XIII. Sensory, Cognitive, and Communication Aids

A. Hearing Impairment

[187] EFFECTS OF STIMULUS VARIABILITY AND LEXICAL DIFFICULTY ON SPEECH RECOGNITION

Donald D. Dirks, PhDSumiko Takayanagi, PhDP; Anahita Moshfegh, MA; Stephan Fausti, PhD; Douglas Noffsinger, PhD
Portland RR&D Center for Rehabilitative Auditory Research, 3710 SW US Veterans Hospital Road, Portland, Oregon, 97207; West Los Angeles VAMC, Wilshire and Sawtelle Blvd., Research - Audiology Bldg., 500, Los Angeles, CA 90073; email: ddirks@ucla.edu; takayana@ucla.edu; moshfegh.anahita@west-la.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C2225-RA)

PURPOSE--The overall goal of these experiments is to examine the role that stimulus variability (e.g., differences due to talkers, speaking rates, dialects, and so forth) and lexical factors have on the speech recognition of normal and hearing-impaired listeners. The specific goal, in the initial experiment, is to determine the relative importance of several lexical factors on speech recognition, including: 1) word frequency; 2) neighborhood density (the number of words that are phonetically similar to the target word, found by deleting, adding or substituting one phoneme for the target); and 3) neighborhood frequency (average word frequency of a target word's lexical neighbors).

METHODOLOGY--Speech recognition was measured for monosyllabic words (consonant-vowel-consonant, CVC) categorized into eight mutually exclusive lexical groups. The word lists were constructed using a computer-readable version of a pocket dictionary containing over 19,000 words, each of which is quantified on the basis of word frequency, number of words in the lexical neighborhood, and the average frequency of occurrence of the lexical neighbors. From a large database of monosyllables, words with low familiarity, as well as pronouns and words with blends and clusters were eliminated. The remaining 1020 CVC words were used in the study. Those words with values above the median for each lexical characteristic were designated as high (H) and those below the median as low (L). The words were then divided into eight experimental categories by combining all words with either H or L values for each variable into mutually exclusive groups. These groups were designated by word frequency, neighborhood density, and neighborhood frequency as follows: HHH, HHL, HLH, HLL, LHH, LHL,LLH, and LLL. From the words available in each group, 50 words were chosen restricted so that the phonemic structure of each group was controlled. Two speakers, a male and a female, recorded the 400 CVCs. A final database of 800 words was digitized, edited and the waveform check for speech sample integrity. The RMS level of each word was measured to assure that the distribution of levels of the words were matched among groups. A spectrally shaped noise matched to the RMS speech spectrum of the talkers was also digitized for use as a masker.

RESULTS--The words were presented to four nonimpaired subjects at a high intensity level in quiet to determine that each word was intelligible. Then estimates were obtained in noise to determine the range of scores that might be anticipated at various signal-to-noise (S/N) ratio presentation levels. It was determined that a S/N of "+1" (speech presented at 75 dB SPL) provided scores from 50-75 percent for the most difficult to easiest conditions respectively. This presentation level was used in the initial experiments. Currently six subjects have been tested, and preliminary results suggest that performance is highest for conditions with high word frequency and low density neighbors and lowest for conditions with low word frequency and high density neighbors. Tests will be completed on 20 normal hearing and ;sl20 hearing-impaired listeners, and a comprehensive analysis will be conducted.

FUTURE PLANS--The initial study will be completed for control subjects and then contrasted with performance among patients with sensorineural hearing loss. Similar experiments will be performed at a low presentation level in quiet. Subsequently, words from lexical groups which provide the lowest and highest scores will be recorded by 10 speakers. Tests will be conducted for multiple versus single speakers in order to determine the effects of stimulus variability on speech recognition of lexically based word lists. Potentially lexically based tests that include multiple speakers may be developed to assess the effects of sensorineural hearing loss on speech recognition.

[188] IS THERE AN 'ACCLIMATIZATION EFFECT' WITH HEARING AIDS?

Carol A. Sammeth, PhD; Julie Patterson, MA; Maureen Coughlin, MA
Roudebush VA Medical Center, Audiology (126), 1481 W. 10th St., Indianapolis, IN 46202; email: Carol.Sammeth@med.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C840-RA)

PURPOSE--In 1992, Gatehouse coined the term "acclimatization effect" to refer to improvement in speech recognition scores seen over time in a small sample of users following the fitting of hearing aids. He believed that the improvements were due to their learning to use the amplified speech cues now available to them, rather than simply to the stimulation of auditorily deprived ears. Since that time, however, conflicting data have been reported as to whether or not such a phenomenon exists. The present study examined this effect with a research design intended to control for test-retest variability of the speech recognition tasks, and by evaluating the effects of changes in preferred volume wheel setting over time, and the use of both linear and nonlinear amplifiers.

METHODOLOGY--Data were obtained on adults with bilaterally symmetrical sensorineural hearing loss who had and had not previously worn amplification. New users were fitted with binaural sets of both linear and nonlinear hearing aids using probe microphone measurements. Aided speech recognition was evaluated using both the adaptive Hearing-In-Noise Test (HINT) to determine signal-to-noise ratio (SNR) for 50 percent correct sentence recognition and the Speech-Perception-In-Noise (SPIN) test at a fixed, individually determined SNR. Half of the subjects wore the linear hearing aid set home and the other half wore the nonlinear set home. A unique aspect of this study was that each subject served as his/her own control. There were several repeat sessions measuring binaurally aided speech recognition during a 2-wk period prior to the subjects taking the hearing aids home, in order to evaluate across-session test/retest reliability. Subsequently, repeat measurements were made at intervals over a 4-mo period, during which subjects wore the hearing aids on a daily basis. At each of these sessions, testing was accomplished both with the volume wheel at user-preferred setting and at the original prescribed gain setting. Abbreviated Profile of Hearing Aid Benefit (APHAB) questionnaire data also were collected repeatedly to examine any change in subjective impressions of hearing aid satisfaction. Experienced users were evaluated on the speech recognition tasks with their own hearing aids at a user-selected volume wheel setting fixed across sessions, and also with repeat APHABs, as an additional means to explore test variability.

PROGRESS--Complete data were collected on a total of 27 subjects (18 new users and 9 experienced users), with hearing losses ranging from mild to severe. The data are now being analyzed. Analysis will include group descriptive statistics and parametric analyses of variance (percent-correct scores will be arcsine transformed), and visual examination of individual data.

PRELIMINARY RESULTS--Although requiring final confirmation, initial examination of the data suggests that neither the objective speech recognition data nor the subjective questionnaire data on the new users show a significant group acclimatization effect. A few individual subjects with the more severe hearing losses (for whom soft, high-frequency speech components would be inaudible without amplification), do appear, however, to show a weak acclimatization effect. A majority of the subjects showed relatively high test/retest variability pre-wearing time, and, surprisingly, few showed any systematic increase in preferred volume wheel setting over time. In addition, some of the experienced users also showed significant test/retest variability.

