Acoustic Psychometric Severity Index of Dysphonia (APSID): Development and Clinical Application

Voice disorder discrimination using vowel acoustic measures in female speakers

BACKGROUND

Sustained vowels are important vocal tasks that have been investigated in discriminating voice disorders using acoustic analysis. To date, no study has combined vowel acoustic measures only that evaluate major aspects of the pathological voice signals in voice disorder discrimination.

AIMS

To investigate the value of vowel acoustic measures that quantify glottal noise, signal stability, signal periodicity, spectral slope and overall voice quality in discriminating female speakers with and without voice disorders.

METHODS & PROCEDURES

Sustained vowel /ɑ/ samples were extracted from 133 voice-disordered female patients and 97 non-voice disordered female speakers and were signal typed prior to analysis. Praat software was used to measure harmonics-to-noise ratio (HNR), glottal-to-noise excitation ratio (GNE), the standard deviation of fundamental frequency (F0SD) and cepstral peak prominence (CPPp); and the Analysis of Dysphonia in Speech and Voice (ADSV) program was used to measure CPPadsv, low/high spectral ratio (LH) and the cepstral/spectral index of dysphonia (CSID). Outcome measures included sensitivity, specificity, and discrimination accuracy.

OUTCOMES & RESULTS

As individual acoustic measures, only spectral-based measures showed good (CPPadsv) and acceptable (CSID) discrimination results. The HNR, GNE and CPPp measures had acceptable sensitivity but poor or non-acceptable specificity and discrimination accuracy. Logistic regression models with all Praat measures (F0SD, HNR, GNE, CPPp) plus ADSV measures (CPPadsv, LH or CSID) provided excellent sensitivity, good-to-excellent specificity and excellent discrimination accuracy. ROC analysis for all individual measures showed that CPPadsv, CSID, CPPp, GNE and F0SD had the highest area under the curve (AUC) values.

CONCLUSIONS & IMPLICATIONS

A combination of acoustic measures that evaluate the major aspects of vocal dysfunction resulted in good to excellent voice discrimination outcomes. Individual acoustic measures had lower discrimination ability than combined measures. The findings implied that acoustic measures extracted from a prolonged vowel were useful in voice disorder discrimination.

WHAT THIS PAPER ADDS

What is already known on this subject Acoustic measures hold great value in discriminating voice disorders from normal voices. However, no study has evaluated discrimination values of a combination of sustained vowel acoustic measures that quantify additive noise, signal stability, signal periodicity, spectral slope and overall voice quality in single-gender cohorts. Previous studies have not used signal typing (the classification of the acoustic signals) for time-based measures, impacting the reliability of discrimination. What this study adds to the existing knowledge This study was the first to implement signal typing to include sustained vowel samples of Types 1 and 2 signals for discrimination statistics. We showed that a combination of vocal acoustic measures using time- and spectral-based extraction from the sustained /ɑ/ vowel evaluating additive noise, signal stability, signal periodicity, spectral slope and overall voice quality resulted in good to excellent sensitivity, specificity and discrimination accuracy. As individual measures, traditional time-based measures such as HNR had rather limited discrimination values whilst spectral-based measures provided higher discrimination values. Measures that are sensitive to signal types have low discrimination ability. What are the potential or actual clinical implications of this work? The sustained vowel /ɑ/ is a relevant, universal vocal task for clinical application using acoustic measures to discriminate female speakers with and without voice disorders if signal typing is implemented. Clinical voice assessment using vowels may not be effective if relying solely on time-based measurements. Spectral-based measures perform better in voice disorder discrimination given their insensitivity to signal types. The most effective voice disorder discrimination could only be obtained using a combination of acoustic measures that quantify major phenomena in the signals of disordered voices. Using measures extracted from both programs, Praat and ADSV, is useful given that specific settings in a program may impact on discrimination accuracy.

Differentiation between depression and bipolar disorder in child and adolescents by voice features

OBJECTIVE

Major depressive disorder (MDD) and bipolar disorder (BD) are serious chronic disabling mental and emotional disorders, with symptoms that often manifest atypically in children and adolescents, making diagnosis difficult without objective physiological indicators. Therefore, we aimed to objectively identify MDD and BD in children and adolescents by exploring their voiceprint features.

