Prediction of speech intelligibility in spatial noise and reverberation for normal-hearing and hearing-impaired listeners

Binaural speech intelligibility of individual listeners under realistic conditions was predicted using a model consisting of a gammatone filter bank, an independent equalization-cancellation (EC) process in each frequency band, a gammatone resynthesis, and the speech intelligibility index (SII). Hearing loss was simulated by adding uncorrelated masking noises (according to the pure-tone audiogram) to the ear channels. Speech intelligibility measurements were carried out with 8 normal-hearing and 15 hearing-impaired listeners, collecting speech reception threshold (SRT) data for three different room acoustic conditions (anechoic, office room, cafeteria hall) and eight directions of a single noise source (speech in front). Artificial EC processing errors derived from binaural masking level difference data using pure tones were incorporated into the model. Except for an adjustment of the SII-to-intelligibility mapping function, no model parameter was fitted to the SRT data of this study. The overall correlation coefficient between predicted and observed SRTs was 0.95. The dependence of the SRT of an individual listener on the noise direction and on room acoustics was predicted with a median correlation coefficient of 0.91. The effect of individual hearing impairment was predicted with a median correlation coefficient of 0.95. However, for mild hearing losses the release from masking was overestimated.     

Spectral integration of speech bands in normal-hearing and hearing-impaired listeners

This investigation examined whether listeners with mild-moderate sensorineural hearing impairment have a deficit in the ability to integrate synchronous spectral information in the perception of speech. In stage 1, the bandwidth of filtered speech centered either on 500 or 2500 Hz was varied adaptively to determine the width required for approximately 15%-25% correct recognition. In stage 2, these criterion bandwidths were presented simultaneously and percent correct performance was determined in fixed block trials. Experiment 1 tested normal-hearing listeners in quiet and in masking noise. The main findings were (1) there was no correlation between the criterion bandwidths at 500 and 2500 Hz; (2) listeners achieved a high percent correct in stage 2 (approximately 80%); and (3) performance in quiet and noise was similar. Experiment 2 tested listeners with mild-moderate sensorineural hearing impairment. The main findings were (1) the impaired listeners showed high variability in stage 1, with some listeners requiring narrower and others requiring wider bandwidths than normal, and (2) hearing-impaired listeners achieved percent correct performance in stage 2 that was comparable to normal. The results indicate that listeners with mild-moderate sensorineural hearing loss do not have an essential deficit in the ability to integrate across-frequency speech information.     

Prediction of speech intelligibility for normal-hearing and cochlearly hearing-impaired listeners

The word recognition ability of 4 normal-hearing and 13 cochlearly hearing-impaired listeners was evaluated. Filtered and unfiltered speech in quiet and in noise were presented monaurally through headphones. The noise varied over listening situations with regard to spectrum, level, and temporal envelope. Articulation index theory was applied to predict the results. Two calculation methods were used, both based on the ANSI S3.5-1969 20-band method [S3.5-1969 (American National Standards Institute, New York)]. Method I was almost identical to the ANSI method. Method II included a level- and hearing-loss-dependent calculation of masking of stationary and on-off gated noise signals and of self-masking of speech. Method II provided the best prediction capability, and it is concluded that speech intelligibility of cochlearly hearing-impaired listeners may also, to a first approximation, be predicted from articulation index theory.     

The ability of listeners to use recovered envelope cues from speech fine structure

Recent work has demonstrated that auditory filters recover temporal-envelope cues from speech fine structure when the former were removed by filtering or distortion. This study extended this work by assessing the contribution of recovered envelope cues to consonant perception as a function of the analysis bandwidth, when vowel-consonant-vowel (VCV) stimuli were processed in order to keep their fine structure only. The envelopes of these stimuli were extracted at the output of a bank of auditory filters and applied to pure tones whose frequency corresponded to the original filters' center frequencies. The resulting stimuli were found to be intelligible when the envelope was extracted from a single, wide analysis band. However, intelligibility decreases from one to eight bands with no further decrease beyond this value, indicating that the recovered envelope cues did not play a major role in consonant perception when the analysis bandwidth was narrower than four times the bandwidth of a normal auditory filter (i.e., number of analysis bands > or =8 for frequencies spanning 80 to 8020 Hz).     

