics.com/pvalues/tdistribution.aspx, accessed June 19, 2018.  41  Chapter 3: Results 3.1 CM amplitude, duration, and CM/V amplitude ratio findings Measurable CMs were present in all ANSD ears (n = 24), including all OAE+ (n=17) and OAE- (n=7). Wave V was identified in n=19 ears, including 70 percent of OAE+ (n=12) and all OAE- (n=7) ears. CM onset, offset, and wave V amplitudes were not tested for significant mean differences between any groups, but these values are presented below for descriptive and informational purposes. Wave V latencies are presented in appendix A, table A-1, but not explored here as this manuscript’s focus is on the CM in ANSD. CM durations were found to range from 2.28ms to 6.37ms in infants with ANSD, with a mean value of 4.197 ± 1.154ms. CM amplitudes were found to range from 0.029µV to 0.685µV with a mean of 0.322 ± 0.173µV. The CM/V ratio results ranged from 0.74 to 14.28 with a mean of 4.921 ± 3.313 and are detailed with OAE+/- data in section 3.1.3, below. Individual waveforms and measures are available in appendix A, table A-1, and figures A-1 through A-5. Median values presented in tables 3-1 through 3-3 were calculated for descriptive and informational purposes. The confidence intervals presented below indicate the differences between sample means (ANSD and normally-hearing infant groups, OAE+ and OAE-) that would be expected to be observed if the populations were sampled again.    3.1.1 NICU and ANSD comparisons Table 3-1 presents one-tailed independent samples t-test results comparing ANSD to normally-hearing NICU ears (Hunter et al., 2018). It was found that CM duration was significantly longer in ANSD ears (p<0.00007) both before and after FDR compared to the 42  normally-hearing NICU group. It was found that mean CM amplitude was significantly larger in ANSD than the NICU group before FDR, but not after (p=0.05642).  Infant Group ABR Variable ANSD NICU (Hunter et al., 2018) p-value (1-tail) CI 5th –95th n Mean SD± Median n Mean SD± Median Latency (ms) CM Onset 24 0.345 0.164 0.315 58 0.46 0.09 0.5 - - CM Offset 24 4.542 1.158 4.59 58 1.28 0.50 1.31 - - CM Duration 24 4.197 1.154 4.11 58 0.82 0.51 0.86 <0.00007 * -3.864, -2.893 Amplitude (µV) Wave V 19 0.069 0.039 0.052 59 0.29 0.10 0.29 - - CM 24 0.322 0.173 0.255 58 0.26 0.13 0.28 0.05642 -0.139, 0.016 Table 3-1. Combined ANSD ears (n=24) compared to normally-hearing infants from the neonatal intensive care unit (NICU; n=58). Statistically significant (p<0.05) results are indicated with (*) symbols. Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again.  3.1.2 Well-baby and ANSD comparisons Table 3-2, below, illustrates the one-tailed independent samples t-test results from normal hearing well-baby ears (Hunter et al., 2018) compared to this study’s ANSD infant findings. Statistically significant differences were found before and after FDR for both CM duration and amplitude (p<0.00007, and p=0.0069, respectively) between ANSD and well-baby groups.       43  Infant Group ABR Variable ANSD WB (Hunter et al., 2018) p-value (1-tail) CI 5th –95th n Mean SD± Median n Mean SD± Median Latency (ms) CM Onset 24 0.345 0.164 0.315 53 0.42 0.11 0.47 - - CM Offset 24 4.542 1.158 4.59 53 1.16 0.28 1.07 - - CM Duration 24 4.197 1.154 4.11 53 0.73 0.3 0.7 <0.00007 * -3.943, -2.990 Amplitude (µV) Wave V 19 0.069 0.039 0.052 53 0.28 0.09 0.28 - - CM 24 0.322 0.173 0.255 53 0.24 0.09 0.23 0.0069 * -0.156, -0.007 Table 3-2. Combined ANSD ears (n=24) compared to normally-hearing well-baby (WB; n=53) ears. Statistically significant (p<0.05) results are indicated with (*) symbols.  Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again.  3.1.3 OAE+ and OAE- comparisons and CM/V amplitude ratio results Table 3-3, below, presents two-tailed independent samples t-test results of comparisons between OAE+ and OAE- groups. Neither CM duration nor CM amplitude reached significance (p=0.16107, and p=0.3453, respectively) when comparing between groups. The CM/V ratio was significantly different between groups (p=0.003). These statistical OAE+/- comparisons should be viewed with caution given the limited sample size and non-normal distribution of OAE- ears.         44  Infant Group ABR Variable OAE+ OAE- p-value (2-tail) CI 5th –95th n Mean SD± Median n Mean SD± Median Latency (ms) CM Onset 17 0.305 0.165 0.28 7 0.443 0.124 0.38 - - CM Offset 17 4.275 1.072 4.03 7 5.189 2.63 4.68 - - CM Duration 17 3.971 1.108 3.68 7 4.746 1.157 4.2 0.16107  -1.839, 0.288 Amplitude (µV) Wave V 12 0.0624 0.029 0.049 7 0.080 0.053 0.063 - - CM 17 0.344 0.168 0.277 7 0.269 0.185 0.115 0.3453 -0.093, 0.243 CM/V Ratio 12 6.602 2.987 5.89 7 2.040 1.112 1.83 0.003* 2.538, 6.586 Table 3-3. Independent t-test results of OAE+ infant ears (n=17) compared to OAE- ears (n=7). Statistically significant (p<0.05) results are indicated with (*) symbols.  Values not tested for significance are indicated by (-) symbols. Confidence intervals (CI) indicate the expected differences between sample means if these populations were sampled again.  All but two of 17 OAE+ ears were identified with “definite ANSD” based on the proposed CM amplitude criteria from table 1-1 of >0.2µV; these two ears fell into the “probable” category (0.1 to 0.2µV) as described by Hyde and colleagues (2016). OAE- ears were more variable, with two of seven indicating “definite”, three of seven “probable”, and two of seven ears “not suspected” for ANSD based on CM amplitude criteria alone (not the ratio); these two “not suspected” ears were also misidentified by CM/V ratio results (Hyde et al., 2016).  When wave V was detectable, all but two of 12 OAE+ ears showed “definite ANSD” based on wave V amplitude criteria (<0.1µV); these two ears fell into the “probable” category (0.1 to 0.2µV) according to criteria (Hyde et al., 2016). Results for OAE- ears were similar to OAE+ for this measure. A total of two of seven ears from two infants were identified with a “probable” ANSD component; the remaining five ears were correctly identified as “definite” (Hyde et al., 2016). No ANSD ears were found to be “not suspected” with values >0.2µV; 45  however, these wave V amplitude results should be viewed with caution, as wave V may be small due an underlying or co-occurring SHL resulting from cochlear hair cell damage or dysfunction (i.e. absence of OAEs). As noted above, CM/V ratios differed significantly between OAE+ (n=17 ears) and OAE- (n=7 ears) groups. OAE+ ears all achieved a “definite” ANSD diagnosis (mean CM/V ratio of 6.6 ± 2.9) based on the Hyde and colleagues (2016) criteria with a ratio of >2. OAE- results were less robust (mean CM/V ratio of 2.04 ± 0.27). Three of seven OAE- ears failed to meet the “probable” ANSD diagnostic criterion (ratio of 1.5 to 2); two of these three OAE- ears belonged to the same subject. The same proportion of OAE- ears (43 percent) fell under the “definite” category with a CM/V ratio of >2 (Hyde et al., 2016). Altogether, 16 of 19 ears from infants with ANSD (OAE+ and OAE-) were correctly identified with ANSD based on the CM/V ratio.        