Title: Precision depth of anaesthesia estimation through machine learning and regression method: an intriguing case study unveiling the future

Authors: Mohammad Reza Gharib; Najmeh Jamali; Atefeh Roustaei; Najme Mohammadyahya

Addresses: Department of Mechanical Engineering, University of Torbat Heydarieh, Torbat Heydarieh, Iran ' Department of Industrial Engineering, University of Torbat Heydarieh, Torbat Heydarieh, Iran ' Department of Industrial Engineering, University of Torbat Heydarieh, Torbat Heydarieh, Iran ' Electrical Engineering Department, University of Houston, Texas, USA

Abstract: We introduce a novel propofol dosage prediction method for anaesthesia induction, emphasising the role of regression. Diverging from conventional reliance on mathematical models, our approach harnesses machine learning and data mining techniques for heightened accuracy. Experiences from past drug dosage errors inform the development of software that autonomously refines predictions. By extracting crucial information from datasets, machine learning techniques predict anaesthesia levels. This study focuses on applying regression techniques to anaesthesia research, mitigating the severe consequences of drug dosage errors. Our method demonstrates high precision, emphasising age and gender as crucial factors in predicting drug dosage. This contributes significantly to anaesthesia depth estimation, enhancing patient safety and reducing costs in Iranian surgical operating rooms. Integrating machine learning into drug dosage prediction shows potential for precision and efficiency in anaesthesia induction. However, thorough validation, considering medical history and concurrent medications, is essential before clinical implementation to ensure safe and effective anaesthesia.

Keywords: regression method; machine learning; optimisation; depth of anaesthesia; artificial intelligence.

DOI: 10.1504/IJAISC.2025.148114

International Journal of Artificial Intelligence and Soft Computing, 2025 Vol.9 No.1, pp.35 - 49

Received: 10 Feb 2024
Accepted: 11 Nov 2024
Published online: 25 Aug 2025 *