Abstract
BACKGROUND: Coronary artery disease (CAD) is a leading cause of death worldwide, and noninvasive diagnostic methods are essential. Although invasive coronary angiography (ICA) is the reference standard, it is invasive and carries procedural risks. Conventional coronary computed tomography angiography (CCTA) is limited by its dependence on electrocardiographic (ECG)-gating, which reduces its feasibility in patients with arrhythmias, high heart rates (HRs), or in emergency settings. Therefore, this study aimed to assess the diagnostic accuracy of a non-ECG-gated CCTA (ECG-less CCTA) protocol for identifying obstructive CAD, using ICA as the reference. METHODS: This retrospective single-center study included 110 patients with suspected CAD undergoing ECG-less CCTA [256-row computed tomography (CT) with simulated ECG signals, automated tube voltage selection (80-120 kV], and tube current modulation [noise index: 20 Hounsfield units (HU)]. Contrast administration (0.6 mL/kg) was optimized via bolus tracking. Images were reconstructed using deep learning (TrueFidelity™) and motion correction (SnapShot Freeze 2). Two blinded radiologists assessed stenosis ≥50% [Society of Cardiovascular Computed Tomography (SCCT) 18-segment model], with non-diagnostic segments classified as positive. Subgroups were stratified by HR [≤75 vs. >75 beats per minute (bpm)] and calcium burden (Agatston ≤400 vs. >400). RESULTS: ECG-less CCTA showed patient-level sensitivity of 92.1% [95% confidence interval (CI): 85.6-96.2%] and specificity of 91.5% (82.3-96.4%). Vessel- and segment-level specificity/negative predictive value (NPV) were 93.6%/95.1% and 96.2%/97.2%, respectively. Non-diagnostic segments (6.4%) were conservatively positive. Radiation dose was 1.4±0.5 mSv. Specificity decreased in Agatston >400 (84.6% vs. 94.1%, P=0.02), whereas HR >75 bpm did not significantly reduce sensitivity (89.7% vs. 94.1%, P=0.12). CONCLUSIONS: ECG-less CCTA achieves high diagnostic concordance with ICA for obstructive CAD, demonstrating excellent specificity/NPV across analysis levels. Its tolerance to variable HRs and streamlined workflow support clinical utility in emergency settings or arrhythmic patients, avoiding ECG dependency and β-blockers.