[Differentiation of infarct patterns in the computerized ECG and confirmation with clinical and coronarographic...

With a view to assessing the validity of ECG criteria for a personal computerization programme, definite external features have been introduced as controls to provide objective data on the patient's disease apart from the ECG picture. To analyse the reliability of this programme, the ECG infarction component of the following groups of patients was considered. 1) 84 ECGs of patients who have suffered myocardial infarction in the more or less distant past and the diagnosis of which was confirmed at the time by chemical and laboratory data. These patients were also submitted to angiocardiography. 2) 120 ECGs of patients who have suffered myocardial infarction in the comparatively recent past and diagnosis of which was confirmed by chemical and laboratory data. These patients were not submitted to angiocardiography. 3) 85 ECGs of definitely not coronaropathic patients in whom angiocardiography was performed. In these clinically well documented cases, computer diagnoses were well correlated with those of treating physicians. It should however be pointed out that in 18% of patients with myocardial infarction, the ECG did not confirm the lesion. In the control group also, ECG alterations were recorded in 18% cases. This is a high percentage because in numerous patients the presence of rheumatic or shunt valvulopathies or myocardiopathies was encountered.