METHODS

This study included a total of 150 participants, with 50 MDD patients, 50 BD patients, and 50 healthy controls aged between 6 and 16 years. After collecting voiceprint data, chi-square test was used to screen and extract voiceprint features specific to emotional disorders in children and adolescents. Then, selected characteristic voiceprint features were used to establish training and testing datasets with the ratio of 7:3. The performances of various machine learning and deep learning algorithms were compared using the training dataset, and the optimal algorithm was selected to classify the testing dataset and calculate the sensitivity, specificity, accuracy, and ROC curve.

RESULTS

The three groups showed differences in clustering centers for various voice features such as root mean square energy, power spectral slope, low-frequency percentile energy level, high-frequency spectral slope, spectral harmonic gain, and audio signal energy level. The model of linear SVM showed the best performance in the training dataset, achieving a total accuracy of 95.6% in classifying the three groups in the testing dataset, with sensitivity of 93.3% for MDD, 100% for BD, specificity of 93.3%, AUC of 1 for BD, and AUC of 0.967 for MDD.

CONCLUSION

By exploring the characteristics of voice features in children and adolescents, machine learning can effectively differentiate between MDD and BD in a population, and voice features hold promise as an objective physiological indicator for the auxiliary diagnosis of mood disorder in clinical practice.

Relationship among five-factor personality traits and psychological distress with acoustic analysis

Objectives

The relationship between personality traits and psychological distress with acoustic characteristics was investigated in the present study, regarding the existence of dysphonia, abnormal overall voice quality (AOVQ), and dysphonia type.

Methods

Fifty-five participants with dysphonia and 64 participants without dysphonia completed NEO Five-Factor Inventory and Depression, Anxiety, and Stress Scale-21. Jitter, shimmer, noise-to-harmonic ratio (NHR), cepstral peak prominence (CPP), and cepstral peak prominence-smoothed (CPPS) were calculated in sustained vowel /a/ by Praat. Three expert speech and language pathologists divided participants with dysphonia into mild, moderate, and severe, based on the AOVQ. Pearson and Spearman correlation tests were performed by IBM SPSS Statistics.

Results

The findings were indicative of large correlations between agreeableness with CPP, conscientiousness with shimmer, depression with jitter and shimmer, and anxiety with shimmer in patients with functional dysphonia (p < 0.05). The results showed small to medium significant correlations between agreeableness with jitter and NHR, conscientiousness with CPP in participants without dysphonia, and depression with jitter in the participants with dysphonia (p < 0.05). Lastly, no significant correlation was observed between personality traits and psychological distress with acoustic characteristics in mild, moderate, and severe AOVQ groups (p > 0.05).

Conclusion

In participants with functional dysphonia, personality traits and psychological distress can provide some information about acoustic characteristics and vice versa.

Level of Evidence

3.

Comprehensive Index of Vocal Fatigue (CIVF): Development and Clinical Validation

OBJECTIVES

In this study, we sought to investigate whether auditory-perceptual and acoustic measures change after a passage-reading task and to develop a multidimensional index of performance fatigue in voice disorders. The clinical usefulness and cut-off scores of the newly developed index were also explored.

METHODS

A total of 150 patients (age = 35.65 ± 11.44 years) with voice disorders and 100 healthy controls (age = 35.04 ± 11.26 years) were enrolled in the study. Overall severity (OS) and acoustic measures of sentence samples were obtained before and after a passage-reading task and compared between the two positions. Acoustic features included cepstral peak prominence (CPP), σCPP, L/H spectral ratio (SR), and σSR. Using the changes (Δ) in those acoustic measures and the scores of the Vocal Fatigue Index (VFI), a predictive model of the ΔOS was established. The clinical cut-off point of the index was also identified.

RESULTS

For the patient group, OS was higher, but CPP and σCPP were lower for the rear position (after reading the passage) compared to the front position (before reading the passage). A significant portion (adjusted R² = 77.7%) of the ΔOS was predicted by ΔCPP, tiredness score of the VFI, ΔσCPP, ΔσSR, and ΔSR. The area under the curve of the newly developed index was .864 (sensitivity, 78.67%; specificity, 78.00%), and the index's criterion was >2.6489.

