Ear and Hearing (Ear Hear)

Publications in this journal

• Article: Editorial: Changes, Transitions, and Impact

  Brenda M. Ryals
  [show abstract] [hide abstract]
  ABSTRACT: An abstract is unavailable. This article is available as HTML full text and PDF.
  Ear and Hearing 01/2011; 32(1):1.
• Article: The Impact of Noise and Hearing Loss on the Processing of Simultaneous Sentences

  Virginia Best, Frederick J. Gallun, Christine R. Mason, Gerald Jr Kidd, Barbara G. Shinn-Cunningham
  [show abstract] [hide abstract]
  ABSTRACT: Objectives: To examine the impact of hearing impairment on a listener's ability to process simultaneous spoken messages. Design: Nine young listeners with bilateral sensorineural hearing loss and nine young listeners with normal hearing participated in this study. Two messages of equal level were presented separately to the two ears. The messages were systematically degraded by adding speech-shaped noise. Listeners performed a single task in which report of one message was required and a dual task in which report of both messages was required. Results: As the level of the added noise was increased, performance on both single and dual tasks declined. In the dual task, performance on the message reported second was poorer and more sensitive to the noise level than performance on the message reported first. When compared to listeners with normal hearing, listeners with hearing loss showed a larger deficit in recall of the second message than the first. This difference disappeared when performance of the hearing loss group was compared to that of the normal-hearing group at a poorer signal to noise ratio. Conclusions: A listener's ability to process a secondary message is more sensitive to noise and hearing impairment than the ability to process a primary message. Tasks involving the processing of simultaneous messages may be useful for assessing hearing handicap and the benefits of rehabilitation in realistic listening scenarios.
  Ear and Hearing 03/2010; 31(2):213-220.
• Article: When an Air-Bone Gap is not a Sign of a Middle Ear Conductive Hearing Loss: RE: Watanabe, T., Bertoli, S., & Probst, R. (2008). Transmission pathways of vibratory stimulation as measured by subjective thresholds and DPOAEs. Ear Hear, 29, 667-673

  Haim Sohmer, Jean-Yves Sichel, Ronen Perez, Cahtia Adelman
  [show abstract] [hide abstract]
  ABSTRACT: An abstract is unavailable. This article is available as HTML full text and PDF.
  Ear and Hearing 01/2009; 30(1):147-148.
• Article: Anatomy and physiology of hearing for audiologists.

  Paul J Abbas
  Ear and Hearing 07/2008; 29(3):476.
• Article: Evaluation of an aided TEN test for diagnosis of dead regions in the cochlea.

  Josephine Marriage, Brian C J Moore, Vivian Ogg, Michael A Stone
  [show abstract] [hide abstract]
  ABSTRACT: To compare results for the original version of the threshold equalizing noise [TEN(HL)] test for diagnosis of dead regions (DRs) in the cochlea, using stimuli presented via headphones, with results for an aided version of the test, the ATEN test, in which subjects listened to stimuli presented in free field using their own hearing aids. The test tones were warble tones for both the TEN(HL) and the ATEN test. Twenty-five subjects (12 males and 13 females), aged between 12 and 19 yr, with severe or profound sensorineural hearing loss were tested. For each test, two levels of the TEN were used, chosen to fall within the comfortable range of levels for the individual subject. A DR was considered to be present when the TEN(HL) produced at least 10 dB of masking and when the masked threshold was at least 10 dB above the nominal TEN(HL) level. Measurements of the outputs of the hearing aids in response to the TEN(HL) plus the test tones were obtained using a KEMAR acoustic manikin to assess the extent to which distortion or compression might have influenced the outcomes. For the TEN(HL) test, the results were often inconclusive, because the TEN(HL) could not be made sufficiently intense to give at least 10 dB of masking. The incidence of these inconclusive cases was markedly reduced for the ATEN test. There were more positive diagnoses of DRs for the ATEN test than for the TEN(HL) test. The KEMAR measurements indicated that distortion, compression, and/or feedback cancellation probably influenced the outcomes in some cases, leading to a moderate incidence of false positives for the ATEN test, and also some "missed" cases. The ATEN test leads to a lower incidence of inconclusive results than the TEN(HL) test in the diagnosis of DRs in people with severe to profound hearing loss. However, for some hearing aids the gain changed rapidly as a function of frequency, which undermined the validity of the ATEN test. Also, some hearing aids introduced distortion that probably affected the outcome of the test and gave misleading results. Hence, the ATEN test cannot be recommended for use in the clinic.
  Ear and Hearing 07/2008; 29(3):392-400.
• Source

  Available from: washington.edu

  Article: Speech evoked potentials: from the laboratory to the clinic.

