ON ESTIMATION AND PREDICTION OF MYOCARDIAL-INFARCTION SIZE USING ARTIFICIAL NEURAL-NETWORK

Myocardial Infarction Size (MIS) is one of the major determinants of l eft ventricular function and is a factor of prognosis after acute myoc ardial infarction. For this reason, quantification of MIS is very impo rtant for the assessment of efficacy of metabolic, pharmacologic and s urgical intervention in reducing infarction size and appropriate selec tion of doses of drug therapy. Two models using neural network have be en presented that can be used for estimating the MIS from serial enzym e measurements in blood serum; one of the models may be used to predic t MIS with early data points.