The evolution strategy--a search strategy used in individual optimization of electrical parameters for therapeutic carotid sinus nerve stimulation.

Abstract

Optimization problems, arising in the search for parameters and/or techniques of functional electrostimulation (FES), disproportionally increase when multiple electrodes, electrode configurations, electrical parameters, and stimulation modes may be applied. When computational or investigational effort precludes systematic studies in FES, we propose to apply and evaluate Rechenberg's evolution strategy, which in technical use and numerical optimization has been valid in comparison to more traditional methods. This strategy implements mutation and selection processes in analogy to biological evolution. The effect of combined multiple input variables on a quality function (Q) is experimentally evaluated. The actual computed value of Q serves as a selection criterion for those input variable combinations which lead Q to approach a target value (maximization), similar to a hill-climbing procedure. In radiofrequency controlled, therapeutic electrical carotid sinus nerve stimulation (CSNS), we varied (mutated) combinations of pulse frequency and pulse amplitude parameters, according to the evolution strategy, in individual patients. CSNS lowers blood pressure and decreases heart rate. Q was computed from blood pressure and heart rate responses to CSNS. The strategy individually optimized electrical parameters to achieve large depressor responses upon CSNS. Although, in contrast to technical usage, only two input variables were investigated, and biomedical experience with the evolution strategy is limited so far, its potential use in other fields of FES, especially when more input variables are to be optimized, is discussed and encouraged.