Can an Optimal Control Model Capture Tracking Kinematics? Insights from Healthy and Post-Stroke Robot Interactions.

Area of Science:

• Neuroscience
• Biomechanics
• Robotics

Background:

• Kinematic deficits are key indicators of motor and cognitive impairments in post-stroke individuals and healthy older adults.
• Robot-based kinematic analysis offers objective measures for evaluating motor performance and recovery.

Purpose of the Study:

• To evaluate an optimal feedback control model's ability to capture and predict kinematic performance in healthy older adults and post-stroke participants.
• To investigate if the model reflects motor learning changes after a robot-based pursuit tracking training task.

Main Methods:

• An optimal feedback control model was fitted to data from 6 healthy older adults and 3 post-stroke participants performing a robot-based pursuit tracking task.
• Kinematic data were collected at baseline (pre-training) and assessment (post-training) time points.
• Model performance was assessed by comparing modeled kinematics to actual participant movements and by analyzing changes in model parameters.

Main Results:

• The optimal control model accurately modeled pursuit tracking behavior in both position and velocity domains for healthy and post-stroke participants.
• Model accuracy significantly improved post-training, with higher correlations between modeled and actual kinematics (position median R = 0.89, velocity median R = 0.68).
• Significant increases in controller gains and decreases in positional noise were observed post-training, suggesting motor learning.

Conclusions:

• The optimal feedback control model effectively captures and reflects motor learning in pursuit tracking tasks for both healthy older adults and post-stroke survivors.
• The model's parameters provide insights into underlying changes in motor control strategies following training.
• This approach holds potential for objective assessment of motor function and rehabilitation progress in clinical populations.