Non-invasive computerised detection of acute coronary occlusion.

The goal of this study was to evaluate the role of a computerised, non-invasive ECG method for detecting acute coronary occlusion (ACO). Ninety-five standard ECG leads were recorded, before and during ACO, from 18 patients undergoing balloon angioplasty. ECG amplitude and derivative parameters were calculated for the ORS, ST and T components of the ECG signal, before and during ACO. Results were obtained for each lead. Sensitivity of the standard visual ECG analysis for detecting ACO was 48%, whereas the percentage of conventional ECG changes during baseline was 14%. For the best ECG parameter, the amplitude parameter of the ORS component, sensitivity was 82%, and the percentage of parameter changes during baseline was 20%. The sensitivity for detecting ACO with five of the six ECG parameters studied was greater than that of the standard visual analysis. Ischaemic changes were detected in 4.3 +/- 1.6 leads per patient using the amplitude parameter of the ORS component, whereas, with the standard visual analysis, 2.5 +/- 2.1 leads demonstrated such changes (p<0.001). Results were then summarized per patient. The standard visual ECG analysis detected ACO in 15 of 18 patients (83%), if at least one lead showed ischaemic changes. The computerised analysis detected ACO in all 18 patients using the same criterion. The sensitivity of the computerised method for detecting ACO in the clinical setting of angioplasty was greater than that of the standard visual analysis. It is suggested that the computerised method may be useful for detecting myocardial ischaemia in other clinical settings of acute myocardial ischaemia.