IMPLICATIONS--In agreement with these preliminary findings, other researchers have also reported, since the initiation of our study, that the acclimatization effect may not be present in patients who have less than severe high-frequency hearing losses and that, when present, it may be small.

[189] EUSTACHIAN TUBE AND MIDDLE-EAR PROSTHESIS IMPLANTS: CHRONIC-EAR DISEASE

Christopher J. Linstrom, MD; Carol A. Silverman, PhD; Ari Rosen, MD; Jed A. Kwartler, MD; Lawrence Meiteles, MD
VA Medical Center, East Orange, NJ 07019; Department of Otolaryngology--Head and Neck Surgery and Communicative Sciences, The New York Eye and Ear Infirmary, New York, NY 10003

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C2017-2RA)

PURPOSE--Our overall objective is to establish the role of the eustachian tube in chronic-ear disease and in the outcome of hearing restoration by surgical implantation of an ossicular replacement prosthesis (ORP). This objective will be met through endoscopic exploration of the eustachian tube at the time of surgery and eustachian-tube tympanometric testing. Another objective is to examine the contribution of various subject and surgical parameters in outcome of hearing restoration by surgical implantation of an ORP.

METHODOLOGY--Patients to be entered into the study comprise adults (18 or more years of age) who will receive surgical ORP implantation in management of chronic-ear disease from chronic-suppurative otitis media, cholesteatoma, or atelectasis. Each subject will receive testing (short-term at 6-8 wk postop, and long-term at 1-yr postop) by: a) conventional audiologic assessment, including multifrequency tympanometry; b) eustachian-tube tympanometric testing, including the inflation-deflation test, Toynbee, and Valsalva tympanometric tests; and c) otolaryngologic evaluation. At the time of the ORP implantation, eustachian-tube and middle-ear endoscopy is performed for the purpose of describing the status of the middle ear and eustachian tube. The results of endoscopy are videotaped. The audiologic, demographic, otologic, and eustachian-tube tympanometric tests will be correlated with the endoscopic findings.

PROGRESS--Instrumentation and database set-up have been completed and the endoscopic instruments have been modified (for improved articulation and durability) in accordance with the results of pilot testing. Preoperative testing has been performed on seven subjects. Short-term postop testing has been completed on four. The 1-yr anniversary has not yet occurred for any subject.

FUTURE PLANS--Additional subjects continue to be recruited into the investigation, and short-term and long-term postop tests are being completed on existing subjects. Data analyses will commence when sufficient data have been collected.

[190] ACCLIMATIZATION IN MONAURALLY FITTED ADULTS

Shlomo Silman, PhD, Carol A. Silverman, PhD, Michele B. Emmer, MS., John Lutolf, PhD
VA Medical Center, East Orange, New Jersey 070l9, Brooklyn College, CUNY, Brooklyn NY 11201, Hunter College, CUNY New York, NY l00l0; VA Medical Center, East Orange, NJ 07019

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C845-RA)

PURPOSE--The purpose of this study is to investigate the aided soundfield performance at each of the 4 discrete VC settings on the hearing aid. This will be done periodically over a 2-yr period in experienced and inexperienced monaurally fitted adults on the two behavioral measures: the low-predictability (LP) sentences of the Speech Perception in Noise (SPIN) test and the nonsense syllable test (NST).

PROGRESS--To date, we have seen 51 subjects. However, statistical analysis was made on 15 subjects, who were seen for an initial, 3-, 6-, and l2-mo follow-up. Analysis of variance showed that there was no significant difference in NST. However, there was a significant improvement in the SPIN.

  The Tukey test indicates that 6 mo for the SPIN was significantly different from the initial test, but not different from the 3-mo and l2-mo. However, trend improvement was seen at 3 and 12 mo on the NST.

  The inconsistency in the statistical findings may be related to power statistics.

FUTURE PLANS--We will continue collecting data to meet the minimum number (at least 80) of subjects to meet the requirement for a moderate effect.

[191] EARLY DETECTION OF OTOTOXICITY BY HIGH-FREQUENCY OBJECTIVE EVALUATION

Stephen A. Fausti, PhD; James A. Henry, PhD; Curtin Mitchell, PhD; Alison Bobal, MA; Christopher Flick, BS; Roger Ellingson, MSCEE
RR&D Center for Rehabilitative Auditory Research, VA Medical Center, Portland, OR 97207; email: henryj@ohsu.edu

Sponsor: Department of Veterans Affairs, VA Medical Research Service, Washington, DC 20420
(Project #0008)

PURPOSE--Behavioral testing of hearing sensitivity is the traditional method to monitor for hearing loss in patients receiving ototoxic medications. However, about one third of these patients are unable to respond reliably to behavioral methods. The focus of this study is to develop objective ototoxicity monitoring protocols, utilizing the auditory brainstem response (ABR) that would be sensitive and time-efficient for behaviorally untestable patients.

METHODOLOGY--Behavioral audiometric thresholds are monitored from patients receiving cisplatin (CDDP) or a specified aminoglycoside antibiotic (AMG). If their thresholds are 100 dB and below through 11.2 kHz, ABR monitoring is performed as follows. A high frequency (6-14 kHz) toneburst algorithm is used to determine the highest frequency toneburst obtaining a repeatable response; that frequency becomes the target frequency used for subsequent monitoring sessions. Additionally, high-frequency clicks (HFCs) are presented during each session by simultaneously presenting a combination of tonebursts to produce wide-band high-frequency acoustical energy (8-14 kHz). HFCs are the high-frequency equivalent of conventional clicks that present energy at 1-4 kHz. The purpose of the HFC is to monitor the region of hearing that is thought to be most susceptible to early ototoxicity.

  Two versions of the HFC are presented: flat and sloped. The flat HFC presents energy equally across frequencies, while the sloped HFC presents energy at different levels across frequencies to correspond with the typical decrease in human auditory sensitivity (3 dB difference per 1,000 Hz). The efficacy of these ABR stimuli for detecting initial ototoxicity will be evaluated in relation to behavioral testing.

  Additionally, we are conducting three substudies: 1) The above ABR stimuli are being used to determine reliability norms in nonimpaired controls; ABRs are obtained over two sessions from subjects in response to the flat and sloped HFCs, tonebursts of 8, 10, 12, and 14 kHz, as well as conventional clicks. 2) A high-frequency toneburst train is being evaluated in patients to determine the efficacy of this new technology to obtain responses to multiple tonebursts, rather than testing one toneburst at a time; tonebursts of 6, 8, 10, 12, and 14 kHz, each presented at three ascending levels, are sequenced to obtain 15 responses. The same tonebursts are presented singly to compare responses between single and multiple methods. 3) Patients are tested with the 15-stimulus train to find the highest frequency at which a response is obtained. That frequency is then presented at five levels that are sequenced together (each separated by 10 dB) in a multiple-intensity train. A multiple-intensity train is also run at 8 kHz, as well as single stimuli at the highest frequency and 8 kHz. Responses to the single and multiple methods will be compared.

