Coronary artery disease (CAD) remains the leading cause of mortality and morbidity worldwide [1]. In the current routine clinical practice, X-ray coronary angiography (CAG) remains the gold standard for the assessment of CAD, especially in the examination of patients with severe coronary stenosis. However, CAG as an invasive procedure is costly and may cause a variety of complications. In addition, coronary computed tomography angiography (CCTA) is typically used as a non-invasive imaging modality in patients with low-to-intermediate probability for CAD [2,3]. But the scanning is still a challenge for the people with hypersensitivity to iodinated contrast agents and other contraindications [4]. Therefore, a reasonable alternative modality for early identification of CAD to minimize and avoid unwarranted risks is clinically needed, which is also useful for the long-term follow-up.

Coronary magnetic resonance angiography (CMRA) can be a potential alternative method for detecting coronary artery stenosis [[5], [6], [7]], with advantages including versatility, excellent display of soft tissue features, and suitability for repeat imaging. As most coronary arteries and their branches are embedded in epicardial fat, effective fat suppression techniques are required for high-quality visualization of the coronary arteries. Dixon water-fat separation utilizes the difference in resonance frequencies of water and fat protons to separate water from fat and can be post-treated with a range of measures to minimize interference from magnetic field inhomogeneities [8]. A previous study proved that the Dixon water-fat separation technique can significantly improve coronary artery image quality without the use of contrast agent at 3.0 T [9]. Meanwhile, the diagnostic performance of 3.0 T Dixon water-fat separation whole heart CMRA in patients with moderate-to-high risk for CAD had been explored already [10,11]. While there lacks reports on diagnostic value of this method in patients with low to intermediate probability of CAD. In fact, non-invasive screenings are more preferred clinically in this group of patients for long-term follow-up. Meanwhile, previous studies just considered the diagnostic performance of CMRA for the presence of at least 50% stenosis, however, the ability to confidently evaluate specific degree of stenosis in this method requires more investigation.

Thus, the current study was aimed to investigate the diagnostic performance of 3.0 T Dixon water-fat separation whole heart CMRA in patients with low to intermediate probability of CAD and its ability to grade the severity of coronary heart disease based on Coronary Artery Disease Reporting and Data System (CAD-RADS) [12].