  Brett A Martin, Kelly L Tremblay, Peggy Korczak
  [show abstract] [hide abstract]
  ABSTRACT: Speech-evoked auditory event-related potentials (ERPs) provide insight into the neural mechanisms underlying speech processing. For this reason, ERPs are of great value to hearing scientists and audiologists. This article will provide an overview of ERPs frequently used to examine the processing of speech and other sound stimuli. These ERPs include the P1-N1-P2 complex, acoustic change complex, mismatch negativity, and P3 responses. In addition, we focus on the application of these speech-evoked potentials for the assessment of (1) the effects of hearing loss on the neural encoding of speech allowing for behavioral detection and discrimination; (2) improvements in the neural processing of speech with amplification (hearing aids, cochlear implants); and (3) the impact of auditory training on the neural processing of speech. Studies in these three areas are reviewed and implications for audiologists are discussed.
  Ear and Hearing 07/2008; 29(3):285-313.
• Article: Frequency-specific electrocochleography indicates that presynaptic and postsynaptic mechanisms of auditory neuropathy exist.

  Catherine M McMahon, Robert B Patuzzi, William P R Gibson, Halit Sanli
  [show abstract] [hide abstract]
  ABSTRACT: The physiological mechanisms underlying auditory neuropathy (AN) remain unclear and it is likely that the multiple disruptions are classified under the broadly defined term. Cochlear implantation is being more widely used in this population to bypass the suspected site-of-lesion although a number of cases have been identified within the Sydney Cochlear Implant Centre where this management strategy has been unsuccessful. It is likely that this relates to the different physiological mechanisms underlying AN. To investigate the site-of-lesion in AN, frequency-specific round window electrocochleography (ECochG) was used to assess local hair-cell, dendritic, and axonal currents generated within the cochlea in 14 subjects with AN and compared with responses from two normally hearing subjects. ECochG results were then compared with electrically evoked auditory brain stem response (EABR) measured after cochlear implantation. The results of this study demonstrate that two dominant patterns of ECochG waveforms (produced by a high-frequency alternating tone burst) can be identified in this population of AN subjects: (a) gross waveform showing a prolonged summating potential (SP) latency that, in most cases, is followed by a small compound action potential; and (b) gross waveform showing a normal latency SP waveform followed by a broad negative potential [assumed to reflect the dendritic potential (DP) identified in anaesthetized guinea-pigs]. This study demonstrates that in most subjects (n = 7) with a prolonged latency SP but no DP, normal morphology EABR waveforms were elicited for all electrode channels. On the other hand, all subjects (n = 7) who showed a normal latency SP followed by a broad negative DP, showed EABR waveforms that were absent or having poor wave V morphology. The authors' interpretation of this is that ECochG results may provide a classification of AN into pre- and postsynaptic lesions. We suggest that a presynaptic and postsynaptic type of AN exist that may have implications for the fitting of cochlear implants.
  Ear and Hearing 07/2008; 29(3):314-25.
• Article: Comparison of speech recognition and localization performance in bilateral and unilateral cochlear implant users matched on duration of deafness and age at implantation.

