Decoding movement intent of patient with multiple sclerosis for the powered lower extremity exoskeleton

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Multiple Sclerosis (MS) significantly impacts mobility and motor control.
• Developing intuitive control systems for assistive devices like exoskeletons is crucial for improving quality of life for MS patients.
• Decoding neuromuscular signals offers a promising avenue for volitional control of powered exoskeletons.

Purpose of the Study:

• To develop and validate an algorithm for recognizing movement intent in patients with Multiple Sclerosis (MS).
• To integrate neuromuscular signals (EMG) with mechanical measurements for enhanced control of lower extremity exoskeletons.
• To assess the accuracy and predictive capabilities of the intent recognition system during various activities.

Main Methods:

• Collected surface electromyographic (EMG) signals from lower extremity muscles.
• Measured ground reaction forces and thigh segment kinematics from a single MS patient (EDSS 6).
• Developed an intent recognition algorithm fusing EMG and mechanical data to classify activities (walking, sitting, standing).

Main Results:

• Observed clear modulation of lower extremity muscle activity during different activities in the MS patient.
• Achieved 98.73% accuracy in classifying intended movements during static states.
• Successfully predicted activity transitions 100-130 ms prior to their occurrence.

Conclusions:

• The developed intent recognition algorithm demonstrates high accuracy in decoding movement intentions for MS patients.
• The system shows significant potential for enabling volitional control of powered lower extremity exoskeletons.
• This approach could lead to more responsive and intuitive assistive technologies for individuals with MS.