Radiology最新文献

{"title":"Accuracy of Dual-Energy CT-derived Fat Maps and Bone Marrow Edema Maps in Pedal Osteomyelitis Diagnosis.","authors":"Christoph Stern, Andrea B Rosskopf, Adrian A Marth, Georg C Feuerriegel, Martin C Berli, Benjamin Fritz, Reto Sutter","doi":"10.1148/radiol.232900","DOIUrl":"https://doi.org/10.1148/radiol.232900","url":null,"abstract":"<p><p><i>\"Just Accepted\" papers have undergone full peer review and have been accepted for publication in <i>Radiology</i>. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content.</i> Background In patients who cannot undergo MRI, dual-energy CT (DECT) with bone marrow edema (BME) maps are used as an approach for diagnosing pedal osteomyelitis, but with lower accuracy. Purpose To compare the diagnostic accuracy of additional bone marrow fat maps with that of DECT with BME maps and MRI for pedal osteomyelitis detection. Materials and Methods In this prospective study, thirty-one participants with clinically suspected osteomyelitis of the mid- and forefoot underwent noncontrast DECT (80 kV/140 kV) and MRI between October 2020 and February 2022. With image postprocessing, DECT-derived BME and fat maps were generated. Four independent readers evaluated 3 different image sets for osteomyelitis: DECT and BME maps (set 1); DECT, BME maps and fat maps (set 2); and MRI (set 3). Sensitivity, specificity and accuracy were calculated for each image set, with clinical and microbiological data as the reference standards. In a subanalysis, the DECT BME map, DECT fat map and DECT erosion map were analyzed for their accuracy in predicting bone marrow fat loss at T1-weighted MRI. Results Of the 31 participants included in the study (mean age, 61.7 years ±14.6 [SD]; 21 males) 17 (55%) had osteomyelitis. Sensitivity, specificity and accuracy for detecting osteomyelitis were 47% (8/17), 79% (11/14), and 61% (19/31) (set 1); 77% (13/17), 86% (12/14) and 81% (25/31) (set 2); and 82% (14/17), 93% (13/14) and 87% (27/31) (set 3), respectively. Thirty-one of 661 individual bones (0.5%) showed bone marrow fat loss on T1-weighted MRI; in the subanalysis, DECT fat map specificity was higher than that of the DECT BME map for predicting bone marrow fat loss in individual bones (97% (612/630) vs. 89% (560/630)) (P<.001). Conclusion Pedal osteomyelitis detection with novel DECT-derived fat map imaging in addition to DECT and BME maps was accurate. See also the editorial by Khurana in this issue.</p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e232900"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep Learning Applications in Imaging of Acute Ischemic Stroke: A Systematic Review and Narrative Summary.","authors":"Bin Jiang, Nancy Pham, Eric K van Staalduinen, Yongkai Liu, Sanaz Nazari-Farsani, Amirhossein Sanaat, Henk van Voorst, Ates Fettahoglu, Donghoon Kim, Jiahong Ouyang, Ashwin Kumar, Aditya Srivatsan, Ramy Hussein, Maarten G Lansberg, Fernando Boada, Greg Zaharchuk","doi":"10.1148/radiol.240775","DOIUrl":"10.1148/radiol.240775","url":null,"abstract":"<p><p>Background Acute ischemic stroke (AIS) is a major cause of morbidity and mortality, requiring swift and precise clinical decisions based on neuroimaging. Recent advances in deep learning-based computer vision and language artificial intelligence (AI) models have demonstrated transformative performance for several stroke-related applications. Purpose To evaluate deep learning applications for imaging in AIS in adult patients, providing a comprehensive overview of the current state of the technology and identifying opportunities for advancement. Materials and Methods A systematic literature review was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A comprehensive search of four databases from January 2016 to January 2024 was performed, targeting deep learning applications for imaging of AIS, including automated detection of large vessel occlusion and measurement of Alberta Stroke Program Early CT Score. Articles were selected based on predefined inclusion and exclusion criteria, focusing on convolutional neural networks and transformers. The top-represented areas were addressed, and the relevant information was extracted and summarized. Results Of 380 studies included, 171 (45.0%) focused on stroke lesion segmentation, 129 (33.9%) on classification and triage, 31 (8.2%) on outcome prediction, 15 (3.9%) on generative AI and large language models, and 11 (2.9%) on rapid or low-dose imaging specific to stroke applications. Detailed data extraction was performed for 68 studies. Public AIS datasets are also highlighted, for researchers developing AI models for stroke imaging. Conclusion Deep learning applications have permeated AIS imaging, particularly for stroke lesion segmentation. However, challenges remain, including the need for standardized protocols and test sets, larger public datasets, and performance validation in real-world settings. © RSNA, 2025 <i>Supplemental material is available for this article.