Date of Award

Summer 2009

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)


Biomedical Engineering

First Advisor

Scheidt, Robert A.

Second Advisor

Schmidt, Brian D.

Third Advisor

Schindler-Ivens, Sheila


Objective: To characterize the two-dimensional spatial distribution of hand forces and abnormal muscle activation patterns during and after passive displacement of the hand in hemiparetic individuals post-stroke. This was done in order to quantify deficits in neuromuscular control relative to healthy age-range matched control subjects. Methods: We characterized how the arm's response to passive stretch varied across the subject's reachable workspace for 10 stroke subjects (SS) and 3 neurologically intact (NI) subjects. The subject's hand was attached to the handle of a two-joint robot that generated controlled movements between 20 spatial locations in the workspace. The subject was coached to relax at all times. The system measured hand forces and electromyographic activity (EMG) during the transition between target locations and for a minimum of 20 seconds thereafter. Target locations and the sequence of transitions between them were chosen so that motion occurred predominantly at either the shoulder or elbow joint (while keeping the other joint angle nearly constant). The subject's hand was moved such that motion at the joint of interest occurred at one of three peak speeds ( computed separately for each subject based on anthropometric measurements): slow (6 degrees/sec), medium (30 degrees/sec) and fast (90 degrees/sec). Most target locations were approached from multiple directions (i.e. elbow extension, elbow flexion, shoulder extension and shoulder flexion). Each stroke survivor participated in at least two experimental sessions performed on separate days in order to assess repeatability of our observations. We analyzed the temporal evolution of hand forces with respect to movement speed in order to determine if and when hand forces might depend on the rate of joint rotation, as would be anticipated for individuals presenting with hyperactive velocity-dependent stretch reflexes. We also analyzed the effect of workspace location on hand forces during and after passive transition of the hand. Results: Passive translation of the hand between workspace locations induced repeatable velocity-dependent reaction forces that decay to steady-state within about 2 seconds after end of transition (EOT). In steady-state, posture-dependent muscle activity in the "relaxed" arm post stroke gave rise to large forces at the hand/handle interface that varied systematically across the workspace. These forces reflect long lasting, posture dependent biases that oppose hand displacements away from the center of its reachable workspace. Conclusion: Future studies should examine the extent to which posture-dependent biases such as those measured here contribute to dysfunction in the control of arm posture and movement after stroke.



Restricted Access Item

Having trouble?