A computerized assessment tool for the upper extremities motor performance in individuals with Parkinson's disease

Objective: Computerized assessment of Parkinson's disease (PD) patients' motor performance can pave the way to subject-specific design of the rehabilitative interventions. The objective is to propose a computerized assessment tool for upper extremities motor performance in PD. It is based on kinematic indices and a virtual reality exergaming system. The reliability, construct and discriminative validity of the proposed tool were investigated.Methods: A set of virtual tasks was designed and integrated to the therapeutic programs of the participants including uni/bimanual, in-phase/anti-phase hand movements on the less/more affected sides. 33 persons with PD participated in a test re-test study to determine the reliability, discriminative validity on the medication states, and validity of the kinematic indices. Clinical assessments, including Box and Block Test, 9-Hole Peg Test, Movement Disorders Society motor section of the Unified PD Rating Scale (MDS-UPDRS III), were performed for validity assessment.Results: Task type has a significant impact on the validity of the indices. Reaching tasks were shown to be superior. Besides, it was demonstrated that a minimum of 3 repeats is needed for reliable results in the proposed setting. Jerk indices, curvature, and spectral speed metrics lack either reliability or validity. Temporal speed metrics have superior reliability (Intraclass Correlation Coefficient ICC > 0.9) and validity (Pearson > 0.6) in reaching tasks.Conclusion: Temporal speed metrics and point reaching virtual tasks were shown to have high potentials for objective assessment of upper extremity (UE) motor performance in PD in the context of a virtual reality exergaming rehab program.