Automatic closed-loop control of mivacurium infusions applyingneural networks

摘要

Recently introduced short-acting blocking agents like mivacuriumprovide benefits for the clinical routine due to a small onset time andhalf life. So for an automatic closed-loop control of the neuromuscularblockade a fast and highly adaptable control unit is needed. Itsdevelopment should not imply costly investigations for determining apharmacological model. The fulfilling of these requirements yields aself-adapting model-based predictive control system applying artificialneural networks. In a clinical study mivacurium was administered to 35patients (ASA I-III). The performance of the novel system (mean of the T₁-error: -0.32±1.7) compares favourably withclosed-loop controllers demonstrated in the past. These promisingresults and the easy adaption to other blocking agents encourage toapply this technology even for delivering hypnotic drugs