Fuzzy Braunwald-modified chest pain assessment for Unstable Angina

A chest pain assessment system for diagnosing the risk of Acute Myocardial Infarction for Unstable Angina based on the fuzzy logic and a modified Braunwald risk classification is proposed in this paper. While the fuzzy logic allows dealing with both the uncertainty and the imprecision in information, the risk chest pain evaluation is carried out by the symptom severity diagnosis derived from the Braunwald classification. This paper extends previous approach by using classes very low, low, medium, high, and very high, instead of three classes (very) low, medium, (very) high, respectively, concerning to recent angina (RA), subacute angina at rest (SAR), and acute angina at rest (AAR), as originally envisaged. The resulting fuzzy mapping between the input variables and the Braunwald-modified risk chest pain output variable enables smoothing the decision (diagnosing) surface. The proposed fuzzy decision support system may prove to be an auxiliary tool in the early diagnosis of Acute Myocardial Infarction to allow a smooth and gradual representation of the patient's clinical condition.