Assessing systematic effects of stroke on motorcontrol by using hierarchical function-on-scalar regression

Area of Science:

• Neuroscience
• Biostatistics
• Rehabilitation Science

Background:

• Stroke significantly impacts population-level motor control.
• Individual variations in motor control post-stroke require careful consideration.
• Understanding these effects is crucial for effective rehabilitation strategies.

Purpose of the Study:

• To develop a statistical model for analyzing stroke effects on upper extremity motor control.
• To differentiate common population-level motor control changes from subject-specific effects.
• To identify relationships between motor impairment severity, target number, and motor control patterns.

Main Methods:

• Utilized bivariate function-on-scalar regression with subject-level random functional effects.
• Employed penalized splines for fixed and random effects modeling.
• Incorporated a Wishart prior for residual correlation and variational Bayes for parameter estimation.

Main Results:

• The proposed Bayesian method accurately estimates parameters and provides reliable inference.
• Application results demonstrate a systematic, population-level component in stroke-induced motor control deficits.
• Identified specific patterns in kinematic data related to stroke severity and target reach.

Conclusions:

• Stroke exerts a consistent influence on upper extremity motor control across individuals.
• The statistical framework effectively disentangles population and individual effects in motor control studies.
• This approach offers valuable insights for targeted stroke rehabilitation and understanding motor recovery.