Contents lists available at ScienceDirect Physica Medica journal homepage: www.elsevier.com/locate/ejmp Review paper Computer-aided diagnosis of congestive heart failure using ECG signals – A review V. Jahmunah a, ⁎ , Shu Lih Oh a , Joel Koh En Wei a , Edward J Ciaccio e , Kuang Chua a , Tan Ru San b , U. Rajendra Acharya a,c,d,1 a Department of Electronics and Computer Engineering, Ngee Ann Polytechnic, Singapore b National Heart Centre, Singapore c Department of Biomedical Engineering, School of Science and Technology, Singapore University of Social Sciences, Singapore d School of Medicine, Faculty of Health and Medical Sciences, Taylor’s University, 47500 Subang Jaya, Malaysia e Department of Medicine – Cardiology, Columbia University, USA ARTICLEINFO Keywords: Computer-aided detection system Congestive heart failure Deep learning Machine learning Statistical analysis ABSTRACT The heart muscle pumps blood to vital organs, which is indispensable for human life. Congestive heart failure (CHF) is characterized by the inability of the heart to pump blood adequately throughout the body without an increase in intracardiac pressure. The symptoms include lung and peripheral congestion, leading to breathing difculty and swollen limbs, dizziness from reduced delivery of blood to the brain, as well as arrhythmia. Coronary artery disease, myocardial infarction, and medical co-morbidities such as kidney disease, diabetes, and high blood pressure all take a toll on the heart and can impair myocardial function. CHF prevalence is growing worldwide. It aficts millions of people globally, and is a leading cause of death. Hence, proper diagnosis, monitoring and management are imperative. The importance of an objective CHF diagnostic tool cannot be overemphasized. Standard diagnostic tests for CHF include chest X-ray, magnetic resonance imaging (MRI), nuclear imaging, echocardiography, and invasive angiography. However, these methods are costly, time-con- suming, and they can be operator-dependent. Electrocardiography (ECG) is inexpensive and widely accessible, but ECG changes are typically not specifc for CHF diagnosis. A properly designed computer-aided detection (CAD) system for CHF, based on the ECG, would potentially reduce subjectivity and provide quantitative as- sessment for informed decision-making. Herein, we review existing CAD for automatic CHF diagnosis, and highlight the development of an ECG-based CAD diagnostic system that employs deep learning algorithms to automatically detect CHF. 1. Introduction Approximately 26 million adults worldwide sufer congestive heart failure(CHF) [1], which is a burgeoning healthcare problem [2]. Be- sides being a primary cause of death, CHF is also universally becoming a main cause of morbidity [3]. 70% of CHF cases are caused by cardi- ovascular ailments such as coronary artery disease [4]. Other causes of CHF include an elevated hemodynamic load, dysfunction related to ischemia, adverse ventricular remodeling, and genetic mutations [5]. Regardless of the etiology, early detection of CHF to avert further structural or functional impairment to the heart is imperative, and can save lives. CHF is a chronic illness that afects the heart chambers. It occurs when the heart is unable to pump blood adequately throughout the body without an increase in intracardiac pressure. The kidneys respond by retaining body fuid, which results in lung congestion and swelling inthearmsandlegs.CHFiscausedbyfunctionalimpairmentoftheleft ventricle (LV), which is the dominant contractile chamber that pumps blood systemically. The systolic contractile function of the LV is con- ventionally quantitated using the LV ejection fraction (EF), defned as the ratio of LV stroke and end-diastolic volumes, with normal LVEF being 50% or more. CHF can be stratifed into two main types: heart failure with reduced (HFrEF) and preserved EF (HFpEF) ejection frac- tion, characterized by predominance of either inadequate LV systolic contraction(EFlessthan50%typically)orinabilityoftheLVtoexpand or fll efciently during diastole, respectively. While classifcation of https://doi.org/10.1016/j.ejmp.2019.05.004 Received 6 March 2019; Received in revised form 2 May 2019; Accepted 4 May 2019 ⁎ Corresponding author. E-mail addresses: e0145834@u.nus.edu (V. Jahmunah), aru@np.edu.sg (U. R. Acharya). 1 Postal address: Department of Electronics and Computer Engineering, Ngee Ann Polytechnic, Singapore 599489, Singapore. Physica Medica 62 (2019) 95–104 1120-1797/ © 2019 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved. T