PROGRESS--To date, 16 hospitalized patients have been enrolled: two receiving AMG and 14 receiving CDDP. Final data analyses will determine the efficacy of ototoxicity monitoring using the ABR protocol. Two groups of 20 controls have completed two of the substudies. From the first substudy, both types of HFCs were reliable within and between testing sessions. Results from the second study demonstrated good reliability with the multiple stimulus method and a significant time-savings during testing. All test stimuli appear to be well suited for serial monitoring during treatment with potentially ototoxic agents. Controls are currently being tested in the third substudy. Additionally, subjects with sensorineural hearing loss will be tested in each substudy.

IMPLICATIONS/FUTURE PLANS--By the end of this study period, a clinically applicable tool will be available for use in early detection of ototoxic hearing loss in patients who are currently untestable by behavioral means. This tool will be generalizable to measure the ototoxic effects of drugs other than AMG and CDDP. In addition, the techniques developed during this study are expected to have a much broader clinical application beyond ototoxicity monitoring.

[192] DEVELOPMENT OF AN AUTOMATED TECHNIQUE FOR THE CLINICAL EVALUATION OF TINNITUS

Stephen A. Fausti, PhD; James A. Henry, PhD; Alison Gilbert, MS; Christopher Flick, BS; Roger Ellingson, MSCEE
RR&D Center for Rehabilitative Auditory Research, VA Medical Center, Portland, OR 97207 email: henryj@ohsu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C891-RA)

PURPOSE--Subjective tinnitus is the perception of sound generated from within the auditory system. Some 50 million Americans experience it, and the condition is severe for at least 7 million. Tinnitus is usually associated with hearing loss, and the veteran population, with its high incidence of noise-induced hearing loss has a correspondingly high incidence of tinnitus. Tinnitus measurement should be a component of diagnosis and treatment, but currently a standardized technique for quantifying tinnitus does not exist. It is the purpose of this project to develop a fully automated system for measuring tinnitus loudness and pitch.

METHODOLOGY--The project was divided into four substudies; two were designed to evaluate the within- and between-session test-retest reliability of hearing threshold measurements. The first of these uses controls with unimpaired hearing, and the second uses individuals with sensori-neural hearing loss. The third substudy was designed to evaluate within- and between-session test-retest reliability of tinnitus loudness matches at 1/3-octave frequencies between 1 and 16 kHz. The fourth is a comparison of two bracketing techniques for obtaining tinnitus pitch matches (and loudness matches at the tinnitus frequency). Based on results of the substudies, a final testing algorithm will be developed, and a large group of subjects will be tested repeatedly to determine reliability of responses and to compare results obtained with the automated system to those obtained when testing is done manually.

PROGRESS--The four substudies have been completed upon 20 subjects each. Results of the first two have shown that hearing thresholds obtained with the automated system are reliable both within and between sessions, using control or hearing-impaired subjects. The third substudy has shown that, focusing only on repeated loudness-match measurements, these measurements are also reliable over repeated tests. The fourth substudy has defined more closely how to design a shortened procedure to obtain tinnitus loudness and pitch matches in a clinical setting. These results are being further analyzed and development is underway to implement a final clinical testing protocol.

IMPLICATIONS/FUTURE PLANS--The current study is expected to provide an automated test procedure that can be adapted for clinical use in measuring tinnitus loudness and pitch. Such methodology could facilitate the implementation of standardized tinnitus evaluation throughout VA Medical Centers. It could also be adapted for use in validating tinnitus claims, contributing significantly to future management and rehabilitation of veterans. Further applications of the technique should be able to assist in differentiating between auditory pathologies causing tinnitus, to track changes in tinnitus over time (e.g., during treatment or to monitor for ototoxicity), and to assist in patient counseling by associating real numbers with their less tangible auditory perceptions.

RECENT PUBLICATIONS FROM THIS RESEARCH

[193] EARLY DETECTION OF HEARING LOSS DUE TO OTOTOXIC AGENTS BY HIGH-FREQUENCY AUDITORY EVALUATION

Stephen A. Fausti, PhD; James A. Henry, PhD; Vernon D. Larson, PhD; Cynthia G. Fowler, PhD; Douglas Noffsinger, PhD
RR&D Center for Rehabilitative Auditory Research, VA Medical Center, Portland, OR 97207; VA Medical Center, Augusta, GA; VA Medical Center, Long Beach, CA; VA Medical Center, West Los Angeles, CA email: henryj@ohsu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C227-2RA)

PURPOSE--Ototoxic drugs such as the aminoglycoside antibiotics (AMGs) and the chemotherapeutic agent cisplatin (CDDP) can cause irreversible hearing loss. Early detection of ototoxicity is essential to identify the need for intervention to prevent or reduce communicatively disabling hearing loss. This laboratory has been involved since 1988 in the prospective evaluation of changes in hearing sensitivity as they occur in patients receiving ototoxic drugs. The aim of the present study was to determine whether ototoxicity is detectable in the high-frequency (HF) range (9-20 kHz) earlier than in the conventional-frequency (CF) range (0.25-8 kHz).

METHODOLOGY--Data were collected at five participating sites: VA Medical Centers at Portland, OR, Augusta, GA, West Los Angeles, CA, and Long Beach, CA, and at Oregon Health Sciences University, Portland, OR. Inpatients receiving CDDP or specified AMGs were recruited as meeting inclusion criteria. During each session, hearing thresholds were evaluated at 0.25, 0.5, 1, 2, 3, 4, 6, 8, 9, 10, 11.2, 12.5, 14, 16, 18 and 20 kHz. Baseline thresholds provided the reference against which all further tests were compared. The criteria for identifying ototoxic threshold changes were consistent with the national standards, developed in this laboratory and published by the American Speech-Language-Hearing Association (ASHA) in 1994.

PROGRESS--Serial monitoring was completed with 452 patients. Using the ASHA criteria, 141 of the 407 qualifying ears (34.6 percent) from AMG-treated patients demonstrated ototoxic hearing loss. Of the 265 qualifying ears from CDDP-treated patients, 193 showed loss (72.8 percent).

RESULTS--To answer the primary study question of whether ototoxicity were detectable in the HF range earlier than in the CF range, ears that demonstrated ototoxic change were analyzed to determine whether change was initially detected in the HF or CF ranges, or in both, concurrently. In the AMG change group, 62 percent of the ears were first detected in the HF, 16 percent in both, and 22 percent in the CF range. From the CDDP ears that changed, 49 percent were first detected in the HF, 41 percent in both, and 10 percent in the CF range. Combining data from HF first and both ranges concurrently, suggests that if only HF had been tested, 78 percent of initial change in AMG ears would have been detected, 90 percent in CDDP ears, and 85 percent of both drug groups combined.

  For patients showing hearing changes due to ototoxicity, a frequency range consisting of five test frequencies was identified for its apparent high sensitivity to initial ototoxicity. This sensitive range is identified according to hearing threshold configuration, and is, therefore, distinct for each individual. From retrospective analysis, if only this target-frequency range had been tested, hearing loss would have been detected in approximately 90 percent of all ears showing change.

