Validation of a computerized QRS criterion for estimating myocardial infarction size and correlation with quantitative morphologic measurements.