Myocardial infarction (MI) is a life threatening heart disease in human beings. Early detection and treatment can save many lives. This paper presents the development of an artificial neural network, a novel non-linear soft computing tool for the detection of myocardial infarction using heart rate data derived from ECG signals of myocardial patients and healthy subjects. This heart rate data has been used to obtain a set of statistical, spectral and spatial parameters. A feedforward backpropagation artificial neural network has been trained to predict the presence or absence of myocardial infarction on the basis of these parameters. The accuracy, specificity and sensitivity of the neural network model in identifying myocardial infarction were 95.74 %, 91.67 % and 100% respectively. The results demonstrate the capability of the neural network model developed in identifying myocardial infarction with significant diagnostic accuracy. The developed system can support physicians in the diagnosis of myocardial infarction.