A Review on Earlier Diagnosis of CVD Using Chest CT Image Segmentation and Classification

Abstract

In recent years, cardiovascular diseases (CVDs) have become a major global cause of death. These illnesses start with mild symptoms that worsen over time. During the early stages of CVD, fatigue, dyspnea, oedema, fluid retention, and other indications are frequently seen. Among the most common CVDs are angina, mitral regurgitation, congenital heart defect (CHD), arrhythmia, cardiomyopathy, and coronary artery disease (CAD). The best tools for identifying CVDs include clinical techniques such as blood tests, electrocardiography (ECG), and medical imaging. Among them, cardiac computed tomography (CT) is increasingly used for CVD diagnosis, tracking, planning, and prediction. Healthcare practitioners find it difficult to diagnose CVD despite the benefits of CT imaging data because of a number of factors, including inadequate contrast and a large amount of data. The diagnosis of CVDs using CT imaging has incorporated machine learning (ML) and deep learning (DL) approaches, and research in this area is now underway. To better understand the potential applications of machine learning (ML) and deep learning (DL) techniques for chest CT image segmentation and classification—with an emphasis on CVD detection—this review presents a thorough overview of studies that have explored these areas. To that end, we have carefully reviewed 77 papers gathered from various databases, such as Google Scholar, PubMed, and Springer, among others. The research under consideration has undergone rigorous evaluation to highlight the key characteristics, advantages, and disadvantages of the numerous methods established to date for using CT scans to identify CVD.