FIELD: medicine. SUBSTANCE: method involves detecting five levels of twenty-five factors of recurrent myocardial infarction like age, risk factors like previous angina pectoris according to class, occurrence of myocardial infarction in the past, myocardial infarction type and localization, the number of hours from the beginning of myocardial infarction attack to admission in hospital, factors evaluated during three days of myocardial infarction illness like anginous pains after the first arrest procedure, maximum respiratory frequency, heart beat rate, minimum systolic blood pressure, shock type, cardiac insufficiency, extrasysolia, paroxysmal tachycardia, atrioventricular block, sinoauricular block, His bundle block, lift or dislocation of T-segment, thromboembolism complications, acute pneumonia, the number of blood leukocytes, blood erythrocyte sedimentation rate, maximum enzyme or fibrinogen content. Artificial neuron network techniques are applied for performing mathematical calculations. Prognosis level being found greater than 50%, recurrent myocardial infarction occurrence is to be predicted. EFFECT: enhanced accuracy of prognosis. 1 dwg, 2 tbl
