A Brief Review of Computation Techniques for ECG Signal Analysis

Automatic detection of life-threatening cardiac arrhythmias has been a subject of interest for many decades. The automatic ECG signal analysis methods are mainly aiming for the interpretation of long-term ECG recordings. In fact, the experienced cardiologists perform the ECG analysis using a strip of ECG graph paper in an event-by-event manner. This manual interpretation becomes more difficult, time-consuming, and more tedious when dealing with long-term ECG recordings. Rather, an automatic computerized ECG analysis system will provide valuable assistance to the cardiologists to deliver fast or remote medical advice and diagnosis to the patient. However, achieving accurate automated arrhythmia diagnosis is a challenging task that has to account for all the ECG characteristics and processing steps. Detecting the P wave, QRS complex, and T wave is crucial to perform automatic analysis of EEG signals. Most of the research in this area uses the QRS complex as it is the easiest symbol to detect in the first stage. The QRS complex represents ventricular depolarization and consists of three consequences waves. However, the main challenge in any algorithm design is the large variation of QRS, P, and T waveform, leading to failure for each method. The QRS complex may only occupy R waves QR (no R), QR (no S), S (no Q), or RSR, depending on the ECG lead. Variations from the normal electrical patterns can indicate damage to the heart, and these variations are manifested as heart attack or heart disease. This paper will discuss the most recent and relevant methods related to each sub-stage, maintaining the related literature to the scope of ECG research.