Development of a muscle electrical stimulation parameter selection method with an intelligent system.

In this study, we propose a methodology for the selection of electrical stimulation parameters for muscle rehabilitation. This methodology uses a particle swarm optimizer to find a set of parameters that maximizes the muscle contraction force during the stimulation session. The optimization makes use of a previously validated mathematical muscle contraction model. To adjust the model to this application, we propose a series of adaptations to the model, such as the inclusion of a muscle fatigue function and different levels of muscle atrophy representations thru parametric variation. We based our model adaptations on the results of parametric sensitivity analysis and the qualitative changes in muscle properties during atrophy and fatigue described in the current literature. With the fatigue function, we successfully recreated the qualitative behavior reported in the literature for muscle fatigue caused by electrical stimulation. The parametric variations proposed following the sensitivity analysis are consistent with the physiological changes occurring during muscle atrophy, and the simulations with the model supported our hypothesis. Finally, we showed that with the model and the particle swarm optimizer, it is possible to find a set of electrical stimulation parameters that maximize the muscle contraction force during the stimulation session.