ABSTRACT
In this chapter, it is discussed how deep learning-based solutions can help to address this challenge. It is shown how two types of transfer learning techniques can be to retrain the pre-trained VGG-16 model to obtain two new networks, namely, VGG-MI1 and VGG-MI2. Specifically, it is discussed that how one can modify the last layer of the VGG-16 and the final layer of the VGG-Net model accordingly to suit our requirements. In addition, it is shown how various functions can be introduced to the model to reduce overfitting. Within the VGG-MI2, one layer of the model is selected as a feature descriptor of the ECG images to describe some of the essential features. Thus, in this chapter, it is shown how it is possible to develop an accurate tool for the analysis of electrocardiograms for the efficient diagnosis of myocardial infarction.
