What AEPs Can (or Could)Tell Us About Auditory (Dys)Synchrony? Robert Burkard Department of Rehabilitation Science University at Buffalo The Ruth Symposium in Audiology & Hearing Science James Madison University Saturday, 10 October 2015, 11:05 AM- 12:05PM Financial Disclosure: Employed by the University at Buffalo and receives a salary; Member of ASHA Health Care Economics Committee; Receiving an Honorarium from James Madison University for the 2015 Ruth Symposium Non-Financial Disclosure: No relevant non-financial relationship exists
1976 Cost of a gallon of gas: $.59 1/17/76: "I Write the Songs" by Barry Manilow hits #1 4/1/76: Stephen Wozniak & Steven Jobs found Apple Computer
1976: Auditory Evoked Potentials Imaging modalities: limited No Universal newborn hearing screening Site-of-lesion tests in Audiology largely behavioral Auditory Brainstem Responses: ABR 1967: Sohmer and Feinmesser: EcochG 1970: Jewett et al.: ABR: Human/animal 1973: Terkildsen et al: ABR: stimulus & recording parameters 1974: Hecox & Galambos: Infant ABR 1976: Starr: ABR in CNS Pathology Hall 1992
Outline Current clinical uses of Auditory Evoked Potentials Mechanisms underlying the effects of stimulus manipulations on AEPs (human/non-human mammals) Auditory neuropathy An animal model of auditory neuropathy? What do we mean by auditory synchrony, and thus auditory dys-synchrony? Can/do AEPs tell us something about auditory (neural) synchrony?
Current Clinical Uses of Auditory Evoked Potentials: Hearing screening/threshold estimation: ABR/ASSR Site of Lesion: ABR/EcochG (imaging?) Intraoperative monitoring: ABR/EcochG/facial nerve monitoring
Why are short latency responses (especially the ABR) used clinically? 1.Not strongly dependent on attention/arousal 2. Strongly dependent on stimulus factors 3. Affected by pathologies affecting the auditory periphery and brainstem 4. Generators of the various peaks fairly well known
Mechanisms underlying the effects of stimulus manipulations on AEPs (in particular, the ABR) What is the ABR?: Auditory Nerve Low Brainstem Rostral Brainstem Burkard and Secor, 2002 Hecox and Burkard 1982
Human Stimulus Level Gerbil: Hood 1998 Adapted from Burkard and Hecox 1983a Burkard and Voigt 1989a
Peak latency and IWI changes with click level Human Gerbil Cox 1985 Burkard and Voigt 1989a
Risetime Burkard and Secor, 2002 Burkard 1984
Barth and Burkard 1993
H U M A N G E R B I L Barth and Burkard 1993 Burkard 1991
Stimulus Level and Risetime: With increasing stimulus level, and decreasing risetime, ABR peaks shift in parallel. That is, the interwave intervals remain more or less constant: the effects of stimulus level and risetime both occur in wave I (i.e., the 8 th nerve response) and are simply reflecting in the later peaks. Is the underlying mechanism leading to the effects of risetime and stimulus level on the AEP the same or different?
