EFFECT OF FORCE REDIRECTION ON UPPER LIMB NET JOINT MOMENTS DURING WHEELCHAIR PROPULSION Joseph M. Munaretto 1 , Jill L. McNitt-Gray 1,2,3 , Henryk Flashner 4 , and Philip S. Requejo 5 1 Department of Biomedical Engineering, 2 Kinesiology, 3 Biological Sciences 4 Aerospace and Mechanical Engineering, University of Southern California, Los Angeles, CA 5 Pathokinesiology Laboratory, Rancho Los Amigos National Rehabilitation Center, Downey, CA E-mail: munarett@usc.edu INTRODUCTION The force applied by the wheelchair user during manual wheelchair (MWC) propulsion often has a radial component that does not contribute to torque generation about the axis of the wheel. The mechanical cost on the body as well as effect on the pushrim while varying force direction has been simulated [1] using an inverse dynamic model. As shown for other multijoint goal directed tasks [2, 3], redirection of the reaction force relative to the body segments redistributes the net joint moments across the joints. While redistribution of the mechanical demand across the upper extremity may be advantageous to avoid overloading of individual musculoskeletal components, the same propulsive torque needs to be applied to the wheel to accomplish the task objective. In this study, we determined the range of net joint moment cost profiles needed to perform the same MWC task by using an experimentally validated 2D inverse dynamic model. Propulsion force was varied in direction, while torque applied to the pushrim was constrained to the experimentally measured condition. The net joint moment cost profiles of two subjects performing the MWC task were compared to illustrate how the mechanical demand imposed on the individual may vary when satisfying the same task objectives. METHODS Two wheelchair users volunteered to participate in this study in accordance with the Institutional Review Board. The subjects propelled at self- selected speeds for 10 seconds. Reflective markers were used to monitor the 3D motion of the hand, forearm, upper arm, and trunk segments (VICON, 50 Hz). The force applied to the wheelchair during propulsion was measured using force transducers (2500 Hz) mounted along the spokes of the wheel. Kinematics were projected into the sagittal plane and translated so that the origin lied at the center of the wheel. A two segment, two dimensional inverse dynamic model of the human body was created using MATLAB. Segments for the forearm and upper arm were represented as rigid bodies and segment kinematics were determined from experimental marker data. The measured reaction force was assumed to act at the distal end of the forearm (wrist marker). Figure 1: Kinematics and NJM at peak push force The location of the wrist/pushrim reaction force on the pushrim was determined from the location of the wrist on the wheel 1 tan W Wheel W y x θ −   =     where x w and y w are the Cartesian coordinates of the wrist. In this case, top dead center (TDC) would be 90°. The torque applied to the wheel was calculated as () ( )cos () ( )sin () z y Wheel x Wheel t F t t F t t τ θ θ = − where F x and F y are the components of the reaction force in the global frame acting on the wheel. The