Predictive simulation of gait in rehabilitation.

This paper aims at demonstrating the capabilities of predictive, computational simulations of gait in clinical applications. In particular, the gait with bilateral and unilateral Ankle Foot-Orthoses (AFO's) is investigated. The problem is formulated in an optimal control framework where optimal motion and neural excitations to the muscles are computed solely on the basis of an assumed optimality criterion and periodicity constraints. The enormous potential of the approach as well as some of the current limitations are discussed on the light of simulation results.