Word Recognition In Noise Research Articles

Variation in normal hearing thresholds predicts word recognition in noise

Analyses of speech identification data collected from young, normal-hearing adults often do not include measures of listeners’ hearing. Sub-clinical variation in hearing, it is assumed, will not significantly affect the identification of speech signals presented well above threshold levels. This study tested that assumption in the context of an experiment that investigated young adults’ recognition of words with few versus many phonological neighbors produced in clear and conversational speaking styles. Words were presented in speech-shaped noise at a + 5 dB SNR. A mixed effects regression analysis was performed to account for the contributions of lexical difficulty, speaking style, order of word presentation, and listeners’ hearing (as measured by pure-tone average (PTA)). This analysis showed, as expected, that correct identification was predicted by word type (easy > hard) and speaking style (clear > conversational). It also showed that identification improved over the course of the experiment. Finally...

Toward a Differential Diagnosis of Hidden Hearing Loss in Humans.

Recent work suggests that hair cells are not the most vulnerable elements in the inner ear; rather, it is the synapses between hair cells and cochlear nerve terminals that degenerate first in the aging or noise-exposed ear. This primary neural degeneration does not affect hearing thresholds, but likely contributes to problems understanding speech in difficult listening environments, and may be important in the generation of tinnitus and/or hyperacusis. To look for signs of cochlear synaptopathy in humans, we recruited college students and divided them into low-risk and high-risk groups based on self-report of noise exposure and use of hearing protection. Cochlear function was assessed by otoacoustic emissions and click-evoked electrocochleography; hearing was assessed by behavioral audiometry and word recognition with or without noise or time compression and reverberation. Both groups had normal thresholds at standard audiometric frequencies, however, the high-risk group showed significant threshold elevation at high frequencies (10–16 kHz), consistent with early stages of noise damage. Electrocochleography showed a significant difference in the ratio between the waveform peaks generated by hair cells (Summating Potential; SP) vs. cochlear neurons (Action Potential; AP), i.e. the SP/AP ratio, consistent with selective neural loss. The high-risk group also showed significantly poorer performance on word recognition in noise or with time compression and reverberation, and reported heightened reactions to sound consistent with hyperacusis. These results suggest that the SP/AP ratio may be useful in the diagnosis of “hidden hearing loss” and that, as suggested by animal models, the noise-induced loss of cochlear nerve synapses leads to deficits in hearing abilities in difficult listening situations, despite the presence of normal thresholds at standard audiometric frequencies.

Outer Hair Cell Damage: A Completely Different Listening Experience

Outer Hair Cell Damage: A Completely Different Listening Experience

Speech intelligibility for native and non-native talkers and listeners

Much of daily communication occurs under a range of suboptimal conditions related to talker, listener, and signal characteristics. These “adverse” conditions (Mattys et al., 2012) increase intelligibility variation and impact perceptual processes, representations, attention, and memory functions. In this talk, I focus on two factors that condition intelligibility variation, namely, when interlocutors are non-native speakers of a target language and when communication occurs in noise. First, I will discuss a set of experiments examining how native and non-native talkers modify their speech with the goal of enhancing their intelligibility and the extent to which these modifications enhance word recognition in noise for native and non-native listeners. I also present results from studies examining how listeners adapt to more and less familiar foreign accents of varying intelligibility when presented in noise. These studies provide insights into the interplay between physical and mental factors in the production and perception of intelligible speech. They further reveal constraints on talker-listener adaptation processes during communication. A detailed understanding of speech intelligibility variation for native and non-native talkers and listeners will enhance our understanding of the compensatory and cognitive mechanisms that allow speech comprehension in naturalistic conditions.

The effect of speaking style adaptations on speech perception in noise by native and non-native listeners

It is well established that non-native listeners perform worse on speech perception in noise tasks compared to native listeners. However, few studies have directly compared the effect of speaking style adaptations on speech perception in energetic and informational maskers for these two listener groups. The present study examined intelligibility of conversational (CO), clear (CS) and noise-adapted (NAS) meaningful sentences mixed with speech shaped noise (SSN) and 2-talker (2T) babble. Sentences were presented at −5 dB signal to noise ratio (SNR) to native and −3 dB SNR to non-native listeners. The results revealed that CS and NAS significantly improved word recognition in noise for both listener groups. However, the gains in intelligibility were substantially greater for native compared to non-native listeners. Word recognition was overall better in 2T babble than in SSN. Our findings confirm that both native and non-native listeners benefited from speaking style adjustments, but native listeners were better able to utilize the intelligibility-enhancing modifications. The results also suggest that informational masking in 2T babble is less disruptive than energetic masking in SSN. This may in part be due to the “glimpsing” windows through which the target speech could become more easily accessible.

