Effect of stimulation rate on cochlear implant recipients' thresholds and maximum acceptable loudness levels.

Clinically, speech processor programs are created using electrical thresholds and maximum acceptable loudness levels (MALs) at several different stimulation rates to determine what rate will provide cochlear implant recipients with the best speech recognition when using fast-rate speech coding strategies. This study was designed to determine the difference in thresholds and MALs (expressed in the clinical unit, Current Level [CL]) for pairs of six rates spanning those available with the Nucleus 24 device (i.e., 250 to 2,400 pps/ch) using monopolar, 25 microsec/phase stimulation. Test-retest measures of threshold and MAL for each rate were obtained from seven adult Nucleus 24 recipients on each of 11 electrodes. The difference in threshold and in MAL between pairs of rates was dependent on the absolute CL. Below approximately 190 CL, thresholds and MALs decreased with increasing rate; above 210 CL, there was little change in threshold or MAL with increasing rate. Based on these findings, an approach to estimating threshold and MAL from one rate to another is suggested, pending further research.

Speech perception without hearing.

In this study of visual phonetic speech perception without accompanying auditory speech stimuli, adults with normal hearing (NH; n = 96) and with severely to profoundly impaired hearing (IH; n = 72) identified consonant-vowel (CV) nonsense syllables and words in isolation and in sentences. The measures of phonetic perception were the proportion of phonemes correct and the proportion of transmitted feature information for CVs, the proportion of phonemes correct for words, and the proportion of phonemes correct and the amount of phoneme substitution entropy for sentences. The results demonstrated greater sensitivity to phonetic information in the IH group. Transmitted feature information was related to isolated word scores for the IH group, but not for the NH group. Phoneme errors in sentences were more systematic in the IH than in the NH group. Individual differences in phonetic perception for CVs were more highly associated with word and sentence performance for the IH than for the NH group. The results suggest that the necessity to perceive speech without hearing can be associated with enhanced visual phonetic perception in some individuals.

Identification of speech by cochlear implant recipients with the multipeak (MPEAK) and spectral peak (SPEAK) speech coding strategies II. Consonants.

OBJECTIVE: The major objective of this study was to evaluate differences in consonant recognition with the Multipeak (MPEAK) and the Spectral Peak (SPEAK) speech coding strategies of the Nucleus-22 Cochlear Implant System. This objective was addressed by comparison of acoustic and electrode activation analyses of consonants with cochlear implant recipients' responses to these same consonant tokens when they used the two speech coding strategies. DESIGN: Nine subjects identified 14 English consonants with the MPEAK and SPEAK speech coding strategies. These strategies were compared with an ABAB design. Evaluation occurred during two weekly sessions after subjects used each strategy for at least 3 wk in everyday life. RESULTS: Group medial consonant [aCa] identification scores with the SPEAK strategy were significantly higher than with the MPEAK strategy (76.2% versus 67.5%; p < 0.001). This improvement was largely due to the significant increase in information transmitted for the place feature (p < 0.001) through accurate tracking of second formant transitions and spectrally specific stimulation patterns to differentiate [s] from [symbol see text] and [n] from [m], and the stop consonant bursts. For this reason, more nasal consonants were correctly identified with SPEAK, but there also were more non-nasal error responses when the nasal murmur was of unusually low amplitude. Consequently, significantly less information was transmitted for the nasality feature with SPEAK than MPEAK (p < 0.001). CONCLUSIONS: Electrical stimulation with the SPEAK strategy provided better spectral representation of the stop consonant bursts, tracking formant transitions into the following vowel, frication in the consonant [symbol see text], and the formants for the nasals [m] and [n] than with the MPEAK strategy. The marked improvement in recognition of the velar consonants, [g] and [k], which cannot be seen during speechreading, should allow greater ease and accuracy of communication with SPEAK than MPEAK.

Comparison of two methods for selecting minimum stimulation levels used in programming the Nucleus 22 cochlear implant.