IMPLICATIONS/FUTURE PLANS--This study has conclusively shown that, in patients meeting study criteria, when ototoxicity occurs it is generally first detectable in the HF range. Retrospective data analysis identified a range of five frequencies, identified on an individual basis, that appeared to have the highest sensivitity to early ototoxicity. In the current continuation study, we shall evaluate whether or not the target-frequency approach will be effective when used on a prospective basis. Subjects from the previous study were required to have HF hearing up to 11.2 kHz. The test frequencies shown to be most sensitive for early detection of ototoxicity were generally at and above 8 kHz, where they are spaced by 1/6-octave intervals. For patients with more hearing loss, the individualized target frequency range will occur primarily at conventional frequencies (<8 kHz), separated by 1/2 octave. It will be investigated whether or not monitoring with higher frequency resoloution (1/6-octave) in the conventional range is necessary to provide the same sensitivity to ototoxicity as was observed at high frequencies.

RECENT PUBLICATIONS FROM THIS RESEARCH

[194] INVESTIGATION OF INDIVIDUALIZED HIGH-FREQUENCY REGIONS FOR EARLY DETECTION OF OTOTOXICITY

Stephen A. Fausti, PhD; James A. Henry, PhD; Douglas Noffsinger, PhD; Gene Bratt, PhD; Wendy Helt, MA; Dan McDermott, MA
RR&D Center for Rehabilitative Auditory Research, VA Medical Center, Portland, OR 97207; VA Medical Center, West Los Angeles, CA; VA Medical Center, Nashville, TN; email: henryj@ohsu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C227-3RA)

PURPOSE--Treatment with therapeutic drugs such as the aminoglycoside antibiotics and the chemotherapeutic agent cisplatin can cause irreversible hearing loss. Early identification of ototoxicity is important to alert physicians of the need to implement treatment modifications to minimize or prevent the progression of hearing loss. Ototoxicity monitoring programs are effective for this purpose, but they are labor intensive and thus are accepted with difficulty in an era of hospital cost cutting. The purpose of the present study is, therefore, to develop auditory monitoring techniques that can be conducted rapidly without compromising the sensitivity to early ototoxicity that is achieved with full-frequency (0.25-20 kHz) behavioral testing. There are two primary objectives: 1) to determine the clinical efficacy of monitoring hearing in a limited range of frequencies spanning one octave and separated by 1/6 octave; and 2) to determine whether distortion-product otoacoustic emissions (DPOAE) testing has utility for objective early detection of ototoxicity.

METHODOLOGY--Hearing data are currently being collected at three participating VA Medical Centers (Portland, OR, West Los Angeles, CA, and Nashville, TN). Qualifying patients receive baseline testing to determine hearing thresholds at all audiometric frequencies (0.25-20 kHz), and to identify the target frequency range (hypothesized to be most sensitive range for early detection of ototoxicity: one octave, 1/6 octave separation between test frequencies). Serial testing throughout and following treatment consists of testing at the target frequencies, followed by all remaining test frequencies in the 0.25-20 kHz range. This order of testing will enable prospective analysis of the sensitivity of testing independently in the target-frequency range, as compared to full-frequency testing. At each test session, DPOAE testing is also done to obtain 1/6-octave DP-grams at three intensity levels.

  At the Portland site, three development studies designed to improve efficiency of the ototoxicity monitoring protocols are also being conducted: 1) A method for rapid identification of the target-frequency range is being evaluated in an effort to reduce baseline testing time; 2) Insert earphones are being evaluated for test-retest reliability of hearing thresholds. (Insert earphones are expected to provide several advantages to currently-used circumaural earphones); and 3) A prototype DPOAE probe system, designed to measure high-frequency DPOAE responses to 16 kHz, is being investigated for validity and reliability of responses.

PROGRESS--To date, 88 patients receiving ototoxic drugs, and 8 control patients, have been entered into the study. Based on retrospective analysis from the previous study, initial detection of ototoxicity is hypothesized to be equally effective at the target frequencies relative to testing all frequencies in about 90 percent of the patients. The development study to evaluate the accuracy of the protocol for rapid identification of the target frequency range has been completed. Results from the rapid identification procedure were compared with results using the longer procedure (testing frequencies at 2-20 kHz). The same target frequency range was indicated for 36 of the 42 ears tested (85.7 percent). Four ears (9.5 percent) revealed a one-frequency difference in the range. One ear (2.4 percent) yielded a two-frequency difference and one ear (2.4 percent) revealed a three-frequency difference.

  Evaluation of the reliability of testing using insert earphones has been completed in 20 controls and in 20 individuals with sensori-neural hearing loss. Comparison of results to previous studies conducted in this laboratory has shown that reliability of thresholds obtained with the insert earphones is as good or better than using conventional circumaural earphones. This provides important documentation for the use of insert earphones for ototoxicity monitoring. Current work is underway to determine the reliability of insert earpones when testing is conducted at bedside.

IMPLICATIONS/FUTURE PLANS--Completion of this study will provide the needed prospective data to validate use of a limited-frequency range in a rapid and sensitive ototoxicity monitoring protocol for most patients receiving treatment with ototoxic drugs. DPOAE measurements will more clearly define the cochlear events that occur during early ototoxicity as they relate to behavioral hearing sensitivity changes.

  With further development of instrumentation and techniques, it is expected that the target-frequency protocol can be used with dedicated instrumentation designed to be carried into hospital rooms and administered by any minimally trained person. The final product envisioned is a portable unit that performs highly refined and automated testing.

  Finally, the technology exists for ototoxicity monitoring to become a centralized operation. That is, data collected from any number of sites, using the type of portable device described above, could be sent to a central site via electronic data transport. An audiologist at the central site would monitor incoming data for changes in hearing sensitivity, and notify attending physicians when significant changes are noted. Such a protocol would minimize resources required at the peripheral sites.

[195] QUANTIFYING AUDITORY PERCEPTUAL LEARNING POST HEARING AID FITTING

Charles T. Nelson, MS; Catherine V. Palmer, PhD; Shelley Myers, MA; Melissa Nascone, MA
VA Pittsburgh Healthcare System, Pittsburgh, PA 15206 School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, PA 15260; email: cvp@vms.cis.pitt.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C2036-2RA)

PURPOSE--The intention of this project is to quantify the extent and rate of adult auditory perceptual learning of previously inaudible consonants across specific conditions of quiet, noise, and reverberation in a carefully defined subject population. The proposed research will address the following specific questions: 1) Do adults with longstanding moderate sensorineural hearing loss that has caused inaudibility of high frequency consonants, improve in their discrimination (in terms of percent correct, confidence, or response time) of these consonants after the specific sounds are made audible again on a day-to-day basis via appropriate amplification? 2) What is the time course (to asymptote) of the above-mentioned auditory perceptual learning? 3) Is the time course impacted by what consonant is measured, adverse listening conditions (i.e., noise and/or reverberation), and/or amount of hearing aid use per day? 4) Can the rate or extent be altered by applying controlled auditory practice similar to the auditory test material?