</i></p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e240775"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143804172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison of Deuterium Metabolic Imaging with FDG PET in Alzheimer Disease.","authors":"Nikolaj Bøgh, Malene Aastrup, Janne K Mortensen, Hanne Gottrup, Jakob U Blicher, Per Borghammer, Mattias H Kristensen, Esben S S Hansen, Michael Vaeggemose, Christoffer Laustsen","doi":"10.1148/radiol.241808","DOIUrl":"10.1148/radiol.241808","url":null,"abstract":"<p><p>Background The approval of amyloid-targeting therapies has made it increasingly important to differentiate Alzheimer disease (AD) from other causes of dementia. Dysfunctional glucose metabolism is a recognized pathophysiological element in AD that may be visualized with spectroscopic MRI of deuterated glucose and its metabolites, also known as deuterium metabolic imaging (DMI). Purpose To explore the potential of DMI as a diagnostic tool for AD. Materials and Methods In this prospective cross-sectional study, participants with newly diagnosed AD and age-matched controls were recruited from April to October 2023. DMI was performed with a 3-T system equipped with a proton/deuterium head coil following oral consumption of 75 g of deuterated glucose. Clinical fluorodeoxyglucose (FDG) PET data were acquired from patient records for comparison. The predefined primary outcome, the ratio between lactate and glutamine plus glutamate (Glx) at DMI, was analyzed using age-corrected linear mixed-effect models. Results Ten participants with AD (mean age, 72 years ± 6 [SD]; six women) and five age-matched healthy controls (mean age, 68 years ± 7; four men) were included. The primary analysis revealed no evidence of a difference in the ratio of lactate to Glx between participants with AD and controls (<i>P</i> = .24 across all regions of interest). Exploratory analyses revealed that participants with AD had reduced signals for medial temporal lactate (0.7 ± 0.2 vs 0.5 ± 0.1, <i>P</i> = .04) and Glx (0.5 ± 0.03 vs 0.48 ± 0.05, <i>P</i> = .03) compared with controls. Finally, a strong correlation (<i>r</i> = 0.73) was observed between DMI and FDG PET. Conclusion This study did not find evidence to support a shift from oxidative to anaerobic metabolism in AD. Exploratory analyses revealed a decrease in glucose metabolism in the medial temporal lobe. In extension hereof, a similar distribution of low DMI metabolism and decreased FDG PET glucose uptake was observed. © RSNA, 2025 <i>Supplemental material is available for this article.</i> See also the article by Liu et al in this issue. See also the editorial by Port in this issue.</p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e241808"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143804170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MRI-based Radiomic Features for Risk Stratification of Ductal Carcinoma in Situ in a Multicenter Setting (ECOG-ACRIN E4112 Trial).","authors":"Kalina P Slavkova, Ruya Kang, Anum S Kazerouni, Debosmita Biswas, Vivian Belenky, Rhea Chitalia, Hannah Horng, Michael Hirano, Jennifer Xiao, Ralph L Corsetti, Sara H Javid, Derrick W Spell, Antonio C Wolff, Joseph A Sparano, Seema A Khan, Christopher E Comstock, Justin Romanoff, Constantine Gatsonis, Constance D Lehman, Savannah C Partridge, Jon Steingrimsson, Despina Kontos, Habib Rahbar","doi":"10.1148/radiol.241628","DOIUrl":"https://doi.org/10.1148/radiol.241628","url":null,"abstract":"<p><p>Background Ductal carcinoma in situ (DCIS) is a nonlethal, preinvasive breast cancer for which breast MRI is best suited for accurate disease extent characterization. DCIS is often overtreated, necessitating robust models for improved risk stratification. Purpose To develop logistic regression models using clinical and MRI-based radiomic features of DCIS and to evaluate the performance of such models in predicting disease upstaging at surgery and DCIS score. Materials and Methods This study is a secondary analysis of dynamic contrast-enhanced (DCE) MRI data from the Eastern Cooperative Oncology Group-American