Synergy-Based Estimation of Balance Condition During Walking Tests

Area of Science:

• Human-machine interface research
• Biomechanics
• Robotics

Background:

• Continuous estimation of Center of Pressure (COP) is crucial for assessing human balance.
• Existing methods may lack accuracy or comprehensive analysis of influencing factors.
• Muscle synergy analysis offers insights into neuromuscular control during movement.

Purpose of the Study:

• To develop a novel synergy-based method for continuous human balance estimation during walking.
• To analyze the impact of electromechanical delay compensation, number of synergies, and walking speed on estimation accuracy.
• To fuse temporal and spatial features for improved COP and Ground Reaction Force (GRF) prediction.

Main Methods:

• Introduction of muscle synergy coherence features and analysis of their variation with balance conditions.
• Fusion of temporal features from a bidirectional long short-term memory (BILSTM) network with spatial features from muscle synergy coherence.
• Continuous estimation of mediolateral COP and GRF using the fused feature set.
• Validation on real-world walking test data.

Main Results:

• A significant correlation was found between muscle synergy coherence features and human balance conditions.
• The proposed method accurately estimates COP and GRF during walking.
• The combined feature approach (muscle synergy coherence + BILSTM) demonstrates high estimation capability.

Conclusions:

• The novel synergy-based method provides accurate continuous estimation of human balance states.
• This approach enhances the safety and applicability of human-machine interfaces.
• Potential applications include advanced exoskeletons and prosthetic devices.