Accurate imaging of coronary artery disease (CAD) plays a vital role in guiding clinical decisions, particularly in patients with high coronary calcium scores. While conventional coronary CT angiography (CCTA) has proven reliable for ruling out CAD in low to intermediate-risk patients, it tends to overestimate the severity of stenosis in the presence of calcified plaques. This is largely due to blooming artefacts, which reduce the precision of luminal measurements.
The development of photon-counting detector (PCD) CT introduces the possibility of achieving significantly higher spatial resolution through the use of smaller detector elements and direct photon conversion. Ultra-high-resolution CCTA (UHR-CCTA) leverages this capability to enhance image detail, with a slice thickness of just 0.2 mm. Comparing its performance against the established standard of invasive quantitative coronary angiography (QCA) provides insight into its potential value in clinical settings.
Sharper Resolution Reduces Overestimation
Higher image resolution achieved through photon-counting technology appears to offer a clear advantage in quantifying coronary stenosis. In a cohort of 49 patients, CCTA images reconstructed at three resolutions—standard (0.6 mm), high (0.4 mm) and ultra-high (0.2 mm)—were assessed. Across 103 coronary segments, measured stenosis percentages consistently declined with increasing spatial resolution. Median stenosis dropped from 61.4% using standard resolution to 50.9% with ultra-high resolution. This trend brought CCTA measurements closer to QCA-derived values, which recorded a median stenosis of 46.4%. The overestimation bias fell from 13.2% with standard resolution to 5.2% with UHR-CCTA, and the limits of agreement narrowed accordingly. Additionally, agreement between readers improved with resolution, as reflected in the intraclass correlation coefficient (ICC), which rose from 0.72 to 0.86.
These improvements were particularly evident in segments with calcified plaques, which comprised 70% of stenoses in the analysis. With increasing resolution, measurements of calcified plaques aligned more closely with QCA results, demonstrating a reduction in bias and improved reliability. In contrast, no significant improvements were observed in partially calcified or non-calcified plaques. The visual delineation of the lumen in softer plaque types remains limited, potentially due to the increased image noise associated with higher spatial resolution. Nonetheless, the refined imaging of calcified segments offers clear diagnostic benefits, especially in populations with high Agatston scores.
Performance across Segment Types and Plaque Composition
Resolution improvements enhanced diagnostic performance consistently across different anatomical regions. Whether evaluating proximal or distal coronary segments, UHR-CCTA provided more accurate stenosis estimates. For both segment types, the overestimation bias and spread of values relative to QCA results decreased as resolution increased. The ICC also improved across both categories, indicating stronger agreement with invasive reference values. When focusing on plaque composition, calcified plaques benefited most from higher resolution. Measurements in this group were brought significantly closer to QCA findings, with mean stenosis levels reduced and the range of variation narrowed. In contrast, results from non-calcified and partially calcified plaques were more variable and less responsive to changes in spatial resolution.
Must Read: Enhanced Coronary Plaque Quantification with Ultrahigh-Resolution PCD CT
The impact on diagnostic performance was also demonstrated through improved specificity and positive predictive value (PPV). For detecting stenosis of at least 50%, per-patient specificity rose from 29.6% with standard resolution to 51.6% with ultra-high resolution. At the 70% stenosis threshold, specificity climbed from 61% to 80.5%. Similarly, PPV improved with resolution, reaching 61.8% for 50% stenoses and 50% for 70% stenoses using UHR-CCTA. These improvements suggest a greater capacity to correctly identify clinically significant stenoses and reduce unnecessary invasive procedures. However, the findings also reveal a plateau in diagnostic benefit when resolution surpasses a certain threshold for softer plaque types.
Clinical Relevance and Future Implications
The ability to minimise overestimation bias has clear clinical implications. Reducing the perceived severity of coronary narrowing, particularly in patients with significant calcifications, can prevent unwarranted referrals for invasive procedures. By improving the accuracy of non-invasive imaging, UHR-CCTA may help streamline care pathways and support more informed treatment decisions. The advantage of photon-counting CT lies in its capacity to depict fine anatomical detail, making it particularly useful in challenging cases with dense calcification.
Despite its strengths, several factors temper the generalisability of the findings. The analysis excluded patients with prior stent placement and focused primarily on individuals with high calcium scores, driven by a threshold-based selection process. As a result, plaque types were unevenly represented, and diagnostic improvements could not be verified across all patient groups. Additionally, the higher image noise inherent in UHR modes may offset gains in resolution when evaluating soft tissue structures. Artificial intelligence techniques that reduce image noise without compromising detail could help address this limitation.
The performance of UHR-CCTA compared favourably with previous findings, particularly in high-risk populations undergoing pre-procedural imaging for interventions such as transcatheter aortic valve replacement. Yet variability in PPV remains a challenge, partly due to the elevated prevalence of CAD in the selected study population. A broader application of photon-counting technology across diverse clinical settings and plaque morphologies will be essential to fully establish its diagnostic role. Large-scale prospective studies that include outcome data are needed to determine whether the improved accuracy translates into measurable clinical benefit, such as fewer downstream tests or better patient outcomes.
Ultra-high-resolution coronary CT using photon-counting detectors significantly reduces the overestimation of coronary stenosis in calcified plaques. The improvement in spatial resolution brings non-invasive imaging results closer to those of invasive angiography, enhancing diagnostic confidence and potentially reducing the need for further invasive testing. While benefits are most evident in calcified lesions, diagnostic performance in non-calcified and mixed plaques remains limited. Wider adoption of this technology, supported by future studies with more diverse populations and clinical endpoints, could help refine CAD diagnostics and improve patient management strategies.
Source: European Journal of Radiology
Image Credit: iStock
