Podium Presentation Abstracts

Sunday, March 7th

Module I Diagnostic: Hearing and Balance

2:45 - 3:00

DPOAE and Middle Ear Status of Normal-Hearing
Children With Sickle Cell Disease

Andrew Stuart, Ph.D.
Letitia Walker, M.S.
East Carolina University, Greenville, NC

Walter B. Green, Ph.D.
Dalhousie University, Halifax, Nova Scotia

Downs, Stuart & Holbert (2000) previously reported DPOAE amplitudes were significantly larger for children with homozygous sickle cell disease (SCD) compared to normal controls. Since the integrity of the middle ear system directly influences OAE characteristics, it was felt that concurrent investigation of DPOAE amplitudes and outer/middle ear function in children with SCD was warranted. Subsequently, DPOAEs and outer/middle ear status of 12 normal-hearing African-American children with SCD and age, gender and ear-matched African-American controls were examined. DPOAEs were evoked by 13 primary tone pairs with f2 frequencies ranging from 1000 to 4500 Hz. Outer/middle ear status was assessed with tympanometry through indices of peak compensated static acoustic admittance, tympanometric width, tympanometric peak pressure, ear canal volume, middle ear resonance frequency. Tympanograms were recorded with probe tone frequencies of 226, 678 and 1000 Hz. DPOAE amplitudes were significantly larger for children with SCD (p < .05). There were no group differences in any of the middle ear indices (p > .05). These findings suggest that increased DPOAE amplitudes for children with SCD cannot be attributed to differences in outer/middle ear function as assessed with tympanometry. A subsequent analysis of children with SCD on a specific medication regime suggests a possible mechanism for and recovery from hyper-responsive DPOAEs.

---

3:00 - 3:15

Classification of Audiograms by Sequential Testing Using ASSRs (CAST-ASSR)

Ozcan Ozdamar, Ph.D.
University of Miami, Coral Gables, FL

Rafael Delgado, Ph.D.
Intelligent Hearing Systems, Miami, FL

Nuri Acikgoz, M.S.
University of Miami, Coral Gables, FL

A method for estimating four-frequency audiogram patterns using auditory steady state responses (ASSR) is developed for rapid testing. The method is based on a similar algorithm developed for behavioral response audiometry to estimate audiogram patterns (Ozdamar et al., JASA, 1990). The Classification of Audiograms by Sequential Testing (CAST) technique was modified for two-ear, multi-frequency ASSR testing and implemented in an automated ASSR system. The CAST algorithm uses a Bayesian probability formula to determine the optimal testing sequence to determine the most likely audiogram pattern from the obtained testing results. Computer simulations with 9 audiogram patterns were conducted to evaluate the feasibility of the algorithm under different response and noise conditions. The automated system was tested using normal and hearing loss adult subjects. The results indicate that CAST-ASSR is a valid technique for estimating four-frequency audiograms rapidly and can potentially be used with patients who require electrophysiological evaluation in a short test period.

---

3:15 - 3:30

DPOAE, ABR, AMLR in Insulin-Dependent Diabetes Mellitus Adolescents

John D. Durrant, Ph.D.
Lisa B. Taubman, M.A.
Christopher M. Ryan, Ph.D.
Dorothy J. Becker, M.D.
University of Pittsburgh, Pittsburgh, PA

There is suspicion of diabetes-related hearing loss, but data in adult patients suffer potential confounds of aging. Juvenile IDDM subjects reduce confounds, and early intervention should avert any IDDM-related loss. Yet, euglycemia is difficult to maintain perfectly with neuropsychological consequences. Previously reported (AAS, 2002) were preliminary results of DPOAE testing from our on-going study of IDDM adolescents, suggesting a subtle trend of decreased high-frequency emissions. Results of the completed study are now presented, yielding more substantial cohorts of IDDM and age-appropriate subjects (mean age ~14.5 years). Additionally, ABR and AMLR results are presented, testing performed to investigate possible neuropathy-like effects. No meaningful audiometric differences were observed between the IDDM and control groups. Results of DPOAE testing continued to demonstrate a slight decrease in the means of the IDDM group above 4 kHz, with outputs more like those of an older control group (~22.5 years)--perhaps a harbinger of accelerated presbycusis? The ABR and AMLR latencies, however, were not notably different, in contrast to findings from other reports in the literature, perhaps reflecting the high homogeneity and high-quality health care of our cohort. Results of yet other analyses are presented and the theoretical framework of the study discussed.