46  Chapter 4: Discussion 4.1 CM Duration in ANSD compared to normally-hearing infants The current study found significant mean differences (p<0.00007) in CM duration between infant ears with ANSD from the current data set and normally-hearing well baby ears from Hunter and colleagues (2018), as well as between ANSD and normally-hearing infant ears from the NICU (Hunter et al., 2018). Such a robust finding is believed to have required considerable skew in the samples for the abnormal distributions reported by Hunter and colleagues (2018) to have been a confounding factor. Furthermore, adding three standard deviations to Hunter and colleagues (2018) mean CM durations (0.73 ± 0.3ms for the well-baby group, and 0.82 ± 0.51ms for the NICU group) results in 99.7 percentile durations of 1.63ms for the well-baby group and 2.35ms for the NICU group. Comparing these values to the CM durations from infants with ANSD in the present study (see appendix A, table A-1, CM duration range 2.28ms to 6.37ms) indicates that only one infant with ANSD falls within the 99.7 percent range of the normally-hearing NICU group, while no infants fall within this range of the normally-hearing well-baby group. Crucially, these results were found using similar extratympanic ABR recording techniques, and similar methods for waveform averaging and CM measurement. The theory of a deficient physiological brake presented in figure 1-2 (above) could explain these findings. This theory is supported by related findings from normally-hearing subjects with CNS disorders as described by Santarelli and colleagues (2006), detailed below.  CM durations for ANSD ears found in the current report (4.197 ± 1.154ms) were notably shorter than those described by Santarelli and colleagues (2006) for 20 children and adults with ANSD (6.77 ± 2.58ms). Their application of a more invasive transtympanic methodology may be the source of these differences. Closer proximity to the cochlea results in higher amplitude, and 47  higher fidelity recordings of cochlear potentials by reducing the distance the electrical signal must travel to the recording electrode, as compared to extratympanic ABR (Katz, 2015). Similar to the findings of the current study, Santarelli and colleagues (2006) reported significantly long CM durations in children and adults with ANSD compared to normally-hearing controls. They also reported significantly long CM durations (7.99 ± 2.49ms) in 187 normally-hearing patients with CNS disorders (e.g. vascular/ischaemic disorders, congenital malformations, syndromes) compared to the normal hearing control group (4.36 ± 1.97ms) who presented with no CNS disorders. Reportedly, 80 percent of the subjects with CNS disorders showed lesions involving cortical and/or subcortical neural structures (Santarelli et al., 2006). The researchers suggested that transient interactions with the MOC may cause long CM durations in ANSD due to dys-synchronous neural activity (Santarelli et al., 2006), while the long durations in the group with CNS disorders were suggested to arise from top-down MOC dysfunction due to the subject’s neurological impairments. In these cases, rather than dysfunction in afferent conduction towards higher centers, dysfunction was suggested to arise from dysfunction in the efferent connections themselves, with poor activity from the cortex and MOC to the OHCs. This dysfunction may result in a similar lack of an efferent brake on the OHC gain mechanism, leading to long CMs.   The CM duration findings of the current report more closely resemble those described by Soares and colleagues (2016) in their systematic review, who combined CM duration results from transtympanic and extratympanic methods. These researchers found that CMs may persist for a mean of 4ms after presentation of a transient stimulus; across studies, this duration was found to be significantly longer in ANSD ears than normally-hearing controls. The results of the current study for infants with ANSD (mean CM duration 4.197 ±1.154ms), showing significantly longer CM durations compared to well-baby (mean CM duration 0.73 ± 0.3ms) and NICU (mean 48  CM duration 0.82 ± 0.51ms) groups nicely reflect the findings in this systematic review (Soares et al., 2016). This is a promising indication for the measures’ clinical utility for ANSD diagnosis. These results indicate that neural dys-synchrony in ANSD may produce exploitable differences in the duration of the CM for diagnostic purposes, as suggested by Santarelli and colleagues (2006), and explored in the MOC efferent brake model in figure 1-2. Future prospective studies investigating the sensitivity and specificity of CM duration are warranted based on these findings. Infants with SHL should specifically be included in such a study, as ANSD and severe to profound SHL populations may present more similarly to each other in high-amplitude ABR assessments due to diminished or absent wave Vs and possibly absent OAEs. Furthermore, ANSD and SHL groups share risk factors for hearing loss, and the two pathologies have been reported to co-occur in the same ear (Hyde et al., 2016). Considering the OHC efferent brake model, if deficient efferent suppression of the OHCs is responsible for long duration CMs, then OHC lesions in SHL may preclude this ringing. That is, without numerous OHCs to act and react on the basilar membrane, CM durations may be significantly shorter in infants with SHL.  Turning to other comparable literature on the CM, Shi and colleagues (2012) did not report duration values; however, they noted that phase-reversing components such as the CM visibly persisted for several milliseconds after a transient stimulus. They also reported that CMs could only be assessed before the onset of ABR wave I in normal hearing infants due to the presence of neural wave I and the presence of latency shifted neural components late in the (R – C) waveform, which obscured the genuine CM response (see figure 1-1, above). Due to this latency shift in neural activation between R and C polarities, the subtraction of neural waves late in the waveform was incomplete, mimicking a long CM in normally-hearing ears. Similarly, Starr and colleagues (2001), found that CMs could not be evaluated beyond 0.7ms in normal 49  hearing controls. It is unclear how Hunter and colleagues (2018) were able to assess the CM beyond this latency range, or if this was an issue in duration measures in their study. Table 4-1 summarizes the discussed results from research on the CM presented in this report, below.  Cited Literature  Study  Method Subjects Summary of Relevant Findings Starr et al., (2001) CMs elicited by 110 dB peSPL, 19/s R&C, ABR clicks through ER3A inserts. Tests included clamp trials. (R–C) was used to measure CM amplitude.  Group A: 33 ANSD subjects (57 ears) aged 4 months to 64 years. Group B: 21 healthy controls (28 ears) aged 0.25 months to 45 years.  Duration: CMs could not be evaluated after 0.7ms in normal ears but persisted for several milliseconds after a transient stimulus in ANSD ears.   Amplitude: For ANSD CMs were 0.42 ± 0.29μV.  Maximum was found at 0.4 to 0.6ms for 50/57 ANSD ears, and all normal ears. CMs were considered normal if within 2 standard errors of the mean. CMs were abnormally large in 21 ears under age 10, and overall larger in ANSD. No correlation was found between degree of hearing loss and CM amplitude.  Santarelli et al., (2006) CMs from R & C clicks via soundfield and transtympanic testing including clamped trials. (R–C) was used for amplitude, oscillation, and duration measures.   Group A: 187 normal-hearing subjects with a CNS pathology (CNS+). Group B: 315 controls with no disorder (CNS-). Group C: 20 OAE+ subjects with ANSD. Groups aged 7 months to 47 years.  Duration: Identified via rapid increases in instantaneous frequency in Fourier analysis. ANSD = 6.77 ± 2.58ms, with CNS+ (7.99 ± 2.49ms) and ANSD both significantly longer than CNS- (4.36 ± 1.97ms).   Amplitude: ANSD = 13.5 ± 26.8μV. No significant difference between ANSD and CNS-. CNS+ results were significantly larger than ANSD and CNS- ears.   