An overview on intra- and inter-patient paradigm for ECG Heartbeat Arrhythmia Classification

Abstract

Electrocardiogram (ECG) is one the most used tool to diagnose the health problems of the heart, but a manual diagnosis of these heart-beat classes by cardiologists could be time-consuming. An automated computer aid remains important in such cases to facilitate some tasks and aid specialists rapidly define and classify arrhythmias. In this paper, we investigated paradigms used for ECG arrhythmia classification of the Intra- and inter-patient paradigm with and without the association for the advancement of medical instrumentation (AAMII) standards and presented several papers that worked on each type. We discussed a variety of limitations and revealed that there is still room for improvement in the classification's performance, especially for the NON-AAMII inter-patient paradigm, to further boost confidence for the applicability of such solutions in clinical diagnosis which is more conforming to the clinical practice in which varieties of ECG signals are collected from different subjects. Such an organized literature survey allows researchers to merge an unobstructed view on all the aspects of ECG classification for the identification of gaps and the research issues unmet so far.