Minimizing Epistemic Uncertainty for Predictive Control of General Anesthesia

Epistemic uncertainty refers to the lack of information from knowledge and in medical field it is often paired with decision making situations and systems largely dependent on context and expertise. Our approach is uncertainty-centric and digitalizes knowledge with high certainty level, in order to improve the efficacy of knowledge-driven decision support system for anesthesia regulation. We propose to digitize some of surgical actions as part of minimizing uncertainty in closed loop control. Notable disturbances are those induced by the surgery actions, but they abide clinical protocols and semantic transformation enables their use in computers. We introduce a predictive control methodology to accommodate the augmented information of the system and evaluate its impact on the closed loop performance. Simulations of closed loop control of anesthesia for various surgery protocols indicate a high relevance and applicability of proposed approach. The proposed control structure allows the use of medical semantics in decision making process of anesthesia regulation.