---

3:30 - 3:45

Identification And Diagnosis of Neonatal Hearing Loss

Barbara S. Herrmann, Ph.D.
Harvard Medical School, Boston, MA

Universal newborn hearing screening has made it possible to identify, and, with timely follow-up, characterize neonatal hearing loss. Over a four-year period, 23,881 infants were screened using automated ABR. The majority, 16,094, were from a major teaching hospital, and the others were from community hospitals. Pass and refer rates from Level 1 and 2 nurseries were similar across all hospitals with 98.4% passing in both ears, 1.3% passing in one ear and 0.3% not passing in either ear. The refer rate from the Level 3 nursery (NICU) was higher with 95.1% passing in both ears, 2.2% passing in one ear and 2.7% not passing in either ear. Follow-up evaluations using toneburst ABR were completed within three weeks of discharge and were available for approximately 90% of the infants from the teaching hospital. Across nurseries, 74% of bilateral refers were diagnosed with hearing loss with 47% demonstrating a sensorineural hearing loss. Eighty percent of unilateral refers had normal hearing on follow-up. Of the ears with sensorineural hearing loss, 44% were in the mild to moderate range. Thirty percent of refers from Level 1 and 2 nurseries had sensorineural loss compared to 60% from Level 3.

---

3:45 - 4:00

Relationship of Behavioral Thresholds and
DPOAE Regions Reflecting Ototoxic Damage

Kelly Deal, M.S.
Dawn Konrad-Martin, Ph.D.
David Phillips, Ph.D.
Wendy Helt, M.A.
Jane Gordon, M.S.
Stephen Fausti, Ph.D.
Portland VAMC; NCRAR, Portland, OR

This study describes the relationship between ototoxic-induced behavioral threshold and distortion-product otoacoustic emission (DPOAE) changes in veterans. Data were obtained as part of a large prospective study investigating methods for early detection of ototoxic hearing change. Subjects were 53 veterans receiving therapeutic treatment with either aminoglycoside antibiotics or platinum-based chemotherapy. Behavioral thresholds were obtained at 0.5-20 kHz. DPOAE primary frequency, f2, was varied from .8-10 kHz with f2/f1=1.22 and L1/L2=65/59 dB SPL, respectively. Results were examined from 89 ears demonstrating ototoxic-induced behavioral threshold shifts. Of these ears, 65/89 (73%) had reduced or absent DPOAE following drug treatment and 24/89 (27%) did not exhibit change. The group with change had significantly better mean behavioral thresholds and more DPOAE responses at baseline than the group with no change. DPOAE change occurred for most ears (54/65, 83%) within an octave of the highest frequency response recorded at baseline. Similar to behavioral thresholds, it appears that DPOAE change initially occurs at cochlear regions coding the uppermost frequencies capable of producing a reliable DPOAE response. However, mechanisms underlying these two measurements are different and thus result in two different sensitive regions for ototoxicity. Spatial and temporal differences between the two sensitive regions will be discussed.

---

4:00 - 4:15

Wideband Reflectance in Infants and Children With Middle Ear Effusion

Lisa L. Hunter, Ph.D.
University of Utah, Salt Lake City, UT

Conventional tympanometry is ineffective in detecting otitis media with effusion in infants younger than 6 months, and in older children, has relatively low sensitivity to middle ear effusion. Wideband reflectance has the potential to improve diagnosis of middle ear disorders since it measures middle ear function across a broad frequency range. This study compared wideband reflectance with tympanometry at 226 and 1000 Hz in children birth to 2 years of age with normal middle ear status (n=30) and middle ear effusion (n=30), determined at surgery. Wideband reflectance (Mimosa Acoustics Power Flow system), tympanometry at 226 and 1000-Hz, and distortion product otoacoustic emissions were completed immediately prior to surgery. Wideband reflectance was completed twice to assess test re-test reliability.

Wideband reflectance showed no significant test re-test differences for either group. Significant age-related changes in wideband reflectance were found in the normal middle ear function group. Distributions of reflectance at several frequencies were done to establish cut-points for ears with and without middle ear effusion. Wideband reflectance showed significantly higher reflectance in ears with middle ear effusion compared to ears with no effusion at surgery and ears with normal middle ear function, and may improve test performance over conventional low frequency tympanometry.

