Correlations between statistical models of robotically collected kinematics and clinical measures of upper extremity...

Area of Science:

• Rehabilitation Robotics
• Neurorehabilitation
• Biomechanics

Background:

• Measuring treatment efficacy in rehabilitation robotics is challenging.
• Existing tools lack efficiency and comprehensive functional data.
• A need exists for reliable, data-rich measurement in upper extremity rehabilitation.

Purpose of the Study:

• To present linear regression models for assessing upper extremity function.
• To correlate kinematic measures from robotic-assisted virtual tasks with functional outcomes.
• To evaluate the models' effectiveness in capturing treatment effects in stroke survivors.

Main Methods:

• Collected kinematic data from 19 individuals with upper extremity hemiparesis post-stroke.
• Utilized haptic robots integrated with virtual rehabilitation environments.
• Developed three linear regression models based on seven kinematic measures.
• Correlated model outputs with scores from the Jebsen Test of Hand Function (JTHF).

Main Results:

• One model showed significant correlation with JTHF scores.
• Two models demonstrated significant correlation with JTHF change scores.
• Kinematic data from virtual tasks effectively reflected functional hand abilities.

Conclusions:

• The developed linear regression models show promise for measuring rehabilitation outcomes.
• Robotic-assisted virtual reality provides a viable platform for objective functional assessment.
• This approach enhances the measurement of treatment effects in upper extremity hemiparesis.