College of Radiology Imaging Network, or ECOG-ACRIN, E4112 trial. Primary analysis focused on predicting disease upstaging (<i>n</i> = 295), and secondary analysis focused on predicting DCIS score (<i>n</i> = 174). Radiologist-drawn lesion segmentations and publicly available Cancer Phenomics Toolkit, or CaPTk, software was used to compute 65 radiomic features. Participants were clustered into groups based on their radiomic features (ie, radiomic phenotypes), and principal component analysis was used to summarize the feature space. Clinical information and qualitative MRI features were available. Associations were tested using χ² and likelihood ratio tests. Data were split into training and test sets with a 3:2 ratio, and model performance was assessed on the test set using the area under the receiver operating characteristic curve (AUC). Results Data from 297 female participants with median age of 60 years (IQR, 51-67 years) were analyzed. Two radiomic phenotypes were identified that were associated with disease upstaging (<i>P</i> = .007). For predicting disease upstaging, the top three radiomic principal components combined with clinical and qualitative MRI predictors yielded the highest AUC of 0.77 (95% CI: 0.65, 0.88) among all tested models (<i>P</i> = .007), identifying 25% more true-negative (49 of 93 true-negative findings, 53% specificity) findings, compared with using clinical information alone (23 of 93 true-negative findings, 28% specificity). Radiomic models were not predictive of the DCIS score (<i>P</i> > .05). Conclusion In patients with DCIS, combining radiomic metrics with clinical information improved prediction of disease upstaging but not DCIS score. ClinicalTrials.gov Identifier: NCT02352883 <i>Supplemental material is available for this article.</i> ©RSNA, 2025 See also the editorial by Kim and Woo in this issue.</p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e241628"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Parallel Detection of Multicontrast MRI and Deuterium Metabolic Imaging for Time-efficient Characterization of Neurologic Diseases.","authors":"Yanning Liu, Henk M De Feyter, Zachary A Corbin, Robert K Fulbright, Scott McIntyre, Terence W Nixon, Robin A de Graaf","doi":"10.1148/radiol.241597","DOIUrl":"10.1148/radiol.241597","url":null,"abstract":"<p><p>Background Deuterium metabolic imaging (DMI) is a novel, MRI-based method to map metabolism noninvasively in vivo and has potential to augment existing clinical MRI with unique metabolic information. However, adding DMI scans to a standard, clinical MRI protocol is challenging due to the relatively long scan time of DMI that can result in decreased patient compliance and increased scanning costs. Purpose To design and evaluate a parallel acquisition strategy, based on the large frequency difference between hydrogen proton (¹H) and deuterium (²H) MRI signals, to obtain metabolic DMI data during a comprehensive, multicontrast (fluid-attenuated inversion recovery [FLAIR] and T1-, T2-, and susceptibility-weighted) MRI protocol without adding scanning time. Materials and Methods A parallel MRI DMI protocol based on four essential MRI types-FLAIR, and T1-, T2-, and susceptibility-weighted MRI-was interwoven with three-dimensional DMI by executing ²H acquisition blocks during the contrast-generating delays intrinsic to MRI. Results Phantom and in vivo human brain data show that MRI scan quality, DMI sensitivity, and information content are preserved in the parallel MRI DMI acquisition method. DMI data acquired in parallel with MRI in a patient with an astrocytoma show unique metabolic image contrast that complements the multicontrast MRI examinations. Conclusion Parallel MRI scan acquisition technology was a practical solution to obtain both high-quality anatomic and metabolic scans without prolonging the scanning duration compared with an MRI-only protocol; the method had high flexibility to upgrade traditional MRI protocols with DMI and will be key for many clinical sites to gain access to DMI and drive its further development and validation by use in larger, diverse patient populations. © RSNA, 2025 <i>Supplemental material is available for this article</i>. See also the article by Bøgh et al in this issue. See also the editorial by Port in this issue.