Key result
Novel entropy-based machine learning model classifies normal ECG, AFIB, and CHF with ~92% accuracy.
Why the study?
An automatic classification method is needed to facilitate the detection of ECG signal abnormalities such as Normal, AFIB, and CHF states.
Does an automatic classification method using FdispEn and first-order statistics combined with SVM accurately classify ECG signals into Normal, AFIB, and CHF?
Does an automatic classification method using FdispEn and first-order statistics combined with SVM accurately classify ECG signals into Normal, AFIB, and CHF?
A novel machine learning approach using FdispEn and SVM can classify ECG signals into normal, AFIB, and CHF with 91.5% accuracy, potentially assisting in automated clinical diagnosis.
Supports automated ECG classification research; leaves open prospective clinical validation before adoption.
The heart is one of the most important organs in the human body. The presence of abnormalities in the heart can be fatal for a person. One way to detect heart abnormalities is an Electrocardiogram (EKG) signal examination. To facilitate the detection of ECG signal abnormalities, an automatic classification method is needed. Therefore, in this study, a method for classifying ECG signals using FdispEn (Fluctuation-based dispersion Entropy) and first-order statistics is proposed. FdispEn measures the uncertainty in the signal and is expected to be able to distinguish the physiological state of the ECG signal time series. In this study FdispEn and statistical computing were used as feature extraction of the ECG signal and combined with the Support Vector Machine (SVM) for the classification process of Normal ECG, AFIB (Atrial Fibrilation), and CHF (Congestive Heart Failure). The method proposed in this study generates an accuracy of 91.5%. The system proposed in this study is expected to assist in the clinical diagnosis of abnormalities in the heart.
No takes yet. Share an insight, caveat, or question.
Jafal et al. (2022) studied Cardiac abnormalities (Normal, Atrial Fibrillation, Congestive Heart Failure). Fluctuation-based dispersion Entropy (FdispEn) and first-order statistics with Support Vector Machine was evaluated on Classification accuracy. A classification method using Fluctuation-based dispersion Entropy, first-order statistics, and Support Vector Machine achieved 91.5% accuracy in distinguishing Normal ECG, AFIB, and CHF.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: