Research on Monitoring Assistive Devices for Rehabilitation of Movement Disorders through Multi-Sensor Analysis...

Area of Science:

• Biomedical Engineering
• Rehabilitation Science
• Artificial Intelligence in Medicine

Background:

• Gait analysis is crucial for evaluating movement disorders and rehabilitation progress.
• Current methods may lack the precision needed for nuanced assessments.
• Integrating advanced computational models can enhance diagnostic capabilities.

Purpose of the Study:

• To develop and validate a novel rehabilitation assessment device.
• To integrate a convolutional neural network (CNN) and Random Forest Model for comprehensive gait analysis.
• To improve the evaluation of rehabilitation progress in patients with movement disorders.

Main Methods:

• A device equipped with accelerometers and six-axis force sensors was utilized.
• Data from normal and abnormal walking groups, with simulated movement disorders, were collected.
• Feature extraction and analysis were performed using a CNN, with Random Forest Model for performance scoring.

Main Results:

• Significant differences in acceleration were noted between moderately and severely abnormal gait groups without assistance (p < 0.05).
• Force sensor data indicated distinct patterns for normal and abnormal walking groups.
• The CNN and Random Forest Model achieved high accuracies (88.4% and 92.3%) in recognizing gait conditions.

Conclusions:

• The integrated CNN and Random Forest Model provide accurate gait evaluations for movement disorder patients.
• The device effectively distinguishes gait characteristics across different walking modes and disorder severity.
• This approach enhances the physician's ability to monitor and evaluate rehabilitation progress.