Right Insert Earone Latency Latency ICP 1 Amplitude ICP 2 Amplitude Left Inferior Colliculus Potential ACP 1 Amplitude ACP 2 Amplitude Left Auditory Cortex Potential Unanaesthetized chinchilla in testing restraint Broadband Click Contralateral (Right) Stimulus Ipsilateral (Left) Stimulus Binaural Stimulus Left Insert Earphone
Effects of Noiseburst Risetime and Level AN IC Phillips, Hall, Guo and Burkard 2001
Slope of Onset: 78 db SPL, 4 ms risetime noiseburst: 78 db SPL = 20log(Pressure/.00002 Pa) Pressure =.1589 Pa 4 ms =.004 s Slope =.1589 Pa/.004 s = 39.73 Pa/s
Slope of Onset Normalizes Risetime Functions AN IC Pa/s Pa/s Phillips, Hall, Guo and Burkard 2001 Pa/s Pa/s
Tursiops truncatus Noiseburst Risetime SAY Finneran, Mulsow, Houser and Burkard 2015
TRO Finneran, Mulsow, Houser and Burkard 2015
Tonebursts: rat Burkard, Feldman and Voigt, 1990
Frequency/ Cochlear place Harrison, 2006
High-pass masking Human.5-1 khz 1-2 khz 2-4 khz 4-8 khz >8 khz Burkard and Hecox 1983
Human Eggermont and Don 1980
High-pass subtractive masking: Tursiops truncatus Finneran, Mulsow, Houser, Burkard 2015
Finneran, Mulsow, Houser, Burkard 2015
The effects of stimulus rate on the (Human) ABR Burkard and Hecox 1987
Human: Burkard and Hecox 1987 Gerbil: Burkard and Voigt 1989a
Human Broadband noise masking CBA Mouse From Burkard and Hecox 1987 Burkard, Durand, Secor & McFadden 2001
Human From Burkard and Hecox 1987 CBA Burkard, Durand, Secor & McFadden 2001
Gerbil: Noise Burkard and Voigt 1989b
Thus, the interwave intervals increase with increasing rate or masking noise level. This is interesting for several reasons: 1. For 8 th nerve and brainstem lesions, there is often an increase in the ABR interwave intervals. So an increase in the IWIs with increase rate/masking noise suggests that, unlike cochlear place, stimulus level and risetime, that at least a portion of the effects of rate/masking on peak latency are central effects. 2. It begs the question: Are rate and masking effects on the ABR due to the same underlying mechanism?
ABR: Interaction of rate and noise level (Human) Burkard and Hecox 1983
Gerbil i v Latency Amplitude Burkard and Voigt 1990
Gerbil: Burkard and Voigt 1990
Effects of covarying click level and masking noise level (Human) Burkard and Hecox 1983
Gerbil i v Latency Amplitude Burkard and Voigt 1990
Gerbil: Burkard and Voigt 1990
We have looked at a variety of stimulus manipulations: Level, Risetime, Frequency/High-Pass Masking, Rate, Ipsilateral-Direct Masking Based on whether the IWIs are constant or increase, and based on the additivity of the effects, we have found 3 primary mechanisms leading to changes in the ABR latency and/or amplitude: Rate of stimulus amplitude change: level/risetime (a peripheral mechanism) Cochlear place: Frequency (a peripheral mechanism) Adaptation/Masking: rate/noise level (a central mechanism)
Auditory Neuropathy Over the past 30 + years, case studies were published, showing abnormal ABR in the presence of otherwise good hearing abilities. With the common use of OAEs and ABR clinically, a pattern of hearing loss emerged that did not fit easily under the traditional headings of conductive, sensorineural or retrocochlear.
Symptoms of Auditory Neuropathy Test Pure tone thresholds Speech recog. (quiet) Speech recog. (noise) Otoacoustic emissions ME Acoustic reflexes Cochlear microphonic Aud. Brainstem resp. Outcome Normal to Profound HL; can be Asymmetrical Variable: slightly to greatly reduced Generally poor Normal Absent Normal Masking level difference None Efferent suppression (TEOAEs) Absent or severely abnormal None Adapted from Hood (2007)
Site of Lesion: Why neuropathy must be due to an IHC or 8 th nerve disorder OAEs and CM (both largely arising from OHC activity) are normal CAP/ABR missing/abnormal- as the auditory nerve generates CAP and wave I of ABR, suggests that site of lesion reflects a disorder in the eighth nerve or more distal Candidates for site of lesion include: IHCs, with 90-95% of auditory nerve afferents synapsing on them (type I afferents), the synapse itself, or the afferent nerve fibers
Carboplatin and IHC loss Carboplatin, a second-generation platinumbased, anti-neoplastic agent, creates selective inner hair cell (IHC) loss in chinchillas (Harrison and colleagues: Wake et al., 1993, 1994; Takeno et al., 1994a, b).
Is IHC loss the (a?) cause of Auditory Neuropathy? In 1998, Harrison suggested that the cause of auditory neuropathy was loss of IHCs, and in chinchillas post carboplatin, loss of IHCs led to elevated ABR thresholds. Berlin and colleagues have suggested that auditory neuropathy is due to a neural dys-synchrony of auditory nerve fibers, resulting in asynchronous firing of the auditory nerve fibers, and leading to absence/abnormality of the CAP and ABR, and resulting in the perceptual results reported previously.