---

4:15 - 4:30

Vestibular Neuritis

Bryan J. Layton, Au.D.
Don W. Worthington, Ph.D.
Jocelyn Monroe
IHC Hearing and Balance Center, Salt Lake City, UT

This presentation will report the results of vestibular function tests performed on a 26-year-old patient with acute vestibular neuritis. Testing with complete ENG and Computerized Dynamic Posturography (CDP) were completed starting on day three from date of onset. Subjectively, the patient reported this day as being the worst of the previous two days. This particular patient had results of testing consistent with a complete loss of peripheral vestibular function unilaterally. The strength of spontaneous/positional nystagmus measured during the ENG was quite impressive, nearly reaching levels of caloric hyperactivity. The patient was follow periodically over an eight month period with vestibular function tests including ENG, Computerized Rotation, and CDP to monitor the status of the vestibular system. From initial diagnostic evaluation, he was enrolled in vestibular rehabilitation for adaptation and functional measures (dynamic gait index, dynamic visual acuity, single limb support, sharpened Rhomberg and Jacobsen Dizziness Inventory) were completed on each visit during the eight-month period. This study will discuss the debilitating effects of vestibular neuritis, patient symptoms, and changes on vestibular function tests and rehabilitation, onset to 8 months post onset. A literature search failed to demonstrate a case study of this type.

---

4:30 - 4:45

Assessment of Falls Risk in a Cohort of Elderly Patients

Gary P. Jacobson, Ph.D.
Devin L. McCaslin, Ph.D.
Vanderbilt Bill Wilkerson Center, Nashville, TN

A contemporary multidimensional assessment of falls risk was administered to a cohort of 50 elderly patients who had fallen or were felt to be at risk for falling. The mean age of the sample was 75 years (sd 10 years, range 40-89 years) and 64% were female. Of this sample 61% had a history of falls either in or outside the home. All of the patients were living independently.

The factors evaluated in the risk of falls assessment included: medical and falls history, medication history, mentation (Mini-Mental Status Examination), depression (Geriatric Depression Scale), gait (Timed "Up and Go" Test), postural hypotension (laying and standing blood pressures), visual (Snellen chart at 20'), somesthetic (vibration threshold at 250 Hz and tibial nerve somatosensory evoked potentials), and vestibular systems function (caloric and multi-frequency rotational examinations), standing balance (Equitest Sensory Organization Test), and automatic postural reflexes (Equitest Motor Control Test).

The results of this descriptive analysis showed that, among other things, 70% of the sample demonstrated impaired vestibular system function, 75% showed impairment in lower extremity somesthesia and 17% of the sample demonstrated best corrected visual acuity (Snellen) of 20/100 or poorer. These findings are interpreted in the context of multisensory control of balance.

---

Sunday, March 7th

Module II
Tinnitus and Speech Perception

2:45 - 3:00

Is Tinnitus a GABA Mediated Phenomenon?

William Hal Martin, Ph.D.
Subramaniam Seetharaman, M.D.
Valerie C. Anderson, Ph.D.
Kim J. Burchiel, M.D.
Baker Y-B. Shi, M.D., Ph.D.
Oregon Health & Science University, Portland, OR

Keith Hildebrand, DVM. Ph.D.
Medtronic Neurological, Minneapolis, MN

The purpose of this pilot study was to determine if enhancement of the auditory inhibitory system through the local, direct application of a GABA agonist could modify the perception of tinnitus.

Tinnitus may result from an imbalance between the excitatory and inhibitory auditory pathways causing oscillatory activity resulting in the persistent perception of phantom sound. It should be possible to modify tinnitus perception by altering the effectiveness of the inhibitory system. Baclofen is a GABA agonist that binds to GABAB receptors, modulating and enhancing inhibitory processes. When applied locally in the intrathecal space, it successfully controls abnormal neural activity related to spasticity.

Baclofen was continually infused at ascending doses over a 72 hour period into the cervical intrathecal spaces of 8 hospitalized, adult subjects pre-selected for stable, severe, chronic tinnitus. A tinnitus evaluation battery was performed before, after and at 12-hour intervals during the study and washout periods. VAS measurements for loudness and annoyance were recorded every 4 hours.

