Adpative Neuro-Fuzzy Inference System Estimation Propofol Dose in the Induction Phase During Anesthesia; Case Study

In this study, the anesthetic drug dose is estimated with respect to patients' physiological parameters. the most critical anesthetic drug, propofol is considered in this modeling. the intravenous propofol is one of the widely used for both induction and maintenance phases of anesthesia. according to a deep uncertainty estimation model, the adaptive neuro-fuzzy inference system is applied to estimate the safe dose of anesthetic drug. the propofol model is estimated based on the patients' physiological parameters (age, weight, height, and gender) and variables (blood pressure, heart rate, and depth of anesthesia) each time. the sensitivity analysis evaluates the validity of the estimation model. at the end, performance of the proposed estimation model is compared to that of the classical pharmacokinetics-pharmacodynamics (pk-pd) model and the data obtained from the patients undergoing surgery. the results showed that the adaptive neuro-fuzzy inference system (anfis) estimation model with accuracy of 0.999 reduces the total amount of propofol dose. the novelty of the proposed model in this paper lies in its estimation of the depth of anesthesia in induction separately from the maintenance phase independent of bispectral index (bis). to validate our methodology, a real case study of mashhad hospital in iran has provided, resulting in a comprehensive discussion and managerial insights.