A novel system for automatic classification of upper limb motor function after stroke: an exploratory study

Area of Science:

• Biomedical Engineering
• Rehabilitation Medicine
• Neuroscience

Background:

• Accurate motor function assessment is crucial in early post-stroke recovery for evaluating interventions.
• Existing methods for motor performance measurement can be subjective and time-consuming.
• There is a need for reliable, automated tools for independent and repeated motor function evaluation.

Purpose of the Study:

• To explore the potential of a newly designed portable motion capture system for automatic assessment of motor impairment after stroke.
• To develop and validate a system for objective, quantitative evaluation of upper limb motor function.
• To assess the system's ability to classify motor impairment based on established functional scales.

Main Methods:

• Development of a portable motion capture system to acquire 3D kinematics of upper limb movements.
• Utilizing an automatic decision tree classifier with features derived from the Wolf Motor Function Test (WMFT) Functional Ability Score (FAS).
• Testing the system on five stroke patients across five tasks, comparing results with a blinded clinician.

Main Results:

• The system demonstrated high agreement with clinician assessments for performance time (mean difference 0.17s).
• The automated system achieved agreement with clinician FAS evaluations in 4 out of 5 patients for two specific tasks.
• The prototype successfully classified upper limb movement according to the WMFT in a clinical setting.

Conclusions:

• The developed portable system can automatically classify upper limb motor function post-stroke using a widely recognized scale (WMFT).
• This technology represents a significant advancement towards precise, independent motor evaluation for stroke patients.
• The system's portability and low cost enhance its potential for use in clinical settings and research trials.