CONCLUSIONS

The newly developed Comprehensive Index of Vocal Fatigue (CIVF) can quantify vocal fatigue in voice disorders, reflecting both acoustic and auditory perceptual changes after a passage-reading task. Further studies should explore the responsiveness of the CIVF and the impact of related factors on the CIVF in various voice disorders.

Relationship Between Aerodynamic Measurement of Maximum Phonation Time With Acoustic Analysis and the Effects of Sex and Dysphonia Type

OBJECTIVES/HYPOTHESIS

This study set out to uncover the correlation between maximum phonation time (MPT) with acoustic and cepstral analysis in the dysphonic and control groups, considering the effects of sex and dysphonia type.

METHODS

For this cross-sectional study, a sample of 179 attendees (141 dysphonic and 38 control) were randomly selected and requested to sustain the vowel /a/ as long as they could with their habitual pitch and loudness. Reading standard sentences and conversational connected speech tasks were obtained too. Using Praat, the MPT, jitter, shimmer, noise-to-harmonic ratio, cepstral peak prominence (CPP), and smoothed cepstral peak prominence (CPPS) were calculated in the target vocal tasks.

RESULTS

There was a very low to low significant correlation (r = 0.00-0.50) between MPT amounts and acoustic analysis in the dysphonic group (P < 0.05), except for between MPT with shimmer (P > 0.05). In contrast, findings showed no significant correlation between MPT and acoustic analysis in the control group, not even separated by sex (P > 0.05). There was a very low to low correlation between MPT amounts and acoustic analysis in the male dysphonic group (P < 0.05), except for the MPT with shimmer (P > 0.05). There was no significant correlation between MPT and acoustic analysis in the female dysphonic group (P > 0.05), except for MPT with CPP (sustained vowel) (P < 0.05). Finally, very low to high correlations between MPT and some of the acoustic analysis in all the different dysphonia types were observed (P < 0.05).

CONCLUSIONS

MPT contains some information about the acoustic features of the dysphonic voice, specifically the CPP and smoothed cepstral peak prominence. The data suggested that the observed relationship between MPT and the acoustic analysis has the capacity to be considered for the development of new multiparametric tests of voice assessment in dysphonia, regarding the sex and dysphonia type.

Cepstral Measures in the Fado Voice: Gender, Age and Phonatory Tasks

Cepstral measures are sensitive to gender, age and phonatory tasks. With a cepstral measure designated as the CPP, it was possible to confirm the vulnerability of the Fado singers' voice. These were established at the vocal pathological threshold, which suggests a need for a direct clinical approach for these voice users.

OBJECTIVES/HYPOTHESIS

This study aimed to characterize cepstral peak prominence (CPP) and cepstral peak prominence smoothed (CPPS) in the Fado singing voice and to determine if there were significant differences in CPP and CPPS measures between spoken and sung tasks, as well as due to singers' gender and age.

METHODS

Forty seven males and 57 females Fado singers, ranging from 18 to 70 years participated in this study. Spoken voice tasks were sustained [a] and reading aloud the phonetically balanced text "O Sol". Sung tasks were sustained [a] of the word [ɐfinaɫ] and the Fado chorus song "Nem às paredes confesso". Acoustic measures included CPP and CPPS. CPP was measured using Analyses of Dysphonia in Speech and Voice software, of Multi-Speech program, Model 3700, by KayPENTAX. CPPS was measured using Praat software (4.2.1/2003). Statistical analysis was performed with an IBM SPSS Statistics version 22 program. CPP and CPPS mean differences of spoken and sung tasks were analyzed using paired samples t-test, with α at .05.

RESULTS

CPP and CPPS values of singers' voice changed according to the gender, age and phonatory tasks. There were significant differences between CPP and CPPS measures (P < 0.05). Generally, young male singers, in their sung task, presented the highest CPP and CPPS values. The highest CPP mean was obtained by older males in sustained spoken [a] and the lowest was obtained by younger males in their reading aloud task. For CPPS, the highest mean value was obtained by younger males during sung [a] and the lowest was by younger males in the reading aloud task.

