Parameter analysis for robotic assessment of impairments in reaching due to stroke

Abstract

Stroke is a condition that arises from partial or complete blockage of blood flow to parts of the brain. Upper limb impairments (sensorimotor deficits) usually occur as a result of stroke. Assessment and diagnosis of the resulting upper limb impairments is important for effective rehabilitation and proper recovery. In terms of assessment, current ordinal scale systems used are inherently subjective and incorporate coarse rating scores (Chedoke McMaster, Fugel Meyer test, etc.). These scores do not provide clinicians with information regarding the underlying impairments which is important for effective rehabilitation therapy. We investigated the use of a reaching task on a bi-lateral upper limb robotic device (KINARM), developed by one of our team members (Stephen Scott), to assess upper limb impairments due to stroke. Control and stroke subjects were instructed to reach with both arms on the robotic device to illuminating spatial targets as quickly and as accurately as possible. We first developed a tool that would aid in locating targets where reaching impairments are visible for stroke subjects. From the Kinematic data collected, we then selected reaction time and first peak velocity of the hand and investigated their potential for separating control and stroke subjects. Results of our analysis showed the potential of both parameters as good quantitative assessment measures. In particular results for reaction time presented symmetrical differences between both arms were of main interest in this thesis. Such differences presented the potential of sensitivity in showing subtle impairments.