METHODOLOGY--A pre-experiment is being conducted to evaluate the test/retest reliability of the Nonsense Syllable Test (NST) in quiet, noise, and noise and reverberation as presented in this particular experiment. Thirty subjects with ages and hearing losses similar to the main experiment groups will complete the computerized adminstration of the NST four times over two test sessions (1 wk apart). Test-retest reliability will be evaluated for the three outcome measures (percent correct, response time, confidence).

PROGRESS--A computer program for the administration, scoring, and analysis of the NST being used as the outcome measure in this study has been completed. The test syllables were digitized in three listening conditions: quiet, noise, noise and reverberation. The program presents the test syllables in random order (within subtests) according to condition. The order in which the conditions are presented also is randomized. The subjects are required to listen to a carrier phrase that contains the test syllable and choose via touch screen, the syllable they believe was the test syllable from a list of seven to nine syllables of the same subtest. The program then lists six confidence ratings ranging from "I am positive I identified the item incorrectly" to "I am positive I identified the item correctly." The subject touches the confidence rating that most closely describes how confident he/she is with the choice. The computer program measures the response time (from the time of presentation to response selection). A second computer module presents the training task. The perceptual fading technique will be used to train subjects to identify a closed-set of nonsense syllables. Perceptual fading is characterized by the combination of an initial reduction in uncertainty through the systematic sequencing of stimuli followed by increased stimulus uncertainty in the later stages of training. A computerized administration of this technique is ideal because of the varying number of exemplars that are presented, based on previous performance.

  To date, 54 individuals have been tested for the purposes of determining their eligibility for the study, and 17 have been enrolled: 6 individuals for the Pre-Experiment and 11 for the Main Experiment.

RESULTS--Six individuals have completed the pre-experiment. This small number does not allow for statistical analysis at this time. Subjects will not begin the main experiment until the pre-experiment data are analyzed. Results from the pre-experiment will be used in the on-line analysis of the main experiment.

[196] FREQUENCY SELECTIVITY ASSESSMENT: IMPLICATIONS FOR AUDITORY REHABILITATION

James K. Kane, PhD
Audiology (126), VA Medical Center, 1540 Spring Valley Drive, Huntington, WV 25704; email: James.Kane@med.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C2093-RA)

PURPOSE--Frequency selectivity (FS) is the basic property of the auditory system whereby a complex stimulus (e.g., speech) is resolved into its component frequencies. The major limitation for assessing FS clinically has been the amount of training required for patients to perform complex psychoacoustic tasks that are time- and cooperation-intensive. Distortion-product otoacoustic emission (DPOAE) assessment, based on the 2f1-f2 distortion product, is an alternative procedure that is noninvasive, objective, time-efficient, and requires no active participation from a patient.

  The purpose of this study is two-fold: 1) to relate salient parameters of DPOAE data to existing knowledge of FS based on psychophysical studies; and 2) to determine if DPOAE tuning changes with hearing threshold level (HTL). Specifically, we are evaluating the relationships between f1 and the maximum distortion-product (fd*) generated by the DPOAE f2/f1 paradigm to the "tip-tail" slope-change frequency (Fxb) in psychophysical tuning curves (PTCs) at 2 kHz and 4 kHz.

METHODOLOGY--In all, 100 subjects (100 ears), 50 males/50 females, will participate in this study representing five categories of HTL at 2 kHz and 4 kHz re ANSI (1996). Three groups (60 ears) will span the sensitivity range (0-25 dB HTL) called "clinically-normal" and two groups (40 ears) will define the mild, sensorineural loss range (30-45 dB HTL). Each frequency will be treated independently. A low-intensity (5 dB SL) and high-intensity PTC (15 dB SL) for each HTL subject group will be generated using a 2-alternative forced-choice adaptive procedure with a forward-masking paradigm. DPOAE magnitude will be assessed as a function of stimulus frequency separation at 2 kHz and 4 kHz at three intensity levels (f2=45, 55, and 65-dB sound pressure level (SPL)) by sweeping f1 from approximately one-half octave below f2 to f2.

PROGRESS--Major efforts to date have involved establishing the research laboratory, developing software for running multiple subjects simultaneously, hiring a research audiologist, and ensuring that pilot work done at the University of Pittsburgh, Experimental Audiology Laboratory could be replicated. Preliminary results (predominantly normally-hearing subjects) are consistent with those obtained at the University of Pittsburgh laboratory, which were as follows: 1) fd* changed with HTL at each test frequency. 2) f1 and Fxb were related across test frequency and SPL.

FUTURE PLANS--We shall complete data collection on all subject groups, analyse the data, and report the findings. These data will increase the size of the reference DPOAE database, which will be important for future studies (e.g., speech processing ability and amplification parameters).

B. Speech Impairment

[197] PROBLEM-SOLVING STRATEGIES OF BRAIN-INJURED AND NORMAL SUBJECTS

Robert C. Marshall, PhD
Department of Communicative Disorders, University of Rhode Island, Kingston, RI 02881; email: rcmarsh@uriacc.uri.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C837-3RA)

PURPOSE--The aims of this study are 1) to examine the problem-solving strategies used by traumatically brain-injured (TBI) and two groups of noninjured subjects (young normals and older normals) on a Twenty Questions (20Q) task and 2) to determine the effects of Strategy Modeling Training (STM) as an intervention on the problem-solving performance of the TBI subjects.

METHODOLOGY--Subjects meeting selection criteria are given a battery of speech, language, and cognitive tests. Subjects also make self-ratings of their problem-solving ability using the Heppner Problem Solving Inventory (PSI). Independent ratings of subjects' problem-solving are made by a family member with a visual analog scale. Following standardized instructions to explain the 20Q game, the subject plays three pretreatment 20Q games. For each game the subject views a 4×8 matrix of pictures of objects belonging to common categories. The subject is told he/she is to identify the picture the examiner is thinking of by asking yes/no questions, and that the game's objective is to identify the target picture with as few questions as possible. Following the pretreatment games, TBI and the older-normal subjects then receive a period of SMT to show them how to solve 20Q problems. Training uses stimuli similiar to, but different than, those used in the experimental task. One week after SMT training, the subject plays three more 20Q games (posttraining). One month later, three final games are played (Maintenance).

PROGRESS--At this point, 19 TBI, 22 young normals, and 20 older normals have entered the study. We have developed a mechanism for quantifying the efficiency which a subject plays the 20Q games. When subjects' 20Q performance is rated using a visual analog scale that ranges from "not at all efficient" to "highly efficient," normal subjects are rated higher than TBI subjects, and young normal subjects are rated higher than older normal ones. Objective indicators for lower ratings relate to 1) the number of constraint-seeking questions used, 2) demonstration of obvious strategies for playing 20Q games (e.g., Is it in the top row?), 3) use of questions to narrow choices after the category containing the target picture is known, and 4) absence of ambiguity in questions asked. STM training appears to improve the performance of TBI subjects to approximate that of the young normal subjects at post training. TBI subjects who appear to have difficulty with the 20Q game appear to have high estimates of their problem-solving skills as estimated by PSI scores. Conversely, the significant others of these individuals rate the problem-solving skills of these subjects low. No maintenance data have been analyzed at this point.

