Batch identification of neuromuscular blockade models

Abstract

This work addresses the problem of identifying neuromuscular blockade models of patients undergoing general surgery. First, a sensitivity analysis is made, exploring the Wiener structure of the system. The outcomes of this analysis are twofold: First, it provides information about the time periods in which data is more informative for parameter estimation. Second, it is the basis of a local identifiability analysis that allows to decide which parameters are to be estimated from data and which are the ones whose values should be a priori selected based on previous insight. The time dependency of sensitivity is then used to adjust the weight of output errors in a Bayesian cost function whose minimization yields parameter estimates: Whenever the sensitivity is low, the weight is reduced. The contribution of the paper consists in the demonstration of this procedure using actual clinical data.