Bio-inspired auto-adaptive framework for optimized movement of passive knee prosthesis

Area of Science:

• Biomedical Engineering
• Robotics
• Biomechanics

Background:

• Amputees face mobility challenges with passive knee prostheses.
• Existing prostheses often lack adaptive capabilities for natural gait.
• Restoring function and comfort for amputees is a significant unmet need.

Purpose of the Study:

• To develop a bio-inspired adaptive framework for passive knee prostheses.
• To mimic natural walking and compensate for lost limb mobility.
• To enhance the quality of life for lower-limb amputees.

Main Methods:

• Developed a framework utilizing femur input power and a damping control mechanism.
• Integrated sensors for real-time data acquisition.
• Employed a deep learning architecture for gait phase event classification.

Main Results:

• Achieved 94.44% classification accuracy for gait phase events using deep learning.
• Demonstrated optimized movement with reduced hip hikes and less fatigue.
• Maintained normal knee flexion (64°±6) and achieved 95% fall prevention.

Conclusions:

• The proposed framework significantly improves passive knee prosthesis functionality and comfort.
• This innovation offers a promising solution for enhancing amputee mobility and safety.
• The research contributes to advancing prosthetic technology for better amputee quality of life.