The influence of age and high-frequency hearing loss on sensitivity to temporal fine structure at low frequencies (L)

Sensitivity to temporal fine structure (TFS) at low frequencies may be adversely affected by hearing loss at high frequencies even when absolute thresholds at low frequencies are within the normal range. However, in several studies suggesting this, the effects of hearing loss and age were confounded. Here, interaural phase discrimination (IPD) thresholds for pure tones at 500 and 750 Hz were measured for 39 subjects with ages from 61 to 83 yr. All subjects had near-normal audiometric thresholds at low frequencies, but thresholds varied across subjects at high frequencies. IPD thresholds were correlated with age. IPD thresholds for the test frequency of 750 Hz were weakly correlated with absolute thresholds at high frequencies, but these correlations became non-significant when the effect of age was partialed out. The results do not confirm that sensitivity to TFS at low frequencies is influenced by hearing loss at high frequencies, independently of age.     

The effects of age and cochlear hearing loss on temporal fine structure sensitivity, frequency selectivity, and speech reception in noise

Temporal fine structure (TFS) sensitivity, frequency selectivity, and speech reception in noise were measured for young normal-hearing (NHY), old normal-hearing (NHO), and hearing-impaired (HI) subjects. Two measures of TFS sensitivity were used: the "TFS-LF test" (interaural phase difference discrimination) and the "TFS2 test" (discrimination of harmonic and frequency-shifted tones). These measures were not significantly correlated with frequency selectivity (after partialing out the effect of audiometric threshold), suggesting that insensitivity to TFS cannot be wholly explained by a broadening of auditory filters. The results of the two tests of TFS sensitivity were significantly but modestly correlated, suggesting that performance of the tests may be partly influenced by different factors. The NHO group performed significantly more poorly than the NHY group for both measures of TFS sensitivity, but not frequency selectivity, suggesting that TFS sensitivity declines with age in the absence of elevated audiometric thresholds or broadened auditory filters. When the effect of mean audiometric threshold was partialed out, speech reception thresholds in modulated noise were correlated with TFS2 scores, but not measures of frequency selectivity or TFS-LF test scores, suggesting that a reduction in sensitivity to TFS can partly account for the speech perception difficulties experienced by hearing-impaired subjects.     

Duration discrimination of speech and tonal complex stimuli by normally hearing and hearing-impaired listeners

The ability to discriminate changes in the length of vowels and tonal complexes (filled intervals) and in the duration of closure in stop consonants and gaps in tonal complexes (unfilled intervals) was studied in three normally hearing and seven severely hearing-impaired listeners. The speech stimuli consisted of the vowels (i, I, u, U, a, A) and the consonants (p, t, k), and the tonal complexes consisted of digitally generated sinusoids at 0.5, 1, and 2 kHz. The signals were presented at conversational levels for each listener group, and a 3IFC adaptive procedure was used to estimate difference limens (DLs). The DLs for speech were similar to those for tonal complex stimuli in both the filled and unfilled conditions. Both normally and impaired-hearing listeners demonstrated greater acuity for changes in the duration of filled than unfilled intervals. Mean thresholds for filled intervals obtained from normally hearing listeners were smaller than those obtained from hearing-impaired listeners. For unfilled intervals, however, the difference between listener groups was not significant. A few hearing-impaired listeners demonstrated temporal acuity comparable to that of normally hearing listeners for several listening conditions. Implications of these results are discussed with regard to speech perception in normally and impaired-hearing individuals.     

Effects of noise and distortion on speech quality judgments in normal-hearing and hearing-impaired listeners

Noise and distortion reduce speech intelligibility and quality in audio devices such as hearing aids. This study investigates the perception and prediction of sound quality by both normal-hearing and hearing-impaired subjects for conditions of noise and distortion related to those found in hearing aids. Stimuli were sentences subjected to three kinds of distortion (additive noise, peak clipping, and center clipping), with eight levels of degradation for each distortion type. The subjects performed paired comparisons for all possible pairs of 24 conditions. A one-dimensional coherence-based metric was used to analyze the quality judgments. This metric was an extension of a speech intelligibility metric presented in Kates and Arehart (2005) [J. Acoust. Soc. Am. 117, 2224-2237] and is based on dividing the speech signal into three amplitude regions, computing the coherence for each region, and then combining the three coherence values across frequency in a calculation based on the speech intelligibility index. The one-dimensional metric accurately predicted the quality judgments of normal-hearing listeners and listeners with mild-to-moderate hearing loss, although some systematic errors were present. A multidimensional analysis indicates that several dimensions are needed to describe the factors used by subjects to judge the effects of the three distortion types.     