Minimum stimulation levels for active electrodes in a Nucleus 22 cochlear implant were set at threshold (clinical default value) and raised levels (M = +2.04 dB) to determine if raised levels would improve recipients' understanding of soft speech sounds with the SPEAK speech coding strategy. Eight postlinguistically deaf adults participated in a 4-phase A1B1A2B2 test design. Speech recognition was evaluated with consonant-vowel nucleus-consonant (CNC) words in quiet and sentences in noise, both presented at 50, 60, and 70 dB SPL during 2 weekly sessions at the end of each phase. Group mean scores were significantly higher with the raised level program for words and phonemes at 50 and 60 dB SPL and for sentences at 50 and 70 dB SPL. All participants chose to use the raised level program in everyday life at the end of the study. The results suggest that clinical use of a raised level program for Nucleus 22 recipients has the potential to make soft sounds louder and, therefore, more salient in everyday life. Further research is needed to determine if this approach is appropriate for other cochlear implant devices.

Adjustment to hearing impairment II: audiological and demographic correlates.

To study adjustment to hearing impairment, clinical records from a five-center consortium (N = 1,008) were used to create a heterogeneous clinical database with results of audiometric tests, demographic and case history information, and responses to the Communication Profile for the Hearing Impaired (CPHI; Demorest & Erdman, 1986, 1987). Normative findings have been described previously (Erdman & Demorest, 1998). Hierarchical regression analyses revealed that audiometric variables were moderately correlated with communication performance, behavioral strategies, and personal adjustment. With hearing impairment controlled statistically, age and education effects were evident in many areas of adjustment; correlations between adjustment and gender were relatively weak; and marital status, employment status, and race/ethnicity were rarely significant correlates.

Adjustment to hearing impairment I: description of a heterogeneous clinical population.

To obtain data on adjustment to hearing impairment and its potential predictors, a consortium of five audiology clinics was established. Clinical records generated over 19-27 months were reviewed, and a database (N = 1,008) was compiled that contained standard audiometric test results, demographic and case history information, and responses to the Communication Profile for the Hearing Impaired (CPHI; Demorest & Erdman, 1986, 1987). Clinic populations differed on audiometric measures, gender, race/ethnicity, educational level, employment, marital, and hearing aid status, and on CPHI profiles, but not on age. Internal consistency of CPHI scales was higher than reported by Demorest and Erdman (1987), and standard errors of measurement were smaller for Communication Performance scales. The consortium database is sufficiently heterogeneous to provide normative data applicable to a wide range of local clinical populations and to support investigation of the correlates of adjustment to hearing impairment (see Erdman & Demorest, 1998).

Relationships between subjective ratings and objective measures of performance in speechreading sentences.

Ninety-six participants with normal hearing and 63 with severe-to-profound hearing impairment viewed 100 CID Sentences (Davis & Silverman, 1970) and 100 B-E Sentences (Bernstein & Eberhardt, 1986b). Objective measures included words correct, phonemes correct, and visual-phonetic distance between the stimulus and response. Subjective ratings were made on a 7-point confidence scale. Magnitude of validity coefficients ranged from .34 to .76 across materials, measures, and groups. Participants with hearing impairment had higher levels of objective performance, higher subjective ratings, and higher validity coefficients, although there were large individual differences. Regression analyses revealed that subjective ratings are predictable from stimulus length, response length, and objective performance. The ability of speechreaders to make valid performance evaluations was interpreted in terms of contemporary word recognition models.

Speech recognition at simulated soft, conversational, and raised-to-loud vocal efforts by adults with cochlear implants.

Ten postlinguistically deaf adults who used the Nucleus Cochlear Implant System and SPEAK speech coding strategy responded to vowels, consonants, words, and sentences presented sound-only at 70, 60, and 50 dB sound-pressure level. Highest group mean scores were at a raised-to-loud level of 70 dB for consonants (73%), words (44%), and sentences (87%); the highest score for vowels (70%) was at a conversational level of 60 dB. Lowest group mean scores were at a soft level of 50 dB for vowels (56%), consonants (47%), words (10%), and sentences (29%); all except subject 7 had some open-set speech recognition at this level. For the conversational level (60 dB), group mean scores for sentences and words were 72% and 29%, respectively. With this performance and sound-pressure level, it was observed that these subjects communicated successfully in a variety of listening situations. Given these subjects' speech recognition scores at 60 dB and the fact that 70 dB does not simulate the vocal effort used in everyday speaking situations, it is suggested that cochlear implant candidates and implantees be evaluated with speech tests presented at 60 dB instead of the customary 70 dB sound-pressure level to simulate benefit provided by implants in everyday life. Analysis of individuals' scores at the three levels for the four speech materials revealed different patterns of speech recognition among subjects (e.g., subjects 1 and 5). Future research on the relation between stimuli, sound processing, and subjects' responses associated with these different patterns may provide guidelines to select parameter values with which to map incoming sound onto an individual's electrical dynamic range between threshold and maximum acceptable loudness level to improve speech recognition.

