Automatic rehabilitation assessment method of upper limb motor function based on posture and distribution force

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Clinical Assessment

Background:

• Traditional clinical assessments for hemiplegic upper limb motor function lack objectivity and uniformity.
• Current methods are often time-consuming and rely on subjective interpretations.
• There is a need for automated, reliable tools to quantify motor function post-stroke.

Purpose of the Study:

• To develop and validate an automated system for assessing upper limb motor function in hemiplegic patients.
• To improve the objectivity, efficiency, and consistency of rehabilitation assessments.
• To correlate automated assessment scores with established clinical scales like the Fugl-Meyer scale.

Main Methods:

• Utilized Azure Kinect and MediaPipe for upper limb and hand movement detection.
• Employed array distributed flexible thin film pressure sensors for hand force measurement.
• Developed a fuzzy logic assessment model using feature ratios from affected and healthy sides.

Main Results:

• Automated measurement of 30 items from the Fugl-Meyer scale.
• High correlation (r=0.99, p<0.001) between automated system scores and clinical Fugl-Meyer total scores.
• Bland-Altman analysis indicated good agreement between physician and system-derived scores.

Conclusions:

• The proposed system accurately and automatically assesses hemiplegic upper limb motor function.
• Posture and distributed force measurements provide objective data for rehabilitation evaluation.
• This technology offers a promising advancement in objective and efficient stroke rehabilitation assessment.