FUTURE PLANS--Our plan is to enroll 30 TBI, 30 young normal, and 30 older normal subjects in the study. We then will refine our scale for rating problem-solving efficiency and determine inter- and intrarater reliability for this measure.

[198] QUANTIFYING SPOKEN LANGUAGE HANDICAP IN APHASIA

Patrick J. Doyle, PhD; Malcolm R. McNeil, PhD
Aphasia Rehabilitation Research Laboratory and Clinic VA Pittsburgh Healthcare System, Pittsburgh PA 15206; email: pdoyle+@pitt.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C894-2RA)

PURPOSE--Current procedures for the assessment, description, and quantification of aphasic language impairment fail to assess either directly or indirectly, the impact of the disordered language on the communicative partner's reception of the message. Because of the lack of attention to the consequences of aphasic individuals' impaired language, in the face of well quantified and elegantly described aphasic performance, the functional impact of the impairment has gone practically undescribed and unquantified. One solution to this problem is to assess the relative effort expended by the listener while processing and comprehending spoken language. A general measure of the effort expended, if shown to be differentially sensitive to varying severity levels, and types of aphasic impairment, would provide a composite measure of all the physical, social, psychological, linguistic and communicative skills, weaknesses, and biases of the listener.

METHODOLOGY--A computer-based interactive assessment tool was designed and developed to quantify objectively the degree of spoken language handicap resulting from different types and frequencies of aphasic language errors as they occur in connected speech. Using resource allocation theory as the guiding theoretical model, a dual-task program for standard laptop computers was developed that requires participants (i.e., normal listeners) to simultaneously perform a visuo-motor tracking task (i.e., the primary task), while listening to the connected speech of persons with aphasia, (i.e., the secondary task). At the completion of the dual task, listeners are required to provide information regarding the content of the aphasic discourse. In this dual-task paradigm, effort is quantified by performance on the visuo-motor task, and information transfer is quantified by performance on the listening task. We propose to demonstrate that the effort expended for the primary task as assessed by the dual-task paradigm will vary greatly by listener. That is, depending on the state of the listener, differential amounts of cognitive resources (effort) will be expended in the processing and comprehending the impaired language. Further, we propose to demonstrate a systematic increase in effort expenditure within listeners from the comprehension of nonimpaired spoken language to the processing and comprehension of spoken language produced by patients with different severity levels and types of aphasic impairment.

PROGRESS--The story retell, information transfer, and manual-tracking components of the software have been completed, and preliminary psychometric studies are ongoing. The analysis component of the dual-task program is undergoing laboratory testing and further development. Initial experiments examined the story retell component of the program for the effects of sampling conditions on objective measures of language within lexical-semantic, syntactic, phonologic, and fluency domains. Data from additional experiments are being examined to determine sample size requirements and stability of objective measures across alternative forms of the story elicitation procedure, and subjective judgments of different levels of discourse severity as determined by objective measures of language. Finally, validity and reliability of the visuomotor tracking task and information transfer task under noncompeting and dual-task conditions are ongoing as necessary for the psychometric development of the tool.

RECENT PUBLICATIONS FROM THIS RESEARCH

C. Vision Impairment

[199] MEASURING LOW VISION READING ASSESSMENTS USING A SCANNING LASER OPHTHALMOSCOPE

Gale R. Watson, MD; Ronald Schuchard, PhD
Rehabilitation Research and Development Center, VA Medical Center, 1670 Clairmont Road, Decatur, GA 30033; email: GW1@delphi.com; schuchard.ronald_a@atlanta.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C849-2RA)

PURPOSE--Our purpose was to evaluate available low vision reading rehabilitation assessment tools using the scanning laser ophthalmoscope (SLO) in terms of five questions:

  1. Do rate, accuracy, and error scores on a visual skills for reading test relate to preferred retinal locus (PRL) ability and scotoma characteristics as measured by the SLO?
  2. Does a follow-up reading training technique in the home efficiently and effectively assist readers with macular loss in improving reading?
  3. Is the subjective report of reading improvement related to objective improvement in reading?
  4. Are changes in reading performance as measured by reading assessment scores related to changes in different PRL abilities/characteristics?
  5. Are improvements in reading performance related to PRL dominance?

METHODOLOGY--We assessed 60 subjects with ARMD using standard clinical testing instruments (e.g., acuity, contrast sensitivity, glare, etc.) and specialized clinical testing, (e.g., SLO macular perimetry, binocular retinal correspondence, and PRL testing, including fixation stability, pursuit ability, saccade ability, and word retinal location. Reading performance was assessed by the Pepper Visual Skills for Reading Test, the Morgan Low Vision Reading Comprehension Assessment and the MNRead Acuity charts. Reading rehabilitation was provided using strategies that assist the subjects in maximizing their visual skills for reading. One-half the subjects received an at-home reading series to continue practicing. After the reading rehabilitation was complete, the standard clinical testing, specialized clinical testing, and reading assessments were repeated. Subjects were asked to indicate their subjective impressions of reading changes with a visual analog scale (VAS).

RESULTS--Subjects with scotoma placement to the right and below the PRL experienced the most reading difficulty; slower rates, lower accuracy, and more errors. Steady fixation ability and smaller fixation degrees were associated with higher reading accuracy.

  Changes in reading ability were not related to changes in PRL abilities, but higher pretest scores in saccade and pursuit abilities were related to the most improvement in rate and comprehension in reading. Subjects' improvement in reading was not related to the at-home reading practice, subjects in both groups changed in reading ability. Subjects' subjective rating of their changes (VAS) in reading ability were unrelated to their objectively measured changes in rate, accuracy and comprehension. Subjects' reading scores showed no impact of binocular vs. monocular or right/left/both eyes.

FUTURE PLANS/IMPLICATIONS--This research suggests that scotoma placement and PRL abilities are strongly related to reading ability for this population. The investigators will explore in future projects whether instructional strategies that change scotoma placement and enhance PRL abilities have positive effects on reading ability. This research further suggests that although some readers improve their reading ability clinically and statistically, the small gains that are made in 5 weeks are not noticeably appreciated by them. Finding a way to assist readers in understanding the progress they are making might be helpful in motivating them to continue a training program at home. This project did not find that there is an advantage to using low vision devices that are binocular, or that reading with either eye or both eyes was an advantage, indicating that this may not be a concern for the prescription of low vision devices for reading.