Generalizability of speechreading performance on nonsense syllables, words, and sentences: subjects with normal hearing.

Ninety-six adults with normal hearing viewed three types of recorded speechreading materials (consonant-vowel nonsense syllables, isolated words, and sentences) on 2 days. Responses to nonsense syllables were scored for syllables correct and syllable groups correct; responses to words and sentences were scored in terms of words correct, phonemes correct, and an estimate of visual distance between the stimulus and the response. Generalizability analysis was used to quantify sources of variability in performance. Subjects and test items were important sources of variability for all three types of materials; effects of talker and day of testing varied but were comparatively small. For each type of material, alternative models of test construction and test-score interpretation were evaluated through estimation of generalizability coefficients as a function of test length. Performance on nonsense syllables correlated about .50 with both word and sentence measures, whereas correlations between words and sentences typically exceeded .80.

Identification of speech by cochlear implant recipients with the Multipeak (MPEAK) and Spectral Peak (SPEAK) speech coding strategies. I. Vowels.

OBJECTIVE: The main objective was to evaluate differences in performance associated with the two speech coding strategies. To achieve this objective, acoustic and electrical analyses of vowels identified by cochlear implant recipients were compared with their responses when they used the Multipeak (MPEAK) and the Spectral Peak (SPEAK) speech coding strategies of the Nucleus Cochlear Implant System. DESIGN: Nine subjects identified pure and r-colored English vowels with the two speech coding strategies. The two processing strategies were compared using an ABAB design. Evaluations were conducted at two weekly sessions after at least 3 wk of use with each strategy. RESULTS: Group vowel identification scores with the MPEAK versus the SPEAK strategy were not significantly different (72.3% and 73.4%, respectively). However, hierarchical loglinear analysis of group data showed that transmitted information of r-color, duration, and second-formant features was significantly better with the SPEAK than with the MPEAK strategy. In contrast, identification of the first formant feature was significantly better with the MPEAK than with the SPEAK strategy. Individual subjects had different error patterns in response to the 14 vowels. CONCLUSIONS: Electrical stimulation with the SPEAK strategy provides clearer spectral representation of second formant and duration information as well as second and third formant change in r-colored vowels than with the MPEAK strategy. Consequently, there was marked improvement in recognition of r-colored vowels with SPEAK compared with MPEAK. In contrast, transmitted information for first-formant features was significantly less with SPEAK than with MPEAK. This may have occurred because four instead of six to eight electrodes were assigned to first formant frequencies with SPEAK versus MPEAK.

Comparison of procedures for obtaining thresholds and maximum acceptable loudness levels with the nucleus cochlear implant system.

Two stimulus paradigms and two presentation methods were combined to form three procedures (keyboard, knob, and ascending loudness judgments with knob [ALJK]) to obtain detection thresholds and maximum acceptable loudness levels (MALs) from 11 adults with the Nucleus cochlear implant. Thresholds at which subjects correctly counted the number of stimulus bursts also were obtained. Keyboard detection thresholds were higher (Scheffé, p = .01) than knob and ALJK detection thresholds. Counted thresholds were obtained most efficiently by using keyboard detection thresholds as the initial level for testing. Keyboard MALs were highest, knob MALs intermediate, and ALJK MALs lowest (Scheffé, p = .001). MALs were obtained most efficiently with the ALJK procedure. Implications of these results for clinical practice are discussed.

Use of test-retest measures to evaluate performance stability in adults with cochlear implants.

