Mathematical models of central pattern generators in locomotion: III. Interlimb model for the cat.

Possible neural connective patterns and functions with respect to interlimb coordination are studied theoretically with a mathematical model of the central pattern generating system for cat locomotion. Activities in populations of neurons controlling limb joint flexors and extensors in all four limbs are represented by a system of nonlinear differential equations. Solutions of the system for various parameter values simulate various gaits of the cat. The model is shown to be capable of generating all gaits of the cat and accounting for corresponding phase changes in interlimb coordination. The model also exhibits smooth changes of gait, and smooth initiation and termination of stepping. Further, within each limb, muscle sequencing, step cycle phases, and flexor-extensor interactions can be studied. The model suggests that one of the simplest mechanisms for a central command system to change the gait is via inhibition of specific interlimb propriospinal pathways. In a final section, properties of both proposed single limb and interlimb models are reviewed with specific reference to planning future experimental and theoretical studies.