[200] EMPLOYMENT OF INFRA-RED BEACONS AS GUIDES FOR PEOPLE WITH A VISUAL DISABILITY: A PILOT STUDY

David A. Ross, MSEE, MED; Bruce B. Blasch, PhD
Atlanta VA Medical Center, Rehabilitation R&D (151-R), Decatur, Georgia 30033; email: DavidRoss1@mindspring.com; Bearconsul@aol.com

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Pilot Project #C2234-PA)

PURPOSE--Our purpose is to evaluate a range of user interface possibilities for guiding veterans with a visual disability across the street at traffic-light-controlled intersections. Being evaluated are 12 variations on an interface tested by subjects in a previous study where Infra-Red (IR) beacons were employed to guide subjects to a specific location by providing an audible tone when the subject was walking in the correct direction. The variations suggested by the subjects include providing them with: 1) a stereo-positioned representation of the target location; 2) verbal directions to the target location, based on the user's current position; and 3) real-time tactile cues indicating the direction of the target relative to the user's current position and body orientation. The presentation of "Walk/Don't Walk/Don't Start" information in these various sound/speech/tactile formats is also being evaluated.

METHODOLOGY--The Center for Education and Research in Safety (CERS) was contracted to replace existing "Walk/Don't Walk" (W/DW) pedestrian traffic lights with their own light-emitting diode (LED) display at two major intersections in Atlanta. These displays present the usual W/DW information; however, they are driven by a power source modulated by a digital signal carrying the ASCII equivalent of the visual display. A hand-held receiver has been designed to decode this signal and speak it to the user. The investigators will design and construct 12 different human interfaces and integrate these interfaces with the existing CERS receivers. These variations will be constructed by modularly combining alternative design choices: 1) sensors mounted on the head versus on the torso; 2) monaural output hardware versus stereo output; and 3) tone versus speech versus tactile output hardware.

  The prototype will be tested by 24 subjects in a 2×2×3 covariable format in a repeated-measures design with pre- and post-baseline trials and randomized test trials of the 12 interfaces. Investigators will record street crossing times, distance by which subjects miss their destination, maximum deviation from a straight-line path, hesitations and missteps taken during each street-crossing. Employing Likert scales, subjects will indicate 1) ease of use, 2) intuitiveness of use, 3) degree of obtrusiveness, and 4) overall rating of each interface. Analysis of covariance with repeated measures will be employed. Covariant measures will include baseline walking speed and the 2×2×3 covaried interface options. Dependent measures will include: time to reach destination, errors made (missteps), maximum deviation from a straight line, distance by which the target was missed, difficulties encountered, self-report Likert scale ratings, and subject comments. Both descriptive and inferential statistics will be used to evaluate the degree to which each interface accommodated the needs of each subject. The three data types (observed measures of performance, self report rating, and subject comments) will be combined to determine overall device usability.

  Based on the results of this study, a set of recommendations for the optimal interface design of a street-crossing system for people with visual impairment will be written, and implications to other potential applications for the disabled will be surmised.

PROGRESS--CERS has installed their system into the traffic lights of two intersections in Atlanta, and initial interface prototypes are being constructed and tested.

RESULTS--None to date.

FUTURE PLANS--Once an optimal street-crossing interface is determined, the researchers plan to investigate the use of this interface as a guide to specific locations such as building entrances, bus stops, elevators, rest rooms, stairwells, and exits.

[201] EVALUATION OF CENTRAL AND PERIPHERAL VISION ENHANCEMENT DEVICES FOR DRIVING

Janet P. Szlyk, PhD; Timothy Mcmahon, OD; William Seiple, PhD; Joan Stelmack, OD
Research and Development, VA Medical Center (West Side) 820 S. Damen Avenue Chicago, IL 60612 Department of Ophthalmology & Visual Sciences, University of Illinois at Chicago, Chicago, IL 60612; Department of Ophthalmology New York University Medical Center New York, NY 10016; Blind Rehabilitation Service, Hines VA Medical Center, Hines, IL 60141; email: janeszly@uicvm.uic.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C754-2RA)

PURPOSE--In the current phase of the project, we tested the hypothesis that a bioptic form of a telescopic lens system could be used effectively by patients with central vision loss. We also tested whether these subjects would benefit from a 13-week training curriculum in the use of the bioptic telescopes for navigation, mobility, and automobile driving.

METHODOLOGY--We recruited 25 subjects (13 men and 12 women) with central vision loss, ranging in age from 16 to 78. Each was assigned to one of the following three groups: Group 1 received the bioptic telescopes and training during the first 3 mo of the study; Group 2 received the lenses and training during the second 3 mo; Group 3 were given the lenses for a 3 mo period without any training. The three groups were statistically equivalent in age, sex, central scotoma size as measured with the Goldmann perimeter using the III-4-e target, visual acuity as measured with the Lighthouse Charts, and letter contrast sensitivity as measured with the Pelli-Robson Charts. The visual acuities of the subjects ranged from 20/50 to 20/200 in their better eyes without the telescope, and 20/20 to 20/60 through the telescopes.

  The Orientation and Mobility (OM) training was accomplished in five weekly sessions, with each week focusing on specific areas of visual function. These areas included locating objects, tracking stimuli, scanning skills, and enhancing visual memory. Following the OM training, subjects began a driving training curriculum consisting of eight weekly sessions on a road course within the grounds of the Hines Veterans Administration Medical Center. Once again, training focused on specific skills related to the use of the bioptic telescopes for driving.

  All were given an assessment battery after discrete time intervals consisting of clinical vision, psychophysical and laboratory, orientation and mobility, and driving assessment tests. Groups 1 and 2 were given an assessment battery at baseline (Test Day 1), at 3 mo (Test Day 2), and at 6 mo (Test Day 3). Group 3 was given an assessment battery at Test Days 1 and 2. The design allowed us to assess: A) The effects of lenses and training (Group 1, Test Day 1 to Test Day 2 and Group 2, Test Day 2 to Test Day 3); B) Test-retest reliability (Group 2, Test Day 1 to Test Day 2); C) The sustained effects of training and lenses (Group 1, Test Day 2 to Test Day 3); and D) The effects of lenses and no training (Group 3, Test Day 1 to Test Day 2).

  For the purpose of subsequent analyses, the individual tasks within the assessment battery were coded according to the primary visual skill involved in the task: recognition (26 tasks), mobility (19), peripheral identification (10), scanning (36), tracking (4), and visual memory (28).

RESULTS--The effects of lenses and training: both Group 1 and 2 showed substantial improvements from baseline to the assessment day following training. There were improvements in all visual skill categories beyond the test-retest reliabilities for these tasks. There were no statistically significant differences between them in the distribution of the percent of tasks improved (chi-square (5)=0.32, p=ns). Test-retest reliability: The percent change of Group 2 across all tasks averaged 11.9 percent. The sustained effects of training and lenses: The effects were sustained across all the visual skills categories following 3 mo of no training for Group 1 (chi-square (5)=2.49, p=ns). The effects of lenses and no training: Trained groups showed significantly greater improvement on the visual skills than did Group 3 in recognition (t(20)=2.0, p<0.05), peripheral identification (t(20)=2.6, p<0.02), and in scanning (t(20)=2.4, p<0.03). Following training in the use of the bioptic telescopes there was improvement in all skill categories beyond the test-retest reliability. The improvement in the skills was maintained for at least 3 mo following training. Training significantly improves performance in the following categories: Mobility, Peripheral Identification, and Scanning.