Both amplitude and duration showed a high degree of variability in ANSD.  Shi et al., (2012) 70-100 dB nHL clicks recorded from ABR using split alternating methods, and clamped tube trials. Compiled CM amplitude input/output functions based on increases from 70 dB nHL to 100 dB nHL in 10 dB increments. Group A: 15 children with OAE+ ANSD (30 ears) aged 3 months to 7 years. Group B: 21 children with OAE- ANSD (30 ears) aged 3 months to 9 years Control: 15 normal hearing infants (30 ears) aged 2 to 12 months.  Duration: Clearest CMs were found from 0.5 to 0.8ms. No significant differences in CM onset between groups. “[CMs] lasted a few milliseconds in (C-R)/2 waves of subjects with ANSD and appeared only before wave I in controls.”  Amplitude: No significant differences between OAE+ and controls. OAE- CMs were significantly smaller than OAE+ and controls. Input/output values (below) showed non-linear increases in OAE+ and controls that were linear in OAE-.   Group CM amplitude means (μV) Stimulation Amplitude (dB nHL) 70 dB nHL 80 dB nHL 90 dB nHL 100 dB nHL OAE+ 0.13 ± 0.05 0.26 ± 0.08 0.4 ± 0.13 0.47 ± 0.15 OAE- 0.04 ± 0.03 0.1 ± 0.06 0.18 ± 0.07 0.24 ± 0.08 Controls 0.14 ± 0.06 0.28 ± 0.11 0.41 ± 0.13 0.45 ± 0.13  50   Cited Literature  Study  Method Subjects Summary of Relevant Findings Hunter et al., (2018) 70 dB nHL 37.1/s clicks using ABR and “split alternating” protocol, incl. clamps. CM onset was identified from (R-C)/2 waves where residual noise was flat compared to CM. A whole-wave correlation of 0.6 required for inclusion.  Group A: 30 infants (53 ears) from the well-baby (WB) nursery. Group B: 32 high-risk infants (59 ears) from the NICU all with normal hearing. 75% of NICU infants showed high-bilirubin (bilirubin counts not defined).  Group were roughly 1.25 months of age, corrected. Duration: No significant differences between WB (0.73 ± 0.3ms) and NICU (0.82 ± 0.51ms) infants.   Amplitude: No significant differences between WB (0.24 ± 0.09μV) and NICU (0.26 ± 0.13μV) infants.    ANSD case study: OAEs present at 3 months, absent at 8 months.  Duration: 1.82ms left, 1.49ms right – abnormally long at 3 months. 1.22ms left, and 0.99ms right at 8 months, within NICU 5th – 95th.  Amplitude: 0.20 μV and 0.14μV, left and right ears, respectively – within normal range at 3 months using 70dB nHL. 0.4μV and 0.5μV left and right respectively - significantly larger at 8 months using 80 dB nHL. Table 4-1. Summary of referenced research studies of the cochlear microphonic  4.2 CM amplitude in ANSD compared to normally-hearing infants Significant mean amplitude differences (p=0.0069) were found between the current study’s ANSD infant subjects and the normally-hearing well-baby infant data presented by Hunter and colleagues (2018). After application of the FDR, CM amplitudes were not significantly different between NICU and ANSD infants (p=0.05642); these non-significant results could be a statistical power issue, as results were significant prior to FDR. That is, the application of the FDR to limit the number false-positives (type-1 errors) may have inadvertently produced a missed true-positive result (type-2 error). Other confounds between data from the current study and values presented by Hunter and colleagues (2018) must be addressed as well and are considered below.    51  Firstly, the data supplied by Hunter and colleagues (2018) was reported to be not normally distributed; this is cause for skepticism regarding any nearly significant results where a normal distribution is assumed in the statistical method. Secondly, it is possible that different stimulus presentation levels between studies could also have led to the significantly different CM amplitude results described above. Hunter and colleagues (2018) presented 70 dB nHL click stimuli, whereas this study examined responses to 90 dB nHL or higher (mean 94 dB nHL). It is unclear if these differences may impact CM duration measures in normal ears, but higher stimulus presentation levels have been shown to produce larger CM amplitude waveforms. For example, Shi and colleagues (2012) demonstrated this increase in CM amplitude with increases in stimulus presentation level. They plotted input/output functions of CM amplitude by stimulation intensity. In their report these researchers describe non-linear increases in CM amplitude in normally-hearing controls from 0.14 ± 0.06µV at 70 dB nHL, to 0.45 ± 0.13 at 100 dB nHL. A similar non-linear CM amplitude growth with increasing click stimulus levels was also found in OAE+ ears with ANSD (Shi et al., 2012). This non-linear growth was not significantly different across OAE+ ears with ANSD and normally-hearing control groups. OAE- subjects with ANSD showed significantly smaller, and more linear CM response-amplitude increases, from 0.04 ± 0.03µV at 70 dB nHL to 0.24 ± 0.08 at 100 dB nHL but CM amplitude increases nonetheless. CM response amplitudes for infants with ANSD in the current study, elicited by 90-to-100 dB nHL clicks (mean 94 dB nHL) are similar to the 90-to-100 dB nHL results reported in Shi and colleagues (2012). These results are compared in detail in the following section examining results from OAE+ and OAE- infants with ANSD. Likewise, Hunter and colleagues’ (2018) reported CM response amplitudes of 0.24 ± 0.09μV for normal-hearing well-baby infants, and 0.26 ± 0.13μV for normally-hearing infants from the NICU fit 52  within the 70-to-80 dB nHL range of CM response amplitudes (0.14 ± 0.06μV to 0.28 ± 0.11μV, respectively) reported by Shi and colleagues (2012). These results from Shi and colleagues (2012) are tabulated in figure 4-1, above. Altogether, the above comparison of CM response amplitudes implies that the significant differences found between well-baby and ANSD infants in the current study may not be significantly different if testing were done at equal click stimulation intensities. The application of a higher click stimulus intensity in normal ears would likely result in increased well-baby CM response amplitudes, and this may result in a non-significant difference in mean CM amplitude values between infants with ANSD and normally-hearing well-babies. That is, equal stimulation intensities, resulting in increased well-baby CM response amplitudes may not differ significantly from the current ANSD results, and may in fact be larger than the reported ANSD CM response amplitudes reported in the current study. The normal hearing CM amplitudes from Shi and colleagues (2012) of 0.41 ± 0.13μV from 90 dB nHL stimuli are larger than those found for the infants with ANSD in the current study of 0.322 ± 0.173μV from a mean of 94 dB nHL. Regarding stimulation intensities in the current study, it is unclear why stimulation intensities do not adhere to the BCEHP protocols, which state that high-intensity click-ABR assessment should be performed at 95 dB nHL (Bremner et al., 2012). Since the date range of ABR data included for analysis spans the years of 2008 through 2015, it is possible that these ABRs were performed prior to the 2012 release date of the protocol. It may also be possible that individual clinical judgement factored in to a higher or lower stimulation intensity. Higher intensity could have been selected for if no response was obtained at 95 dB nHL, or if a low amplitude response did present and the clinician hoped to see this response grow with increased intensity, as may be done at lower intensities for frequency-specific tone-ABR testing. Clinicians 53  may have applied lower click stimulation intensity than 95 dB nHL for concerns of over stimulation and potential noise damage to ears with present OAEs. In the protocol, the application of high amplitude click stimulation (95 dB nHL) is restricted to rare occasions where ANSD must be ruled out as a contributing factor for hearing loss for these very reasons (Brenmer et al., 2012). High-intensity stimulation is restricted to these occasions to avoid risk of undue noise damage to healthy ears with intact OHCs. It is possible this same concern that prompted the use of 70 dB nHL clicks in the report by Hunter and colleagues (2018); however, this is only speculation.       