Acoustic Profiling and Modeling of Mandarin-Speaking Children With Ankyloglossia: Implications for Therapy.

PURPOSE: Acoustic profiling and modeling serve as valuable tools to detect subtle speech disorders in children, complementing existing diagnostic criteria and enhancing treatment planning. This study aims to (a) examine the acoustic characteristics of Mandarin-speaking children with and without ankyloglossia across consonants, vowels, tones, and voice quality, and to (b) evaluate the relative weighting of these acoustic cues in distinguishing this clinical population.

METHOD: Both perceptual (n = 21) and production tasks (n = 26) were included. Sixteen acoustic parameters were analyzed in Mandarin-speaking tongue-tied children (TT, n = 13) and typically developing children (TD, n = 13), which were further used for automatic recognition of tongue-tied speech.

RESULTS

The preliminary results indicated that neither free tongue classification nor age reliably predicted TT children's speech intelligibility, and the TT group exhibited significant acoustic deviations from the TD group across multiple acoustic measurements. Classification accuracy reached 85.1% for the TT group and 89.7% for the TD group, with the most robust acoustic cues being intensity, energy, and center of gravity.

CONCLUSIONS

The observed heterogeneity in this study does not suggest that ankyloglossia necessarily leads to speech misarticulation. Rather, the findings underscore the value of a noninvasive and child-friendly acoustic approach that incorporates tongue function into assessment. The investigation is particular relevant for evaluating pathological voices that can vary from mild-severe with no agreed-upon standard of diagnosis.