Sentence Intelligibility Scores Research Articles

Development of a Patient-Centered Functional Outcomes Questionnaire in Head and Neck Cancer

Incorporation of patient perspectives, or patient-reported outcomes, in functional outcome measures has been gaining prominence in the literature on reconstructive surgery. To create and validate an instrument for measuring the main functional areas of concern for patients with head and neck cancer. This 4-phase mixed-methods qualitative study was conducted from July 1, 2013, to June 30, 2016, in a quaternary head and neck oncology center in Edmonton, Alberta, Canada. Patients were recruited from 3 Head and Neck Research Network sites: University of Alberta (Edmonton, Canada), Mount Sinai Health Network (New York, New York), and University of Turku Hospital (Turku, Finland). The inclusion criteria included 18 years of age or older, diagnosis of squamous cell carcinoma involving the subsites of the head and neck (ie, oral cavity, oropharynx, hypopharynx, and larynx), and at least 1 year since treatment completion. Those patients who were undergoing additional active treatment or with evidence of disease recurrence were excluded. Data were analyzed from July 1, 2013, to June 30, 2016. The primary outcome measures were the clinical correlation of the Edmonton-33 instrument scores with swallowing, speech, dry mouth, and chewing assessment outcomes. In total, 10 patients with head and neck cancer (mean age, 59.6 years; 6 men [60%]) were included in phase 1 of the study, 5 patients (mean age, 55.2 years) were included in phase 2, 10 patients were included in phase 3, and 25 patients with head and neck cancer (mean age, 62.6 years; 14 men [56%]) participated in the phase 4 validation. The Edmonton-33 instrument scores correlated strongly with the swallowing scores of the MD Anderson Dysphagia Inventory (r = 0.77; 95% CI, 0.49-1.0), the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Head and Neck 35 (EORTC QLQ-H&N35) (r = -0.73; 95% CI, -1.0 to -0.44), and the modified barium swallow test (r = -0.60; 95% CI, -0.94 to -0.25). The instrument scores were also strongly correlated with the Speech Handicap Index scores (r = -0.64; 95% CI, -0.97 to -0.31), word intelligibility scores (r = 0.61; 95% CI, 0.27-0.95), and sentence intelligibility scores (r = 0.55; 95% CI, 0.19-0.91). A moderate to strong correlation was observed between the Edmonton-33 instrument and the EORTC QLQ-H&N35 scores in the dry mouth (r = -0.54; 95% CI, -0.91 to -0.18) and chewing (r = -0.45; 95% CI, -0.84 to -0.06) domains. The factor loading values for the domains of swallowing, speech, dry mouth, and chewing were all greater than 0.3. The mean factor loading values for the items related to swallowing were 0.71 (95% CI, 0.62-0.80) and for the items related to speech were 0.76 (95% CI, 0.72-0.80). The mean factor loading values for the items related to dry mouth were 0.71 (95% CI, 0.59-0.83) and for those related to chewing were 0.77 (95% CI, 0.69-0.85). The Edmonton-33 appears to be a validated instrument that will allow patients with head and neck cancer to assess and report their own functional outcomes. It could serve as a single comprehensive measure for functional outcomes.

A deep learning based segregation algorithm to increase speech intelligibility for hearing-impaired listeners in reverberant-noisy conditions.

Recently, deep learning based speech segregation has been shown to improve human speech intelligibility in noisy environments. However, one important factor not yet considered is room reverberation, which characterizes typical daily environments. The combination of reverberation and background noise can severely degrade speech intelligibility for hearing-impaired (HI) listeners. In the current study, a deep learning based time-frequency masking algorithm was proposed to address both room reverberation and background noise. Specifically, a deep neural network was trained to estimate the ideal ratio mask, where anechoic-clean speech was considered as the desired signal. Intelligibility testing was conducted under reverberant-noisy conditions with reverberation time T 60 = 0.6 s, plus speech-shaped noise or babble noise at various signal-to-noise ratios. The experiments demonstrated that substantial speech intelligibility improvements were obtained for HI listeners. The algorithm was also somewhat beneficial for normal-hearing (NH) listeners. In addition, sentence intelligibility scores for HI listeners with algorithm processing approached or matched those of young-adult NH listeners without processing. The current study represents a step toward deploying deep learning algorithms to help the speech understanding of HI listeners in everyday conditions.