Dose related changes in tinnitus perception, including decreases in loudness and annoyance and increased residual inhibition, presented in 6 of 8 subjects. The remaining 2 subjects were notable because they exhibited no residual inhibition at any time during the study.

---

3:00 - 3:15

Low-cost Tinnitus Masking Using MP3 Technology

David Zapala, Ph.D.
Mayo Clinic - Jacksonville, Jacksonville, FL

The purpose of this presentation is to describe a low-cost and effective method of developing and delivering customized sounds for tinnitus masking and habituation purposes. The method entails the use of sound recording software that is available for free over the Internet, and low-cost personal MP3 players. To generate sounds having a specified spectrum, the software program entitled 'audacity' is used. Audacity is a free, cross-platform audio editor that is available at <http://audacity.soundforge.net>. The program can be used to generate white noise that can be subsequently shaped using digital filters provided in the program. Once the desired sound spectrum has been created, the file can be translated into MP3 format. From there, the file can be transferred to a personal MP3 player or to any MP3 formatted compact disc for distribution. Coupler comparisons between MP3 formatted sound files and commercially available maskers will be presented, as will early clinical outcomes. Additionally, the use of these files as ambient masking sounds will be described.

---

3:15 - 3:30

Audiometric Profiles of Professional Musicians,
Audio Engineers, & Music Industry Personnel

Rachel J. Cruz, M.A.
Laurel Fisher, Ph.D.
House Ear Institute, Los Angeles, CA

According to the National Institute on Deafness and other Communication Disorders, over 30 million Americans are regularly exposed to hazardous sound levels. Of the total 28 million Americans with permanent hearing loss, about one third lost their hearing either partially or entirely due to Noise-Induced Hearing Loss (NIHL).

Our study examined over seven years of hearing screenings characterizing the prevalence of hearing loss across a unique vocational group, who are at risk for Noise-Induced Hearing Loss (NIHL). A retrospective analysis of 4600 audiograms of professional audio engineers, musicians, and music industry personnel stratified hearing as a function of age, gender, employment (type and duration), and recreational noise exposure. Obtained thresholds were compared to age- and gender-matched thresholds from a normative database and from both individuals with no history of vocational noise exposure as well as those with a history of mechanical vocational noise exposure, rather than musical. A predictive model for noise-induced hearing loss in this population was also created.

---

3:30 - 3:45

Free Field and Interaural Noise Correlation: Effects on Speech Intelligibility

David Gnewikow, Ph.D.
Vanderbilt Bill Wilkerson Center, Nashville, TN

This study evaluated the effects noise correlation on speech intelligibility. Four experiments were completed, evaluating speech intelligibility both in free field and under headphones. The Hearing in Noise Test (HINT) was administered under correlated and uncorrelated conditions for both white noise and cafeteria babble noise.

Initially, Binaural Intelligibility Level Difference (BILD) was measured under headphones and, in agreement with previous work, subjects performed better under uncorrelated conditions. The free field experiments were designed to determine if the same effect of noise correlation could be demonstrated in free field. The variables of noise modulation, monaural/binaural listening and calibration method were also analyzed. Results from the free field studies indicated no significant effect of noise source correlation. Thus, the effect of noise correlation shown under headphones could not be duplicated in free field. No significant effect of modulation was found. Subjects' overall performance was significantly better binaurally than in the monaural conditions; however, binaural listening did not interact significantly with the effect of noise correlation. Finally, significant differences in the pattern of results were found depending on calibration method. The results of the experiments will be discussed relative to previous research on release from masking and design of speech intelligibility-in-noise experiments.

---

3:45 - 4:00

Relevant Modulation Frequencies for Consonant Identification as Revealed By MDI

Frederic Apoux, Ph.D.
Sid P. Bacon, Ph.D.
Arizona State University, Tempe, AZ

Previous psychoacoustic experiments on modulation detection interference (MDI) indicate that processing of amplitude modulation (AM) in one spectral region can be disrupted by simultaneous AM in another region. This effect is maximal when the masker and the target carriers are modulated at similar rates. In the present study, the role of different modulation frequencies in the speech envelope was studied by means of this disruption. Consonant identification for a band of speech located in either the low- or high-frequency region was measured in presence of a 4-ERB wide noise masker located in the opposite frequency region. The masker was unmodulated or amplitude modulated sinusoidally (SAM) or randomly (RAM). The center frequency of the masker modulation was 4, 16, or 64 Hz. Interference was observed with maskers modulated at 4 Hz when the speech was in the low-frequency region and at 4 or 16 Hz when it was in the high-frequency region. The effects were similar with both SAM and RAM. These results suggest that relevant modulation frequencies may differ depending upon the region of the speech spectrum. Results from a second experiment in which seven masker modulation rates ranging from 5 to 65 Hz will also be presented.

