Biomechanical Signal Reorganization and Compensation Patterns in Ankle-Constrained Gait: A Time-resolved Systems Analysis
Year: 2026
Type: conference-paper
Study design: Observational Study
Abstract: Human locomotion is a sophisticated mechatronic process governed by complex neural control loops and structural interactions. Traditional biomechanical analysis often collapses this rich temporal data into discrete scalar metrics, potentially missing the subtle nuances of pathological movement. This study proposes a robust, time-resolved multi-sensor fusion framework designed for high-fidelity feature extraction in pathological gait, utilizing Ankle Dorsiflexion Limitation (ADL) as a representative case study. We synchronized 10-camera infrared motion capture (200 Hz), multi-axis force platforms (2000 Hz), and wireless surface electromyography (sEMG, 2000 Hz) to monitor 30 male participants. These participants were categorized into Congenital Dorsiflexion Limitation (CDFL, n=10), Acquired Dorsiflexion Limitation (ADFL, n=10), and healthy control (CG, n=10) cohorts. By implementing One-dimensional Statistical Parametric Mapping (SPM1d), the framework analyzed continuous vector fields of joint kinematics, kinetics, and muscle activation throughout the 100% stance phase. Results identified distinct, etiology-specific "supra-threshold clusters". The CDFL system implemented a "distal-driven" strategy, significantly intensifying push-off power at the terminal phase (85–99% stance). In contrast, the ADFL system adopted a "proximalized" compensation, redirecting mechanical power to the knee and hip actuators (16–96% stance) to bypass distal constraints. These divergent control templates underscore the necessity for etiology-specific, phase-dependent interventions in wearable robotics. This framework establishes a quantitative foundation for designing adaptive control logic in intelligent exoskeletons and smart orthoses, enabling true human-in-the-loop gait optimization.
Authors: Jiachao Cai, Tianle Jie, Tibor Goda
Institution: Obuda University
Language: English
Subfield: Biomedical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204165845