Radiology最新文献

Background Conventional helical CT systems are limited by spatial resolution, which can be insufficient to detect subtle pulmonary abnormalities. A novel prototype static CT system with a dual-ring structure and 24 x-ray sources has been developed to enable ultra-high-spatial-resolution lung imaging. Purpose To assess whether static CT improves the visibility of lung structures compared with conventional helical CT and to determine whether enhanced image quality has a diagnostic impact. Materials and Methods In this prospective study approved by the local ethics committee, 30 patients who had undergone chest CT within the prior 12 months were enrolled between April and December 2024. Each patient underwent dose-matched scanning with the prototype system. Images were reconstructed with a 3072 × 3072 matrix and 0.165-mm section thickness. Three thoracic radiologists independently rated the visibility of 11 lung structures on a five-point Likert scale (from -2 to 2). Objective metrics, including modulation transfer function, signal-to-noise ratio, contrast-to-noise ratio, and the number of visible bronchial generations, were assessed. Interobserver agreement was evaluated with Fleiss κ coefficients, and Wilcoxon signed rank tests were used for comparisons (P < .01 was indicative of a statistically significant difference). Results Among the 30 patients who were included (mean age, 55 years ± 9; 18 men), the static CT system demonstrated improved visibility in eight of 11 pulmonary structures (all P < .01). Objective analysis confirmed a higher spatial resolution for the prototype (10% modulation transfer function: 25.5 vs 18 line pairs per centimeter) and visualization of more distal bronchial generations (median, 9 vs 7), albeit with lower signal-to-noise ratio and contrast-to-noise ratio (both P < .001). Although visibility scores were higher than 0, they remained below the "diagnostic impact" threshold of 2 (all P < .01). Interobserver agreement was excellent (Fleiss κ = 0.828). Conclusion The prototype stationary multisource CT improved the visualization of fine pulmonary structures at similar radiation doses. © RSNA, 2026 Supplemental material is available for this article.

Background Data on the detection of enhancement-related malignant features and applicable scenarios for low-dose ultrahigh-resolution photon-counting CT (PCCT) in lung cancer are lacking. Purpose To compare the benefits of PCCT versus energy-integrating detector (EID) CT at contrast-enhanced chest CT for different populations with lung cancer. Materials and Methods This prospective study enrolled 459 consecutive participants with lesions who underwent either low-dose contrast media PCCT (1 mL/kg at 2 mL/sec) or standard-dose EID CT (1.2 mL/kg at 3 mL/sec) from June to December 2024. The EID CT group was selected using propensity score matching. Images were reconstructed into five section thicknesses (PCCT: 5 mm, 1 mm, 0.4 mm; EID CT: 5 mm, 1 mm). Adverse reactions and contrast-induced acute kidney injury were recorded. Lesion image quality, malignant radiologic features, diagnostic confidence, and subgroup (body mass index [calculated as weight in kilograms divided by height in meters squared], tumor size) were analyzed using a five-point Likert scale. Statistical data were compared using one-way analysis of variance and post hoc comparisons. Results Among 200 participants after propensity score matching (mean age, 61.66 years ± 9.60 [SD]), PCCT reduced radiation and iodine exposure by 66.34% (effective dose, 1.36 mSv vs 4.04 mSv) and 26.57% (iodine load, 20.62 mg vs 28.08 mg; P < .001), respectively, lowering adverse reactions and contrast-induced acute kidney injury (2% [two of 100 participants] vs 9% [nine of 100], P = .03; 1% [one of 100] vs 7% [seven of 100], P = .03). PCCT at 0.4 mm showed higher detection and diagnostic confidence for enhancement-related malignant features (PCCT vs EID CT, 291-340 findings vs 194-255 findings [Likert 5 {IQR, 4-5} vs Likert 3 {IQR, 3-4}]; P < .001) and yielded higher overall image quality (Likert 5 [IQR, 5-5]; P < .05) in normal weight participants and structures within enhanced lesions of less than or equal to 3 cm (Likert 4 [IQR, 3-4] vs Likert 3 [IQR, 3-3]; P < .001). Conclusion PCCT reduced radiation exposure, adverse reactions, and contrast-induced acute kidney injury compared with EID CT. With a 0.4-mm section thickness, PCCT improved overall image quality, detection of enhancement-related malignant features, and diagnostic confidence, making it suitable for various body mass index and small lesions (T1 stage). © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license. Supplemental material is available for this article.

