BACKGROUND: A recent simulation study proposed that stenosis measurements on coronary computed tomography (CT) angiography are influenced by the improved spatial resolution of photon-counting detector (PCD)-CT. The aim of the current study was to evaluate the impact of ultrahigh-spatial-resolution (UHR) on coronary stenosis measurements and Coronary Artery Disease Reporting and Data System (CAD-RADS) reclassification rates in patients undergoing coronary CT angiography on both PCD-CT and energy-integrating detector (EID)-CT and to compare measurements against quantitative coronary angiography. METHODS: Patients with coronary calcification on EID-CT (collimation, 192×0.6 mm) were prospectively enrolled for a research coronary CT angiography with UHR PCD-CT (collimation, 120×0.2 mm) within 30 days (between April 1, 2023 and January 31, 2024). PCD-CT was acquired with the same or lower CT dose index and equivalent contrast media volume as EID-CT. Percentage diameter stenosis (PDS) for calcified, partially calcified, and noncalcified lesions were compared between scanners. Patient-level reclassification rates for CAD-RADS were evaluated. The accuracy of PDS measurements was validated against quantitative coronary angiography in patients who underwent invasive coronary angiography. RESULTS: In total, PDS of 278 plaques were quantified in 49 patients (calcified, 202; partially calcified, 51; noncalcified, 25). PCD-CT-based PDS values were lower than EID-CT measurements for calcified (45.1±20.7 versus 54.6±19.2%; P <0.001) and partially calcified plaques (44.3±19.6 versus 54.9±20.0%; P <0.001), without significant differences for noncalcified lesions (39.1±15.2 versus 39.0±16.0%; P =0.98). The reduction in stenosis degrees led to a 49.0% (24/49) reclassification rate to a lower CAD-RADS with PCD-CT. In a subset of 12 patients with 56 lesions, UHR–based PDS values showed higher agreement with quantitative coronary angiography (mean difference, 7.3%; limits of agreement, −10.7%/25.2%) than EID-CT measurements (mean difference, 17.4%; limits of agreement, −6.9%/41.7%). CONCLUSIONS: Compared with conventional EID-CT, UHR PCD-CT results in lower PDS values and more accurate stenosis measurements in coronary plaques with calcified components and leads to a substantial Coronary Artery Disease Reporting and Data System reclassification rate in 49.0% of patients.

Previously healthy individuals with prolonged cardiorespiratory symptoms after SARS-CoV-2 infection who were not hospitalized at any disease stage had no signs of active cardiac inflammation at car... Background Myocardial injury and inflammation at cardiac MRI in patients with COVID-19 have been described in recent publications. Concurrently, a chronic COVID-19 syndrome (CCS) after SARS-CoV-2 i...

The diagnostic performance of cardiac MRI feature-tracking strain analysis for acute myocarditis varied among myocardial layers, which might be explained by the pathophysiology of myocarditis. Purpose To evaluate the diagnostic performance of layer-specific cardiac MRI feature-tracking (FT) strain analysis in patients with acute myocarditis. Materials and Methods Seventy patients (mean a...

Objectives: To analyse the diagnostic impact of dual energy computed tomography (DECT) in acute gout flares and acute calcium pyrophosphate (CPP) crystal arthritis when compared to the gold standard of arthrocentesis with compensated polarised light microscopy. Microscopy results were also compared to musculoskeletal ultrasound (MUS), conventional radiographs, and the suspected clinical diagnosis (SCD). Methods: Thirty-six patients with a suspected gout flare (n = 24) or acute CPP crystal arthritis (n = 11, n = 1 suffered from neither) who received a DECT and underwent arthrocentesis were included. Two independent readers assessed DECT images for signs of monosodium urate crystals or calcium pyrophosphate deposition. Results: Sensitivity of DECT for gout was 63% (95% CI 0.41–0.81) with a specificity of 92% (0.41–0.81) while sensitivity and specificity for acute CPP arthritis were 55% (0.23–0.83) and 92% (0.74–0.99), respectively. MUS had the highest sensitivity of all imaging modalities with 92% (0.73–0.99) and a specificity of 83% (0.52–0.98) for gout, while sensitivity and specificity for acute CPP crystal arthritis were 91% (0.59–1.00) and 92% (0.74–0.99), respectively. Conclusion: DECT is an adequate non-invasive diagnostic tool for acute gout flares but might have a lower sensitivity than described by previous studies. Both MUS and SCD had higher sensitivities than DECT for acute gout flares and acute CPP crystal arthritis. Key Points• DECT offers a lower sensitivity for acute gout flares than previously described.• DECT sensitivity for acute CPP crystal arthritis is less than the already validated ultrasound.

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.