Goal-Directed Therapy for General ICU Patients Using Aggregated Multi-Objective Optimisation

Abstract

SOPAVent (Simulation ofPatients under Artificial Ventilation) is a blood gas model able to simulate patient ventilatory parameters subject to mechanical ventilation in the intensive care unit (ICU). In this paper, the SOPAVent model is further developed into a continuously updated model of patient ventilatory condition. The extended version of the model is used as the core component for a goal-directed optimisation strategy which aims to provide adaptive decision support for ventilatory therapy. Two objective functions are aggregated in an effort to interpret medical goals into an optimisation problem. The settings of the optimisation strategy are fine-tuned based on medical-goals and medical prioritization rather than mathematical optimality of solutions. The final decision support system is tested via a series of closed-loop simulations by assuming different clinical scenarios. Results show that the decision support system provides the correct advice for ventilator settings following adequate prioritization of competing medical goals.