Turning briefly to a consideration of CM/V amplitude ratios from the normally hearing infant data presented by Hunter and colleagues (2018), group means both fell within the “not suspected” category (Hyde et al., 2016). The group mean CM/V ratio for normally-hearing infants from the NICU was found to be 0.897 ± 0.11, while the mean for the normally hearing well-baby group was found to be 0.857 ± 0.09. Group means cannot indicate how the CM/V would have performed on these infants individually; however, the standard deviations from the mean suggest that neither group is at risk of reaching the >1.5 criterion for “probable ANSD” (Hyde et al., 2016). Results of the CM/V amplitude ratio are explored in detail for ANSD ears from the present study in the following section.   4.3 CM duration, amplitude, and CM/V amplitude ratios in OAE+ and OAE- infants For the ANSD data analyzed in the current study, no significant differences in CM duration were found between OAE+ and OAE- ears with ANSD. It appears that ringing CMs may present regardless of the source of dys-synchrony or possible degree of co-occurring SHL, assuming that the mechanism of dys-synchrony in OAE+ and OAE- groups arise from different 54  sources. Because CM duration was significantly longer in the ANSD population than in normally-hearing infant ears, regardless of OAE status, CM duration might be a useful measure for ANSD diagnosis even in OAE- infants suspected of having ANSD. While this is a preliminary study on the infant CM, an in-depth analysis of sensitivity and specificity could indicate an ideal cut-off duration for normal vs abnormal ears in the future. Ideally such an analysis would occur in the presence of more corroborating research to support the current study’s results, along with comparisons to infants with SHL.  Turning to other literature concerning CM amplitude measures, this study’s findings were close to those reported by Shi and colleagues (2012) comparisons of OAE+, and OAE- subjects.  Mean OAE+ amplitudes of 0.344 ± 0.168μV in the current study obtained at a mean of 94 dB nHL were lower than their 90 dB nHL results (0.4 ± 0.13μV), but notably higher than their 80 dB nHL results (0.26 ± 0.08 μV). Mean OAE- results of the current study (0.269 ± 0.185μV) from the mean of 94 dB nHL click stimuli were higher than their reported 100 dB nHL (0.24 ± 0.08μV) values (Shi et al., 2012). Standard deviations of the CM amplitudes in the current study approximated Shi and colleagues (2012) results for OAE+ but were notably larger for OAE-. It can only be speculated why OAE- results of the current study were more variable and visibly larger in CM amplitude; however, it may be possible that the degree of hair cell dysfunction or cochlear lesion in the present study was more extreme or more variable than in the OAE- ANSD sample presented by Shi and colleagues (2012). The current study’s standard deviations for CM amplitude were also larger than those reported by Hunter and colleagues (2018), suggesting more variability in the current study’s data, which is to be expected because Hunter and colleagues (2018) had a waveform correlation criterion for inclusion. These researchers implemented such a requirement for inclusion of infant data to their study in an attempt to equalize electrical noise 55  across waveforms, described as a whole-waveform correlation coefficient of 60 percent or greater (Hunter et al., 2018). No such controls for electrical noise were implemented here aside from the subtraction of average clamped-artifacts from the waveforms. Further, such abnormal waveforms are typical to the ANSD population, which are reported to present with irregular and difficult to interpret ABR waveforms (Hyde et al., 2016; Starr et al., 2001; Berlin et al., 2010). Imposing a waveform correlation criterion in the current study would not be indicative of real-world clinical results (Santarelli et al., 2006; Hyde et al., 2016). Despite the larger standard deviations, mean OAE- CM amplitudes were within 0.02μV of OAE+ findings in the current study, which is consistent with the differences found between normally hearing NICU and well-baby populations reported by Hunter and colleagues (2018). Therefore, CM amplitudes for OAE+ and OAE- infants might be considered similar from a clinical perspective, at least for the purposes of classifying them as the same population. As with CM duration, it appears that measures of CM amplitude may present similarly regardless OAE status; however, this may depend on the level of hair cell dysfunction, as indicated by the differences found between the current studies results for OAE- CM amplitude and those reported by Shi and colleagues (2012), as discussed above.  Mean values for the CM/V amplitude ratios were found to differ significantly between OAE+ and OAE- groups. While mean values for both groups indicated definite ANSD, significant differences on this measure suggest that OAE+/- might not be considered similar from a clinical population perspective as indicated by CM duration, and by CM amplitude alone. For this study’s ANSD data, 16 of 19 ears with ANSD were correctly identified with ANSD under the categorical diagnostic structure presented by Hyde and colleagues (2016). This includes all OAE+ ears with an identifiable wave V (12 ears, 10 infants), and four OAE- ears 56  from four infants with ANSD. Three OAE- ears from two infants were misidentified as “not suspected” for ANSD due to CM/V amplitude ratios falling of <1.5. Interestingly, all OAE- ears with ANSD presented with identifiable wave Vs. Since five of the 17 OAE+ infants presented without an identifiable wave V, it seems plausible that these different physiological indicators may arise from different sites-of-lesion in ANSD. Criteria such as OAE+/- and wave V+/- may be useful to track in the future towards identifying which indicators do well with a given intervention (e.g. cochlear implantation for OAE- cases). It is important to note that between groups comparisons of OAE+ and OAE- should be viewed with caution given the restricted sample size of the OAE- group. The inclusion of 2 ears stimulated with high rates of presentation (61.1/s) in this group is further cause for skepticism; however, this rate is believed to only have impacted the neural components of the ABR, and not measures of the CM. High stimulation rates are known to diminish neural waveform amplitudes as a function of the periods of time necessary for individual nerve fibers repolarize (i.e. refractory periods). In contrast, phase-following CM potentials have been shown to persist far above the upper firing rate limits of nerve fibers (Rance & Starr, 2015). It is believed that this high rate (61.1/s) should have led to a smaller denominator (wave V) with approximately equal numerator (CM) compared to a lower rate of stimulation, producing a larger CM/V