Dark-blood Imaging in Coronary CT Angiography using Dual-layer Detector Spectral CT: Effects on Image Quality and Vessel Wall Visibility

INTRODUCTION: To explore the potential of a newly developed subtraction technique in coronary computed tomography angiography (CCTA) for improving image quality and vessel wall visibility without introducing misregistration artifacts. METHODS: Fifty-six patients who underwent CCTA scans using dual-layer detector spectral CT (SDCT) were retrospectively enrolled. Dark-blood images were generated by subtracting virtual non-contrast (VNC) datasets from 70-keV datasets. Qualitative evaluation of dark-blood images included assessments of image quality, inner-wall visualization, and outer-wall visualization. Quantitative parameters were compared between conventional CCTA images and dark-blood images. The quantitative assessment involved evaluating contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR). SNRwall, SNRlumen, SNRperiaortic fat, CNRwall-lumen, and CNRwall-periaortic fat were calculated. Two experienced radiologists independently evaluated the images, and inter-rater variability was assessed. RESULTS: Patients were categorized into three groups based on plaque types: Group A (calcified plaques, n=88), Group B (non-calcified plaques, n=15), and Group C (vessels without plaque, n=56). Dark-blood images of non-calcified plaques and vessels without plaque exhibited higher image quality and inner-wall visualization scores compared to calcified plaques (all p < 0.05). The subjective scores of radiologists showed good consistency (all kappa values > 0.7). Compared to conventional images, dark-blood images demonstrated higher quantitative scores in terms of SNRwall, SNRlumen, SNRperiaortic fat, CNRwall-lumen, and CNRwall-periaortic fat (all p < 0.001). DISCUSSION: The dark-blood technique enabled superior coronary wall assessment without misregistration artifacts, overcoming a key limitation of prior subtraction CCTA. RCA motion artifacts remain a technical challenge that warrants phase-specific protocol optimization in our study. CONCLUSION: Dark-blood images derived from SDCT demonstrated improved image quality of coronary arteries without misregistration artifacts and enhanced visualization of the coronary vessel wall.