Machine learning in healthcare toward early risk prediction: A case study of liver transplantation

Abstract

Healthcare paradigms have always focused into the domain of early prediction of diseases. Especially in field of chronic diseases, the spotlight is always on the aspect of early detection and prevention by controlling the key risk factors in a comprehensive and integrated manner. In this endeavor the colossal power of health data comes into consideration; clubbed with the advanced techniques of computational intelligence to harvest the health data in the best possible way, the aim lies at the prediction of risks or deciphering interesting patterns and early signs of the diseases hidden in the health data. The output obtained from the intelligent analysis of the health data provides seminal insights to the design of more efficient treatment strategies. This work highlights some aspects of artificial intelligence in healthcare, illustrating through a case study of liver transplantation program, where the patient cohort could be interestingly separated into contrasting groups in a pretransplant scenario using machine learning, evincing a relationship with their respective posttransplant risks. In addition to relating the risk groups before liver transplantation with cardiometabolic risks through vascular age, this study accentuates the foundation of Clinical Decision Support System in transplantations, an assistive tool for the medical personnel to computationally analyze and visualize the comprehensive health situation of patients from the perspective of risks.