A Synergy-Coherence Integration Method for Real-Time Knee Contact Force Estimation under Multiple Gait Patterns.

Area of Science:

• Biomechanics
• Neuroscience
• Rehabilitation Engineering

Background:

• Accurate real-time medial knee contact force (KCF) estimation is crucial for knee osteoarthritis (OA) research and rehabilitation.
• Existing methods often require numerous sensors and lack robustness, limiting their practical application.
• There is a need for a more practical and accurate KCF estimation framework.

Purpose of the Study:

• To develop and validate a novel real-time medial KCF estimation framework.
• To integrate muscle synergies and intermuscular coherence for enhanced KCF prediction accuracy.
• To enable high-accuracy KCF estimation using a minimal number of EMG sensors across diverse gait patterns.

Main Methods:

• Muscle synergy extraction was used for dimensionality reduction, selecting five key muscles.
• Intermuscular coherence was computed to characterize neuromuscular coordination during different gaits.
• A dual-branch LSTM network processed temporal and frequency domain features for KCF prediction.

Main Results:

• The framework achieved strong performance (R² = 0.87 ± 0.07) across five gait conditions on a public dataset.
• Leave-one-gait-out testing confirmed superior generalization and robustness compared to baseline methods.
• Real-world validation showed peak force estimation errors of no more than 0.18 BW during walking tests.

Conclusions:

• The proposed framework offers a practical and robust solution for real-time knee force monitoring with minimal sensors.
• This technology has significant potential for motor function assessment and personalized rehabilitation.
• The findings support the integration of this framework with intelligent assistive devices for enhanced patient care.