ratio; the finding that two of the three misidentified ANSD ears were stimulated by this rate was unexpected. This infant was also misidentified by raw CM amplitude criteria as not suspected. Crucially, this infant was diagnosed with SHL based on the source ABR for the click waveform files analyzed in the current study. This SHL diagnosis suggests that any middle-ear complications had been accounted for in the diagnosis, though middle-ear dysfunction may still be playing a role in these findings by possibly reducing the amplitude for the CM response. This infant received a diagnosis of ANSD from an ABR at 57  13.2 months of age (corrected). This follow-up test showed markedly clearer CMs; however, the time of this second assessment fell beyond the target age for this study. CM durations for this infant were measured as 4.2ms in the left ear, and 5.38ms in the right, both well beyond the NICU 99.7 percentile range of up to 2.35ms, indicating that CM duration may have been an effective indicator of ANSD in this infant despite the CM/V and CM amplitude results. It is also possible that CM morphology changed in this infant in the time between ABR assessments due to increasing myelination of the brainstem or some related developmental mechanism. Such development could lead to greater hyperpolarizing action of the efferent system, as proposed by Starr and colleagues (2001) and explored in figure 2-1, despite what dys-synchrony may be present due to ANSD. For example, Hunter and colleagues (2018) presented a case study of an infant with ANSD in which CM durations shortened from abnormally long at 3 months of age to within the 95 percent confidence intervals for NICU infants at 8 months, possibly due to improved braking action of the efferent system. This infant’s CM amplitudes increased significantly in this timeframe, from within the normal-hearing NICU range at 3 months to above the normal-hearing NICU range at 8 months. These results are similar to those of the misidentified infant from the current study, though the researchers noted that a 10 dB nHL higher stimulus was used at 8 months than had been used for the initial assessment at 3 months of age. Interestingly, OAEs were present at 3 months and absent at 8 months in Hunter and colleagues (2018) ANSD case study, despite larger CM amplitudes at 8 months. This finding may further suggest that the presence or amplitude of OAEs and the amplitude of CM responses are not directly correlated, possibly due to CM amplitude contributions from the IHCs.   The remaining OAE- ear of the three misidentified by CM/V ratio in the current study arose from an infant with unilateral ANSD. ABR responses for this infant showed a present CM 58  and a poorly replicating but consistent wave V in the recordings. With regards to the CM/V ratio in this particular case, the presence of a robust wave V in the denominator led to the infant’s misidentification. This infant may be comparable to the 20 percent of subjects reported on by Starr and colleagues in their 2001 analysis, who showed an irregular but repeatable wave V. As suggested by Santarelli and colleagues (2006), this different pattern of results may arise from a different site-of-lesion than those identified correctly by the ratio. As noted above, clinical documentation of OAE and wave V status in ANSD may provide further insight into what interventions and outcomes tend to perform best in subcategories of ANSD (e.g. OAE+/-, wave V+/-) in the future. Interestingly, this infant’s CM duration was found to be 5.9ms, which is well beyond the NICU 99.7 percentile range of up to 2.35ms, suggesting again that CM duration may be a reliable indicator for ANSD in infants.   4.4 Strengths and limitations of the current study The major strength of this study is the age range of the participants (<12 months old), which resembles the target population that is most dependent on click-ABR measures for ANSD diagnosis. Previous research had limited inclusion of infant data, and thus had limited validity in generalizing to the infant population. This is problematic, because age has been shown to impact physiological measures, as described above. Because early infant diagnosis of ANSD relies on physiological measures, providing clear evidence of electrophysiological differences between ANSD and normal-hearing infants is highly important. Reliable physiological indicators of ANSD may enable earlier intervention and habilitation techniques towards possibly improving hearing and language outcomes in infants with ANSD. This study also benefited from the 59  availability of later behavioral and supporting data for diagnosis of ANSD in OAE- ears, allowing for more confidence in this complex diagnosis.  Due to limited sample size, lack of normal distribution in the comparison data (Hunter et al., 2018), and the retrospective nature of the analysis, results of this study should be viewed with caution. Despite representing approximately seven years of data collection, the study still suffered from having a small sample size for all groups, and especially for the OAE- ears. This low sample size is due to the low prevalence of ANSD, and due to data being collected from only the BCEHP. This analysis was also subject to potential confounds due to variations in clinical settings, recording parameters, infant temperament, sedation, and comorbidities, which were not be proactively controlled for. All these factors could have confounded the results. Furthermore, as discussed in previous sections of this manuscript comparing CM amplitude across ANSD and normally-hearing ears, stimulation intensities differed between normally-hearing and ANSD ABR recordings. This stimulation intensity difference may have resulted in different CM responses being elicited; however, the high intensity stimuli for ANSD assessment carry the risk for noise exposure in healthy cochlea, and would likely wake a normally-hearing infant, making assessment with equal stimulation intensities across these groups impractical.  Considering subject data, instances of missing diagnostic results was a limitation in this study as well. Charts for ANSD subjects were not always complete, so varying middle-ear status may have played a confounding role in waveform presentation. Any data with suspected middle-ear involvement was excluded; however, this information was not reliably available for all ears at the time of click-ABR recording, including 10 ears from 7 infants with no indications of middle-ear status in the form of scanned tympanograms or written clinical reports. It is possible that the presence of a conductive hearing impairment at the time of the ABR recording may have 60  impacted measurements of the CM. Similarly, even a mild SHL may have impaired OHC function in the OAE- group, resulting in smaller CMs due to a reduction in the number hair cells contributing to the response. For this reason, combining the OAE+ and OAE- groups with ANSD together in the comparisons to the normally-hearing infant groups was a limitation. It is possible that an analysis with the OAE+ group may have resulted in significantly larger CM amplitudes than both well-baby and NICU groups, as opposed to just the well-baby group as reported above with the OAE+ and OAE- groups combined. Across subjects, behavioral auditory thresholds and configuration of hearing thresholds could not be adequately equalized across ears, because ear-specific threshold data were not consistently available. It is possible that different hearing thresholds due to different degrees