Cingulo-Opercular Function During Word Recognition in Noise for Older Adults with Hearing Loss

Background/Study Context: Adaptive control, reflected by elevated activity in cingulo-opercular brain regions, optimizes performance in challenging tasks by monitoring outcomes and adjusting behavior. For example, cingulo-opercular function benefits trial-level word recognition in noise for normal-hearing adults. Because auditory system deficits may limit the communicative benefit from adaptive control, we examined the extent to which cingulo-opercular engagement supports word recognition in noise for older adults with hearing loss (HL).Methods: Participants were selected to form groups with Less HL (n = 12; mean pure tone threshold, pure tone average [PTA] = 19.2 ± 4.8 dB HL [hearing level]) and More HL (n = 12; PTA = 38.4 ± 4.5 dB HL, 0.25–8 kHz, both ears). A word recognition task was performed with words presented in multitalker babble at +3 or +10 dB signal-to-noise ratios (SNRs) during a sparse acquisition fMRI experiment. The participants were middle-aged and older (ages: 64.1 ± 8.4 years) English speakers with no history of neurological or psychiatric diagnoses.Results: Elevated cingulo-opercular activity occurred with increased likelihood of correct word recognition on the next trial (t(23) = 3.28, p = .003), and this association did not differ between hearing loss groups. During trials with word recognition errors, the More HL group exhibited higher blood oxygen level-dependent (BOLD) contrast in occipital and parietal regions compared with the Less HL group. Across listeners, more pronounced cingulo-opercular activity during recognition errors was associated with better overall word recognition performance.Conclusion: The trial-level word recognition benefit from cingulo-opercular activity was equivalent for both hearing loss groups. When speech audibility and performance levels are similar for older adults with mild to moderate hearing loss, cingulo-opercular adaptive control contributes to word recognition in noise.

Task-Related Vigilance During Word Recognition in Noise for Older Adults with Hearing Loss

Background/Study Context: Vigilance refers to the ability to sustain and adapt attentional focus in response to changing task demands. For older adults with hearing loss, vigilant listening may be particularly effortful and variable across individuals. This study examined the extent to which neural responses to sudden, unexpected changes in task structure (e.g., from rest to word recognition epochs) were related to pupillometry measures of listening effort.Methods: Individual differences in the task-evoked pupil response during word recognition were used to predict functional magnetic resonance imaging (MRI) estimates of neural responses to salient transitions between quiet rest, noisy rest, and word recognition in unintelligible, fluctuating background noise. Participants included 29 older adults (M = 70.2 years old) with hearing loss (pure tone average across all frequencies = 36.1 dB HL [hearing level], SD = 6.7).Results: Individuals with a greater average pupil response exhibited a more vigilant pattern of responding on a standardized continuous performance test (response time variability across varying interstimulus intervals r(27) = .38, p = .04). Across participants there was widespread engagement of attention- and sensory-related cortices in response to transitions between blocks of rest and word recognition conditions. Individuals who exhibited larger task-evoked pupil dilation also showed even greater activity in the right primary auditory cortex in response to changes in task structure.Conclusion: Pupillometric estimates of word recognition effort predicted variation in activity within cortical regions that were responsive to salient changes in the environment for older adults with hearing loss. The results of the current study suggest that vigilant attention is increased amongst older adults who exert greater listening effort.

The Temporal Dynamics of Spoken Word Recognition in Adverse Listening Conditions.

This study examined the temporal dynamics of spoken word recognition in noise and background speech. In two visual-world experiments, English participants listened to target words while looking at four pictures on the screen: a target (e.g. candle), an onset competitor (e.g. candy), a rhyme competitor (e.g. sandal), and an unrelated distractor (e.g. lemon). Target words were presented in quiet, mixed with broadband noise, or mixed with background speech. Results showed that lexical competition changes throughout the observation window as a function of what is presented in the background. These findings suggest that, rather than being strictly sequential, stream segregation and lexical competition interact during spoken word recognition.