Background Current risk stratification models fail to effectively integrate a broad range of parameters to predict major adverse cardiovascular events (MACE) in patients with ST-segment elevation myocardial infarction (STEMI). Purpose To develop and externally test a machine learning (ML) model integrating comprehensive clinical data and cardiac MRI parameters for long-term MACE prediction in patients with STEMI. Materials and Methods This retrospective study included data from patients with STEMI who underwent clinically indicated cardiac MRI within 7 days after percutaneous coronary intervention, with data from one center composing the training set (September 2015 to September 2023) and data from another center composing the external test set (January 2015 to July 2023). The primary end point was MACE, defined as a composite of cardiovascular death, recurrent myocardial infarction, unplanned coronary revascularization, stroke, and rehospitalization for heart failure or arrhythmia. Sixty-seven variables were initially evaluated to inform the ML model. The final model included established clinical predictors combined with features selected using recursive feature elimination. Model performance was assessed using integrated area under the receiver operating characteristic curve (AUC). Results A total of 1066 patients with STEMI (mean age, 58.15 years ± 11.40 [SD]; 904 male patients) were included: 682 in the training set and 384 in the external test set. During a median follow-up of 40 months (IQR, 22-55 months), 142 patients in the training set and 81 in the external test set experienced MACE. In the external test set, the ML model achieved an integrated AUC for MACE prediction of 0.91, compared with an integrated AUC of 0.86 for a clinical model (P < .001), 0.86 and 0.89 (P = .005) for Cox regression models, 0.66 for Global Registry of Acute Coronary Events score (P < .001), and 0.62 for Thrombolysis in Myocardial Infarction score. The model effectively stratified patients into distinct risk groups (log-rank P < .001). Conclusion An ML model integrating cardiac MRI and clinical data demonstrated excellent long-term prognostic performance compared with traditional models and aided individualized risk stratification in patients with STEMI. © RSNA, 2026 Supplemental material is available for this article. See also the editorial by Garot and Duhamel in this issue.

Background α-Synuclein (α-Syn) aggregation and iron deposition drive synucleinopathies, such as Parkinson disease (PD) and multiple system atrophy (MSA), but noninvasive imaging techniques for α-Syn and differentiating between synucleinopathies remain challenging. Purpose To explore whether MRI-based subvoxel quantitative susceptibility mapping (QSM) can help simultaneously measure α-Syn aggregation and iron deposition in patients with synucleinopathies and differentiate between different synucleinopathies. Materials and Methods Phantom and animal (one C3H mouse) experiments were performed to measure the magnetic susceptibility properties of α-Syn. In the prospective study, subvoxel QSM images were acquired from healthy controls (HCs) and participants with PD and MSA who were recruited from October 2021 to September 2024 to assess α-Syn aggregation and iron deposition via diamagnetic and paramagnetic values, respectively. Correlation analysis was performed between subvoxel QSM and symptom scores. Receiver operating characteristic curve analysis was conducted to evaluate subvoxel QSM performance to differentiate participants with synucleinopathies and HCs. Results The diamagnetic property of α-Syn was confirmed. A total of 273 participants (mean age, 59 years ± 8 [SD]; 148 female participants; 62 with MSA, 107 with PD, and 104 with HC) were evaluated. Substantia nigra pars compacta of participants with synucleinopathies showed α-Syn aggregation (mean absolute diamagnetic values: MSA, 0.0109 ppm ± 0.0004; PD, 0.0100 ppm ± 0.0002; HC, 0.0088 ppm ± 0.0003; P < .001) and iron deposition (mean paramagnetic values: MSA, 0.0992 ppm ± 0.0037; PD, 0.0897 ppm ± 0.0018; HC, 0.0762 ppm ± 0.0018; P < .001). The MSA group showed abnormalities in various brain regions, which correlated with symptoms (r > .26; P < .05). Subvoxel QSM enabled synucleinopathy differentiation (areas under receiver operating characteristic curve: PD vs HC, 0.87 [95% CI: 0.82, 0.92]; MSA vs HC, 0.94 [95% CI: 0.90, 0.98]; and PD vs MSA, 0.96 [95% CI: 0.93, 0.98]). Conclusion Subvoxel QSM enabled simultaneous measurement of α-Syn aggregation and iron deposition in patients with synucleinopathies, effectively distinguishing PD from MSA and both from HCs. © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license. Supplemental material is available for this article.