OBJECTIVE: This study was designed to evaluate performance stability with test-retest measures of electrical thresholds, electrical maximum acceptable loudness levels, sound-field thresholds, and audition-only speech tests. The hypothesis was that the standard error of measurement differs from one individual to another for each test, and therefore test-retest measures need to be obtained from each subject. DESIGN: Test-retest data were obtained in three sets. For the first two sets, data were obtained over 4 successive weeks to determine day-to-day variability. The third set was obtained on 2 successive days, 2 to 10 months later at each subject's annual evaluation. The seven subjects were recipients of the Nucleus cochlear implant and used the Mini Speech Processor at least one year. They were selected because they were available for research testing. RESULTS: The ranges of electrical thresholds and maximum acceptable loudness levels obtained from each subject during the first two sets were substantially larger on some electrodes than on others. Between set 1 and set 2 for each subject, there were significant differences between thresholds and between maximum acceptable loudness levels on some electrodes. For thresholds as well as maximum acceptable loudness levels, the standard error of measurement across the three sets and 10 electrodes differed among subjects. The group standard error of measurement for sound-field thresholds was small (2 dB). The group standard error of measurement for the NU-6 word test scored according to phonemes (2.75%) was less than half that for the vowel (6.06%) and consonant (5.67%) tests. The standard error of measurement for each speech test differed among subjects. CONCLUSIONS: The standard error of measurement for thresholds as well as maximum acceptable loudness levels varies among electrodes in the same subject and among subjects across electrodes. The standard error of measurement also varies among speech tests for the group as well as among subjects for each speech test. There is little variation in test-retest measures of sound-field thresholds among subjects. These results suggest the clinical importance of obtaining test-retest measures for evaluating the stability of electrical thresholds, electrical maximum acceptable loudness levels, and speech tests.

A computational approach to analyzing sentential speech perception: phoneme-to-phoneme stimulus-response alignment.

A solution to the following problem is presented: Obtain a principled approach to studying error patterns in sentence-length responses obtained from subjects who were instructed to simply report what a talker had said. The solution is a sequence comparator that performs phoneme-to-phoneme alignment on transcribed stimulus and response sentences. Data for developing and testing the sequence comparator were obtained from 139 normal-hearing subjects who lipread (speechread) 100 sentences and from 15 different subjects who identified nonsense syllables by lipreading. Development of the sequence comparator involved testing two different costs metrics (visemes versus Euclidean distances) and two related comparison algorithms. After alignments with face validity were achieved, a validation experiment was conducted for which measures from random versus true stimulus-response sentence pairs were compared. Measures of phonemes correct and substitution uncertainty were found to be sensitive to the nature of the sentence pairs. In particular, correct phoneme matches were extremely rare in random pairings in comparison with true pairs. Also, an information-theoretic measure of uncertainty for substitutions in true versus random pairings showed that uncertainty was always higher for random than for true pairs.

Vocal communication in the first 18 months of life.

Fifty-one normally developing infants aged birth to 18 months, 10 or 11 in each of five age groups, were videorecorded in their homes before and after an expected change in the form of their vocalizations and under a set of conditions that reflected common daily occurrences. The vocalizations produced were coded according to their communicative contexts, defined in nonvocal behavioral terms. Communicative codes were assigned to seven major categories. The distribution of codes across categories was found to be different for different age groups. It varied between the first and second observations; however, the pattern of change differed across age groups. Data from individuals were transformed to proportions, to control for individual differences in productivity. They were then found to reflect differences in level of development of vocal communication. It was concluded that vocal communication follows an orderly developmental sequence in normally developing infants in the first 18 months of life.

Sources of variability in speechreading sentences: a generalizability analysis.

Generalizability theory (Cronbach, Gleser, Nanda, & Rajaratnam, 1972) was used to estimate the percentage of variance explained by three sources of variability in speechreading sentences: the subject, the talker, and the sentence materials. Videodisc recordings of the 100 CID Everyday Sentences (Davis & Silverman, 1970), spoken by a male and a female talker, were presented to 104 subjects with normal hearing. For performance on individual sentences (total number of words correct), the most important systematic sources of variability were the sentence (26.3%), the speechreader (10.5%), the talker (4.9%), and the interaction of talker and sentence (5.1%). Residual error accounted for 51.2% of the variance. Generalizability functions are presented, as a function of test length, for five models of test administration and interpretation. For 10-, 50-, and 100-item lists, generalizability is predicted to be .70, .92, and .96, respectively, for a single talker. Psychometric characteristics of these recordings of the CID sentences are also presented.

Lipreading sentences with vibrotactile vocoders: performance of normal-hearing and hearing-impaired subjects.

