Automated Measurement of Geniohyoid Muscle Thickness During Speech Using Deep Learning and Ultrasound

TL;DR

This paper introduces SMMA, an automated deep learning framework for measuring geniohyoid muscle thickness during speech from ultrasound, enabling large-scale, accurate analysis of speech motor control and disorders.

Contribution

It presents a novel fully automated method combining deep learning segmentation with skeleton-based quantification for muscle analysis during speech.

Findings

Near-human accuracy in muscle measurement (Dice = 0.9037)

Systematic differences in muscle thickness between vowels /a:/ and /i:/

Sex differences in muscle thickness consistent with anatomical scaling

Abstract

Manual measurement of muscle morphology from ultrasound during speech is time-consuming and limits large-scale studies. We present SMMA, a fully automated framework that combines deep-learning segmentation with skeleton-based thickness quantification to analyze geniohyoid (GH) muscle dynamics. Validation demonstrates near-human-level accuracy (Dice = 0.9037, MAE = 0.53 mm, r = 0.901). Application to Cantonese vowel production (N = 11) reveals systematic patterns: /a:/ shows significantly greater GH thickness (7.29 mm) than /i:/ (5.95 mm, p < 0.001, Cohen's d > 1.3), suggesting greater GH activation during production of /a:/ than /i:/, consistent with its role in mandibular depression. Sex differences (5-8% greater in males) reflect anatomical scaling. SMMA achieves expert-validated accuracy while eliminating the need for manual annotation, enabling scalable investigations of speech motor control and objective assessment of speech and swallowing disorders.