Speech intelligibility of English, Polish, Arabic and Mandarin under different room acoustic conditions

This paper examines the impact of room acoustic conditions on the speech intelligibility of four languages (English, Polish, Arabic and Mandarin). Listening test scores (diagnostic rhyme tests, phonemically balanced word tests and phonemically balanced sentence tests) of the four languages were compared under four room acoustic conditions defined by their speech transmission index (STI=0.2, 0.4, 0.6 and 0.8). The results obtained indicated that there was a statistically significant difference between the word intelligibility scores of languages under all room acoustic conditions, apart from the STI=0.8 condition. English was the most intelligible language under all conditions, and differences with other languages were larger when conditions were poor (maximum difference of 29% at STI=0.2, 33% at STI=0.4 and 14% at STI=0.6). Results also showed that Arabic and Polish were particularly sensitive to background noise, and that Mandarin was significantly more intelligible than those languages at STI=0.4. Consonant-to-vowel ratios and languages’ distinctive features and acoustical properties explained some of the scores obtained. Sentence intelligibility scores confirmed variations between languages, but these variations were statistically significant only at the STI=0.4 condition (sentence tests being less sensitive to very good and very poor room acoustic conditions). Overall, the results indicate that large variations between the speech intelligibility of different languages can occur, especially for spaces that are expected to be challenging in terms of room acoustic conditions. Recommendations solely based on room acoustic parameters (e.g. STI) might then prove to be insufficient for designing a multilingual environment.

Chinese Word Identification and Sentence Intelligibility in Primary School Classrooms

Abstract The Chinese word identification and sentence intelligibility are evaluated by grades 3 and 5 students in the classrooms with different reverberation times (RTs) from three primary school under different signal-to-noise ratios (SNRs). The relationships between subjective word identification and sentence in- telligibility scores and speech transmission index (STI) are analyzed. The results show that both Chinese word identification and sentence intelligibility scores for grades 3 and 5 students in the classroom in- creased with the increase of SNR (and STI), increased with the increase of the age of students, and decreased with the increase of RT. To achieve a 99% sentence intelligibility score, the STIs required for grades 3, grade 5 students, and adults are 0.71, 0.61, and 0.51, respectively. The required objective acoustical index determined by a certain threshold of the word identification test might be underestimated for younger children (grade 3 students) in classroom but overestimated for adults. A method based on the sentence test is more useful for speech intelligibility evaluation in classrooms than that based on the word test for different age groups. Younger children need more favorable classroom acoustical environment with a higher STI than older children and adults to achieve the optimum speech communication in the classroom.

Importance of envelope modulations during consonants and vowels in segmentally interrupted sentences

The present study investigated the importance of overall segment amplitude and intrinsic segment amplitude modulation of consonants and vowels to sentence intelligibility. Sentences were processed according to three conditions that replaced consonant or vowel segments with noise matched to the long-term average speech spectrum. Segments were replaced with (1) low-level noise that distorted the overall sentence envelope, (2) segment-level noise that restored the overall syllabic amplitude modulation of the sentence, and (3) segment-modulated noise that further restored faster temporal envelope modulations during the vowel. Results from the first experiment demonstrated an incremental benefit with increasing resolution of the vowel temporal envelope. However, amplitude modulations of replaced consonant segments had a comparatively minimal effect on overall sentence intelligibility scores. A second experiment selectively noise-masked preserved vowel segments in order to equate overall performance of consonant-replaced sentences to that of the vowel-replaced sentences. Results demonstrated no significant effect of restoring consonant modulations during the interrupting noise when existing vowel cues were degraded. A third experiment demonstrated greater perceived sentence continuity with the preservation or addition of vowel envelope modulations. Overall, results support previous investigations demonstrating the importance of vowel envelope modulations to the intelligibility of interrupted sentences.