Photomicrograph of surface preparation of chinchilla cochlea following carboplatin
OAEs are Normal in Selective IHC Loss Following selective IHC loss, distortion product otoacoustic emissions (DPOAEs) are either largely unchanged or actually increased in magnitude (Trautwein et al., 1996; Jock et al., 1996; Wang et al., 1997; Burkard et al., 1997; Hofstetter et al., 1997; Liberman et al., 1997).
From Trautwein et al. (1996)
Effects of IHC Loss and Click Level from the IC from Burkard et al. (1997)
Maximum Length Sequences: MLS Burkard et al. 1990 Human Chinchilla: Burkard et al. 1999
Effects of IHC Loss and Click Rate on IC Response from Burkard et al. (1997)
Effects of IHC Loss and Masking Noise from IC From Burkard et al. (1997)
IHC Loss: Noiseburst Level and Risetime Phillips, Hall, Guo and Burkard 2001
IHC Loss: Noiseburst Level and Risetime Phillips Hall, Guo and Burkard 2001
Does Selective IHC Loss Disrupt Any Auditory Function? Masking level difference (MLD) : We can detect signals in noise better using two ears, under certain binaural conditions. Broad band noise Broad band noise Signal in phase Signal out of phase Guo and Burkard 2003 N0S0 N0S
6 chinchillas, Right AC gross electrode recording, TDT-BioSig, filter 10-3000 Hz, gain 10,000, 9.1 Hz rate, 100 sweeps, 40 ms time window, two repetitions, randomize recording order. Carboplatin: 75 mg/kg, IP, pre-and post-recording. Toneburst series (toneburst ): 500 Hz alternating phase TB (2 ms risetime, 1 ms plateau), 80 db to -10 db pspl, -10 step; le, re, b0 and b conditions. MLD series (toneburst + noise): 500 Hz TB (2 ms risetime, 1 ms plateau), 70 db pspl, continuous BBN 40-90 db SPL, 5 db step; s0n0, s n0, sln0,slnl, srn0 and srnr conditions. Dependent variables: AC latency and AC amplitudes (AP1, AP2). Late ncy AP 1 Measurements of AC latency, AP1 and AP2 AP2 Mean Inner Hair Cell Loss (N=6), Left Ears 0.1 1 10 100 IHC OHC1,OHC2,OHC3 80 60 40 20 0 100 80 60 40 20 Guo and Burkard 2003 0 0 20 40 60 80 100 % Distance from apex Mean Inner Hair Cell Loss (N=6), Right Ears 0.1 1 10 IHC OHC1,OHC2,OHC3 0 20 40 60 80 100 % Distance from apex
Amplitudes vs masked levels (MLD series) S N0 vs S0N0 Guo and Burkard 2003
Does IHC loss lead to Neural Dys-synchrony? In chinchilla post-carboplatin, near-field responses are normal in morphology, with changes in response amplitude and modest changes in latency. There are no substantial changes in responses to stimulus manipulations that are designed to stress neural synchrony including reducing level, increasing rate, increasing background noise, increasing risetime. There does appear to be a degradation in the nearfield-ac MLD response.
If, as Chuck Berlin and colleagues suggest, auditory neuropathy is the result of auditory dys-synchrony, then: Exactly what do we mean by auditory synchrony?
