Prediction of difficulties in Intubation using an Expert system

Abstract

Expert anesthesiologist inserts a tube into the respiratory passage of the patient who is undergoing surgery in Intensive Care Unit or Operation Theater. The patient is unable to breath on their own during the surgery. This process helps in providing artificial assistance in breathing and prevents suffocation. Any interruption in oxygen supply or difficulties in intubation may result in acute internal body damages or may lead to death of the patient. Authors have carried out a literature review and field study to identify list of risk parameters leading to difficulty in intubation. Only 1 to 5 percentage of patients who are undergoing intubation suffer from difficult intubation. There is an acute shortage of expert anesthesiologist in hospital and employing an expert anesthesiologist will be very expensive. In US anesthesiologist have to perform at least 150 successful intubations in first pass to consider him as an expert. The author's main aim for designing an expert system using machine learning algorithms was to predict the difficulties in securing airway and also create an allocation system which allocates an expert anesthesiologist for difficult cases based on the results produced by the system. This paper discusses the procedure to carry out Intubation to emulate the high cognitive process of an expert anesthesiologist. The outcomes of the prediction process divides intubation into easy, difficult and impossible. The authors have designed framework of data sets, risk parameters, rules and algorithms. The expert system gives the prediction results of difficulty in intubation which will be validated by expert anesthesiologist.