Heartbeat Pattern and Arrhythmia Classification: A Review

Abstract

In today’s era, the lifestyle of people has become much more sophisticated due to the involvement of stress, anxiety, and depression in the daily routine of human beings. In such a scenario, cardiac diseases are growing rapidly in youngsters and senior citizens. It is also observed that cardiac diseases are crucial and sensitive, including life-threatening chances. So, it is essential to detect and prevent such cardiac disorders within the required time for recovery. Since there has been a lot of research in the prediction and prevention of cardiac disorders, cardiac arrhythmia is also one of the majorly occurring diseases in the bulk of the population. The electrocardiogram is the cheap and best way to diagnose the problem of cardiac arrhythmia, and a huge amount of data is collected daily in hospitals and pathological centers. Previously, various automated models were developed for detecting cardiac arrhythmia using deep learning approaches and machine learning. In this work, we have reviewed recently developed automated models and evaluated their performance based on specific parameters like deployed datasets, variation of input data, applied application, methodology, and results obtained by the developed model. The limitations of reviewed papers are also mentioned in addition to their future scope for improvement.