</p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e241597"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143804175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Agenesis of the Intrahepatic Inferior Vena Cava with Pulmonary Venous Fistula.","authors":"Xiaoxu Guo, Yuhan Zhou","doi":"10.1148/radiol.242069","DOIUrl":"https://doi.org/10.1148/radiol.242069","url":null,"abstract":"","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e242069"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754249","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MRI-based Radiomics for Ductal Carcinoma in Situ: Enhancing Its Role in Risk Stratification.","authors":"Soo-Yeon Kim, Ok Hee Woo","doi":"10.1148/radiol.250085","DOIUrl":"https://doi.org/10.1148/radiol.250085","url":null,"abstract":"","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e250085"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754275","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Beyond Proprietary Models: The Potential of Open-Source Large Language Models in Radiology.","authors":"Satvik Tripathi, Ali S Tejani, Tessa S Cook","doi":"10.1148/radiol.242454","DOIUrl":"https://doi.org/10.1148/radiol.242454","url":null,"abstract":"","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e242454"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual-Energy CT and Pedal Osteomyelitis.","authors":"Bharti Khurana","doi":"10.1148/radiol.250212","DOIUrl":"https://doi.org/10.1148/radiol.250212","url":null,"abstract":"<p><p>\u0000 <i>\"Just Accepted\" papers have undergone full peer review and have been accepted for publication in <i>Radiology</i>. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content.</i>\u0000 </p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e250212"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accuracy of Large Language Model-based Automatic Calculation of Ovarian-Adnexal Reporting and Data System MRI Scores from Pelvic MRI Reports.","authors":"Rajesh Bhayana, Ankush Jajodia, Tanya Chawla, Yangqing Deng, Genevieve Bouchard-Fortier, Masoom Haider, Satheesh Krishna","doi":"10.1148/radiol.241554","DOIUrl":"https://doi.org/10.1148/radiol.241554","url":null,"abstract":"<p><p>Background Ovarian-Adnexal Reporting and Data System (O-RADS) for MRI helps assign malignancy risk, but radiologist adoption is inconsistent. Automatic assignment of O-RADS scores from reports could increase adoption and accuracy. Purpose To evaluate the accuracy of large language models (LLMs), after strategic optimization, for automatically calculating O-RADS scores from reports. Materials and Methods This retrospective single-center study from a large quaternary care cancer center included consecutive gadolinium chelate-enhanced pelvic MRI reports with at least one assigned O-RADS score from July 2021 to October 2023. Reports from January 2018 to October 2019 (before O-RADS MRI implementation) were randomly selected for additional testing. Reference standard O-RADS scores were determined by radiologists interpreting reports. After prompt optimization using a subset of reports, two LLM-based strategies were evaluated: few-shot learning with GPT-4 (version 0613; OpenAI) prompted with O-RADS rules (\"LLM only\") and a hybrid strategy leveraging GPT-4 to classify features fed into a deterministic formula (\"hybrid\"). Accuracy of each model and originally reported scores were calculated and compared using the McNemar test. Results A total of 284 reports from 284 female patients (mean age, 53.2 years ± 16.3 [SD]) with 372 adnexal lesions were included: 10 reports in the training set (16 lesions), 134 reports in the internal test set 1 (173 lesions; 158 O-RADS assigned), and 140 reports in internal test set 2 (183 lesions). For assigning O-RADS MRI scores, the hybrid model accuracy (97%; 168 of 173) outperformed LLM-only model (90%; 155 of 173; <i>P</i> = .006). For lesions with an originally reported O-RADS score, hybrid model accuracy exceeded that of reporting radiologists (97% [153 of 158] vs 88% [139 of 158]; <i>P</i> = .004). Hybrid model also outperformed LLM-only model for 183 lesions from before O-RADS implementation (95% [173 of 183] vs 87% [159 of 183], respectively; <i>P</i> = .01). Conclusion A hybrid LLM-based application, combining LLM feature classification with deterministic elements, accurately assigned O-RADS MRI scores from report descriptions, exceeding both an LLM-only strategy and the original reporting radiologist. © RSNA, 2025 <i>Supplemental material is available for this article.</i></p>","PeriodicalId":20896,"journal":{"name":"Radiology","volume":"315 1","pages":"e241554"},"PeriodicalIF":12.1,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143754335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}