Quantitative Screening Support for Adolescent Idiopathic Scoliosis Based on Surface Electromyography

This study developed a quantitative assessment framework using multi-domain surface electromyography (sEMG) features to support muscle asymmetry screening in adolescent idiopathic scoliosis (AIS). sEMG signals were collected from 14 patients with mild-to-moderate AIS and 13 adolescents without AIS during three tasks: standing, stepping, and forward bending. Time-, frequency-, and nonlinear dynamical features were extracted and integrated using a structured fusion strategy to quantify paraspinal muscle asymmetry. The forward-bending task showed the highest sensitivity in revealing abnormal activation patterns and demonstrated clear correlations between the feature sets and task types. A two-stage feature fusion classification model achieved an accuracy of 91.5% and an AUC of 94.4% in discriminating patients with AIS from adolescents without AIS, supporting its utility for cross-sectional detection of mild-to-moderate AIS. As a noninvasive and resource-efficient method, the proposed framework complements static radiographic assessment (e.g., X-ray) by providing quantitative functional evaluation of spinal musculature across standardized motor tasks in adolescents with mild-to-moderate AIS.