  Camille C Dunn, Richard S Tyler, Sarah Oakley, Bruce J Gantz, William Noble
  [show abstract] [hide abstract]
  ABSTRACT: The purpose of this investigation was to compare speech recognition and localization performance of subjects who wear bilateral cochlear implants (CICI) with subjects who wear a unilateral cochlear implant (true CI-only). A total of 73 subjects participated in this study. Specifically, of the 73 subjects, 64 (32 CICI and 32 true CI-only) participated in the word recognition testing; 66 (33 CICI and 33 true CI-only) participated in the sentence recognition testing; and 24 (12 CICI and 12 true CI-only) participated in the localization testing. Because of time constraints not all subjects completed all testing. The average age at implantation for the CICI and true CI-only listeners who participated in the speech perception testing was 54 and 55 yrs, respectively, and the average duration of deafness was 8 yrs for both groups of listeners. The average age at implantation for the CICI and true CI-only listeners who participated in the localization testing was 54 and 53 yrs, respectively, and the average duration of deafness was 10 yrs for the CICI listeners and 11 yrs for the true CI-only listeners. All speech stimuli were presented from the front. The test setup for everyday-sound localization comprised an eight-speaker array spanning, an arc of approximately 108 degrees in the frontal horizontal plane. Average group results were transformed to Rationalized Arcsine Unit scores. A comparison in performance between the CICI score and the true CI-only score in quiet revealed a significant difference between the two groups with the CICI group scoring 19% higher for sentences and 24% higher for words. In addition, when both cochlear implants were used together (CICI) rather than when either cochlear implant was used alone (right CI or left CI) for the CICI listeners, results indicated a significant binaural summation effect for sentences and words. The average group results in this study showed significantly greater benefit on words and sentences in quiet and localization for listeners using two cochlear implants over those using only one cochlear implant. One explanation of this result might be that the same information from both sides are combined, which results in a better representation of the stimulus. A second explanation might be that CICI allow for the transfer of different neural information from two damaged peripheral auditory systems leading to different patterns of information summating centrally resulting in enhanced speech perception. A future study using similar methodology to the current one will have to be conducted to determine if listeners with two cochlear implants are able to perform better than listeners with one cochlear implant in noise.
  Ear and Hearing 07/2008; 29(3):352-9.
• Article: Auditory evoked potentials: basic principles and clinical application.

  Anu Sharma
  Ear and Hearing 07/2008; 29(3):476-7.
• Article: Cochlear implants, second edition.

  Robyn J Chapman
  Ear and Hearing 07/2008; 29(3):477-8.
• Article: Cochlear implantation results in patients with Kearns-Sayre syndrome.

  Sipke Pijl, Brian D Westerberg
  [show abstract] [hide abstract]
  ABSTRACT: To examine the evidence for a cochlear, retrocochlear, or central site of lesion for deafness in two cochlear implant recipients with Kearns-Sayre Syndrome (KSS). Speech perception data and electrically evoked Auditory Brainstem (eABR) and Middle Latency Responses were obtained in two patients with KSS and compared with a group of non-KSS implant recipients. Speech perception data and electrophysiological responses for the patients with KSS were similar to those obtained in non-KSS patients. Results are consistent with an initial cochlear site of lesion for deafness in KSS, and with relative sparing of the central auditory pathway early in the disease.
  Ear and Hearing 07/2008; 29(3):472-5.
• Article: Music perception of cochlear implant users compared with that of hearing aid users.

  Valerie Looi, Hugh McDermott, Colette McKay, Louise Hickson
  [show abstract] [hide abstract]
  ABSTRACT: To investigate the music perception skills of adult cochlear implant (CI) users in comparison with hearing aid (HA) users who have similar levels of hearing impairment. It was hypothesized that the HA users would perform better than the CI recipients on tests involving pitch, instrument, and melody perception, but similarly for rhythm perception. Fifteen users of the Nucleus CI system and 15 HA users participated in a series of music perception tests. All subjects were postlingually deafened adults, with the HA subjects being required to meet the current audiological criteria for CI candidacy. A music test battery was designed for the study incorporating four major tasks: (1) discrimination of 38 pairs of rhythms; (2) pitch ranking of one-octave, half-octave, and quarter-octave intervals; (3) instrument recognition incorporating three subtests, each with 12 different instruments or ensembles; and (4) recognition of 10 familiar melodies. Stimuli were presented via direct audio input at comfortable presentation levels. The test battery was administered to each subject on two separate occasions, approximately 4 mo apart. The results from the rhythm test were 93% correct for the CI group and 94% correct for the HA group; these scores were not significantly different. For the pitch test, there was a significant difference between the HA group and the CI group (p < 0.001), with higher mean scores recorded by the HA group for all three interval sizes. The CI subject group was unable to rank pitches a quarter-octave apart, only scoring at chance level for this interval size. In the instrument recognition test, although there was no significant difference between the mean scores of the two groups, both groups obtained significantly higher scores for the subtest incorporating single instrument stimuli than those incorporating multiple instrumentations (p < 0.001). In the melody test, there was a significant difference between the implantees' mean score of 52% correct and the HA group's mean of 91% (p < 0.001). As hypothesized, results from the two groups were almost identical for the rhythm test, with the HA group performing significantly better than the CI group on the pitch and melody tests. However, there was no difference between the groups in their ability to identify musical instruments or ensembles. The results of this study indicate that HA users with similar levels of hearing loss perform at least equal to, if not better than, CI users on these music perception tests. However, despite the differences between scores obtained by the CI and HA subject groups, both these subject groups were largely unable to achieve accurate or effective music perception, regardless of the device they used.
  Ear and Hearing 07/2008; 29(3):421-34.
• Source

  Available from: rush.edu

  Article: Identification of environmental sounds with varying spectral resolution.