Predicting the intelligibility of reverberant speech for cochlear implant listeners with a non-intrusive intelligibility measure

Reverberation is known to reduce the temporal envelope modulations present in the signal and affect the shape of the modulation spectrum. A non-intrusive intelligibility measure for reverberant speech is proposed motivated by the fact that the area of the modulation spectrum decreases with increasing reverberation. The proposed measure is based on the average modulation area computed across four acoustic frequency bands spanning the signal bandwidth. High correlations (r=0.98) were observed with sentence intelligibility scores obtained by cochlear implant listeners. Proposed measure outperformed other measures including an intrusive speech-transmission index based measure.

Perceptual measures of speech from individuals with Parkinson's disease and multiple sclerosis: intelligibility and beyond.

The primary purpose of this study was to compare percent correct word and sentence intelligibility scores for individuals with multiple sclerosis (MS) and Parkinson's disease (PD) with scaled estimates of speech severity obtained for a reading passage. Speech samples for 78 talkers were judged, including 30 speakers with MS, 16 speakers with PD, and 32 healthy control speakers. Fifty-two naive listeners performed forced-choice word identification, sentence transcription, or visual analog scaling of speech severity for the Grandfather Passage (Duffy, 2005). Three expert listeners also scaled speech severity for the Grandfather Passage. Percent correct word and sentence intelligibility scores did not cleanly differentiate speakers with MS, PD, or control speakers. In contrast, both naive and expert listener groups judged reading passages produced by speakers with MS and PD to be more severely impaired than reading passages produced by control talkers. Scaled estimates of speech severity appear to be sensitive to aspects of speech impairment in MS and PD not captured by word or sentence intelligibility scores. One implication is that scaled estimates of speech severity might prove useful for documenting speech changes related to disease progression or even treatment for individuals with MS and PD with minimal reduction in intelligibility.

Speech and swallowing after surgical treatment of advanced oral and oropharyngeal carcinoma: a systematic review of the literature

Purpose of this review is the evaluation of speech and swallowing function after surgical treatment for advanced oral and oropharyngeal carcinoma. A systematic literature search (1993–2009), yielding 1,220 hits. The predefined criteria for inclusion in this systematic review were oral or oropharyngeal cancer, surgical treatment, speech and/or swallow function outcome, T-stage ≥ 2, patient cohort > 20, adequate description of the patient cohort in terms of tumor (sub) site, and low risk of bias (Cochrane criteria). Twelve studies fulfilled the predefined criteria. The results for speech more than 1 year after resection of oral or oropharyngeal cancer are reported to be moderate to good; although in the majority of patients speech is experienced as deviant. Overall sentence intelligibility scores are normal (92–98%). Swallowing is reported to be often already disturbed before treatment and is even more severely compromised after treatment. Aspiration rates of liquids vary from 12 to 50% and especially after oropharyngeal resection, pharyngeal transit times are delayed. Postoperative radiotherapy further increases function disturbances significantly. Critical subsites with regard to speech are the mobile tongue, and the soft palate and for swallowing, the floor of the mouth, the posterior base of tongue and the hard and soft palate. Prosthetic appliances (e.g., obturators, palatal augmentation prostheses) can diminish function losses considerably. Surgery for oral and oropharyngeal cancer yields function deficits, most notably with regard to swallowing. Series are small and outcome measurements vary. Therefore, to optimize pre-operative risk assessment, there is a need for internationally standardized outcome measurements.

The impact of alphabet supplementation and word prediction on sentence intelligiblity of electronically distorted speech

Alphabet supplementation is a low-tech augmentative and alternative communication (AAC) strategy that involves pointing to the first letter of each word spoken. Sentence intelligibility scores increased an average of 25% ( Hanson et al., 2004) when speakers with moderate and severe dysarthria (a neurologic speech impairment) used alphabet supplementation strategies. This project investigated the impact of both alphabet supplementation and an electronic word prediction strategy, commonly used in augmentative and alternative communication technology, on the sentence intelligibility of normal natural speech that was electronically distorted to reduce intelligibility to the profound range of <30%. Results demonstrated large sentence intelligibility increases (average 80% increase) when distorted speech was supplemented with alphabet supplementation and word prediction.