Lexical and age effects on word recognition in noise in normal-hearing children

ObjectivesThe purposes of the present study were (1) to examine the lexical and age effects on word recognition of normal-hearing (NH) children in noise, and (2) to compare the word-recognition performance in noise to that in quiet listening conditions. MethodsParticipants were 213 NH children (age ranged between 3 and 6 years old). Eighty-nine and 124 of the participants were tested in noise and quiet listening conditions, respectively. The Standard-Chinese Lexical Neighborhood Test, which contains lists of words in four lexical categories (i.e., dissyllablic easy (DE), dissyllablic hard (DH), monosyllable easy (ME), and monosyllable hard (MH)) was used to evaluate the Mandarin Chinese word recognition in speech spectrum-shaped noise (SSN) with a signal-to-noise ratio (SNR) of 0dB. A two-way repeated-measures analysis of variance was conducted to examine the lexical effects with syllable length and difficulty level as the main factors on word recognition in the quiet and noise listening conditions. The effects of age on word-recognition performance were examined using a regression model. ResultsThe word-recognition performance in noise was significantly poorer than that in quiet and the individual variations in performance in noise were much greater than those in quiet. Word recognition scores showed that the lexical effects were significant in the SSN. Children scored higher with dissyllabic words than with monosyllabic words; “easy” words scored higher than “hard” words in the noise condition. The scores of the NH children in the SSN (SNR=0dB) for the DE, DH, ME, and MH words were 85.4, 65.9, 71.7, and 46.2% correct, respectively. The word-recognition performance also increased with age in each lexical category for the NH children tested in noise. ConclusionsBoth age and lexical characteristics of words had significant influences on the performance of Mandarin-Chinese word recognition in noise. The lexical effects were more obvious under noise listening conditions than in quiet. The word-recognition performance in noise increased with age in NH children of 3–6 years old and had not reached plateau at 6 years of age in the NH children.

The Effect of a High Upper Input Limiting Level on Word Recognition in Noise, Sound Quality Preferences, and Subjective Ratings of Real-World Performance.

One important factor that plays a role in front-end processing is the analog-to-digital converter within current hearing aids. The average input dynamic range of hearing aids is 96 dB SPL with an upper input limiting level (UILL) of 95-105 dB SPL. The UILL of standard hearing aids could distort loud signals, such as loud speech or music, which have root-mean-square values of 90 and 105 dB SPL with crest factors of 12 dB SPL to 14-20 dB SPL, respectively. This indicates that these loud sounds could create a distorted signal for patients when the input limiting level is reached. To examine if significant differences in word recognition in noise, sound quality preferences, and subjective ratings of real-world performance exist between conventional and high UILL hearing aids. Words in noise and sound quality preferences were assessed using recordings on a Knowles Electronic Manikin for Acoustic Research with conventional and high UILL hearing aids, different microphone modes, and listening conditions. Participants wore the hearing aids for 2 mo and completed questionnaires on subjective performance. Ten adults with bilateral slight to moderately severe sensorineural hearing loss were recruited. A four-factor repeated-measures analysis of variance (ANOVA) revealed significant differences between the conventional and high UILL across microphone modes and listening conditions for words in noise [F(2, 18) = 6.0; p < 0.05]. A three-factor repeated-measures ANOVA for sound quality preferences revealed a significant difference only for presentation level [F(1, 9) = 81.0; p < 0.001]. A one-factor ANOVA did not reveal significant differences between the conventional and high UILL on subjective ratings of real-world performance. Word recognition and sound quality preferences revealed significant differences between the conventional and high UILL; however, there were no differences in subjective ratings of real-world performance. One participant preferred the conventional UILL, two the high UILL, and seven thought performance was equal, which may be due to the listening environments participants encountered, as evidenced by datalogging.

How "proficient" is proficient? Bilingual listeners' recognition of English words in noise.

