Identifying Children's Hidden Balance Strategy in mCTSIB using Clustering Techniques

Abstract

Traditional pediatric balance assessments often rely on subjective observation, which inherently fails to capture the intricate, high-dimensional kinematics of postural control. This study introduces an objective, data-driven framework utilizing unsupervised machine learning to classify balance strategies among 44 children during mCTSIB protocols. Kinematic data, captured via the Xsens system, underwent rigorous normalization and multicollinearity mitigation before being processed through four unsupervised clustering algorithms. The optimal four-cluster solution was validated through internal clustering metrics and PCA visualizations. Our analysis successfully delineated four distinct postural archetypes: distal-dominant (ankle), proximal-heavy (hip), and integrated multi-joint synergies. Notably, the results revealed a sensory-dependent shift in motor control logic; while participants favored a localized Ankle Strategy during baseline and visual-deprivation tasks on stable surfaces, the removal of somatosensory cues forced a strategic shift toward a Mixed Strategy (Hip-Ankle). In the most demanding scenarios - where both visual and somatosensory inputs were compromised - the majority of children transitioned to a complex Integrated Strategy, necessitating synchronized Hip, Knee, and Ankle adjustments to maintain stability. Furthermore, the model identified a rare rigid stiffening pattern - an emergency recovery response that remains largely invisible to manual clinical observation. By providing an objective diagnostic taxonomy, this framework enables clinicians to bypass the limitations of subjective scoring, paving the way for highly individualized and precision-targeted rehabilitation in pediatric clinical care.