High-level motor planning allows flexible walking at different gait patterns in a neuromechanical model

Abstract

Humans are able to adopt almost any desired gait pattern on the fly when walking. We postulate that this flexibility in humans is partially due to the ability to control the whole body during walking as a volitional, goal-directed movement that can be planned and changed, rather than having to rely on habitual, reflexive control that is adapted over long time-scales. Here we present a neuromechanical model that accounts for this flexibility by combining movement goals and motor plans on a kinematic task level with low-level spinal feedback loops. We show that the model is able to walk at a wide range of different gait patterns by choosing a small number of high-level control parameters representing a movement goal. A larger number of parameters governing the low-level reflex loops in the spinal cord, on the other hand, remains fixed. We also show that the model is able to generalize the learned behavior by re-combining the high-level control parameters and walk with gait patterns it had not encountered before. Furthermore, the model can transition between different gaits without loss of balance by switching to a new set of control parameters in real time.