Background Recently introduced photon-counting CT may improve noninvasive assessment of patients with high risk for coronary artery disease (CAD). Purpose To determine the diagnostic accuracy of ultrahigh-resolution (UHR) coronary CT angiography (CCTA) in the detection of CAD compared with the reference standard of invasive coronary angiography (ICA). Materials and Methods In this prospective study, participants with severe aortic valve stenosis and clinically indicated CT for transcatheter aortic valve replacement planning were consecutively enrolled from August 2022 to February 2023. All participants were examined with a dual-source photon-counting CT scanner using a retrospective electrocardiography-gated contrast-enhanced UHR scanning protocol (tube voltage, 120 or 140 kV; collimation, 120 × 0.2 mm; 100 mL of iopromid; no spectral information). Subjects underwent ICA as part of their clinical routine. A consensus assessment of image quality (five-point Likert scale: 1 = excellent [absence of artifacts], 5 = nondiagnostic [severe artifacts]) and a blinded independent reading for the presence of CAD (stenosis ≥50%) were performed. UHR CCTA was compared with ICA using area under the receiver operating characteristic curve (AUC). Results Among 68 participants (mean age, 81 years ± 7 [SD]; 32 male, 36 female), the prevalence of CAD and prior stent placement was 35% and 22%, respectively. The overall image quality was excellent (median score, 1.5 [IQR, 1.3-2.0]). The AUC of UHR CCTA in the detection of CAD was 0.93 per participant (95% CI: 0.86, 0.99), 0.94 per vessel (95% CI: 0.91, 0.98), and 0.92 per segment (95% CI: 0.87, 0.97). Sensitivity, specificity, and accuracy, respectively, were 96%, 84%, and 88% per participant (n = 68); 89%, 91%, and 91% per vessel (n = 204); and 77%, 95%, and 95% per segment (n = 965). Conclusion UHR photon-counting CCTA provided high diagnostic accuracy in the detection of CAD in a high-risk population, including subjects with severe coronary calcification or prior stent placement. Published under a CC BY 4.0 license. Supplemental material is available for this article. See also the editorial by Williams and Newby in this issue.

The aim of this paper is to evaluate the diagnostic image quality of spectral dual-source photon-counting detector coronary computed tomography angiography (PCD-CCTA) for coronary artery disease in a multicenter study. The image quality (IQ), assessability, contrast-to-noise ratio (CNR), Agatston score, and radiation exposure were measured. Stenoses were quantified and compared with invasive coronary angiography, if available. A total of 92 subjects (65% male, age 58 ± 14 years) were analyzed. The prevalence of significant coronary artery disease (CAD) (stenosis ≥ 50%) was 17% of all patients, the range of the Agatston score was 0–2965 (interquartile range (IQR) 0–135). The IQ was very good (one, IQR one–two), the CNR was very high (20 ± 10), and 5% of the segments were rated non-diagnostic. The IQ and assessability were higher in proximal coronary segments (p < 0.001). Agatston scores up to 600 did not significantly affect the assessability of the coronary segments (p = 0.3). Heart rate influenced assessability only at a high-pitch mode (p = 0.009). For the invasive coronary angiography (ICA) subgroup (n = nine), the diagnostic performance for CAD per segment was high (sensitivity 92%, specificity 96%), although the limited number of patients who underwent both diagnostic modalities limits the generalization of this finding at this stage. PCD-CCTA provides good image quality for low and moderate levels of coronary calcifications.

There was no difference in diagnostic accuracy for detecting hemodynamically relevant coronary stenosis with CT myocardial perfusion versus CT-derived fractional flow reserve.

