Applications of Artificial Intelligence in Coronary Computed Tomography Angiography: Progress and Challenges

CCTA (coronary computed tomography angiography) is an important tool for evaluating patients with suspected stable coronary artery disease. Recently, the development of artificial intelligence (AI), including machine learning in data analytics and deep learning in image processing, is reshaping the landscape of CCTA in clinical practice. The noise, radiation dose, and motion artifacts have been largely reduced. More advanced algorithms have been proposed for geometric construction, including image segmentation and centerline extraction. Based on the improved image quality, the assessment of different components (calcification, plaque, stenosis, myocardium, and pericardial fat) has achieved higher accuracy. Computational simulation can estimate hemodynamic parameters like fractional flow reserve. These new applications enable clinicians to improve the accuracy of diagnosis and treatment of coronary artery disease. This chapter summarizes the state-of-the-art methods of AI in CCTA, providing an updated reference for biomedical engineers, health professionals, and policymakers.