Assisted Spatial Sit-to-Stand Prediction—Part 2: Virtual Injured Elderly Individuals

Abstract

Abstract Sit-to-stand (STS) is a challenging task, especially for injured elderly in their daily life. This paper presents Part 2 of the assisted spatial STS computational motion prediction to investigate the effect of a unilateral grab-rail bar placed on the right-hand side on STS motion performance for injured virtual individuals (computational digital human model-DHM) with either right or left knee injury. The motion prediction formulation was constructed as a nonlinear optimization and validated in Part 1 of the assisted spatial STS motion prediction (Yang, J., and Ozsoy, B., 2021, “Assisted Spatial Sit-to-Stand Prediction—Part 1: VirtualHealthy Elderly Individuals,” ASME J. Comput. Inf. Sci. Eng., 21(4), p. 041002). Injuries are implemented into the formulation with reduction rates of 75% in a single knee‘s extension torque limits. Two different objective functions are tested: The first one is just minimizing the dynamic effort where the second one is minimizing the difference between right and left side support reaction forces in addition to the dynamic effort. Computational simulations resulted in that systematic changes are seen in the joint coordination better with the first objective function for the elderly group with either right or left knee injury. Since only one of the knees is modeled to experience a strength loss, virtual-individuals tend to flex their trunk to the intact side whilst rotating the upper-body to the right side with both objective functions due to holding the grab-rail bar, which was placed to the right side of the biomechanical model. Based on the peak values of the center of mass velocity and displacement profiles in both the medial-lateral axis, virtual-individuals prefer to use lateral plane motions less and transverse plane motions more to minimize the difference between right and left side-vertical support reaction forces with the second objective function.