Fuzzy Cluster Analysis

Abstract

uzzy Cluster Analysis (FCA) has been shown to be better than ST depression in predicting the extent of coronary disease in patients with positive stress tests. 1 Derived from fuzzy set theory (fuzzy logic), the method combines information from 6 stress test variables, each of which may have a range of abnormality from very mild to very severely abnormal. The resulting output classifies any positive stress test as either mildly, moderately, or severely abnormal.Several studies have attempted to combine exercise variables to improve the prognostic value of treadmill testing. 2,3 We conducted this study to see if FCA is useful in predicting cardiac outcomes in outpatients with positive stress tests. Also we again looked at how well FCA predicted the extent of coronary artery disease using a larger database than in our previous study. 1 Methods Study GroupStandard Bruce protocol treadmill tests from 1,955 stable outpatients were reviewed. ST-segment deviation during exercise was measured 0.06 sec after the J point, and the largest absolute value for ST-segment deviation in any lead except aVR was recorded. Patients were selected for entry into the study if they had 0.5 mm or more of exerciseinduced ST-segment depression or 1 mm or more of exercise-induced ST-segment elevation. Patients were excluded if they had had coronary angioplasty or coronary bypass surgery at any time before the stress test, if they had known significant valvular, hypertensive, or congenital heart disease, if they had significant chronic lung disease or peripheral vascular disease, if they had left bundle branch block at rest or during exercise, or if they were taking digoxin at the time of the exercise test. Patients were not excluded if they had resting ST-segment depression, Q waves, T-wave inversions, or right bundle branch block.Two hundred and thirty-two patients (182 men and 50 women, aged 30-79 years) meeting these criteria formed the database. Typical angina was present in 130 patients, atypical symptoms were present in 74, and 28 were asymptomatic. There was a history of previous myocardial infarction in 25 patients. On the day of the stress test, 28 patients had taken nitrates, 17 had taken calcium-channel blockers, 6 had taken beta-blockers, and 41 had taken aspirin. The patients were followed up to 96 months after the stress test, with an average follow-up of 65 months (range, 8-96 months).The time between the stress test and each patient's first cardiac event was recorded. The cardiac outcomes included in this study were: coronary bypass surgery, percutaneous translumiual coronary angioplasty (PTCA), coronary stent placement, myocardial infarction (either Q wave or nontransmural with confirmatory cardiac enzyme elevation), cardiac death, hospitalization for unstable angina, and hospitalization for congestive heart failure. Between the Jpn Circ J 1998; 62: 750 -754 (Received March 24, 1998; revised manuscript received May 18, 1998; accepted May 19, 1998 Several studies have shown that combining the change in the ST-segment with ...

Fuzzy Cluster Analysis of Positive Stress Tests: Comparison with Stress Echocardiography and Nuclear Perfusion Imaging in Patients with Triple Vessel and Left Main Coronary Disease

“…Thus the results of this study, along with the ability of FCA to predict extent of coronary disease and future cardiac events better than ST-segment depression [1, 2], make FCA a useful tool for the clinician in the evaluation of patients with positive stress tests.…”

“…The details of FCA have been described in previous reports [1, 2]. In brief, a fuzzy set is created for each of six stress test variables (ST-segment change, difference between resting systolic and peak exercise systolic blood pressure, total treadmill time, peak exercise heart rate as a percentage of 100% predicted maximum for age, time to onset of angina, and duration of repolarization abnormalities) with the degree of membership ranging from 0.1 (mildest abnormality) to 1.0 (most severe abnormality).…”