Validation of the Acoustic Voice Quality Index (AVQI) Version 03.01 in Italian

Objective: The aim of the present study is to validate the Acoustic Voice Quality Index version 03.01 in the Italian language (AVQIv3-IT).

Methods: A total of 150 native Italian speakers with normal voices (n = 50) and with various voice disorders (n = 100) were enrolled. Voice samples of a sustained vowel (SV) [a:] and five phonetically balanced continuous speech (CS) samples were recorded. The most appropriate syllable number for a standardized voiced CS approximating 3 seconds was identified. Perceptual evaluations of the overall voice quality were performed by three expert voice clinicians using the G score of the GRBAS scale. AVQIs were calculated using a 3 seconds mid-vowel selection of the SV [a:] and the standardized syllable number of the CS. Finally, concurrent validity and diagnostic accuracy of AVQIv3-IT were analysed.

Results: The most appropriate syllable number for a standardized CS approximating 3 seconds in Italian was identified as 25. The perceptual ratings showed robust intra- and inter-rater reliability. A strong correlation was found between AVQI scores and overall voice quality perceptual evaluations (r = 0.81, P < 0.001). The best diagnostic outcome for AVQIv3-IT was found for a threshold of 2.35 (sensitivity of 90% and specificity of 92%).

Conclusions: AVQIv3-IT was demonstrated to be a valid and robust tool for quantifying overall acoustic voice quality in the Italian speaking population.