Shi (2011, 2013) obtained sensitivity/specificity measures of bilingual listeners' English and relative proficiency ratings as the predictor of English word recognition in quiet. The current study investigated how relative proficiency predicted word recognition in noise. Forty-two monolingual and 168 bilingual normal-hearing listeners were included. Bilingual listeners rated their proficiency in listening, speaking, and reading in English and in the other language using an 11-point scale. Listeners were presented with 50 English monosyllabic words in quiet at 45 dB HL and in multitalker babble with a signal-to-noise ratio of +6 and 0 dB. Data in quiet confirmed Shi's (2013) finding that relative proficiency with or without dominance predicted well whether bilinguals performed on par with the monolingual norm. Predicting the outcome was difficult for the 2 noise conditions. To identify bilinguals whose performance fell below the normative range, dominance per se or a combination of dominance and average relative proficiency rating yielded the best sensitivity/specificity and summary measures, including Youden's index. Bilinguals' word recognition is more difficult to predict in noise than in quiet; however, proficiency and dominance variables can predict reasonably well whether bilinguals may perform at a monolingual normative level.

Assessing speech perception in children with hearing loss: what conventional clinical tools may miss.

This study tested the hypothesis that word recognition in a complex, two-talker masker is more closely related to real-world speech perception for children with hearing loss than testing performed in quiet or steady-state noise. Sixteen school-age hearing aid users were tested on aided word recognition in noise or two-talker speech. Unaided estimates of speech perception in quiet were retrospectively obtained from the clinical record. Ten parents completed a questionnaire regarding their children's ease of communication and understanding in background noise. Unaided performance in quiet was correlated with aided performance in competing noise, but not in two-talker speech. Only results in the two-talker masker were correlated with parental reports of their children's functional hearing abilities. Speech perception testing in a complex background such as two-talker speech may provide a more accurate predictor of the communication challenges of children with hearing loss than testing in steady noise or quiet.

Bone conductive implants in single-sided deafness

Conclusion: Bone conduction implants (BCIs) have been shown to partially restore some of the functions lost when binaural hearing is missing, such as in subjects with single-sided deafness (SSD). The use of a single BCI needs to be recommended by a clinician based on thorough counselling with the SSD subject. Objectives: To perform an overview of the present capabilities of BCIs for SSD and to evaluate the reliability of the audiological evaluation for assessing speech recognition in noise and sound localization cues, which are major problems related to the loss of binaural hearing. Methods: Nine subjects with SSD who received BCI implants underwent a preoperative audiological evaluation that included sound field speech audiometry, word recognition score (WRS) testing and sound localization testing in quiet and in noise. They were also tested for the accuracy of their directional word recognition in noise and their subjective perceptions of their hearing difficulties using the APHAB questionnaire. Results: The mean maximum accuracy of word discrimination was 65.5% in the unaided condition and 78.9% in the BCI-aided condition. Sound localization in noise was better with the BCI than in the unaided condition, especially when the stimulus and noise were presented on the same side as the implanted ear. The accuracy of directional word recognition showed an improvement with the BCI with respect to the unaided condition on the BCI side, with either the stimulus in the implanted ear and the noise in the contralateral ear or with both the stimulus and noise presented to the implanted ear.

Intelligibility of emotional speech in younger and older adults.

