Utilizing PTB XL+ Dataset for Advanced Cardiac Disease Prediction with Deep Neural Networks

One of the main components of early diagnosis and treatment, heart disease prediction affects very much the result of the patient and the healthcare systems. In this work, the authors investigate using ECG images deep neural network-based prediction of cardiac diseases. In many cardiovascular diseases, electrocardiogram findings provide particular information on the electrical activity of the heart and thereby constitute a useful diagnostic tool. Several methods accessible in deep neural networks are used to build carefully planned architectures and train them on annotated ECG picture datasets to acquire the ability to identify unique traits indicative of different cardiac illnesses. Currently under development the DNN model makes use of this dataset. The tests run on several criteria reached an 86.12% accuracy.