of SHL and cochlear dysfunction may have resulted in different CM responses. Furthermore, different configurations of hearing losses, such as a sloping high-frequency loss may have delayed the onset of the CM by requiring the sound stimulus to travel further within the cochlea before activating hair cells for the production of a CM response. Regarding other missing subject data, variable clamped-tube recording procedures made the inversion of R and C polarities necessary, which also impacted the consistency of the study. Similarly, significant amounts of data were not accepted for analysis due to a lack of clamped-tube recordings. Clamped tube trials were often performed for only one ear, despite sound transduction occurring from two transducers that may produce different electrical noise. Infants reported to have cochlear nerve deficiency were excluded from analysis as complete agenesis of the nerve could not be ruled out with the available chart data. As with the absent middle-ear assessment data, it could also not be guaranteed that the chart data for the included infants had complete imaging results, which would have been the gold-standard diagnostic cochlear nerve 61  deficiency. It remains possible that the included clinical reports did not mention imaging results, or that clinical reports mentioning such results were themselves absent from the chart data, or that a diagnosis of cochlear nerve deficiency had not been received during the span of years of the data included in this study. Ultimately, it cannot be guaranteed that all infants included in this study did not have cochlear nerve deficiency, which is a further limitation of this retrospective analysis.    Potential maturational effects on CM measures and/or neural ABR components were not quantified in any manner nor were these variables controlled for within this study. While all infants with ANSD were less than 12 months of age (mean 3.5 months corrected), the control group of normal-hearing infants from the Hunter et al. (2018) study had an average age of 1.25 months, corrected. Maturation effects such as myelination and cortical development from birth to the infant’s age at testing may have been different across groups. For instance, wave V amplitudes may increase, and latencies may shift earlier with brainstem maturation and developing myelination (Katz, 2015). While wave V measures in the current study did include latency values, the results here are limited in that later ABRs were not formally assessed, so no comparisons can made between earlier and later ABR results that could help verify if similar changes in wave V occur in infants with ANSD. As the focus of this study was the CM in ANSD, such an assessment was beyond the scope of the current paper but could make for interesting research in the future. CM amplitudes have been shown to increase up to five years of age (Santarelli et al., 2006), and Hunter and colleagues (2018) reported significant CM duration and amplitude changes in an infant with ANSD between the ages of 3 months and 8 months, as described above. As with wave V metrics, CMs were not measured in later ABRs of infants with ANSD in the current study in the current study. As such, it is not entirely clear what impact this 62  development would have on CM duration, but maturation of the efferent connections to the OHCs from the MOC may increase the efficacy of the proposed physiological braking mechanism discussed in figure 1-2. In theory, a more effective brake may lead to shorter duration CMs with increasing age due to brainstem maturation. Thus, the differences in CM durations between normally-hearing and ANSD infants could be even greater (and more significant) if testing of normally-hearing infants was conducted closer to 3.5 months of age. Likewise, physical growth of the ear canal may have shifted infant’s hearing thresholds, requiring more physical sound pressure for equal cochlear responses to take place. This physical growth could lead to smaller amplitude CM and wave V responses by increasing the volume in which sound pressure must act within the developing infant ear canal. Ultimately, maturation might have been a confounding variable here and should be investigated or controlled for in future studies.  Finally, treating a spectrum disorder like ANSD as a single disorder, as opposed to a spectrum of complications resulting in similar diagnostic results may be limiting. As suggested by Santarelli and colleagues (2006), different sites-of-lesion may produce systematically different waveforms as elicited by click stimuli. It is possible that until definitive diagnostic criteria are determined for the multiple possible sites-of-lesion in ANSD, grouping the diagnostic results indicative of the disorder may be the best option; however, all attempts should be made to disentangle such differences in the ANSD populations in the future.   4.5 Potential applications of the current findings The results of the current study are early steps towards understanding how we can clinically use the CM response to aid in ANSD diagnoses. The significant differences in CM 63  duration between ANSD and normal-hearing NICU and well-baby infants, indicate that this metric could be useful for early and reliable diagnosis of ANSD. Finding no significant differences between OAE+/- ears further indicates that the measure may be sensitive to ANSD regardless of OAE status. This is a critical finding, as there is need for early diagnostic indicators that can inform management decisions in the first months of an infant’s life. The ability to reliably identify ANSD in infants should allow for earlier and more appropriate interventions to be put in place. In complex cases of ANSD, delaying or cautiously implementing interventions until behavioural auditory assessments become developmentally appropriate may negatively impact speech and language outcomes; avoiding this delay has been the central goal of early hearing programs (Bremner et al., 2012; Hyde et al., 2016). Earlier intervention can enable improved speech and language outcomes in children, based on better informed clinician decisions (Yoshinaga-Itano et al., 1998). For instance, decisions regarding the level of amplification to provide an infant showing absent or abnormal ABR could be better informed by reliable indicators that ANSD is present. Better informed clinicians can better prepare the parents of these infants for what habilitation techniques may be warranted. These may include traditional amplification and what limitations amplification may have (Sharma & Cardon, 2015). Interventions may also include earlier consideration of cochlear implantation (Teagle et al., 2010), or earlier implementation of manually signed language.  Regarding the efficacy of the CM/V categories, it is possible that a CM/V ratio of different specifications may add more sensitivity to identifying OAE- ears with ANSD, while remaining suitably specific to detecting ANSD. As Starr and colleagues (2001) have suggested, levels of ANSD severity may be graded. It is possible that different levels of severity may be identifiable through these measures with more discerning criteria. As it currently stands, CM 64  amplitude criteria and CM/V ratios were accurate in identifying 21 of 24 ears with ANSD and found the group means of normally-hearing ears to be “not suspicious” for the disorder. These findings could be beneficial to clinicians struggling with ANSD diagnoses in difficult cases when using non-invasive, and widely available tools and techniques.   