Little is known about the influence of vocal emotions on speech understanding. Word recognition accuracy for stimuli spoken to portray seven emotions (anger, disgust, fear, sadness, neutral, happiness, and pleasant surprise) was tested in younger and older listeners. Emotions were presented in either mixed (heterogeneous emotions mixed in a list) or blocked (homogeneous emotion blocked in a list) conditions. Three main hypotheses were tested. First, vocal emotion affects word recognition accuracy; specifically, portrayals of fear enhance word recognition accuracy because listeners orient to threatening information and/or distinctive acoustical cues such as high pitch mean and variation. Second, older listeners recognize words less accurately than younger listeners, but the effects of different emotions on intelligibility are similar across age groups. Third, blocking emotions in list results in better word recognition accuracy, especially for older listeners, and reduces the effect of emotion on intelligibility because as listeners develop expectations about vocal emotion, the allocation of processing resources can shift from emotional to lexical processing. Emotion was the within-subjects variable: all participants heard speech stimuli consisting of a carrier phrase followed by a target word spoken by either a younger or an older talker, with an equal number of stimuli portraying each of seven vocal emotions. The speech was presented in multi-talker babble at signal to noise ratios adjusted for each talker and each listener age group. Listener age (younger, older), condition (mixed, blocked), and talker (younger, older) were the main between-subjects variables. Fifty-six students (Mage= 18.3 years) were recruited from an undergraduate psychology course; 56 older adults (Mage= 72.3 years) were recruited from a volunteer pool. All participants had clinically normal pure-tone audiometric thresholds at frequencies ≤3000 Hz. There were significant main effects of emotion, listener age group, and condition on the accuracy of word recognition in noise. Stimuli spoken in a fearful voice were the most intelligible, while those spoken in a sad voice were the least intelligible. Overall, word recognition accuracy was poorer for older than younger adults, but there was no main effect of talker, and the pattern of the effects of different emotions on intelligibility did not differ significantly across age groups. Acoustical analyses helped elucidate the effect of emotion and some intertalker differences. Finally, all participants performed better when emotions were blocked. For both groups, performance improved over repeated presentations of each emotion in both blocked and mixed conditions. These results are the first to demonstrate a relationship between vocal emotion and word recognition accuracy in noise for younger and older listeners. In particular, the enhancement of intelligibility by emotion is greatest for words spoken to portray fear and presented heterogeneously with other emotions. Fear may have a specialized role in orienting attention to words heard in noise. This finding may be an auditory counterpart to the enhanced detection of threat information in visual displays. The effect of vocal emotion on word recognition accuracy is preserved in older listeners with good audiograms and both age groups benefit from blocking and the repetition of emotions.

The Effects of Linguistic Context on Word Recognition in Noise by Elderly Listeners Using Spanish Sentence Lists (SSL).

This study evaluated the effects of the linguistic context on the recognition of words in noise in older listeners using the Spanish Sentence Lists. These sentences were developed based on the approach of the SPIN test for the English language, which contains high and low predictability (HP and LP) sentences. In addition, the relative contribution of peripheral hearing sensitivity, measured by pure-tone hearing thresholds (PTA), to the performance on both types of sentences was assessed in a regression analysis. The results showed that older listeners obtained benefits on word recognition from the linguistic context. PTA contributed significantly to explaining the variance in performance on both HP and LP sentences.

Reducing cognitive demands on listeners by speaking clearly in noisy places

Listeners have more difficulty identifying spoken words in noisy environments when those words have many phonological neighbors (i.e., similar-sounding words in the lexicon) than when they have few phonological neighbors. This difficulty appears to be exacerbated in old age, where reductions in inhibitory control presumably make it more difficult to cope with competition from similar-sounding words. Fortunately, word recognition in noise can generally be improved for a wide range of listeners (e.g., younger and older adults, individuals with and without hearing impairment) when speakers adopt a clear speaking style. This study investigated whether clear speech, in addition to generally increasing speech intelligibility, also reduces the inhibitory demands associated with identifying lexically difficult words in noise for younger and older adults. The results show that, indeed, the difference between rates of identification for words with many versus few neighbors was eliminated when those words were produced in clear speech. Data on the roles of individual differences (e.g., hearing, working memory, and inhibitory control) that may contribute to word identification in noise will also be presented.

Influence of Tinnitus on Auditory Spectral, Temporal Resolution, and Speech Perception Ability in Tinnitus Patients

Objectives: (1) Evaluate the influence of tinnitus upon the auditory spectral and temporal resolution. (2) Evaluate the effect of tinnitus on speech perception ability in noise. Methods: To exclude the effect of decreased hearing threshold, unilateral idiopathic tinnitus patients with symmetric hearing threshold were enrolled. The subjects were 13 patients with symmetric hearing loss &gt;20 dB HL and binaural difference &lt;10 dB at 0.25, 0.5, 1, 2, 3, 4, and 8 kHz, or who have normal hearing threshold and binaural difference &lt;10 dB at 0.25, 0.5, 1, 2, 3, 4, and 8 kHz, and threshold discrepancies &lt;15 dB at 9, 11.2, 12.5, 14, 16, 18, and 20 kHz. A total of 13 volunteers with normal hearing were enrolled as a control group. Four different psychoacoustic measurements were performed: (1) spectral-ripple discrimination, (2) temporal modulation detection, (3) Schroeder-phase discrimination, and (4) word recognition in noise. Results: There were no significant differences in spectral-ripple threshold, temporal modulation detection threshold, and Schroeder-phase discrimination between affected sides and non-affected sides of unilateral tinnitus patients. For the word in noise test, affected sides showed worse signal-to-noise ratio compared to non-affected side. Conclusions: We could not find any evidence that the tinnitus-affected ears show worse spectral and temporal processing compared with nontinnitus ears in unilateral tinnitus patients. The spectral ripple discrimination data suggests that the tinnitus-affected ear does not have more damage in its cochlea (especially the outer hair cell) causing the auditory filters to broaden compared with the nonaffected ear with the same hearing threshold.