  Valeriy Shafiro
  [show abstract] [hide abstract]
  ABSTRACT: This study investigated the identification of familiar environmental sounds with varying spectral resolution to establish (1) the number of frequency channels needed to perceive a large heterogeneous set of familiar environmental sounds, (2) the role of cross-channel asynchrony in identification performance, and (3) the acoustic correlates of the spectral resolution required for identification. In experiment 1, 60 normal-hearing listeners identified environmental sounds in a 60-alternative closed--set response task as a function of six spectral resolution conditions (i.e., 2, 4, 8, 16, 24, and 32 frequency channels) obtained with an envelope-vocoder. In experiment 2, identification accuracy for varying amounts of cross-channel asynchrony was determined for sounds with preserved and degraded fine spectral structure in 10 normal-hearing listeners. Experiment 3 examined identification performance of 72 listeners across six spectral resolution conditions as in experiment 1, but using three different signal processing methods designed to minimize cross-channel asynchrony across channels. Follow-up acoustic and discriminant analyses were carried out to identify parameters that can distinguish environmental sounds based on required spectral resolution. Identification accuracy tended to improve with increasing spectral resolution reaching the maximum of 76%. However, in experiment 1, performance did not change significantly beyond eight channels, whereas identification accuracy of some sounds declined with increasing spectral resolution. In experiment 2, increases in cross-channel asynchrony for sounds with preserved fine spectra had a small, but significant negative effect on identification. However, minimizing the amount of asynchrony had no significant effect on the overall identification of spectrally degraded sounds in experiment 3. Acoustic analysis indicated several spectral and temporal measures that differed significantly between sounds that required eight or fewer channels and those that required 16 or more channels for 70% correct identification. Discriminant analysis revealed that the sounds could be classified into high- and low-required spectral resolution groups with 83% accuracy based on only two acoustic parameters: the number of bursts in the envelope and the standard deviation of spectral centroid velocity. Increasing spectral resolution generally had a positive effect on identification of familiar environmental sounds. However, across conditions performance accuracy remained well-below that of control stimuli with preserved fine spectra, despite becoming asymptotic above eight channels. Cross-channel asynchrony introduced during vocoder processing, although detrimental for some sounds, was not a major factor that prevented further improvement in overall accuracy. A spectral resolution greater than 32 channels, along with additional fine spectral and temporal information may be required for identification of a number of environmental sounds. This study provides a preliminary basis for optimizing environmental sound perception by cochlear implant users by highlighting the role of several acoustic factors important for environmental sound identification.
  Ear and Hearing 07/2008; 29(3):401-20.
• Article: Test-retest repeatability of distortion product otoacoustic emissions.