Effects of frequency response characteristics on speech discrimination and perceived intelligibility and pleasantness of speech for hearing-impaired listeners

Frequency response characteristics were selected for 14 hearing-impaired ears, according to six procedures. Three procedures were based on MCL measurements with speech bands of three bandwidths (1/3 octave, 1 octave, and 1 2/3 octaves). The other procedures were based on hearing thresholds, pure-tone MCLs, and pure-tone LDLs. The procedures were evaluated by speech discrimination testing, using nonsense syllables in noise, and by paired comparison judgments of the intelligibility and pleasantness of running speech. Speech discrimination testing showed significant differences between pairs of responses for only seven test ears. Nasals and glides were most affected by frequency response variations. Both intelligibility and pleasantness judgments showed significant differences for all test ears. Intelligibility in noise was less affected by frequency response differences than was intelligibility in quiet or pleasantness in quiet or in noise. For some ears, the ranking of responses depended on whether intelligibility or pleasantness was being judged and on whether the speech was in quiet or in noise. Overall, the three speech band MCL procedures were far superior to the others. Thus the studies strongly support the frequency response selection rationale of amplifying all frequency bands of speech to MCL. They also highlight some of the complications involved in achieving this aim.     

The effect of competing melodies on melody recognition by hearing-impaired and normal-hearing listeners

For a group of 30 hearing-impaired subjects and a matched group of 15 normal-hearing subjects (age range 13-17) the following data were collected: the tone audiogram, the auditory bandwidth at 1000 Hz, and the recognition threshold of a short melody presented simultaneously with two other melodies, lower and higher in frequency, respectively. The threshold was defined as the frequency distance required to recognize the test melody. It was found that, whereas the mean recognition threshold for the normal-hearing subjects was five semitones, it was, on the average, 27 semitones for the hearing-impaired subjects. Although the interindividual spread for the latter group was large, it did not correlate with the subjects' auditory bandwidth, nor with their musical experience or education.     

The interpretation of speech reception threshold data in normal-hearing and hearing-impaired listeners: steady-state noise

Speech-in-noise-measurements are important in clinical practice and have been the subject of research for a long time. The results of these measurements are often described in terms of the speech reception threshold (SRT) and SNR loss. Using the basic concepts that underlie several models of speech recognition in steady-state noise, the present study shows that these measures are ill-defined, most importantly because the slope of the speech recognition functions for hearing-impaired listeners always decreases with hearing loss. This slope can be determined from the slope of the normal-hearing speech recognition function when the SRT for the hearing-impaired listener is known. The SII-function (i.e., the speech intelligibility index (SII) against SNR) is important and provides insights into many potential pitfalls when interpreting SRT data. Standardized SNR loss, sSNR loss, is introduced as a universal measure of hearing loss for speech in steady-state noise. Experimental data demonstrates that, unlike the SRT or SNR loss, sSNR loss is invariant to the target point chosen, the scoring method or the type of speech material.     

Detection of a temporal gap in low-frequency narrow-band signals by normal-hearing and hearing-impaired listeners

Temporal processing ability in the hearing impaired was investigated in a 2IFC gap-detection paradigm. The stimuli were digitally constructed 50-Hz-wide bands of noise centered at 250, 500, and 1000 Hz. On each trial, two 400-ms noise samples were paired, shaped at onset and offset, filtered, and presented in the quiet with and without a temporal gap. A modified up-down procedure with trial-by-trial feedback was used to establish threshold of detection of the gap. Approximately 4 h of practice preceded data collection; final estimate of threshold was the average of six listening blocks. There were 10 listeners, 19-25 years old. Five had normal hearing; five had a moderate congenital sensorineural hearing loss with relatively flat audiometric configuration. Near threshold (5 dB SL), all listeners performed similarly. At 15 and 25 dB SL, the normal-hearing group performed better than the hearing-impaired group. At 78 dB SPL, equal to the average intensity of the 5-dB SL condition for the hearing impaired, the normal-hearing group continued to improve and demonstrated a frequency effect not seen in the other conditions. Substantial individual differences were found in both groups, though intralistener variability was as small as expected for these narrow-bandwidth signals.     

Effects of degradation of intensity, time, or frequency content on speech intelligibility for normal-hearing and hearing-impaired listeners

Many hearing-impaired listeners suffer from distorted auditory processing capabilities. This study examines which aspects of auditory coding (i.e., intensity, time, or frequency) are distorted and how this affects speech perception. The distortion-sensitivity model is used: The effect of distorted auditory coding of a speech signal is simulated by an artificial distortion, and the sensitivity of speech intelligibility to this artificial distortion is compared for normal-hearing and hearing-impaired listeners. Stimuli (speech plus noise) are wavelet coded using a complex sinusoidal carrier with a Gaussian envelope (1/4 octave bandwidth). Intensity information is distorted by multiplying the modulus of each wavelet coefficient by a random factor. Temporal and spectral information are distorted by randomly shifting the wavelet positions along the temporal or spectral axis, respectively. Measured were (1) detection thresholds for each type of distortion, and (2) speech-reception thresholds for various degrees of distortion. For spectral distortion, hearing-impaired listeners showed increased detection thresholds and were also less sensitive to the distortion with respect to speech perception. For intensity and temporal distortion, this was not observed. Results indicate that a distorted coding of spectral information may be an important factor underlying reduced speech intelligibility for the hearing impaired.     