4.6 Future research directions Targets of future research could include an analysis of sensitivity and specificity of CM duration and CM/V ratios for discriminating between ANSD and normally-hearing NICU and well-baby ears. Ideally such an analysis would occur in the presence of more corroborating research showing long CM durations in ANSD using non-invasive ABR to support the current results. Prospective research designs are warranted based on the results obtained here. Beyond normally-hearing ears, future studies comparing CM duration, amplitude, and CM/V ratios should include SHL infant populations to determine which measures could best differentiate from infants with ANSD. With the CM/V criteria, it was found that CM/V ratios were able to accurately categorize 16 of 19 ears with ANSD and an identifiable wave V as well as the group means of normally-hearing ears. More research is needed to determine whether or not CM/V ratios can correctly categorize infants with ANSD from infants with SHL in the presence of similarly diminished neural ABR waves. Even with extensive SHL, regions of the cochlea may have preserved IHCs or OHCs, and recordable CMs. For example, a profound high frequency hearing loss may retain good cochlear hair cell function in the low frequencies. It is unclear what degree of cochlear hair cell loss would be required to diminish CMs to the point that CM amplitudes can reliably separate SHL and ANSD populations, or how extensive this loss must be to reduce the CM to undetectable levels for duration measures. It is also possible that CM 65  durations in infants with SHL may fall within the normally-hearing range despite reduced CM response amplitudes. Without intact OHC function, the basilar membrane would may not be amplified with the same feed-forward pattern proposed in figure 1-2, above. Further research with the inclusion of an SHL group is necessary to determine the true efficacy of these metrics moving forwards. Other future research could include the creation of an age-normative dataset and documentation of CM morphology in typically developing infants with normal hearing. Such a dataset would allow for more age appropriate comparisons between ANSD and normally-hearing infants, and to infants with SHL. As suggested above, cochlear and neural responses could also vary depending on levels of neural synchrony, which may differ across presynaptic-IHC, dendritic, spiral ganglion, and auditory nerve sites of lesion. For instance, more extensive demyelination of the auditory nerve may increase the level of dys-synchrony present in the neural system. Currently available options for identifying ANSD subtypes based on site-of-lesion may include categorizing OAE+ and OAE- infants separately, with considerations for wave V presence and absence. Monitoring these physiologically measurable differences in infants with ANSD may help in identifying best practices for interventions in these infants moving forwards. For example, infants presenting with absent OAEs and a diminished but present wave V may make ideal candidates for cochlear implantation if the site-of-lesion indeed lies within the cochlea as indicated by OAE absence. By tracking these factors, clinicians might be better able to prognosticate which interventions (amplification, cochlear implantation, sign language) would be best for each ANSD subcategory (e.g. OAE+, with absent wave V, or OAE- with present wave V).  In conclusion, our understanding of ANSD is improving but we still have much to learn. The research presented here identified that CM duration and CM/V amplitude ratio are metrics 66  that can be used to diagnose ANSD in infants less than 12 months of age. Such a tool has the potential to help clinicians make early and confident decisions regarding interventions and could help improve later speech and language outcomes for infants with ANSD.  67  Bibliography Berlin, C. I., Hood, L. J., Morlet, T., & Wilensky, D. (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; 30-43; 43.  Berlin, C. I. (1998). Auditory neuropathy. Current Opinion in Otolaryngology & Head and Neck Surgery, 6(5), 325-329. doi:10.1097/00020840-199810000-00008 Bremner, D., Davies, D., Hatton, J., Hyde, M., Ishida, I., Janssen, R., Van Maanen, A. (2012). Audiology assessment protocol. Paper presented at the BC Early Hearing Program, 4.1. Retrieved from http://www.phsa.ca/Documents/bcehpaudiologyassessmentprotocol.pdf, May 2018. Elgoyhen, A. B., & Katz, E. (2012). The efferent medial olivocochlear-hair cell synapse. Journal of Physiology, Paris, 106(1-2), 47.  Hunter, L. L., Blankenship, C. M., Gunter, R. G., Keefe, D. H., Feeney, M. P., Brown, D. K., & Baroch, K. (2018). Cochlear microphonic and summating potential responses from click-evoked auditory brain stem responses in high-risk and normal infants. Journal of the American Academy of Audiology, 29(5), 427. doi: https://doi.org/10.3766/jaaa.17085 Hyde, M., Bagatto, M., Martin, V., Pigeon, M., Scollie, S., & Witte, J. (2016). Protocol for auditory brainstem response – based audiological assessment (ABRA). Ontario Ministry of Children and Youth Services Ontario Infant Hearing Program, 2. Retrieved from https://www.mountsinai.on.ca/care/infant-hearing-program/documents/protocol-for-auditory-brainstem-response-2013-based-audiological-assessement-abra, May, 2018. Katz, J. (Ed.). (2015). Handbook of clinical audiology (7th ed ed.). Philadelphia, PA: Wolkers Kluwer Health. 68  Kujawa, S. G., & Liberman, M. C. (2009). Adding insult to injury: Cochlear nerve degeneration after "temporary" noise-induced hearing loss. Journal of Neuroscience, 29(45), 14077. doi:10.1523/JNEUROSCI.2845-09.2009 Lichtenhan, J. T., Wilson, U. S., Hancock, K. E., & Guinan, J. J. (2016). Medial olivocochlear efferent reflex inhibition of human cochlear nerve responses. Hearing Research, 333, 216-224. doi: 10.1016/j.heares.2015.09.001 Michalewski, H. J., Starr, A., Nguyen, T. T., Kong, Y., & Zeng, F. (2005). Auditory temporal processes in normal-hearing individuals and in patients with auditory neuropathy. Clinical Neurophysiology, 116(3), 669-680. doi: 10.1016/j.clinph.2004.09.027 Moser, T., & Starr, A. (2016). Auditory neuropathy-neural and synaptic mechanisms. Nature Reviews.Neurology, 12(3), 135. doi:10.1038/nrneurol.2016.10 Narne, V. K., Prabhu, P., Chandan, H. S., & Deepthi, M. (2014). Audiological profiling of 198 individuals with auditory neuropathy spectrum disorder. Hearing, Balance and Communication, 12(3), 112-120. doi:10.3109/21695717.2014.938481 Prabhu, P., Narne, V., & Barman, A. (2014). Long ringing cochlear microphonics - not unique to auditory neuropathy spectrum disorder in children. J Otol Rhinol, 3(6). doi:10.4172/2324-8785.1000196 Rance, G., & Starr, A. (2015). Pathophysiological mechanisms and functional hearing consequences of auditory neuropathy. Brain, 138(Pt 11), 3141-3158. doi:10.1093/brain/awv270 Riazi, M., & Ferraro, J. A. (2008). Observations on mastoid versus ear canal recorded cochlear microphonic in newborns and adults. Journal of the American Academy of Audiology, 19(1), 46-55. doi:10.3766/jaaa.19.1.5 69  Rodríguez Domínguez, F. J., Cubillana Herrero, J. D., Cañizares Gallardo, N., & Pérez Aguilera, R. (2007). Prevalence of auditory neuropathy: Prospective study in a tertiary-care center. Acta Otorrinolaringológica Española, 58(6), 239.  Santarelli, R., Scimemi, P., Dal Monte, E., & Arslan, E. (2006). Cochlear microphonic potential recorded by transtympanic electrocochleography in normally-hearing and hearing-impaired ears. Acta Otorhinolaryngologica Italica : Organo Ufficiale Della Società Italiana Di Otorinolaringologia E Chirurgia Cervico-Facciale, 26(2), 78.  