The leg drop pendulum test in spastic cerebral palsy: artificial feedback around a passive, non-linear, second order plant

Reports on ongoing work in the modeling of the leg drop pendulum test as a means of assessment in spastic cerebral palsy. In previous papers it was demonstrated that a passive model, is not adequate for modeling the limb of a subject with spastic cerebral palsy. Succeeding papers showed that the addition of active elements to the passive model, in the form of square wave force inputs, greatly improved the model's fit. This paper extends the work by replacing the active, square wave elements, with a set of feedback loops based on functions of the limb's velocity. These functions were optimized so that model data fit actual data taken from three subjects. The results demonstrate several important points. First, the artificial feedback model gives a better fit to real data than any combination of the previous passive model elements with or without active elements. Second, the addition of velocity feedback loops, which are physiologically justifiable, allows the removal of passive elements from the previous model which were not physiologically justifiable.