Can Smartphones Be Used to Record Children's Voices for Acoustic Analysis?

Abstract

Objectives: While technological advancements have enabled the utilization of smartphones for acoustic voice analysis, existing studies have predominantly focused on the adult population. However, dysphonia is prevalent in children, and their anatomy and physiology are different from those of adults. Thus, this paper aims to investigate the feasibility of using smartphones to record children's voices for acoustic voice analysis.

Study design: A methodological study.

Methods: This study involved 29 children, aged 4-10 years, who had healthy voices. Voice recordings of sustained phonation and reading a sentence were obtained using four devices (1-AKG Micromic C520 headset microphone connected to a computer with the computerized speech lab (CSL), 2-Samsung S9 Plus, 3-iPhone 13 Mini, and 4-Huawei Y9 Prime). Then, all the recorded voice samples were analyzed using CSL, examining a total of 13 acoustic parameters. Pearson correlation analysis, Bland-Altman analysis, and uncertainty of measurement analysis were conducted to assess consistency and agreement across devices.

Results: The highest correlation and accuracy between the smartphone measurements and the reference recording system were found for the fundamental frequency (F0) (r > 0.90, P < 0.01) in both speech samples. For other parameters, limited reliability was observed; while some showed weak correlations, others had low accuracy and consistency. There were, however, moderate-to-excellent correlations for the most measurements and a nonsignificant bias according to the Bland-Altman analysis.

Conclusion: This study is the first to investigate the feasibility of using smartphones for acoustic voice analysis focusing specifically on children. The results indicate that smartphone voice recordings can be used to reliably measure F0. More research is needed to improve measurement reliability for other parameters. Nonetheless, the findings demonstrate the potential for smartphones to enable accessible and reliable voice assessment in the pediatric population.