Santarelli, R., Del Castillo, I., Rodríguez-Ballesteros, M., Scimemi, P., Cama, E., Arslan, E., & Starr, A. (2009). Abnormal cochlear potentials from deaf patients with mutations in the otoferlin gene. Journal of the Association for Research in Otolaryngology: JARO, 10(4), 545-556. doi:10.1007/s10162-009-0181-z Sharma, A., & Cardon, G. (2015). Cortical development and neuroplasticity in auditory neuropathy spectrum disorder. Hearing Research, 330(Pt B), 221-232. doi: 10.1016/j.heares.2015.06.001 Shi, W., Ji, F., Lan, L., Liang, S., Ding, H., Wang, H., Wang, Q. (2012). Characteristics of cochlear microphonics in infants and young children with auditory neuropathy. Acta Oto-Laryngologica, 132(2), 188-196. doi:10.3109/00016489.2011.630016 Soares, I. d. A., Menezes, P. d. L., Carnaúba, A. T. L., de Andrade, K. C. L., & Lins, O. G. (2016). Study of cochlear microphonic potentials in auditory neuropathy. Brazilian Journal of Otorhinolaryngology, 82(6), 722-736. doi: 10.1016/j.bjorl.2015.11.022 Stapells, D. R. (2000). Threshold estimation by the tone-evoked auditory brainstem response: A literature meta-analysis. Journal of Speech-Language Pathology and Audiology/Revue D'Orthophonie Et D'Audiologie, 24(2), 74-83.  70  Starr, A., Picton, T. W., Sininger, Y., Hood, L. J., & Berlin, C. I. (1996). Auditory neuropathy. Brain: A Journal of Neurology, 119 ( Pt 3)(3), 741-753. doi:10.1093/brain/119.3.741 Starr, A., Sininger, Y., Nguyen, T., Michalewski, H. J., Oba, S., & Abdala, C. (2001). Cochlear receptor (microphonic and summating potentials, otoacoustic emissions) and auditory pathway (auditory brain stem potentials) activity in auditory neuropathy. Ear and Hearing, 22(2), 91-99. doi:10.1097/00003446-200104000-00002 Teagle, H. F. B., Roush, P. A., Woodard, J. S., Hatch, D. R., Zdanski, C. J., Buss, E., & Buchman, C. A. (2010). Cochlear implantation in children with auditory neuropathy spectrum disorder. Ear and Hearing, 31(3), 325-335. doi:10.1097/AUD.0b013e3181ce693b Vlastarakos, P. V., Nikolopoulos, T. P., Tavoulari, E., Papacharalambous, G., & Korres, S. (2008). Auditory neuropathy: Endocochlear lesion or temporal processing impairment? implications for diagnosis and management. International Journal of Pediatric Otorhinolaryngology, 72(8), 1135-1150. doi: 10.1016/j.ijporl.2008.04.004 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. doi:10.1542/peds.102.5.1161  71  Appendices  Appendix A     Subject Age Cor (mo) Age Uncor (mo) Stimulus Intensity (dB nHL) Stimulus Rate (per second) CM Onset (ms) CM Offset (ms) CM Duration (ms) CM Amplitude (µV) Latency of CM Amplitude (ms) Wave V Latency (ms) Wave V Amplitude (µV) CM/V Amplitude Ratio      OAE+ 1 4 7.5 99 19.5 * | 0.88 * | 3.8 *| 2 .92 * | 0.143 * | 1.4 * | 8.45 * | 0.039 * | 3.67 2 1.5 2.6 95 19.1 0.23 | 0.35 5.13 | 2.8 4.9 | 2.45 0.349 | 0.366 1.48 | 1.03 * | * * | * * | * 3 1.5 4.5 90 | 100 19.3 0.33 | 0.23 3.9 | 3.43 3.57 | 3.2 0.159 | 0.272 1.13 | 1.5 * | 8.9 * | 0.038 * | 7.16 4 2.25 6 90 19.3 0.28 | * 3.03 | * 2.75 | * 0.255 | * 0.95 | * 7.6 | * 0.043 | * 5.93 | * 5 2.25 4 95 19.3 0.2 | 0.25 4.9 | 5.45 4.7 | 5.2 0.404 | 0.573 1.3 | 1.35 7.33 | 7.53 0.075 | 0.080 5.39 | 7.16 6 11.7 14.7 95 19.3 0.38 | 0.33 4.03 | 6.08 3.65 | 5.75 0.277 | 0.521 0.98 | 1.8 8.28 | 8.97 0.077 | 0.052 3.60 | 10 7 * 7 95 19.3 0.13 | 0.25 4.13 | 2.23 4 | 2.28 0.205 | 0.138 1.35 | 0.53 7 | * 0.035 | * 5.86 | * 8 * 3.9 95 19.3 * | 0.13 * | 4.8 * | 4.67 * | 0.243 * | 3.6 * | 7.58 * | 0.046 * | 5.28 9 0.35 2.15 95 19.1 0.35 | 0.28 4.03 | 4.93 3.68 | 4.65 0.535 | 0.685 1.38 | 1.38 7.55 | 7.3 0.123 | 0.105 4.35 | 6.52 10 3.7 5.2 90 19.3 0.28 | 0.3 3.58 | 6.13 3.3 | 5.83 0.202 | 0.514 1.73 | 1.73 * | 7.35 * | 0.036 * | 14.28      OAE- 11 8 11.5 95 61.1 0.48 | 0.3 4.68 | 5.68 4.2 | 5.36 0.036 | 0.029 0.28 | 1.7 7.1 | 6.56 0.025 | 0.029 1.44 | 1.00 12 * 2.4 95 19.3 0.63 | * 6.53 | * 5.9 | * 0.102 | * 0.85 | * 6.43 | * 0.137 | * 0.74 | * 13 2 * 90 19.3 0.38 | * 4.5 | * 4.12 | * 0.138 | * 1.1 | * 8.33 | * 0.049 | * 2.82 | * 14 3 6.25 95 19.3 * | 0.35 * | 3.15 * | 3.15 * | 0.627 * | 0.58 * | 6.85 * | 0.162 * | 3.87 15 * 3.25 95 19.3 0.58 | * 4.68 | * 4.1 | * 0.248 | * 1.48 | * 6.15 | * 0.096 | * 2.58 | * 16 2.3 2.4 90 19.3 * | 0.38 * | 6.75 * | 6.37 * | 0.115 * | 1.23 * | 6.05 * | 0.063 * | 1.83 Table A- 1. Subject measures and recording parameters. Corrected and uncorrected age (in months, mo) is indicated by (Cor), and (Uncor), respectively. Numbers appearing in pairs correspond to left and right ears, respectively. Values that could not be evaluated are indicated with (*) symbols. Amplitude is denoted in microvolts (µV), and latency in milliseconds (ms).   72   Figure A- 1a. OAE+ infant CM evaluation (continued in figure A- 1b, below). (Left) Subject identifiers (n=5 subjects, 8 ears) and overlaid R and C plots for CM duration measures. (Right) [(R - C)/2 CM] averaged waveforms for CM amplitude measures in line with their corresponding duration and subject identifiers. Identifiers span subject’s left and right ears where data was present bilaterally. Blue tracings indicate left ears, red indicates right. The y-axis presents amplitude at an appropriate scale for the wave. CM-bars above and beside tracing indicate duration or amplitude of the CM, respectively.  73    Figure A- 1b. OAE+ infant CM evaluation. (Left) Subject identifiers (n=5 infants, 9 ears) and R and C plots for CM duration measures. (Right) [(R - C)/2 CM] waveforms for CM amplitude measures in line with corresponding subject data. Identifiers span subject’s left and right ears where data was present bilaterally. Blue tracings indicate left ears, red indicates right. The y-axis presents amplitude. CM-bars above and beside tracing indicate duration or amplitude of the CM, respectively.  74      Figure A- 2. OAE- CM evaluation. (Left) Subject identifiers (n=6 infants, 7 ears) and overlaid R and C plots for CM duration measures. (Right) [(R - C) CM] averaged waveforms for CM amplitude measures in line with corresponding duration and subject identifiers. Identifiers span subject’s left and right ears where subject data was present bilaterally. Blue tracings indicate left ears, red indicates right.  The y-axis presents amplitude at an appropriate scale for the wave. CM-bars above and beside tracing indicate duration or amplitude of the CM, respectively.  75   Figure A- 3. Best estimates of OAE+ neural components from (R + C)/2 waves. Subject identifiers (n=9 infants, 12 ears) span left and right ears where data was present bilaterally. Blue tracings indicate left ears, red indicates right. The y-axis presents amplitude. 76   Figure A- 4. Best estimates of OAE- neural components from (R + C)/2 waves. Subject identifiers (n=6 infants, 7 ears) span left and right ears where subject data was present bilaterally. Blue tracings indicate left ears, red indicates right. The y-axis presents amplitude.   Assessed Values All ANSD OAE+ OAE- Duration (ms) Amplitude (µV) Duration (ms) Amplitude (µV) Duration (ms) Amplitude (µV) Kurtosis -0.889 -0.757 -1.040 -0.817 -1.280 4.388 Skew 0.178 0.575 0.186 0.983 0.202 2.032 Mean-Median 4.197 - 4.11 0.322 -0.255 3.97 -3.68 0.344 -0.277 4.74 - 4.2 0.269-0.115 Table A- 2. Measures of central tendency and normality of the distibution of all ANSD ears combined (n=24), and OAE+ (n=17), and OAE- (n=7) ears separately. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0372052/manifest

Comment

Related Items