Detecting Heart Failure from Chest X-Ray Images Using Deep Learning Algorithms

Compared to other diseases, cardiovascular dis- eases are one of the most common causes of death. Actually, new advances in Deep Learning are providing excellent performances in medical image analysis. It has been used in the majority of research to diagnose Lung Nodule disease utilizing Chest Radiograph. In fact, we took advantage of the rare application of Deep Learning to predict heart failure from chest X-ray images, as well as its performance. In this research, we performed several deep Convolutional Neural Network (CNNs) architectures to predict heart failure using ChestX-ray8 database, and then compared their performances. The experiment results showed that the Xception model gives significant results. Therefore, this model is the most suitable for our research topic compared to the others in terms of accuracy.

I. INTRODUCTION

Heart disease can affect the heart’s functionality directly or indirectly, leading to myocardial infarction, heart attack, and a variety of other diseases. Our research focuses on heart failure, which is also a type of heart disease, and is defined as the inefficient pumping of blood, which can lead to heart system disease. Furthermore, because there is no cure for heart failure, a patient must be treated for the rest of his life. As a result, early detection of heart failure is crucial, as it affects approximately 23 million people worldwide, making it a major public health issue [1].