Marzieh Mohammadpour - M.S of Control Engineering Islamic Azaad University, Bojnord Branch

since the science of medicine has equipped with intelligent instruments in the diagnosis and treatment of diseases, physicians’ errors and life and property loss can be greatly reduced. Problems occurred for patients due to the improper management of anesthesia confirm that making decision based on traditional and experimental methods is not reliable and hence, applying scientific and efficient methods seems necessary for accurate adjustment of the dosage of the drugs. Given that the studied system works manually, it is necessary to conduct the preliminary studies on a model. The patient's vital signs change during anesthesia and based on this, anesthesiologist estimates the depth of anesthesia and decides on decreasing or increasing the dosage of the drug. In this paper, pharmacodynamic -pharmacokinetic compartmental model was used and then, through performing simulation by MATLAB software, the patient’s depth of anesthesia was controlled (BIS index) during operation. To control the amount of anesthetic drug in the body, the reference model of high-order artificial neural networks-based controller has been used. Studies indicated that high-order neural networks have an effective capability of storing, calculating and learning. This controller compensates the system delay so well. The system response has the short setting time and high speed and the patient’s depth of anesthesia is intelligently and automatically controlled during surgery in less than 16 minutes.

depth of anesthesia, control, artificial neural networks, pharmacodynamic-pharmacokinetic compartmental model