Enhanced Identification of Chronic Ankle Instability Under Different Conditions: A New Evaluation Framework Based on...

Area of Science:

• Biomechanics and Neuromuscular Control
• Biomedical Engineering
• Machine Learning in Healthcare

Background:

• Chronic ankle instability (CAI) often involves altered neuromuscular control and compensatory movement strategies.
• Surface electromyography (sEMG) and non-negative matrix factorization (NNMF) can reveal muscle synergy patterns.
• Deep learning and machine learning offer advanced methods for analyzing complex biomechanical data.

Purpose of the Study:

• To develop and validate a synergy-informed evaluation framework integrating NNMF-derived muscle synergies with deep feature fusion and machine learning.
• To investigate neuromuscular control differences between individuals with CAI and healthy controls during landing tasks.
• To enhance the detection of CAI-related neuromuscular deficits using a novel computational approach.

Main Methods:

• Collected sEMG data from nine lower-limb muscles during anticipated and unanticipated landing tasks in 30 CAI patients and 30 healthy controls.
• Applied NNMF to extract muscle synergy features from sEMG signals.
• Utilized a convolutional neural network (CNN) for hierarchical feature integration and a random forest (RF) classifier for outcome prediction.
• Evaluated model performance based on classification accuracy and F1-score, comparing different task conditions and fusion strategies.

Main Results:

• Individuals with CAI exhibited altered muscle synergy structures and a proximal recruitment shift, especially during unanticipated landings.
• The CNN-RF-Unant model achieved the highest classification accuracy (0.96) and F1-score (0.95), outperforming other configurations.
• Features derived from unanticipated tasks demonstrated superior informativeness for detecting neuromuscular deficits.

Conclusions:

• The proposed framework combining NNMF-based synergy analysis with CNN-driven fusion significantly improves the detection of neuromuscular deficits in CAI.
• Assessing neuromuscular control during unanticipated tasks provides ecologically valid and sensitive insights into functional impairments.
• This approach offers a clinically valuable tool for identifying latent functional deficits and guiding personalized rehabilitation strategies for CAI.