Upper-limb kinematic reconstruction during stroke robot-aided therapy.

Area of Science:

• Rehabilitation Engineering
• Biomechanics
• Neurorehabilitation

Background:

• Stroke survivors often experience upper-limb motor deficits, necessitating effective rehabilitation.
• Robot-aided therapy offers a controlled environment for motor recovery.
• Accurate kinematic assessment is crucial for understanding and tailoring rehabilitation strategies.

Purpose of the Study:

• To develop and validate a novel, accurate, and unobtrusive method for reconstructing upper-limb kinematics in stroke patients during robot-aided rehabilitation.
• To enable a deeper understanding of stroke patients' motor planning strategies in joint space.
• To facilitate the development of personalized therapies based on residual motor capabilities.

Main Methods:

• A robust analytic inverse kinematics procedure utilizing robot-provided hand pose data and upper arm acceleration measurements.
• Elbow position constraint computed via acceleration data for task space augmentation.
• Experimental validation comparing the method with optoelectronic motion capture and applying it to healthy and stroke subjects during circle-drawing tasks using the InMotion2 robot.

Main Results:

• Upper-limb joint angles reconstructed with a Root Mean Square Error (RMSE) of 8.3 × 10⁻³ radians.
• Distinct differences in joint space features (mean values, standard deviations) observed between healthy and stroke subjects.
• The method effectively assesses inter- and intra-subject variability, discriminating between healthy and stroke behaviors.

Conclusions:

• The proposed method provides accurate upper-limb kinematic reconstruction for stroke patients in robot-aided therapy.
• Findings enable evaluation of therapeutic outcomes, re-planning of robotic treatments, and understanding of pathology-related motor strategies.
• This contributes to a more personalized and effective approach to neurorehabilitation.