CONCLUSION

Males presented higher cepstral measures than females. Young singers presented higher cepstral measures than older. Sung tasks had higher cepstral measures than spoken tasks. CPPS means are overall higher than CPP means. This study reinforces the need for a clinical prevention approach directed at potential vocal disorders in Fado singers.

Reliability of Acoustic Measures in Dysphonic Patients With Glottic Insufficiency and Healthy Population: A COVID-19 Perspective

OBJECTIVES

The COVID-19 pandemic has affected the voice assessment protocols for dysphonic patients. In this study, we compared the changes in acoustic measures of the healthy population as well as dysphonic patients due to glottic insufficiency between the pandemic period requiring face masks and the prepandemic period when the masks were not essential. The clinical reliability of the acoustic measures with and without face masks was explored.

METHODS

A total of 120 patients (age = 42.3 ± 11.9 yrs) with glottic insufficiencies such as UVFP and sulcus vocalis and 40 healthy population (age = 40.5 ± 11.2 yrs) cohorts were enrolled during the pandemic period. Age- and gender-matched 120 patients and 40 healthy population cohorts who underwent voice assessment without face masks before the pandemic were enrolled as prepandemic controls. Acoustic measures and overall severity estimates of vowel and speech samples were compared, which included cepstral peak prominence (CPP), L/H spectral ratio (SR), their standard deviations, F0, jitter percent (Jitt), shimmer percent (Shim), noise-to-harmonic ratio (NHR), Cepstral Spectral Index of Dysphonia (CSID), and Acoustic Psychometric Severity Index of Dysphonia.

RESULTS

Both patients and healthy cohorts showed higher SRv and SRs but lower CSIDv during the pandemic compared to the prepandemic period. F0 of the healthy male controls during the pandemic was higher than during the prepandemic periods, while the CSIDs was lower for the pandemic period. The pandemic patient cohort showed lower σSRs compared to the prepandemic patient cohort. When the acoustic measures of patients were compared to the healthy population cohort, the patient cohort showed lower CPP and σCPPs, while higher σCPPv, Jitt, Shim, and NHR during both pandemic and prepandemic period. Overall, the area under the curve of the acoustic measures and overall severity estimates was similar between the mask and non-mask groups, although the AUC of the SR measures was poor.

CONCLUSIONS

Wearing face masks during the pandemic did not compromise the overall reliability of the acoustic analysis in patients with glottic insufficiency, suggesting the current protocol of acoustic analysis can be carried out reliably while wearing a mask to ensure safety in the pandemic era.

Accuracy of Acoustic Measures of Voice via Telepractice Videoconferencing Platforms

Purpose Telepractice improves patient access to clinical care for voice disorders. Acoustic assessment has the potential to provide critical, objective information during telepractice, yet its validity via telepractice is currently unknown. The current study investigated the accuracy of acoustic measures of voice in a variety of telepractice platforms. Method Twenty-nine voice samples from individuals with dysphonia were transmitted over six video conferencing platforms (Zoom with and without enhancements, Cisco WebEx, Microsoft Teams, Doxy.me, and VSee Messenger). Standard time-, spectral-, and cepstral-based acoustic measures were calculated. The effect of transmission condition on each acoustic measure was assessed using repeated-measures analyses of variance. For those acoustic measures for which transmission condition was a significant factor, linear regression analysis was performed on the difference between the original recording and each telepractice platform, with the overall severity of dysphonia, Internet speed, and ambient noise from the transmitter as predictors. Results Transmission condition was a statistically significant factor for all acoustic measures except for mean fundamental frequency (f o). Ambient noise from the transmitter was a significant predictor of differences between platforms and the original recordings for all acoustic measures except f o measures. All telepractice platforms affected acoustic measures in a statistically significantly manner, although the effects of platforms varied by measure. Conclusions Overall, measures of f o were the least impacted by telepractice transmission. Microsoft Teams had the least and Zoom (with enhancements) had the most pronounced effects on acoustic measures. These results provide valuable insight into the relative validity of acoustic measures of voice when collected via telepractice. Supplemental Material https://doi.org/10.23641/asha.14794812.