Analysis of Disease-Induced Changes in Human Locomotor Patterns Through the Co-Joint Synergistic Attention Algorithm

Area of Science:

• Biomechanics
• Human Movement Analysis
• Computational Neuroscience

Background:

• Disease significantly impacts human locomotion, altering biomechanical patterns.
• Understanding these alterations is crucial for targeted rehabilitation and improving patient outcomes.
• Current methods may not fully capture the nuanced changes in joint coordination.

Purpose of the Study:

• To develop and validate a novel vision-based algorithm for quantifying disease-induced alterations in human movement.
• To analyze co-joint synergy patterns in locomotion and identify specific changes associated with conditions like stroke.
• To provide a quantitative measure of joint synergy variability for assessing disease impact.

Main Methods:

• Recruited 30 participants (15 post-stroke, 15 healthy) for 3D visual motor data collection.
• Employed a serial attention module for joint feature coupling and a dual-stream classification module for spatio-temporal analysis.
• Utilized a looping mask module to extract co-joint synergy patterns and calculate a synergy variability score.

Main Results:

• The co-joint synergistic attention algorithm revealed significant differences in joint synergy patterns between post-stroke patients and healthy individuals.
• Quantitative analysis demonstrated the specific effects of diseases on joint synergies in both patient and healthy groups.
• Results were validated against established methods like Non-negative Matrix Factorization (NMF) and Muscle Synergy Fractionation (MSF).

Conclusions:

• Specific diseases demonstrably alter human movement patterns by affecting joint synergies.
• The co-joint synergistic attention algorithm effectively analyzes these alterations and quantifies the importance of different synergy groups.
• This approach offers a new, targeted protocol for rehabilitation by identifying disease-specific biomechanical changes.