Three vibrotactile vocoders were compared in a training study involving several different speech perception tasks. Vocoders were: (1) the Central Institute for the Deaf version of the Queen's University vocoder, with 1/3-oct filter spacing and logarithmic output scaling (CIDLog) [Engebretson and O'Connell, IEEE Trans. Biomed. Eng. BME-33, 712-716 (1986)]; (2) the same vocoder with linear output scaling (CIDLin); and (3) the Gallaudet University vocoder designed with greater resolution in the second formant region, relative to the CID vocoders, and linear output scaling (GULin). Four normal-hearing subjects were assigned to either of two control groups, visual-only control and vocoder control, for which they received the CIDLog vocoder. Five normal-hearing and four hearing-impaired subjects were assigned to the linear vocoders. Results showed that the three vocoders provided equivalent information in word-initial and word-final tactile-only consonant identification. However, GULin was the only vocoder significantly effective in enhancing lipreading of isolated prerecorded sentences. Individual subject analyses showed significantly enhanced lipreading by the three normal-hearing and two hearing-impaired subjects who received the GULin vocoder. Over the entire training period of the experiment, the mean difference between aided and unaided lipreading of sentences by the GULin aided hearing-impaired subjects was approximately 6% words correct. Possible explanations for failure to confirm previous success with the CIDLog vocoder [Weisenberger et al., J. Acoust. Soc. Am. 86, 1764-1775 (1989)] are discussed.

Scope and design of the clinical trial of the Nucleus multichannel cochlear implant in children.

Although much has been learned in the 5 yr since the first child received a multichannel cochlear implant, we remain at the very early stages of this endeavor. The current evaluation procedures and performance data detailed in subsequent chapters represent the continuation of the efforts described above.

Speechreading sentences with single-channel vibrotactile presentation of voice fundamental frequency.

The main goal of this study was to investigate the efficacy of four vibrotactile speechreading supplements. Three supplements provided single-channel encodings of fundamental frequency (F0). Two encodings involved scaling and shifting glottal pulses to pulse rate ranges suited to tactual sensing capabilities; the third transformed F0 to differential amplitude of two fixed-frequency sinewaves. The fourth supplement added to one of the F0 encodings a second vibrator indicating high-frequency speech energy. A second goal was to develop improved methods for experimental control. Therefore, a sentence corpus was recorded on videodisc using two talkers whose speech was captured by video, microphone, and electroglottograph. Other experimental control issues included use of visual-alone control subjects, a multiple-baseline, single-subject design replicated for each of 15 normal-hearing subjects, sentence and syllable pre- and post-tests balanced for difficulty, and a speechreading screening test for subject selection. Across 17 h of treatment and 5 h of visual-alone baseline testing, each subject performed open-set sentence identification. Covariance analyses showed that the single-channel supplements provided a small but significant benefit, whereas the two-channel supplement was not effective. All subjects improved in visual-alone speechreading and maintained individual differences across the experiment. Vibrotactile benefit did not depend on speechreading ability.

Genetic implications of gender differences in the prevalence of spontaneous otoacoustic emissions.

Spontaneous otoacoustic emissions (SOAEs), which occur in about 40% of normal-hearing humans, do not have a firm explanation in auditory theory nor are their distributional properties well understood. To enhance our understanding of SOAEs, we have pooled data from three reports (Hammel, 1983; Strickland, Burns, & Tubis, 1985; Zurek, 1981) to assemble a large enough sample to assess the relevant hypotheses about the effects of ear and gender on their presence. The results, based on loglinear analyses of the pooled sample of 131 normal-hearing subjects, indicated that (a) the prevalence of SOAEs for female subjects [P(S/F) = .533] was significantly higher than that for male subjects [P(S/M) = .268]; (b) the percentage of right ears with SOAEs (36.6%) was significantly greater than that of left ears (25.2%); and (c) right and left ears were not independent with respect to the presence of SOAEs. These results can be explained by assuming that the (a) tendency to exhibit emissions is inherited, perhaps as a sex-linked trait and (b) ears are asymmetric with respect to the anatomical anomalies of the apical portion of the organ of Corti that may precipitate SOAEs.

Factor structure of the communication profile for the hearing impaired.

Factor structure of 23 scales of the Communication Profile for the Hearing Impaired (CPHI) (Demorest & Erdman, 1986, 1987) was examined in a sample of 1,226 patients who participated in the Aural Rehabilitation Program at Walter Reed Army Medical Center from 1985 through 1987. Five factors were identified: Adjustment, Reaction, Interaction, Communication Performance, and Communication Importance. To enhance clinical interpretation of the CPHI, it is recommended that factor scores be added to the communication profile.