FUTURE PLANS--We are presently conducting Phase 3 of this project in which we are investigating the effectiveness of prism systems for driving and other functional tasks in patients with hemianopic visual field loss.

RECENT PUBLICATIONS FROM THIS RESEARCH

[202] MICROPHOTODIODE TILES AS IMPLANTS FOR THE RESTORATION OF VISUAL FUNCTION

Neal S. Peachey, PhD; Alan Y. Chow, MD; Machelle T. Pardue, PhD; Jay I. Perlman, MD, PhD; Gholam A. Peyman, MD
Hines VA Medical Center, Hines, IL 60141; Departments of Neurology and Ophthalmology, Loyola University Medical Center, Maywood, IL 60153; Optobionics Corporation, Wheaton, IL 60187; Louisiana State University Eye Center, New Orleans, LA 70112; email: npeache@orion.it.luc.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C1569-2RA)

PURPOSE--In blinding disorders such as age-related macular degeneration and retinitis pigmentosa, the degenerative process is often restricted to the photoreceptor cells, sparing the inner retinal layers. As a result, restoration of photoreceptor function could restore vision to a retinal area that is clinically blind. In the report period, we have carried out studies concerning whether a semiconductor-based device implanted in the subretinal space could achieve this purpose.

METHODOLOGY--Implants were constructed from doped silicon using standard semiconductor fabrication techiques. The devices generate electrical current in response to light stimulation without any external power supply. Devices were surgically implanted into the subretinal space of normal adult cats under halothane (0.9 percent) anesthesia.

  Post-operative studies were made at monthly intervals from cats sedated with ketamine (11 mg/kg) and xylazine (2 mg/kg). The retina was evaluated using fundus photography, and electrophysiological recordings were made at the retinal and cortical levels using visible light flashes, as well as infrared light stimuli provided by a light-emitting diode that were used to isolate implant function. At the completion of these studies, the retina was analyzed by light microscopy.

PROGRESS--Surgical techniques have been developed to routinely implant devices into the cat subretinal space. Implanted animals followed with a battery of post-operative electrophysiological assays to assess implant function have shown that the implant makes a functional connection with the visual system.

RESULTS--These studies demonstrated that in each animal successfully implanted with a microphotodiode-based device, infrared stimulation of the implant evoked a slow wave cortical potential that persisted for several months post-operatively; that implants retain a stable location within the subretinal space; that these devices will respond electrically to incident light with a response that is graded with intensity; that implants will continue to function within the subretinal space for many months without any external power supply; that retinal locations away from the implant respond normally to white light and retain a normal histologic appearance; that a model of photoreceptor degeneration may be produced by the solid disk implant in the subretinal space; and that the inner retinal layers overlying the implant were relatively spared.

FUTURE PLANS--We plan to investigate factors that affect the long-term production of cortical responses to improve device biocompatibility and function. We will also investigate the implant-retina interface using an isolated retina preparation that will allow a variety of electrophysiological and pharmacological techniques to be applied. Finally, we plan to utilize animal models of retinitis pigmentosa to assess implant function in retinas with pre-existing photoreceptor degeneration.

RECENT PUBLICATIONS FROM THIS RESEARCH

[203] FACTORS LEADING TO DISUSE OF LOW VISION DEVICES BY VISUALLY IMPAIRED ELDERLY

Gary L. Mancil, OD, FAAO; Gale Watson, MAEd; William de l'Aune, PhD; Rickilyn M. Mancil, MA, COMS; Jim Boyette, MS; Melissa Mitchell, BS
Vision Rehabilitation Research Laboratory, Salisbury VA Medical Center, Salisbury, NC 28146; Atlanta Rehabilitation Research and Development Center; Atlanta VA Medical Center; Decatur, GA 30033; email: mancil.gary@forum.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #C814-2RA)

PURPOSE--In this 3-year study, the project team evaluated long-term use patterns of low vision devices by older veterans to identify causes of disuse of previously successful devices. The following goals were accomplished: 1) identification of a representative sample of older adults who have low vision; 2) prescription of low vision devices to address specific goals according to standardized protocols of comprehensive interdisciplinary low vision care; 3) collection of pretest/posttest data on selected ocular, physical, and psycho-social characteristics; and 4) implementation and evaluation of a monitoring system to document usage patterns over time.

METHODOLOGY--Of 140 enrolled subjects, 116 cognitively intact adults age 63-92 completed protocols and contributed data on long-term use of their devices. Ocular health, psycho-social, and low vision rehabilitation evaluations were completed. A "criterion referenced" training protocol was developed for the project, and 626 low vision devices were prescribed for stated goals. After pretest assessment, subjects were monitored for a mean of 15.96 months using large print post cards with a visual analog scale to rate satisfaction with each device mailed monthly. Reported problems triggered a return to the clinic for posttest assessment ("unsuccessful users"). The accuracy of the monitoring system was evaluated by post-testing 25 randomly selected successful users, and by selecting 7 of the 25 at random to undergo a home visit to confirm functional use.

PROGRESS--During the current year, we 1) collected posttest data any time dissatisfaction with a device was indicated; 2) analyzed data to identify factors associated with disuse; and 3) convened a panel of experts to evaluate the disuse results and make recommendations pertinent to the activities of manufacturers, clinicians, and agencies that fund research. Data collection continued through July 1998, when an expert panel meeting was held to review results of preliminary data. The experts included consumers, low vision therapists, an optometrist, an ophthalmologist, researchers, university personnel, and industry representatives.

RESULTS--A total 626 device-user pairs resulted from the 116 subjects who completed the project protocols and entered the follow-up program. A 94 percent rate of success was reported, compared to the current literature ranging from 13 to 85 percent success. There were 38 (6 percent) returned devices because of dissatisfaction (increase or decrease in vision, cosmesis, difficulty in use, discontinuation of the prescription goal, and so forth) A total of 588 (94 percent) device-user pairs were successful and remained in the project as successful users, including 63 (10 percent) device-user pairs in which the subjects died during the course of the investigation. Of those successful pairs, 15 devices broke and were replaced with the same device. An average of four devices was prescribed per subject, and the device-user pairs were followed for a mean of 15.96 mo. Twenty (3.2 percent) device-user pairs withdrew from the project or stopped returning their follow-up cards.

  The expert panel meeting provided several insights into the project: the removal of cognitively impaired subjects from the subject pool may have caused the increase over the current literature's findings of success; the definition of low vision may have varied from that used in previous studies; the removal of non-optical devices in the data analysis might further align the data (e.g., of 298 optical device-user pairs, 23 users returned devices and 6 withdrew from the study; representing a 12.7 percent disuse rate for optical devices alone).

FUTURE PLANS--Future plans include evaluating criterion referenced training protocols through controlled studies as a possible contributor to the high rate of subject success.

Go to top.

Previous
Contents
Next

Last revised Thu 04/29/1999