Predicting Arm Nonuse in Individuals with Good Arm Motor Function after Stroke Rehabilitation: A Machine Learning...

Area of Science:

• Neurorehabilitation
• Stroke Recovery
• Motor Function Assessment

Background:

• Many stroke survivors exhibit persistent arm nonuse despite achieving good arm motor function post-rehabilitation.
• This discrepancy highlights a gap in understanding the factors contributing to functional arm disuse after stroke.
• Identifying predictors is crucial for tailoring interventions to improve upper limb engagement.

Purpose of the Study:

• To identify predictors of arm nonuse in stroke survivors with good motor function.
• To develop predictive models for post-stroke arm nonuse.
• To inform the design of individualized stroke rehabilitation programs.

Main Methods:

• Retrospective secondary analysis of 78 stroke survivors.
• Categorization into groups based on Fugl-Meyer Assessment Upper Extremity (FMA-UE) and Motor Activity Log Amount of Use (MAL-AOU).
• Feature selection to identify top 5 predictors and subsequent model building using four algorithms.

Main Results:

• Key predictors identified: preintervention FMA-UE, MAL-Quality of Movement, Wolf Motor Function Test-Quality, MAL-AOU, and Stroke Self-Efficacy Questionnaire.
• Predictive models achieved high accuracy (0.75-0.94) and AUC (0.77-0.97).
• Arm motor function, daily activity use, and self-efficacy significantly predict post-stroke arm nonuse.

Conclusions:

• Measures of arm motor function, daily arm use, and self-efficacy are crucial for predicting post-stroke arm nonuse.
• Prioritizing these assessments can help design targeted rehabilitation strategies.
• Interventions should focus on improving both motor function and functional arm engagement to reduce nonuse.