Selected topics in laryngeal, perceptual and acoustic assessments of human voice: Videokymographic evaluations of vocal folds and investigations of teachers' voices

Ketaki PHADKE

Doctoral thesis

Investigations of the vocal fold vibrations using videokymography

Abstract:

The present thesis is based on four publications. In the first part of the thesis, the clinical value of the VKG method as an additional tool to stroboscopy (gold standard) was evaluated for diagnosing and treating various voice disorders (manuscript I). An exploratory questionnaire was designed for this purpose and used to evaluate results of examination of outpatients in a laryngology department. The results showed that VKG was useful in 95 % of cases, by either confirming the stroboscopic diagnosis (in 31 % cases), or making the diagnosis more accurate (44 %) or adjusting the treatment recommendations (20 %). After VKG examination, the diagnostic confidence improved in 68 % cases. VKG provided insights into the vibration characteristics of the vocal folds, and helped the clinicians to take some important diagnostic and treatment decisions when the diagnosis based on stroboscopy was uncertain. Analysis of the results also showed that the shape of lateral peaks (sharp versus rounded) and missing/reduced mucosal waves were the most helpful visual features in VKG images for obtaining the final diagnosis and providing insights into the health and pliability of vocal fold mucosa. The shape of the lateral peaks in VKG images was therefore in focus of the manuscript II which aimed at quantifying the sharpness of the lateral peak using automatic image analysis methods. Open Time Percentage Quotients (OTQ) and Plateau Quotients (PQ) were defined as two types of parameters which were expected to capture the shape of the lateral peaks in the vocal fold contours. The OTQ parameters were derived as fractions of the period and PQ as a fraction of the open phase during which the vocal folds displacement exceeded a pre-determined percentage of the vibratory amplitude. Results showed that the OTQ and PQ parameters derived at 95 % (OTQ95, PQ95) and at 80 % of the vocal fold amplitude (OTQ80, PQ80), had strong and significant correlations with the visual ratings of lateral peak sharpness (P<0.001). Therefore these quotients were considered to be the best objective parameters for quantification of lateral peak sharpness. The quotients increased their values when the shape of the lateral peak changed from sharp to round. In the second part of the thesis, teachers' voices were investigated using perceptual (subjective) evaluation and acoustic (objective) voice measures. Teachers are sensitive to voice attrition as a result of teaching in vocally demanding conditions and therefore investigations of teachers are important for better understanding of the voice properties and conditions influencing teachers' voice quality. In manuscript III, the participants were Finnish teachers who had no voice complaints, but in some of them laryngeal pathology was detected laryngoscopically. The changes in the values of non-smoothed and smoothed cepstral peak prominence (CPP & CPPS), voice sound pressure level (SPL) and perceptual ratings (vocal quality and vocal firmness) for three voice tasks (comfortable vowel phonation, comfortable and loud speech samples) were investigated and the influence of laryngeal pathology on these measures was studied. The results showed that CPP, CPPS and SPL values were significantly higher for vowels and loud speech than for comfortable speech (P<0.001). Significant correlations were found between SPL and cepstral measures. Loud speech was perceived to be firmer and have a better voice quality than comfortable speech. No significant relationships of the laryngeal pathology status with cepstral values, perceptual ratings, or voice SPL were found, however (P>0.05). It was concluded that neither the acoustic measures (CPP, CPPS, and SPL) nor the perceptual evaluations could clearly distinguish between healthy and disordered larynges when the pathologies are not self-perceived negatively by the teachers. Considering no vocal complaints of the subjects, the acoustic data are considered representative for teachers with functionally heal

Abstract:

The present thesis is based on four publications. In the first part of the thesis, the clinical value of the VKG method as an additional tool to stroboscopy (gold standard) was evaluated for diagnosing and treating various voice disorders (manuscript I). An exploratory questionnaire was designed for this purpose and used to evaluate results of examination of outpatients in a laryngology department. The results showed that VKG was useful in 95 % of cases, by either confirming the stroboscopic diagnosis (in 31 % cases), or making the diagnosis more accurate (44 %) or adjusting the treatment recommendations (20 %). After VKG examination, the diagnostic confidence improved in 68 % cases. VKG provided insights into the vibration characteristics of the vocal folds, and helped the clinicians to take some important diagnostic and treatment decisions when the diagnosis based on stroboscopy was uncertain. Analysis of the results also showed that the shape of lateral peaks (sharp versus rounded) and missing/reduced mucosal waves were the most helpful visual features in VKG images for obtaining the final diagnosis and providing insights into the health and pliability of vocal fold mucosa. The shape of the lateral peaks in VKG images was therefore in focus of the manuscript II which aimed at quantifying the sharpness of the lateral peak using automatic image analysis methods. Open Time Percentage Quotients (OTQ) and Plateau Quotients (PQ) were defined as two types of parameters which were expected to capture the shape of the lateral peaks in the vocal fold contours. The OTQ parameters were derived as fractions of the period and PQ as a fraction of the open phase during which the vocal folds displacement exceeded a pre-determined percentage of the vibratory amplitude. Results showed that the OTQ and PQ parameters derived at 95 % (OTQ95, PQ95) and at 80 % of the vocal fold amplitude (OTQ80, PQ80), had strong and significant correlations with the visual ratings of lateral peak sharpness (P<0.001). Therefore these quotients were considered to be the best objective parameters for quantification of lateral peak sharpness. The quotients increased their values when the shape of the lateral peak changed from sharp to round. In the second part of the thesis, teachers' voices were investigated using perceptual (subjective) evaluation and acoustic (objective) voice measures. Teachers are sensitive to voice attrition as a result of teaching in vocally demanding conditions and therefore investigations of teachers are important for better understanding of the voice properties and conditions influencing teachers' voice quality. In manuscript III, the participants were Finnish teachers who had no voice complaints, but in some of them laryngeal pathology was detected laryngoscopically. The changes in the values of non-smoothed and smoothed cepstral peak prominence (CPP & CPPS), voice sound pressure level (SPL) and perceptual ratings (vocal quality and vocal firmness) for three voice tasks (comfortable vowel phonation, comfortable and loud speech samples) were investigated and the influence of laryngeal pathology on these measures was studied. The results showed that CPP, CPPS and SPL values were significantly higher for vowels and loud speech than for comfortable speech (P<0.001). Significant correlations were found between SPL and cepstral measures. Loud speech was perceived to be firmer and have a better voice quality than comfortable speech. No significant relationships of the laryngeal pathology status with cepstral values, perceptual ratings, or voice SPL were found, however (P>0.05). It was concluded that neither the acoustic measures (CPP, CPPS, and SPL) nor the perceptual evaluations could clearly distinguish between healthy and disordered larynges when the pathologies are not self-perceived negatively by the teachers. Considering no vocal complaints of the subjects, the acoustic data are considered representative for teachers with functionally heal

Keywords

Videokymography, clinical value, sharpness of lateral peak, cepstral analysis, teachers' voice, noise

Language used: Czech

Date on which the thesis was submitted / produced: 30. 11. 2018

Thesis defence

• Supervisor: RNDr. Jan Švec, Ph.D., RNDr. Jan Švec, Ph.D.

PALACKÝ UNIVERSITY OLOMOUC

Faculty of Science

Doctoral programme / field:
Physics / Biophysics