Artificial Intelligence in Cardiac CT: Automated Calcium Scoring and Plaque Analysis
- PDF / 2,151,208 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 12 Downloads / 190 Views
CARDIAC COMPUTED TOMOGRAPHY (A. CREAN AND G. SMALL, SECTION EDITORS)
Artificial Intelligence in Cardiac CT: Automated Calcium Scoring and Plaque Analysis Heon Lee 1,2 & Tilman Emrich 1,3,4 & U. Joseph Schoepf 1,5 & Verena Brandt 1,6 & Tyler J. Leonard 1 & Hunter N. Gray 1 & Vincent M. Giovagnoli 1 & Danielle M. Dargis 1 & Jeremy R. Burt 1 & Christian Tesche 1,7,8 Accepted: 25 September 2020 / Published online: 2 October 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Purpose of Review To review the current applications of artificial intelligence (AI) for coronary artery calcium scoring (CACS) and plaque analysis with their achievements and potential clinical impacts. Recent Findings Recent advances of AI-based technologies especially deep learning (DL) approaches in medical imaging have achieved substantial progress in automated detection and characterization of coronary atherosclerotic plaques, providing promising results for AI application in diagnosis and management of coronary artery disease. Summary To date, many studies investigated the potential role of DL in CACS and showed promising results for clinical application in a variety of CT examinations, demonstrating excellent agreement compared with manual scoring. DL-based approaches have also provided considerable progress in automated analysis of non-calcified plaque. Many investigations have shown that automated localization and classification of non-calcified plaque and luminal stenosis is feasible in cardiac CT angiography, although still challenging. Keywords Coronary artery disease . Coronary calcium scoring . Atherosclerotic plaques . Cardiac CT . Machine learning . Deep learning
Introduction Coronary artery disease (CAD) is the leading cause of mortality and morbidity worldwide [1, 2]. The complications of the disease that decrease life expectancy include ischemia, myocardial infarction, heart failure, and sudden cardiac death. Many efforts have been made to prevent or delay these critical events over the past decades. Recent clinical studies have
proved that early identification of the disease is crucial for preventing deaths and may have a substantial impact on patient care by establishing risk factors and assigning risk levels to individuals [3, 4]. Such risk stratification can play an important role in patient management as they often raise the need for initiating preventive therapies in asymptomatic patients and changing of treatment plans in a timely manner, based on objective assessment [2, 5–7]. The use of imaging for
This article is part of the Topical Collection on Cardiac Computed Tomography * U. Joseph Schoepf [email protected] 1
2
3
Division of Cardiovascular Imaging, Department of Radiology and Radiological Science, Medical University of South Carolina, Charleston, SC, USA Division of Cardiovascular Imaging, Department of Radiology, Soonchunhyang University Hospital Bucheon, Bucheon, Republic of Korea Department of Diagnostic and Interventional Radiology, University Medical Center Mainz, Mainz, G
Data Loading...
