This paper mainly focus on health care monitoring and prediction of heart attack using Artificial intelligence (AI) model using Deep (D) Learning (L) methods. Artificial intelligence has been increasing fast in recent years in terms of software algorithms, hardware implementation, and applications in a wide areas. This research shows that our original model's accuracy is about 0.98, and that using Deep Learning Methods can boost accurateness and open up current research possibilities. The AUC of the rhythm classes is shown in the results, and we can see that the accuracy is higher than the annotations. It is realized, there is a possibility for a more complete end-to-end methodology to be developed that employs deep learning using CNN methods and other neural network methods. ECG is a breakthrough concept in the field of cardiology and has paved way for lot of advanced techniques to expedite the diagnosis cardiological issues ensuring in-time treatment. With the beginning of High performance computing (HPC) and easy access to such infrastructure it has opened new doors to integrate new computational paradigm with ECG where we can implement various Machine Learning and Deep Learning Techniques to understand and analyze ECG in a better way. One of the challenges were the complexity of the data, the peak detection in the waveform and learning the patterns. Our current work, emphasizes on these aspects and we have proposed a Deep Learning based algorithm which can predict / classify arrhythmia with a greater accuracy.

Healthcare, deep learning, artificial intelligence, machine learning, ECG, arrhythmia