  Wolfgang Wagner, Guido Heppelmann, Reinhard Vonthein, Hans Peter Zenner
  [show abstract] [hide abstract]
  ABSTRACT: Distortion product otoacoustic emissions (DPOAE) have become part of routine audiological diagnostics. The large scale of clinical DPOAE applications, such as screening of hearing in infants, objective estimation of hearing status, distinction between cochlear and retrocochlear origin of sensorineural hearing loss, exclusion of psychogenic hearing loss, monitoring of hearing during administration of ototoxic drugs, and others illustrates the significance of this audiological tool. In all diagnostic tests, knowledge about the procedure's test-retest repeatability is of crucial importance, to allow for distinction between measurement deviations and true physiological or pathological changes in monitoring over time. Measurements of DPOAE were performed in triplicate in 80 normally hearing ears of 40 subjects. Both immediate remeasurements with the ear probe left in place [single-fit mode (SF-mode)] and remeasurements after approximately 5 to 10 days [multiple-fit mode (MF-mode)] were included. DPOAE primary tone levels were varied in 5 dB steps from L2 = 60 to 20 dB SPL (L1 = L2 x 0.4 + 39 dB SPL) and within the frequency range f2 = 1 to 6 kHz. Repeatability of DPOAE was evaluated by the standard error of measurement (Sm), reliability (Cronbach alpha), absolute differences between measurements, 95% confidence intervals, and repeatability standard deviations. Sm averaged 0.67 dB over all frequencies and primary tone levels in the SF-mode, and 1.44 dB in the MF-mode, respectively. As expected, test-retest repeatability declined with decreasing primary tone levels; however, repeatability values were still mostly satisfactory with the lower primary tone levels. For the exemplary primary tone level combination of L1/L2 = 63/60 dB SPL, which is close to common clinical paradigms, the difference between two DPOAE measurements under the reported test conditions could be considered statistically significant (p = 0.05) if it exceeded 0.7 to 1.3 dB in the range 1 to 5 kHz and 2.3 dB for 6 kHz in the SF-mode, when compared with 1.8 to 2.7 dB for 1 to 5 kHz and 3.7 dB for 6 kHz in the MF-mode. Signal to noise ratio (SNR) did not seem to have a large influence on repeatability, as long as SNR was within 6 to 35 dB, which covers the range of most clinical DPOAE measurements. The DPOAE-test-retest study presented here is to our knowledge the first, which combines variation of primary tone levels, assessment of both SF- and MF-modes, and comparison of the two modalities within the same subjects. Although the measurements were conducted under practical conditions resembling the clinical setting, repeatability was generally good. The widely used minimum SNR of 6 dB seems to be a recommendable criterion when considering both practicability and measurement quality under clinical conditions. The current findings underline the suitability of DPOAE as a monitoring tool of cochlear status over time. The data are intended to assist the clinician and the scientist in the correct interpretation of DPOAE level changes in the test-retest situation.
  Ear and Hearing 07/2008; 29(3):378-91.
• Article: Establishment of age-specific normative data for the canadian French version of the hearing in noise test for children.

  Véronique Vaillancourt, Chantal Laroche, Christian Giguère, Sigfrid D Soli
  [show abstract] [hide abstract]
  ABSTRACT: A Canadian French version of the Hearing in Noise Test (HINT) has been developed to assess children's ability to recognize speech in noise. To avoid testing a large number of children in each clinical test site to establish soundfield norms, a protocol based on the use of correction factors has been proposed and validated in the current study. More specifically, the objective of this study was to provide a protocol for the establishment of age-specific normative data for the Canadian French HINT for children to facilitate its clinical use and allow comparing an individual child's performance with that of age-matched normal hearing children. Using the proposed protocol, a limited number of normal hearing adults are tested in each HINT condition to correct the adult headphone norms for the soundfield in question, and the correction factors established in the current study are then applied to generate age-specific soundfield norms. Mean adult performance values obtained in a given soundfield are entered into the HINT software, which automatically derives the soundfield adult norms, age-specific children norms, and percentile rankings. Speech reception thresholds (SRT) for sentences were measured in 70 native French-speaking subjects to establish mean performances across various age groups, and correction factors were calculated by comparing performance in each age group with adult performance. To validate the normalization protocol, 28 additional subjects were tested in a new soundfield. The correction factors were applied to adult performance (N = 15) and the resulting predicted scores were compared with measured performance in a group of 9-yr olds (N = 13). Statistical analyses indicate that SRTs decrease with age and reach adult values in older children (12-yr olds). Correction factors are therefore provided for children 6 to 12 yrs old. Spatial separation advantage, the improvement in SRT when speech and noise are spatially separated, also improves with age. The correction factors were effective in predicting mean SRTs for a previously untested age group in all HINT conditions apart from the quiet condition. The difference between predicted and measured performances was less than 0.5 dB for the noise conditions but exceeded 4 dB in the quiet condition. The reliability of SRT measures was determined, with an overall within-subjects SD of repeated measurements of 0.7 dB for the noise front condition. No learning effect was found in the current data. Correction factors can be used to predict performance on the HINT in a group of normal-hearing children in all HINT conditions, apart from quiet. Findings of the current study concur with the literature on age effects in auditory processing abilities, where performance on a variety of auditory tasks has been demonstrated to increase with age to reach adult-like values in adolescence or past 10 yrs.
  Ear and Hearing 07/2008; 29(3):453-66.
• Article: Psychophysical versus physiological spatial forward masking and the relation to speech perception in cochlear implants.

