Sensitivity and specificity of magnetocardiography, using computerized classification of current density vectors maps, in ischemic patients with normal ECG and echocardiogram

Abstract

Abstract The present study aimed to develop a computerized classification of magnetocardiograms (MCG) on the basis of current density vectors (CDV) map analysis for the diagnosis of coronary artery disease (CAD). The study included 123 patients with angina and angiographically documented CAD but with normal ECG at rest and normal left ventricular function. The control group consisted of 124 normals. 4-channel SQUID-magnetometer systems located in an unshielded hospital environment was used. MCG recordings were taken at 36 pre-thoracic sites over the pericardial area and reconstruction of CDV maps within the ST–T interval was applied. Each CDV map was classified automatically with a scale from 0 (normal) to 4 (grossly abnormal). The classification number depends on the number of large vectors directed left–downward and the presence of additional clusters. The mean class was calculated for each subject in order to discriminate between groups. Using a mean class value of 1.75 as threshold for the discrimination between healthy persons and CAD patients, 76% sensitivity and 81% specificity were achieved. The computerized classification of CDV maps seems to be a useful tool for the diagnosis of CAD.