Acoustic Voice Quality Index for Discriminating Across Normal and Different Vocal Pathological Conditions

Abstract

The Acoustic Voice Quality Index (AVQI) is a multiparametric measure to assess the overall voice quality using both sustained and continuous speech. The study aimed to compare the AVQI values across normophonic voice and dysphonic voice secondary to different pathological conditions and to compare the values obtained from constituent parameters of AVQI across normophonic and different vocal pathological conditions. Seventy-four participants in the dysphonic group and twenty-eight in the normophonic group were considered. The auditory-perceptual analysis was done by three Speech- Language Pathologists using the Grade, Roughness, Breathiness, Asthenia, and Strain (GRBAS) scale to classify the participants into normophonic and dysphonic groups. Phonation samples of /a/ and reading samples were recorded using the Praat program, and AVQI was calculated using Praat AVQI script v.2.03. The results revealed that among the dysphonic group, higher AVQI values were seen in unilateral vocal fold (VF) palsy, fol- lowed by bilateral mass lesion, unilateral mass lesion, MTD-I, MTD-II & III, and acute laryngitis. CPPS and HNR values were lowest in the unilateral VF palsy group indicating high breathiness and noise component respectively. Shimmer local and shimmer local dB values were high in the unilateral VF palsy group and mass lesion group suggesting maximum aperiodic vibration of vocal folds in these groups. To conclude, AVQI and constituent parameters might help in discriminating vocal pathological conditions acoustically. As the present study is a preliminary attempt, future studies can be carried out with larger sample size, restricted age range, and due consideration of the auditory-perceptual dysphonia severity.