Vowel errors and sentence intelligibility in Spanish-accented English speech

It has been shown that foreign-accented speech is more difficult for listeners to process than native speech when presented under noisy listening conditions. It has also been shown that errors in vowel productions are more damaging to connected speech intelligibility than are errors or distortions in consonant production. This study compared the intelligibility of Spanish-accented English vowels in isolated words with that of short samples of connected speech. Five native Spanish speakers who had learned English as adults read a list of 42 monosyllabic English words selected so that 14 Spanish-accented English monophthongs could be presented to native listeners in a three-alternative forced-choice identification task. In each trial the target and two nearby (in the vowel space) distracters were presented. Talkers also read a set of 20 phonetically balanced sentences. Ten native English listeners identified the vowels in the isolated word task and judged the intelligibility of the sentences in a simple write-down procedure. While overall sentence intelligibility scores are not very well predicted by word intelligibility scores when the full set of vowel contrasts is included, predictability improves when a subset of the vowels is chosen. [Work supported by Purdue Research Foundation.]

Forced-Choice Analysis of Segmental Production by Chinese-Accented English Speakers

This study describes the development of a minimal-pairs word list targeting phoneme contrasts that pose difficulty for Mandarin Chinese-speaking learners of English as a second language. The target phoneme inventory was compiled from analysis of phonetic transcriptions of about 800 mono- and polysyllabic English words with examples of all the vowels, diphthongs, and syllable onsets and codas of the language. The Mandarin-specific minimal-pairs list derived from the phonetic transcription analyses contains 190 items. Tape recordings were made of 8 Mandarin speakers reading a randomized version of target items from the minimal-pairs list and a set of 20 sentences. Listeners who were native American English speakers judged the words in a forced-choice task and wrote down what they understood of the sentences. Correlations between listener responses on the forced-choice task and the sentence intelligibility scores showed differences in the strength of the relationship with sentence intelligibility across categories of minimal-pairs contrasts. Multiple regression analysis found listener responses on the minimal-pairs task to account for approximately 76% of the variance in speakers' sentence intelligibility scores, showing that performance on the minimal pairs of the probe list does predict connected speech intelligibility. Analyses of individual contrasts indicate target phonemes most often misperceived by native listeners.

Measures for assessing architectural speech security (privacy) of closed offices and meeting rooms

Objective measures were investigated as predictors of the speech security of closed offices and rooms. A new signal-to-noise type measure is shown to be a superior indicator for security than existing measures such as the Articulation Index, the Speech Intelligibility Index, the ratio of the loudness of speech to that of noise, and the A-weighted level difference of speech and noise. This new measure is a weighted sum of clipped one-third-octave-band signal-to-noise ratios; various weightings and clipping levels are explored. Listening tests had 19 subjects rate the audibility and intelligibility of 500 English sentences, filtered to simulate transmission through various wall constructions, and presented along with background noise. The results of the tests indicate that the new measure is highly correlated with sentence intelligibility scores and also with three security thresholds: the threshold of intelligibility (below which speech is unintelligible), the threshold of cadence (below which the cadence of speech is inaudible), and the threshold of audibility (below which speech is inaudible). The ratio of the loudness of speech to that of noise, and simple A-weighted level differences are both shown to be well correlated with these latter two thresholds (cadence and audibility), but not well correlated with intelligibility.

