Prediction of Difficulties in Airway Management using Machine Learning Algorithms

Maintenance of the patient respiratory passage (airway) open and unobstructed breathing is a foremost duty of an anesthesiologist or other physicians who are involved in patient care under emergency situations. One of the main reasons for respiratory related problems is unanticipated difficult airway (unexpected difficulty in maintaining the face mask or passing an endotracheal to maintaining the continuous opening of the airway). If the doctor under duty fails to keep the airway open, it may lead to death or permanent bodily harm of patient, because of inadequate oxygenation. There is no readily available software solutions at present for predicting the exact probability of difficulty in mask ventilation and/or endotracheal intubation. Hence there is an urgent need to develop an expert system solution to support the anesthesiologists in predicting the risk and think about the remedial approaches to prevent the future consequences. The hard dealing with the advancement of a uniform categorization of patient safety has hindered the current advances in patient safety. This paper presents a prediction system to deal with difficulties in airway management using machine learning algorithms implemented in Python tools. This expert system aims to support the doctors in the case of prediction of difficult parameters of the patients associated with the airway management.