Wearable Sensor-Based Fine-Grained and Comprehensive Upper Extremity Motor Function Assessment for Poststroke...

Area of Science:

• Biomedical Engineering
• Rehabilitation Science
• Wearable Technology

Background:

• Stroke frequently causes upper extremity motor dysfunction, impacting patient independence.
• Current assessment methods like the Fugl-Meyer Assessment for Upper Extremity (FMA-UE) can be subjective and lack fine-grained detail.
• Objective, quantitative measures are needed for effective stroke rehabilitation and recovery monitoring.

Purpose of the Study:

• To develop and validate a wearable sensor-based framework for quantitative evaluation and prediction of upper extremity motor function in post-stroke patients.
• To enable fine-grained, action-level scoring of motor function.
• To reduce assessment burden while maintaining clinical consistency.

Main Methods:

• A validation cohort study involving 80 post-stroke patients with upper extremity motor dysfunction.
• Collection of upper extremity kinematic data using a wearable motion acquisition system during routine FMA-UE evaluations.
• Development of fine-grained scoring models for 8 critical FMA-UE items using feature engineering and machine learning, followed by a regression model to predict total FMA-UE scores.

Main Results:

• Wearable sensor data correlated strongly with clinician-rated FMA-UE scores.
• The developed framework achieved high prediction accuracy for FMA-UE total scores (R² = 0.950, Spearman correlation = 0.972).
• Fine-grained, action-level scoring models demonstrated enhanced discriminative capability.

Conclusions:

• A wearable sensor-based framework offers objective, high-resolution evaluation of post-stroke upper extremity motor function.
• The system accurately predicts standard FMA-UE scores, utilizing a minimal set of key items.
• This approach reduces assessment burden and maintains clinical consistency, supporting stroke recovery monitoring.