Towards an Intelligent Preoperative Surgical Decision Support System: A machine learning-based approach

Abstract

Artificial Intelligence has experienced a new impetus since the beginning of this decade under reasons interpreted by
the increase in computing capacities, the emergence of distributed massive data processing methods, and powerful machine learning algorithms.Different sectors are showing signs of investing in this new technology, particularly the surgical field.In this area, the managers have found that the operating room is one of the most decisive resources in a hospital establishment,and that the automation and optimization of its process is a primary priority, in particular the preoperative phase, a phase deterministic in relation to the importance of patient care, as well as in relation to good management of the operating room.Through this article, theidea leads us to embark on the trail of proposing an intelligent decision support system for the surgical preoperative phase, based on Machine Learning models, for the reason of the ability to its algorithms to make precise and interpretable classifications, our choice fell on the "Random Forest" algorithm, while exploiting preoperative predictive data,namely in current surgery (blood pressure, oxygen pressure,cerebral activity, body temperature, blood sugar, hematocrit) or to come (detection of circulating tumor cells, .. etc).The proposed model will be validated through a Dataset approved by a renowned institutional review board, in the Python environment version 3.9.16 with Google Colab.