A planar piecewise continuous lumped muscle parameter model for prediction of walking gait.

Goal: This work aims to develop a planar piecewise continuous lumped muscle parameter (PPCLMP) model that can utilize inputs that can be obtained in a clinical or home setting using simple tools (e.g. video cameras and inertial sensors) to predict human walking gait.

Methods: The model characterizes the sagittal-plane movement of the lower limbs during the single stance phase as an inverted pendulum, the double stance phase as a kinematic chain, and the swing phase as a double pendulum. The joint angles and angular velocities at the end of one phase are used as the initial conditions of the next phase. The model predicts the gait cycle based on the initial joint angles and angular velocities via forward dynamics. The errors between the initial and end conditions are minimized by changing the input initial joint angles and angular velocities of the gait cycle.

Results: Sensitivity analysis showed that the errors between the initial and end conditions of a gait cycle were sensitive to the initial joint angles. The step length was sensitive to subject stature. The model only works for a certain range of initial conditions.

Conclusions: The model can predict gait cycles based on forward dynamics and selects initial conditions that minimize the errors between the initial and end conditions of the gait cycle. The model utilizes 2-D representations of lower limbs and simplified representations of joint torques to reduce the required inputs for gait prediction and builds the foundation of gait assessment tools.