ABSTRACT
INTRODUCTION: Benign and malignant vocal fold lesions are growths that occur on the vocal folds. However, the treatments for these two types of lesions differ significantly. Therefore, it is imperative to use a multidisciplinary approach to properly recognize suspicious lesions. This study aims to determine the important acoustic characteristics specific to benign and malignant vocal fold lesions. METHODS: The acoustic model of voice quality was utilized to measure various acoustic parameters in 157 participants, including individuals with normal, benign, and malignant conditions. The study comprised 62 female and 95 male participants (43 ± 10 years). Voice samples were collected at the Shanghai Eye, Ear, Nose and Throat Hospital between May 2020 and July 2021.The acoustic variables of the participants were analyzed using Principal Component Analysis to present important acoustic characteristics that are specific to normal vocal folds, benign vocal fold lesions, and malignant vocal fold lesions. The similarities and differences in acoustic factors were also studied for benign conditions including Reinke's edema, polyps, cysts, and leukoplakia. RESULTS: Using the Principal Component Analysis method, the components that accounted for the variation in the data were identified, highlighting acoustic characteristics in the normal, benign, and malignant groups. The analysis indicated that coefficients of variation in root mean square energy were observed solely within the normal group. Coefficients of variation in pitch were found to be significant only in benign voices, while higher formant frequencies and their variability were identified as contributors to the acoustic variance within the malignant group. The presence of formant dispersion as a weighted factor in Principal Component Analysis was exclusively noted in individuals with Reinke's edema. The amplitude ratio between subharmonics and harmonics and its coefficients of variation were evident exclusively in the polyps group. In the case of voices with cysts, both pitch and coefficients of variation for formant dispersion were observed to contribute to variations. Additionally, higher formant frequencies and their coefficients of variation played a role in the acoustic variance among voices of patients with leukoplakia. CONCLUSION: Experimental evidence demonstrates the utility of the Principal Component Analysis method in the identification of vibrational alterations in the acoustic characteristics of voice affected by lesions. Furthermore, the Principal Component Analysis analysis has highlighted underlying acoustic differences between various conditions such as Reinke's edema, polyps, cysts, and leukoplakia. These findings can be used in the future to develop an automated malignant voice analysis algorithm, which will facilitate timely intervention and management of vocal fold conditions.
ABSTRACT
OBJECTIVE: Airway glottic insufficiency, or glottal gap, may lead to a breathy voice quality. It is hypothesized that a glottal gap may be a source of nonlinearity in speech production. This study aims to gain a chaotic and acoustic profile of glottal gap voice provided by phonation of excised larynges subjected to the insertion of a metal shim in the posterior glottis. STUDY DESIGN: Nonrandomized quasi-experimental study. METHODS: Posterior glottal gap varied from 0 to 3.5 mm in 0.5 mm intervals. Each treatment was investigated independently in a sample population of eight excised canine larynges. Phonation of the larynges for each treatment was recorded and analyzed for the cepstral peak prominence (CPP), harmonics-to-noise ratio (HNR), and correlation dimension. RESULTS: Kruskal-Wallis rank-sum tests yielded significant differences across shim groups for all parameters. Dunn-Bonferroni post-hoc tests revealed that the control group differed significantly from the 1.5, 2, 2.5, 3, and 3.5 mm groups for all metrics. Moreover, Kendall correlation tests indicated a moderately positive correlation between glottal gap size and correlation dimension, a moderately negative correlation between glottal gap size and CPP and between glottal gap size and the HNR. CONCLUSIONS: Glottic insufficiency provides a source of nonlinearity in phonation. Nonlinear dynamic analysis provides quantitative insight into glottal gap voice. This study encourages future studies to further evaluate the relationship between glottal gap and correlation dimension.
