Fuzzy Decision Support Method for Medical Diagnoses

The most medical knowledge is characterised by uncertainty, imprecision and vagueness. To formalise this knowledge is an actual task. One of the possible ways is to use fuzzy logic-based systems for diagnostic decision support. Fuzzy discrimination and connectivity analyses are the basis for the presented method, which helps physicians to make a decision, to avoid fault in the diagnoses. The method allows to model the physicians' behaviour and mentality To find the weight of each symptom, to choose a set of representative symptoms, to separate diseases which look like each other, but different in the main point is a part of examples of senior clinicians knowledge formalising in this method. The method was realised on IBM PC computer, using Pascal language. It was tested for venal diseases and showed a good correlation with traditional diagnostically methods.