Kidney allocation expert system with case-based reasoning

Experts in fields of organ transplantation have to make crucial decisions based on medical, social, moral and ethical factors. The paper describes the kidney allocation expert system KARES-CBR. Our system consists of two main components. Traditional expert system part uses rule-based knowledge. These rules contain general medical knowledge, such as blood group and antigens matching rules. Another group of rules contains decision making rules related to social and ethical information about organ transplantation. After decision making, the expert system component generates the list of medically suitable patients for the transplantation of an available organ. Another part of the system collects the archive of previous successful organ transplantation cases. This historical data are used in case-based reasoning part of the system. Using different metrics, the system calculates the similarity values for kidney transplantation of selected patients relative to the previous successful transplantations. As a result, it selects the best suitable patient for organ transplantation. The system is implemented as an expert system shell and the different criteria and parameters can be easily modified. The system was tested on real data in the University hospital.