Voice Quality Assessment over Telepractice: Evaluating the Impact of Zoom Transmission
Objective: Telepractice, as an emerging approach to speech therapy, has reduced healthcare costs and enabled online treatment options during the COVID-19 pandemic. While substantial research has demonstrated the efficacy of telepractice in delivering effective speech therapy, there is limited exploration of its potential effects on the assessment and reliability of voice disorder evaluations. The purpose of the present study was to evaluate the impact of telepractice on voice quality by comparing original recordings with audio signals transmitted through the widely used telepractice platform, Zoom, using both auditory-perceptual voice assessments and acoustic analyses.
Methods / Design: This experiment was conducted as a controlled study. A total of 45 normophonic and 45 dysphonic voice samples were randomly selected from the Perceptual Voice Qualities Database (PVQD), which is a publicly available database of disordered voices. All voice samples were standardized in terms of recording conditions, duration, and volume to ensure consistency. The selected samples were then transmitted through Zoom telepractice to simulate remote communication.
Auditory-perceptual assessments were conducted by 3 speech-language pathologists based on GRBAS. Acoustic analyses were performed using Praat, and a range of acoustic parameters including mean, maximum, minimum and standard deviation values of fundamental frequency (F0), jitter (local and ppq5), shimmer (local and apq5), harmonics-to-noise ratio (HNR) and cepstral peak prominence (CPP) were obtained.
Results: Preliminary acoustic analyses revealed significant changes in certain voice quality parameters following Zoom transmission. CPP significantly decreased and HNR significantly increased, reflecting spectral degradation and reduced noise components, even when audio intervention had been minimized through settings. These acoustic deviations may pose challenges for Zoom-based voice assessments, as they may obscure clinically relevant vocal characteristics.
Conclusions: The findings indicate that Zoom transmission may affect specific acoustic parameters of voice, as shown by a significant decrease in CPP and a significant increase in HNR. These changes suggest that certain critical features of the original recordings may be insufficiently preserved in remote assessments, potentially impacting the reliability of perceptual judgments. These results highlight the need for adjustments in telepractice protocols and compensatory technological solutions to enhance the fidelity of voice quality evaluations in remote settings.