Speech Intelligibility and Quality of Life in Head and Neck Cancer Survivors

Poor speech intelligibility adversely affects quality of life self-assessment in long term survivors of head and neck cancer treatment. Observational case series including both objective clinical speech testing and subjective quality of life questionnaire administration. Five-year head and neck cancer survivors were recruited to study the association between speech intelligibility and quality of life. Survivors were analyzed as an entire group, and also subdivided into laryngectomees and non-laryngectomees. Objective testing included sentence and word intelligibility. Subjective testing included quality of life questionnaires (UWQOL, FACT, FACT-head and neck, and PSS-HN) and a locally prepared "cancer concern" question. Associations were sought between intelligibility, quality of life and demographics. Sixty-two survivors underwent testing. Lower sentence intelligibility and word intelligibility scores were associated with diminished self-perceived UWQOL Speech (P = .0001 and P = .0001, respectively) and PSS-HN Understandability of Speech (P = .009 and P = .005). Decreased word intelligibility was additionally associated with decreased UWQOL Chewing (P = .003), UWQOL Swallowing (P = .02), UWQOL Recreation (P = .05), PSS-HN Willingness to Eat in Public (P = .03), and PSS-HN Normalcy of Diet (P = .0001). The associations continued even after patients who had undergone laryngectomy were excluded. Long-term survivors of head and neck cancer continue to have both objective and subjective deficits in speech parameters five years after treatment. Objective deficits are associated with subjective concerns about speech, eating, and recreation. Understanding how communication deficits affect quality of life in long-term head and neck cancer survivors may allow more effective therapies to modulate these concerns in the recovery period.

Dysarthric sentence intelligibility: contribution of iconic gestures and message predictiveness.

This study examined changes in the sentence intelligibility scores of speakers with dysarthria in association with different signal-independent factors (contextual influences). This investigation focused on the presence or absence of iconic gestures while speaking sentences with low or high semantic predictiveness. The speakers were 4 individuals with dysarthria, who varied from one another in terms of their level of speech intelligibility impairment, gestural abilities, and overall level of motor functioning. Ninety-six inexperienced listeners (24 assigned to each speaker) orthographically transcribed 16 test sentences presented in an audio + video or audio-only format. The sentences had either low or high semantic predictiveness and were spoken by each speaker with and without the corresponding gestures. The effects of signal-independent factors (presence or absence of iconic gestures, low or high semantic predictiveness, and audio + video or audio-only presentation formats) were analyzed for individual speakers. Not all signal-independent information benefited speakers similarly. Results indicated that use of gestures and high semantic predictiveness improved sentence intelligibility for 2 speakers. The other 2 speakers benefited from high predictive messages. The audio + video presentation mode enhanced listener understanding for all speakers, although there were interactions related to specific speaking situations. Overall, the contributions of relevant signal-independent information were greater for the speakers with more severely impaired intelligibility. The results are discussed in terms of understanding the contribution of signal-independent factors to the communicative process.

Intelligibility and Nonspeech Orofacial Strength and Force Control Following Traumatic Brain Injury

Objective measures of nonspeech orofacial strength and force control were obtained for 20 individuals with traumatic brain injury. The dynamic and static force generating abilities of the upper lip, lower lip, tongue, and jaw were assessed. Based on sentence intelligibility scores, the subjects were divided into two groups, more and less intelligible. Force measures included reaction time, slope, derivative, peak overshoot, and first- and second-half mean hold and standard deviation. Groups differed only in the ability to sustain the 2 N force level with the tongue. Other potential contributors to the differences in intelligibility are discussed.

A comparison of speech synthesis intelligibility with listeners from three age groups

The use of computers in augmentative and alternative communication and in educational applications allows for the utilization of speech synthesis to instruct learners and to communicate messages. Of primary concern in these applications is the intelligibility of the synthesized voice for the individuals who will be potential listeners. To address this concern, the present investigation was designed to compare both single word and sentence intelligibility for four different types of voices (three speech synthesizers [Echo II+, Votrax Personal Speech System, and DECtalk] and a natural speaker) by listeners from three ages groups (children 6-8 years of age, children 10-12 years of age, and adults). The results indicated that single word intelligibility scores were lower than sentence intelligibility scores for all of the speech synthesizers but not the natural speaker. For the single word intelligibility task, there were no significant differences in intelligibility scores across the age groups, but signific...