The Coronary Artery Disease-Reporting and Data System (CAD-RADS) 2.0 offers standardized guidelines for interpreting coronary artery disease in cardiac CT. Accurate and consistent CAD-RADS 2.0 scoring is crucial for comprehensive disease characterization and clinical decision-making. This study investigates the capability of large language models (LLMs) to autonomously generate CAD-RADS 2.0 scores from cardiac CT reports. A dataset of cardiac CT reports was created to evaluate the performance of several state-of-the-art LLMs in generating CAD-RADS 2.0 scores via in-context learning. The tested models comprised GPT-3.5, GPT-4o, Mistral 7b, Mixtral 8 ​× ​7b, Llama3 8b, Llama3 8b with a 64k context length, and Llama3 70b. The generated scores from each model were compared to the ground truth, which was provided by two board-certified cardiothoracic radiologists in consensus based on the reports. The final set comprised 200 cardiac CT reports. GPT-4o and Llama3 70b achieved the highest accuracy in generating full CAD-RADS 2.0 scores including all modifiers with a performance rate of 93 ​% and 92.5 ​%, respectively, followed by Mixtral 8 ​× ​7b with 78 ​%. In contrast, older LLMs, such as Mistral 7b and GPT-3.5 performed poorly (16 ​%) and Llama3 8b demonstrated intermediate results with an accuracy of 41.5 ​%. LLMs enhanced with in-context learning are capable of autonomously generating CAD-RADS 2.0 scores for cardiac CT reports with excellent accuracy, potentially enhancing both the efficiency and consistency of cardiac CT reporting. Open-source models not only deliver competitive accuracy but also present the benefit of local hosting, mitigating concerns around data security.

Background Coronary CT angiography (cCTA) is a class 1 recommendation in the current guidelines by the European Society of Cardiology (ESC) for excluding significant coronary artery stenosis. To achieve optimal image quality at a low radiation dose, the imaging physician may choose different acquisition modes. Therefore, the consensus guidelines by the Society of Cardiovascular Computed Tomography (SCCT) provide helpful guidance for this procedure. Method The article provides practical recommendations for the application and acquisition of cCTA based on the current literature and our own experience. Results and Conclusion According to current ESC guidelines, cCTA is recommended in symptomatic patients with a low or intermediate clinical likelihood for coronary artery disease. We recommend premedication with beta blockers and nitrates prior to CT acquisition under certain conditions even with the latest CT scanner generations. The most current CT scanners offer three possible scan modes for cCTA acquisition. Heart rate is the main factor for selecting the scan mode. Other factors may be coronary calcifications and body mass index (BMI). Key Points: Citation Format

Background Transcatheter aortic valve implantation (TAVI) has evolved into an alternative procedure to surgical valve replacement for high-risk patients with aortic valve stenosis. Despite technical innovations, there is still a risk of complications during and after the intervention. After a TAVI procedure, ECG-gated multidetector computed tomography (MDCT) plays an important role in the early diagnosis of local complications. In this article, we explain for the first time how the technical acquisition of MDCT in the region of the aortic root is performed as post-interventional control of the TAVI prosthesis. In the second part normal post-interventional findings of different prosthetic valves as well as classic and uncommon complications in the implant area will be illustrated in several case studies. Methods In this review the current literature from PubMed about ECG-gated MDCT after TAVI is summarized and structured. It is supplemented by several case studies from our institution. Results and Conclusion Using retrospectively ECG-gated MDCT, an aortic valve prosthesis after TAVI can be visualized with high spatial resolution in several phases of the cardiac cycle. Images of the implanted aortic valve at all time points of the cardiac cycle enable a functional analysis of prosthetic leaflets similar to echocardiography. MDCT is superior to transthoracic echocardiography with respect to the direct detection of prosthetic leaflet thrombosis. The position of the device in relation to the coronary ostia and correct unfolding of the stent frame need to be evaluated. There are different types of stents carrying the valve leaflets with distinct ideal positions. Any stent should cover the left ventricular outflow tract (LVOT) along its whole circumference. Life-threatening complications in the implant area, such as annulus rupture, can be diagnosed reliably with CT. Key points Citation Format