Spatial release from masking in normally hearing and hearing-impaired listeners as a function of the temporal overlap of competing talkers

Listeners with sensorineural hearing loss are poorer than listeners with normal hearing at understanding one talker in the presence of another. This deficit is more pronounced when competing talkers are spatially separated, implying a reduced "spatial benefit" in hearing-impaired listeners. This study tested the hypothesis that this deficit is due to increased masking specifically during the simultaneous portions of competing speech signals. Monosyllabic words were compressed to a uniform duration and concatenated to create target and masker sentences with three levels of temporal overlap: 0% (non-overlapping in time), 50% (partially overlapping), or 100% (completely overlapping). Listeners with hearing loss performed particularly poorly in the 100% overlap condition, consistent with the idea that simultaneous speech sounds are most problematic for these listeners. However, spatial release from masking was reduced in all overlap conditions, suggesting that increased masking during periods of temporal overlap is only one factor limiting spatial unmasking in hearing-impaired listeners.     

Benefit of temporal fine structure to speech perception in noise measured with controlled temporal envelopes

Previous studies have assessed the importance of temporal fine structure (TFS) for speech perception in noise by comparing the performance of normal-hearing listeners in two conditions. In one condition, the stimuli have useful information in both their temporal envelopes and their TFS. In the other condition, stimuli are vocoded and contain useful information only in their temporal envelopes. However, these studies have confounded differences in TFS with differences in the temporal envelope. The present study manipulated the analytic signal of stimuli to preserve the temporal envelope between conditions with different TFS. The inclusion of informative TFS improved speech-reception thresholds for sentences presented in steady and modulated noise, demonstrating that there are significant benefits of including informative TFS even when the temporal envelope is controlled. It is likely that the results of previous studies largely reflect the benefits of TFS, rather than uncontrolled effects of changes in the temporal envelope.     

Algorithms for separating the speech of interfering talkers: evaluations with voiced sentences, and normal-hearing and hearing-impaired listeners

Two signal-processing algorithms, derived from those described by Stubbs and Summerfield [R.J. Stubbs and Q. Summerfield, J. Acoust. Soc. Am. 84, 1236-1249 (1988)], were used to separate the voiced speech of two talkers speaking simultaneously, at similar intensities, in a single channel. Both algorithms use fundamental frequency (FO) as the basis for segregation. One attenuates the interfering voice by filtering the cepstrum of the signal. The other is a hybrid algorithm that combines cepstral filtering with the technique of harmonic selection [T.W. Parsons, J. Acoust. Soc. Am. 60, 911-918 (1976)]. The algorithms were evaluated and compared in perceptual experiments involving listeners with normal hearing and listeners with cochlear hearing impairments. In experiment 1 the processing was used to separate voiced sentences spoken on a monotone. Both algorithms gave significant increases in intelligibility to both groups of listeners. The improvements were equivalent to an increase of 3-4 dB in the effective signal-to-noise ratio (SNR). In experiment 2 the processing was used to separate voiced sentences spoken with time-varying intonation. For normal-hearing listeners, cepstral filtering gave a significant increase in intelligibility, while the hybrid algorithm gave an increase that was on the margins of significance (p = 0.06). The improvements were equivalent to an increase of 2-3 dB in the effective SNR. For impaired listeners, no intelligibility improvements were demonstrated with intoned sentences. The decrease in performance for intoned material is attributed to limitations of the algorithms when FO is nonstationary.     

Upward spread of masking, hearing loss, and speech recognition in young and elderly listeners

Upward spreading of masking, measured in terms of absolute masked threshold, is greater in hearing-impaired listeners than in listeners with normal hearing. The purpose of this study was to make further observations on upward-masked thresholds and speech recognition in noise in elderly listeners. Two age groups were used: One group consisted of listeners who were more than 60 years old, and the second group consisted of listeners who were less than 36 years old. Both groups had listeners with normal hearing as well as listeners with mild to moderate sensorineural loss. The masking paradigm consisted of a continuous low-pass-filtered (1000-Hz) noise, which was mixed with the output of a self-tracking, sweep-frequency Bekesy audiometer. Thresholds were measured in quiet and with maskers at 70 and 90 dB SPL. The upward-masked thresholds were similar for young and elderly hearing-impaired listeners. A few elderly listeners had lower upward-masked thresholds compared with the young control group; however, their on-frequency masked thresholds were nearly identical to the control group. A significant correlation was found between upward-masked thresholds and the Speech Perception in Noise (SPIN) test in elderly listeners.     