Speechreading supplemented with auditorily presented speech parameters.

Results are reported from two experiments in which the benefit of supplementing speechreading with auditorily presented information about the speech signal was investigated. In experiment I, speechreading was supplemented with information about the prosody of the speech signal. For ten normal-hearing subjects with no experience in speechreading, the intelligibility score for sentences increased significantly when speechreading was supplemented with information about the overall amplitude of the speech signal, information about the fundamental frequency, or both. Binary information about voicing appeared not to be a significant supplement. In experiment II, the best-scoring supplements of experiment I were compared with two supplementary signals from our previous studies, i.e., information about the sound-pressure levels in two 1-oct filter bands centered at 500 and 3160 Hz, or information about the frequencies of the first and second formants from voiced speech segments. Sentence-intelligibility scores were measured for 24 normal-hearing subjects with no experience in speechreading, and for 12 normal-hearing experienced speechreaders. For the inexperienced speechreaders, the sound-pressure levels appeared to be the best supplement (87.1% correct syllables). For the experienced speechreaders, the formant-frequency information (88.6% correct), and the fundamental-frequency plus amplitude information (86.0% correct), were equally efficient supplements as the sound-pressure information (86.1% correct). Discrimination of phonemes (both consonants and vowels) was measured for the group of 24 inexperienced speechreaders. Percentage correct responses, confusion among phonemes, and the percentage of transmitted information about different types of manner and place of articulation and about the feature voicing are presented.

Contextual effects in the measurement of hearing-impaired speakers' intelligibility.

Previous research has indicated that hearing-impaired speakers' intelligibility scores are better when sentences are used than when word lists are used as speech material in word identification tests. The speech intelligibility of 20 hearing-impaired speakers was measured with word identification tests using isolated words (W-22 monosyllables) and words in sentence context (CID sentences). Analysis of individual speakers' intelligibility data revealed that sentence intelligibility scores were higher than word intelligibility scores only for the better speakers and that no differences were apparent between sentence and single-word intelligibility for the poorer speakers. These findings agree with the results of research with normal speech degraded in intelligibility by noise or filtering and indicate that an interaction may exist between context and overall intelligibility in which only speech that has a certain degree of overall intelligibility may show further intelligibility improvement with increased contextual clues.

Influence of context on measurement of hearing‐impaired speakers' intelligibility

Previous research has indicated that hearing‐impaired talkers' intelligibility scores are better when sentences are used than when word lists are used as speech material in word identification tests. The speech intelligibility of 20 hearing‐impaired talkers was measured with word identification tests using isolated words (W‐22 monosyllables) and words in sentence context (CID sentences). Analysis of individual talkers' intelligibility data revealed that sentence intelligibility scores were higher than word intelligibility scores only for the better talkers and that no differences were apparent between sentence and single word intelligibility for the poorer talkers. These findings agree with the results of research with normal speech that has been degraded in intelligibility by noise or filtering and indicate that an interaction may exist between context and overall intelligibility in which only speech that has a certain degree of overall intelligibility may show further intelligibility improvement with increased contextual clues. [Work supported by U.S. Department of Education.]

Sentence intelligibility as a function of key word selection.

The purpose of this study was to investigate the relationship between word predictability and sentence intelligibility. This relationship was examined by comparing intelligibility scores obtained with frequency-filtered sentences (CID Sentence Lists B and D and Revised List C) which were scored using key words empirically determined to be representative of three different degrees of predictability. We hypothesized that the scores obtained would be a function of the predictability status of the key words used in scoring. Results indicated significant differences between the three scoring procedures for each sentence list under two filtering conditions (420-Hz and 360-Hz low pass) and these differences were in the hypothesized direction. The results suggest that use of “easy-to-predict” words for scoring purposes will increase sentence intelligibility scores, while use of “difficult-to-predict” words will depress scores. We concluded that word predictability is a factor influencing sentence intelligibility and that careful selection of key words, based on their predictability status, is a method of controlling or influencing the intelligibility of sentences.