Related Experiment Video
Updated: Aug 6, 2026
09:42
Motor Imagery Brain-Computer Interface in Rehabilitation of Upper Limb Motor Dysfunction After Stroke
Published on: September 1, 2023
Predicting Motor FIM at Discharge for Recovery Stroke Patients in a Rehabilitation Ward Using an AI Predictive
Kazuyuki Morita1, Senshu Abe2,3, Keisuke Ono2
1Occupational Therapy Section, Rehabilitation Department, Tokachi Rehabilitation Center, Hokuto Social Medical Corporation.
Japanese Journal of Comprehensive Rehabilitation Science
|July 17, 2026
Summary
An artificial intelligence (AI) tool accurately predicts motor Functional Independence Measure (FIM) scores for stroke patients at discharge. This AI model shows performance comparable to therapists, offering consistent predictions for rehabilitation outcomes.
Area of Science:
- Rehabilitation Medicine
- Artificial Intelligence
- Health Informatics
Background:
- Stroke rehabilitation requires accurate prediction of patient functional status at discharge.
- The Functional Independence Measure (FIM) is a key metric for assessing disability and outcomes.
- Traditional prediction methods may lack consistency or require extensive clinical experience.
Purpose of the Study:
- To develop and validate an artificial intelligence (AI) predictive analysis tool for motor FIM scores at discharge in stroke patients.
- To compare the accuracy of the AI tool against therapist predictions.
Main Methods:
- An AI prediction model was developed using a training dataset of 457 stroke patients.
- Model accuracy was validated on a separate dataset of 43 stroke patients.
- Statistical comparison of residuals, correlation coefficients, and coefficients of determination between AI and therapist predictions.
Main Results:
- The AI prediction model achieved a coefficient of determination (R²) of 0.799.
- AI predictions had residuals of 5.43 ± 5.93, comparable to therapist predictions (6.42 ± 5.52).
- AI and therapist predictions showed similar correlation coefficients (0.939 vs. 0.943) and R² values (0.881 vs. 0.889).
Conclusions:
- The developed AI model demonstrates high accuracy in predicting motor FIM at discharge for stroke patients.
- The AI tool's predictive performance is comparable to that of experienced therapists.
- This AI tool offers a consistent and potentially valuable method for predicting rehabilitation outcomes, irrespective of clinician experience.