  Michelle L Hughes, Lisa J Stille
  [show abstract] [hide abstract]
  ABSTRACT: The primary goal of this study was to determine if physiological forward masking patterns in cochlear implants are predictive of psychophysical forward masking (PFM) patterns. It was hypothesized that the normalized amount of physiological masking would be positively correlated with the normalized amount of psychophysical masking for different masker-probe electrode separations. A secondary goal was to examine the relation between the spatial forward masking patterns and speech perception performance. It was hypothesized that subjects with less channel interaction overall (either psychophysically or physiologically) would have better speech perception ability because of better spectral resolution. Data were collected for 18 adult cochlear implant recipients [N = 9 Clarion CII or HiRes 90K, N = 9 Nucleus 24R(CS)]. Physiological spatial forward masking patterns were obtained with the electrically evoked compound action potential (ECAP) through the implant telemetry system. PFM patterns were obtained using a three-interval, two-alternative forced-choice adaptive procedure. Both measures used a fixed probe electrode with varied masker location. For each subject, spatial forward masking patterns were obtained for three probe electrodes with five masker locations per probe. On an individual basis, the correlation between ECAP FM and PFM was strong for 10 subjects (r = 0.68-0.85, p <or= 0.02), moderately strong for two subjects (r = 0.54-0.55, p = 0.06-0.07), and poor for six subjects (r = 0.13-0.45, p > 0.14). Results across subjects and electrodes showed a highly significant correlation between ECAP FM and PFM (r = 0.55, p < 0.0001); the correlation was strongest for basal electrodes. There was no significant correlation between speech perception and ECAP FM or PFM. Subjects whose ECAP FM patterns correlated well with PFM patterns generally had the poorest speech perception and subjects with the poorest correlations had the best speech perception. ECAP FM and PFM patterns correlated well for two-thirds of the subjects. Although the group correlation was statistically significant, ECAP FM patterns only accounted for 30% of the variance in the PFM measures. This suggests that the ECAP measures alone are not sufficient for accurately predicting PFM patterns for individual subjects.
  Ear and Hearing 07/2008; 29(3):435-52.
• Article: Lexical tone recognition with an artificial neural network.

  Ning Zhou, Wenle Zhang, Chao-Yang Lee, Li Xu
  [show abstract] [hide abstract]
  ABSTRACT: Tone production is particularly important for communicating in tone languages such as Mandarin Chinese. In the present study, an artificial neural network was used to recognize tones produced by adult native speakers. The purposes of the study were (1) to test the sensitivity of the neural network to speaker variation typically in adult speaker groups, (2) to evaluate two normalization procedures to overcome the effects of speaker variation, and (3) to compare tone recognition performance of the neural network with that of the human listeners. A feedforward multilayer neural network was used. Twenty-nine adult native Mandarin Chinese speakers were recruited to record tone samples. The F0 contours of the vowel part of the 1044 monosyllabic words recorded were extracted using an autocorrelation method. Samples from the F0 contours were used as inputs to the neural network. The efficacy of the neural network was first tested by varying the number of inputs and the number of neurons in the hidden layer from 1 to 16. The sensitivity of the neural network to speaker variation was tested by (1) using the raw F0 data from speech tokens of a number of randomly drawn speakers that varied from 1 to 29, (2) using the raw F0 data from speech tokens of either male-only or female-only speakers, and (3) using two sets of normalized F0 data (i.e., tone 1-based normalization and first-order derivative) from speech tokens from a number of randomly drawn speakers that varied from 1 to 29. The recognition performance of the neural network under several experimental conditions was compared with the corresponding recognition performance of 10 normal-hearing, native Mandarin Chinese speaking adult listeners. Three inputs and four hidden neurons were found to be sufficient for the neural network to perform at about 85% correct using speech samples without normalization. The performance of the neural network was affected by variation across speakers particularly between genders. Using the tone 1-based normalization procedure, the performance of the neural network improved significantly. The recognition accuracy of the neural network as a whole or for each tone was comparable with that of the human listeners. The neural network can be used to evaluate the tone production of Mandarin Chinese speaking adults with human listener-like recognition accuracy. The tone 1-based normalization procedure improves the performance of the neural network to human listener-like accuracy. The success of our neural network in recognizing tones from multiple speakers supports its utility for evaluating tone production. Further testing of the neural network with hearing-impaired speakers might reveal its potential use for clinical evaluation of tone production.
  Ear and Hearing 07/2008; 29(3):326-35.