Computerized electrocardiogram diagnosis: fuzzy approach.

This paper investigates the computerized analysis of electrocardiographic signals. The biological variability, the lack of standards in the definition of measurements and of diagnostic criteria make the classification problem a complex task. Two basic methods of the diagnostic process are described: the statistical model and the deterministic approach. In particular, a model for ECG classification will be illustrated where the imprecise knowledge of the state of cardiac system and the vague definition of the pathological classes are taken care of by means of the fuzzy set formalism.