Regulation of Depth of Hypnosis through Optimized AI based Conditioned Super Twisting SMC

Abstract

Anesthesia infusion has major role in regulating and maintaining the depth of hypnosis of a patient during surgeries. The idea of anesthesia is to move the patient to an unconscious state and make the surgery painless. For regulating depth of hypnosis of a patient during surgical procedures, many linear control algorithms have been introduced but they are not good enough to deal with disturbances and uncertainties present in the system. In this research work, we have proposed optimized conditioned super twisting sliding mode controller to get the desired Bi-spectral Index Scale (BIS) value and to improve the system’s overall performance as well as to remove chattering. Moreover, the controller is robust against inter-patient variability dynamics and the surgical stimuli that help in moving the patient towards conscious state. For regulation of hypnosis level of a patient, the reference value of BIS is generated through a trained Artificial Neural Network. The real set of data of 8 patients undergoing surgery is used to examine the controller’s performance. Lyapunov stability criterion is used for ensuring stability of the system. MATLAB/Simulink ODE 45 environment is used for verifying the performance of proposed controller. The proposed controller is then compared with already proposed super twisting sliding mode controller (STSMC). The simulation results demonstrate the improved dynamic behavior and robustness with conditioned STSMC. Real-time hardware-in-loop (HIL) testing is used to further verify the system’s performance.