Extension of a temporal model of frequency discrimination: intensity effects in normal and hearing-impaired listeners

The effects of intensity on the difference limen for frequency (DLF) in normal-hearing and in hearing-impaired listeners are incorporated into the temporal model of frequency discrimination proposed by Goldstein and Srulovicz [Psychophysics and Physiology of Hearing, edited by E. F. Evans and J.P. Wilson (Academic, New York, 1977)]. A simple extension of the temporal mode, which includes the dependence of phase locking on intensity, is sufficient to predict the effects of intensity on the DLF in normal-hearing listeners. To account for elevated DLFs in hearing-impaired listeners the impairment is modeled as a reduction in the synchrony of the discharge from VIIIth-nerve fibers that innervate the region of hearing loss. Constraints on the optimal processor and the validity of the temporal model at high frequencies are discussed.     

The impact of reverberant self-masking and overlap-masking effects on speech intelligibility by cochlear implant listeners (L)

The purpose of this study is to determine the relative impact of reverberant self-masking and overlap-masking effects on speech intelligibility by cochlear implant listeners. Sentences were presented in two conditions wherein reverberant consonant segments were replaced with clean consonants, and in another condition wherein reverberant vowel segments were replaced with clean vowels. The underlying assumption is that self-masking effects would dominate in the first condition, whereas overlap-masking effects would dominate in the second condition. Results indicated that the degradation of speech intelligibility in reverberant conditions is caused primarily by self-masking effects that give rise to flattened formant transitions.     

Place-pitch sensitivity and its relation to consonant recognition by cochlear implant listeners using the MPEAK and SPEAK speech processing strategies

Two related studies investigated the relationship between place-pitch sensitivity and consonant recognition in cochlear implant listeners using the Nucleus MPEAK and SPEAK speech processing strategies. Average place-pitch sensitivity across the electrode array was evaluated as a function of electrode separation, using a psychophysical electrode pitch-ranking task. Consonant recognition was assessed by analyzing error matrices obtained with a standard consonant confusion procedure to obtain relative transmitted information (RTI) measures for three features: stimulus (RTI stim), envelope (RTI env[plc]), and place-of-articulation (RTI plc[env]). The first experiment evaluated consonant recognition performance with MPEAK and SPEAK in the same subjects. Subjects were experienced users of the MPEAK strategy who used the SPEAK strategy on a daily basis for one month and were tested with both processors. It was hypothesized that subjects with good place-pitch sensitivity would demonstrate better consonant place-cue perception with SPEAK than with MPEAK, by virtue of their ability to make use of SPEAK's enhanced representation of spectral speech cues. Surprisingly, all but one subject demonstrated poor consonant place-cue performance with both MPEAK and SPEAK even though most subjects demonstrated good or excellent place-pitch sensitivity. Consistent with this, no systematic relationship between place-pitch sensitivity and consonant place-cue performance was observed. Subjects' poor place-cue perception with SPEAK was subsequently attributed to the relatively short period of experience that they were given with the SPEAK strategy. The second study reexamined the relationship between place-pitch sensitivity and consonant recognition in a group of experienced SPEAK users. For these subjects, a positive relationship was observed between place-pitch sensitivity and consonant place-cue performance, supporting the hypothesis that good place-pitch sensitivity facilitates subjects' use of spectral cues to consonant identity. A strong, linear relationship was also observed between measures of envelope- and place-cue extraction, with place-cue performance increasing as a constant proportion (approximately 0.8) of envelope-cue performance. To the extent that the envelope-cue measure reflects subjects' abilities to resolve amplitude fluctuations in the speech envelope, this finding suggests that both envelope- and place-cue perception depend strongly on subjects' envelope-processing abilities. Related to this, the data suggest that good place-cue perception depends both on envelope-processing abilities and place-pitch sensitivity, and that either factor may limit place-cue perception in a given cochlear implant listener. Data from both experiments indicate that subjects with small electric dynamic ranges (< 8 dB for 125-Hz, 205-microsecond/ph pulse trains) are more likely to demonstrate poor electrode pitch-ranking skills and poor consonant recognition performance than subjects with larger electric dynamic ranges.