Kinematic Implications of Learned Non-Use for Robotic Therapy

Abstract

Learned non-use (LNU) may be a measure of carry-over and a potential measure for determining the efficacy of robot-based training for stroke rehabilitation. Identifying sensitive and reliable measures for detecting LNU would help evaluate whether robot training consistently reduced both motor impairment and functional disability after stroke-induced hemiplegia. We report on investigations into the relationship between LNU and four kinematic metrics commonly used to characterize impaired arm movements in Cartesian space during functional reaching. Results indicate that movement time, peak velocity, total displacement and movement smoothness accounted for less that 17% of the variability in LNU scores. A metric, derived from changes in impaired arm kinematic scores between unilateral and bilateral functional reaches, did not improve the prediction of LNU. Objectively quantifying LNU may require kinetic and kinematic metrics that measure proximal and distal sensory and motor impairment.