MEDBOLI: Medical Diagnosis Based on Ontologies and Logical Inference Alejandro Rodríguez¹, Myriam Mencke¹, Giner Alor-Hernandez², Ruben Posada-Gomez², Juan Miguel Gomez¹, Alberto A. Aguilar-Lasserre², ¹Departamento de Informática Universidad Carlos III de Madrid, Spain {alejandro.rodriguez, myriam.mencke, juanmiguel.gomez }@uc3m.es ²Division of Research and Postgraduate Studies, Instituto Tecnologico de Orizaba, Mexico {galor, rposada, aaguilar}@itorizaba.edu.mx Abstract The Differential diagnosis (ddx) is a systematic method to identify unknowns. This method, essentially a process of elimination, is used by taxonomists to identify living organisms and by physicians to diagnose the specific disease in a patient. One important point is the enormous amount of knowledge that doctors are required to possess, in order to be able to make a correct ddx. The main problem is that the number of diseases which exist worldwide is enormous, and it is beyond the scope of the human brain to remember them all. For this reason, in this paper we present MEDBOLI, an ontology-driven medical diagnosis system which applies the use of ontologies combined with logical inference and computation of probabilities to establish the probability of the diagnosis. 1. Introduction Nowadays there is an abundant number of medical diagnosis systems in existence which use computational intelligence to carry out differential diagnosis. The peak of the development of such systems was witnessed in the 1970s, during which many expert systems were developed. The objective of this paper is to offer an alternative to these expert systems, that in the great majority used obsolete technologies. In this paper we offer a solution using Semantic Web Technologies to develop a software that allows the users to make differential diagnosis. Furthermore, this software permits the user to introduce various features, such as displaying the probability of the diagnosis, calculated using biomedical statistics techniques and the possibility of making a diagnosis using drug interactions. 1. Motivating Scenario Differential diagnosis (or ddx) is a medical term which refers to a systematic method to identify a patient’s illness based on determined signs and symptoms, by routinely assessing the symptoms and eliminating diagnoses until the most likely diagnosis which matches the symptoms is deduced [6]. The principal objective of the research was to build a system which combines the strengths of probabilistic techniques, an ontology and logical inference, in order to aid a physician in the above process. In particular, throughout time many cases have emerged where incorrect medical diagnoses were made. This was a forceful motivation for the construction of the current system. Medical practitioners have the ability to make judgments based on their skills and experience, however their decisions are subjective and are not aided by an objective algorithmic decision making procedure, which is supported by mathematical computation and therefore always returns the same result. There are various practical applications of the system. Systems such as the current one may be viewed in the context of a diagnostic tool for medical practitioners, as well as a support tool for students of medicine, for example, students preparing for the MIR (Médico Interno Residente) exam in Spain. Therefore, the systems usefulness may be demonstrated in a scenario where, for example, a doctor is faced with a number of possible choices of diagnoses which he decided himself without the help of a computational system. In this case, it is the system which assigns a statistical probability to each possible diagnosis, enabling the medical doctor to conclude his decision. International Conference on eHealth, Telemedicine, and Social Medicine 978-0-7695-3532-6/09 $25.00 © 2009 IEEE DOI 10.1109/eTELEMED.2009.43 233