Large language models (LLMs) have transformative potential in radiology, including textual summaries, diagnostic decision support, proofreading, and image analysis. However, the rapid increase in studies investigating these models, along with the lack of standardized LLM-specific reporting practices, affects reproducibility, reliability, and clinical applicability. To address this, reporting guidelines for LLM studies in radiology were developed using a two-step process. First, a systematic review of LLM studies in radiology was conducted across PubMed, IEEE Xplore, and the ACM Digital Library, covering publications between May 2023 and March 2024. Of 511 screened studies, 57 were included to identify relevant aspects for the guidelines. Then, in a Delphi process, 20 international experts developed the final list of items for inclusion. Items consented as relevant were summarized into a structured checklist containing 32 items across six key categories: general information and data input; prompting and fine-tuning; performance metrics; ethics and data transparency; implementation, risks, and limitations; and further/optional aspects. The final FLAIR (Framework for LLM Assessment in Radiology) checklist aims to standardize reporting of LLM studies in radiology, fostering transparency, reproducibility, comparability, and clinical applicability to enhance clinical translation and patient care. © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license. Supplemental material is available for this article.

Background In pediatric biliary cirrhosis secondary to biliary atresia (BA), hepatic lobules are disrupted to form pseudolobules. How the hepatic circulation is reorganized remains a poorly characterized issue. Purpose To reproduce the three-dimensional (3D) structural and remodeling alterations of BA-induced pseudolobules via phase-contrast CT (PCCT) and reveal their circulatory self-rescue mechanism. Materials and Methods In this retrospective study, residual donor normal liver tissue samples and BA-affected liver samples collected between November 2013 and April 2024 were imaged with PCCT. Combined with 3D visualization technology, the spatial anatomic morphologic characteristics of the veins, arteries, and sinusoidal system within the hepatic lobules were reproduced based on liver histopathologic examination. Key indexes such as the number of inlet and outlet venules, sinusoidal volume fraction, and anisotropy, were determined to characterize the structural remodeling of the circulatory pathway, from portal and arterial inflow through the sinusoidal exchange network to hepatic venous outflow, in patients with BA. All indicators were analyzed with generalized estimating equations, and the multiple test correction was performed by conducting Holm-Bonferroni correction, with P < .05 indicating statistical significance. Results Normal livers from three donors (median age, 36 months; IQR, 12-132 months; two male donors) and 25 patients with BA (median age, 7.00 months; IQR, 5.8-10 months; 17 male patients) were analyzed. The mean number of inlet venules decreased from 253.07 ± 70.99 (SD) in normal lobules to 138.80 ± 28.12 in pseudolobules (P < .001). The mean number of inlet arterioles increased from 3.93 ± 1.39 in normal lobules to 5.65 ± 2.43 in pseudolobules (P < .001). The number of outlet venules in pseudolobules decreased by 35% compared with that in normal lobules (mean, 105.02 ± 42.47 vs 161.96 ± 42.47; P < .001). In the sinusoidal system, the volume fraction in pseudolobules increased by 51% compared with that in normal lobules (mean, 40.42 ± 6.28 vs 26.72 ± 6.32; P < .001), especially in the inlet area, where this value nearly doubled (mean, 44.47 ± 4.73 vs 23.44 ± 5.95; P < .001). Mean anisotropy decreased from 0.53 ± 0.02 in normal lobules to 0.51 ± 0.03 in pseudolobules (P < .001). Conclusion The lobular circulation remodeling mechanism in BA-induced pseudolobules, from the inlet channel to the exchange network and outlet channel, was comprehensively revealed via PCCT. © RSNA, 2026 Supplemental material is available for this article.