---

4:00 - 4:15

High-Frequency Hearing Loss: The Utility
of High and Low-Frequency Information

Benjamin W. Y. Hornsby, Ph.D.
Todd A. Ricketts, Ph.D.
Vanderbilt Bill Wilkerson Center, Nashville, TN

The utility of frequency specific speech information was assessed in 10 persons with sloping high-frequency hearing loss (HI group) and 10 persons with normal hearing (NH). Sentence recognition in noise (+6 dB SNR) was measured using the Connected Speech Test. To quantify audibility, masked thresholds were measured for both groups in the speech noise used during sentence testing. Sentence recognition was measured in multiple low and high pass filter condition. The utility of high- and low-frequency speech information was assessed by comparing crossover frequencies derived from the filtered speech testing, as well as, absolute performance, and performance improvements, between groups. Additionally, the Speech Intelligibility Index (SII) was used to account for residual differences in audibility between groups. Crossover frequency analysis suggests that HI are less able to utilize HF information than NH participants. Persons with HFHL showed less benefit from HF speech information than NH persons listening under comparable acoustic conditions. SIIs results, however, suggest that the lower crossover frequencies observed for the HI may be due primarily to better than predicted performance under most low pass conditions, in conjunction with rather limited utility of HF information. These results will be contrasted to those from persons with "flat" SNHL.

---

4:15 - 4:30

Rapid Speech Processing in Older Listeners: Cognitive Effects

Nancy Vaughan, Ph.D.
Daniel Storzbach, Ph.D.
Portland VA Medical Center, Portland, OR

Older adults often complain that speech is too fast or that the environment is too noisy. Hearing loss alone does not account for typical related speech understanding difficulties of many older adults. In the past decade, age effects in speech understanding have been associated with age-related cognitive decline. A study is being conducted to identify specific cognitive deficits that are associated with declines in speech understanding of older listeners, especially in demanding listening situations.

Sentences that were speeded at different rates of time compression to increase processing demands were administered to participants, ages 50 to 75 years of age. A battery of selected cognitive tests aimed at identifying processing deficits associated with speech understanding was administered to the same participants.

Results of preliminary data reveal that performance on specific neurocognitive tests is associated with performance on time-compressed speech recognition tests at certain rates of speech. At lower rates of time compression, hearing loss accounts for a significant portion of the variance of the speech recognition scores, but at higher rates age is the significant predictor variable.

These results suggest that time-compressed speech tests may be appropriate clinical tools if sentence materials and rates of speech are carefully chosen.

---

4:30 - 4:45

Measuring Dead Regions Clinically: Is it Worth the Time?

Jill E. Preminger, Ph.D.
University of Louisville School of Medicine, Louisville, KY

A clinical test is available which can identify dead regions along the basilar membrane, the Threshold Equalizing Noise Test (TEN). It is not clear whether it is worth the time to give the TEN test to all patients or to a sub-set of patients whose audiograms suggest that dead regions may be present (Moore, 2001). If listeners with dead regions perform more poorly than expected on typical clinical tests of speech recognition, and if these listeners cannot be identified based on their audiograms alone, then it may be worthwhile to identify these patients. The purpose of this study was to compare the audiograms between listeners with and without dead regions, and to determine whether listeners with dead regions performed differently than listeners without dead regions on a typical clinical speech intelligibility task (QuickSIN). Individuals with hearing loss (98 ears) with at least two pure tone thresholds greater than 50 dB HL and no pure tone thresholds greater than 85 dB HL participated. The results indicate that subjects with high frequency dead regions could not be identified based on pure tone thresholds alone. Subjects with dead regions performed significantly more poorly on the QuickSIN than listeners with no dead regions.