Timecourse of word recognition for listener-oriented speaking style adaptations

Previous research has found that listener-oriented speaking style adaptations, such as Infant Directed Speech (IDS) and Clear Speech (CS), aid perception and development and improve intelligibility for children and adults, respectively. However, less is known about whether these different speaking styles enhance intelligibility in general, or if young children benefit most from IDS features aimed at capturing and maintaining attention. Additionally, information about time course differences in word recognition for these speech modifications has not been investigated. This study investigates both questions by presenting young adult listeners with semantically anomalous and meaningful sentences produced in IDS, CS, and Conversational (CO) Speech. Listeners participated in (1) a word-recognition-in-noise test and (2) a visual word recognition paradigm tracking eye movements. Results showed facilitated word recognition in noise for both IDS and CS compared to CO. Both speaking style adjustments also resulted in increased speed of word recognition in the visual paradigm. Semantic context provided additional facilitation only when more exaggerated acoustic cues were present (in IDS and CS). Preliminary analyses show that even for adult listeners IDS provided the greatest perceptual benefits, indicating that established advantages of IDS for young listeners cannot be mainly attributed to affect in addition to enhanced acoustic cues.

Effects of hearing loss and linguistic context for phoneme and word recognition in noise

This project measured whole-word and phoneme recognition in normally hearing (NH) and hearing-impaired (HI) participants listening in steady-state noise. Stimuli were modeled after those used in Olsen et al. [EarHear 1997] including 10-word lists of consonant-vowel-consonant (CVC) monosyllables containing 10 vowels and 20 consonants. Words were presented in isolation and in contextually correct sentences. Speech-shaped noise was presented at a range of signal-to-noise ratios, from −9 to + 3 dB SNR for NH listeners, and from −6 to + 6 dB SNR for HI listeners. Results suggested that HI listeners needed an increased proportion of components (pp) to recognize a whole word (pw) relative to their NH counterparts (e.g., a higher j value) when stimuli were presented at unfavorable SNRs. For -3 and 0 dB SNR conditions, j factors for NH listeners (0 dB SNR: j = 2.17) were consistently lower than those for HI listeners (0 dB SNR: j = 2.52). However, for the 3 dB SNR conditions, j factors were higher for NH listeners (j = 3.06) than those for HI listeners (j = 2.58). This suggests an unexpected interaction between hearing acuity, background noise, and the recognition of phonemes versus whole words. [Research supported by DC008306 to Peggy Nelson.]

Regional linguistic variations in Canadian French: Do they affect performance on speech perception in noise?

Many audiologists working with a Canadian French population use word recognition tests to measure speech perception abilities in noise. The TMB test (Test de Mots dans le Bruit) includes four lists of 35 words presented in babble noise. The test is intended to measure the pre-cognitive perceptual stage of auditory processing and does not require understanding of the phonetic differences between speech sounds at a cognitive level. Previous studies examining performance on auditory tests similar to the TMB showed differences between populations speaking the same language but with different accentuations, such as the English spoken in the United States versus the United Kingdom. Variations in performance were attributed to accentuation differences between the speaker and the listener. To the authors' knowledge, no study appears to have investigated the effect of regional linguistic variations of Canadian French on word recognition in noise. Normative data for the TMB are being collected in three regions of Canada: Moncton, Montreal and Ottawa. Participants are all native speakers of Canadian French, but there are important linguistic variations across the three regions. Knowledge of the effect of regional linguistic variations on the TMB performance will help refine interpretation of test results in the audiology clinics.