M. Elsonbaty, A. Elsonbaty, Islam Rasmy, Yasmin Elbasha, Shaimaa Wahba, Mohamed Abdelghany, Ahmed Hazem and et al

Abstract

Background and Aim: The neural-fuzzy systems used in modeling patient dynamics are usually complex while the Wiener models adopt a single anesthetic input for single output. Therefore, the research aimed at providing a real-time, multi-input and multi output model for anesthetic administration among pediatric patients that can be used in place of other complex models, for example, the neural fuzzy models which are complex as they contain many parameters. Methods: The study included 100 pediatric patients scheduled for herniorrhaphy surgery, which was via intravenous unconscious sedation. Two groups were used in the study, which were Group N (sevoflurane and rocuronium anesthesia was administered, n=50) and Group M (propofol and rocuronium anesthesia was administered, n=50) to establish their effects on the participants. The patients were taken to an operating room, and started by a face mask hooked at an electrocardiogram (ECG) monitor, and a non-invasive cuff which was subsequently placed on the contralateral arm where the cuff cycle was calibrated to measure the blood pressure after every 180 seconds, taking an average time of 154 minutes. Heart rates and BIS levels were also measured. Results: For Group M, 42%, 28%, and 30% of patients had normal, high, and low blood pressure levels, respectively. For Group N, 34%, 30%, and 36% of patients had normal, high, and low blood pressure levels in that order. For Group M, 24%, 34%, and 42% of patients had normal, high, and low heart rate levels in succession. For Group N, 20%, 34%, and 46% of patients had normal, high, and low heart rate levels correspondingly. For Group M, 12%, 46%, and 42% of patients had normal, high, and low BIS values concurrently. For Group N, 16%, 42%, and 42% of patients had normal, high, and low BIS values in that order. Conclusion: Real-time, multi-input and multi output model for anesthetic administration among pediatric patients is more effective compared to neural-fuzzy systems used in modeling patients’ dynamics, which are usually complex while the Wiener models were ineffective as they adopted a single anesthetic input for single output.