ABRs, and many other AEPs are onset responses: They are typically only recorded to an abrupt change in the stimulus envelope. All neurons in the auditory periphery, and many at higher levels, show sustained discharge in response to a continuous stimulus Harrison, 2006
Yet, even to much longer stimuli, short-latency AEPs are only seen at the onset (and perhaps the offset) of a long duration stimulus. Responses to Noiseburst Gaps Fig. 1 Paired noiseburst, 50 ms duration, 80 db SPL, gap time 32 ms. Onset 1 Offset 1 Onset 2 Offset 2 IC evoked by 32ms gap shows onset and offset Onset 1 Offset 1 Onset 2 Offset 2 AC evoked by 32ms gap show onset and offset Fig. 2 Gap time 32 ms, IC response shows onset 1, offset 1, onset 2 and offset 2 components. Fig. 3 Gap time 32 ms, AC response shows onset 1, offset 1, onset 2 and offset 2 components. From Guo and Burkard (2001)
So, for AEPs such as the ABR, neural synchrony means that a substantial number of auditory nerve (and/or brainstem) fibers must discharge nearly simultaneously in order to observe a response. Antoli-Candela and Kiang, 1978
Can AEPs tells us something about auditory (dys)synchrony? So, the cochlear delay leads to a desynchronization of the ABR (and other AEPs), which likely precludes seeing subtle desynchronization that may exist in, e.g., milder forms of auditory neuropathy, or, perhaps, auditory processing disorders(???) To observe more subtle dys-synchrony, we would need: 1. A method to account for the cochlear delay line 2. A method to assess subtle dys-synchronization of the non-delayed response
One method to account for the cochlear delay is the Stacked ABR approach The work of Don and colleagues, in Burkard and Don 2007
Another method to account for the cochlear delay line Torsten Dau was the first to use an upwardsweeping chirp to attempt to offset the cochlear delay to lower frequency (more apical) cochlear regions. Chirp 5 Chirp 4 Chirp 3 Chirp 2 Chirp 1 Click -15-10 -5 0 5 Time [ms] Burkard and Don 2015 (reprinted, with permission from Elberling et al. 2010)
500 nv ABRs to Clicks and Chirps 20 dbnhl 40 dbnhl 60 dbnhl Chirp 5 Chirp 4 Chirp 3 Chirp 2 Chirp 1 Click 0 5 10 15 Time [ms] 0 5 10 15 Time [ms] 0 5 10 15 Time [ms] From Burkard and Don 2015 (reprinted, with permission, from Elberling et al. 2010)
800 ABR amplitude [nv] 700 600 500 400 300 200 db nhl 40 60 20 100 0 Click 1 2 3 4 5 Chirps 0 1 2 3 4 5 6 Change of delay [ms] From Burkard and Don 2015 (reprinted, with permission, from Elberling et al. 2010)
A method to assess subtle dys-synchronization of the response? The Stacked ABR can be used to identify the sometimes subtle ABR amplitude decrease caused by small 8 th nerve tumors. Is the stacked ABR actually measuring subtle dys-synchrony caused by the tumor? From Burkard and Don 2015, reprinted with permission from Don et al. 2005
How to assess subtle dys-synchronization of the response? Decreasing stimulus level, increasing risetime, high stimulus rates or the presence of background noise increases ABR peak latencies and decreases peak amplitudes. Based on IWIs, stimulus level and risetime are peripheral effects while rate and masking noise are central effects. These manipulations do appear to reduce auditory (neural) synchrony, thus resulting in the latency increases and amplitude decreases, but these more subtle (dys)synchrony effects may be masked by the substantial dys-synchrony created by the cochlear delay line. Perhaps accounting for cochlear delay using the Stacked ABR approach, or using chirps, combined with low stimulus levels, longer stimulus risetimes, high stimulus rates and/or ipsilateraldirect masking noise will sensitize the ABR for the detection of disorders that represent more subtle dys-synchrony.
A method to assess subtle dys-synchronization of the response? I have been showing you AEP data in humans, and non-primate mammals. A good animal model of auditory neuropathy (as IHC loss does not appear to cause much dys-synchrony ) would be invaluable, both to produce animals with more subtle neural dys-synchrony, and in testing whether the combination of Stacked ABR/chirps and low stimulus level/longer risetime/high stimulus rates/masking noise can be used to identify changes in auditory synchrony in the ABR, or other AEPs.
Acknowledgements Co-authors/Co-investigators: Kurt Hecox, Carrie Secor, Craig Barth, Yuqing Guo, Susan Hall, Dennis Phillips, Marty Feldman, Dick Salvi, Herb Voigt, Shi Ying, Dalian Ding, Patty Trautwein, Blanca Durand, Sandra McFadden, James Finneran, Jason Mulsow, Dorian Houser Others: Clarke Cox, Linda Hood, Bob Harrison, Chuck Berlin, Nelson Kiang, Betty Kwong, Francisco Antoli-Candela, Phil Hofstetter, J. Wang, A. Nostrant My colleagues who have helped me with the elusive concept of auditory dys-synchrony: Manny Don Torsten Dau Jos Eggermont Claus Elberling
Questions?