Radiology-Artificial Intelligence最新文献

{"title":"A Taxonomy of Machine Hallucination in Radiology.","authors":"Frank J Brooks, Mark A Anastasio","doi":"10.1148/ryai.250203","DOIUrl":"10.1148/ryai.250203","url":null,"abstract":"<p><p>Measuring the rate of machine hallucination is critical to assessing the utility and trustworthiness of generative artificial intelligence (AI) deployed in radiology; however, there are multiple widely used notions of what constitutes hallucination. This ambiguity pervades industry, academia, and regulatory discourse, such that even an experienced radiologist cannot be certain that a generative model has been evaluated using a definition of hallucination aligned with clinical use. As a result, the same generative AI system may be characterized as either never hallucinating or always hallucinating, depending on perspective. This article provides a brief, nontechnical explanation of the fundamental disparity between two seemingly incongruous notions of machine hallucination. Using terms familiar to radiologists, a taxonomy of machine hallucination is proposed that explicitly delineates output contingencies for deployed AI systems. By clarifying what constitutes hallucination under different interpretations, this framework aims to reduce ambiguity and facilitate clearer communication among users, developers, vendors, and administrators regarding the performance of generative AI in radiology. <b>Keywords:</b> Case-based Reasoning, Convolutional Neural Network © RSNA, 2026.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e250203"},"PeriodicalIF":13.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13019329/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147436170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Body Charts from CT Segmentations across the Adult Lifespan: Large-scale Cross-sectional and Longitudinal Analyses.","authors":"Christian Wachinger, Bernhard Renger, Christopher Späth, Marcus R Makowski","doi":"10.1148/ryai.250506","DOIUrl":"10.1148/ryai.250506","url":null,"abstract":"<p><p>Purpose To model the distribution of CT-derived whole-body anatomic volumes across adulthood and establish comprehensive cross-sectional and longitudinal reference charts, addressing the current lack of nonbrain CT-based whole-body standards. Materials and Methods Retrospective CT scans acquired from March 2017 to April 2025 (189 710 scans, 106 563 patients) from the institutional picture archiving and communication system and two external datasets (19 393 and 1158 patients, respectively) were automatically segmented into 104 structures (totaling 7.8 million volumes). An automated quality control pipeline, incorporating a novel outlier removal strategy based on strong correlation between organ sizes, ensured data reliability. Cross-sectional normative models were constructed using generalized additive models for location, scale, and shape to capture nonlinear age effects through fractional polynomial functions. A generalized additive mixed model was used for longitudinal analyses to assess within-patient changes over follow-up visits. Results All anatomic structures followed complex, nonlinear age trajectories, with marked sex differences and distinct CT contrast material effects on vascular structures. Bootstrap resampling confirmed the stability and precision of these volume trajectories in both central tendency and variability. An exemplary cardiomegaly case-control analysis showed significantly increased centile scores (<i>P</i> < .001) for heart volume. The longitudinal analysis further revealed significant age-sex interactions influencing within-patient trajectories. Conclusion Cross-sectional and longitudinal reference models were developed from CT-derived anatomic volumes that map the trajectories of structural body changes across adulthood. These body charts facilitate robust quantification of individual deviations via centile scores. <b>Keywords:</b> Anatomy, Statistics, Multivariate Adaptive Regression Splines (MARS), Regression Algorithms, Machine Learning Algorithms, Segmentation <i>Supplemental material is available for this article.</i> © RSNA, 2026. See also commentary by Chai and Shi in this issue.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e250506"},"PeriodicalIF":13.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145821197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-Supervised Text-Vision Alignment for Automated Brain MRI Abnormality Detection: A Multicenter Study (ALIGN Study).","authors":"David A Wood, Emily Guilhem, Sina Kafiabadi, Ayisha Al Busaidi, Kishan Dissanayake, Ahmed Hammam, Nina Mansoor, Matthew Townend, Siddharth Agarwal, Yiran Wei, Asif Mazumder, Gareth J Barker, Peter Sasieni, Sébastien Ourselin, James H Cole, Nikhil Nair, Anil Geetha, Chike Onyekwuluje, Rob Dineen, Permesh Dhillon, Carolyn Costigan, Kavi Fatania, Mark Igra, Rebecca Nichols, Janak Saada, Arne Juette, Ramona-Rita Barbara, Hilmar Spohr, Thomas C Booth","doi":"10.1148/ryai.240619","DOIUrl":"10.1148/ryai.240619","url":null,"abstract":"<p><p>Purpose To develop a self-supervised text-vision framework to detect abnormalities on brain MRI scans by leveraging free-text neuroradiology reports, eliminating the need for expert-labeled training datasets. Materials and Methods This retrospective and prospective multicenter study included 81 936 brain MRI examinations and corresponding radiology reports for adult patients at two UK National Health Service hospitals from January 2008 to December 2019 for training and internal testing and 1369 prospectively collected examinations between March 2022 and March 2024 from four separate National Health Service hospitals for external testing (ClinicalTrials.gov no. NCT04368481). A neuroradiology language model (NeuroBERT) was trained using self-supervised tasks to generate report embeddings. Convolutional neural networks (one per MRI sequence) were trained to map scans to embeddings by minimizing mean squared error loss. The framework then detected abnormalities in new examinations by scoring scans against query sentences using text-image similarity. Model diagnostic performance was assessed using the area under the receiver operating characteristic curve (AUC). Results The framework achieved an AUC of 0.95 (95% CI: 0.94, 0.97) for normal versus abnormal classification and generalized to external sites with examination-level AUCs of 0.90 (95% CI: 0.86, 0.93) in Bedford, 0.87 (95% CI: 0.83, 0.90) in Nottingham, 0.86 (95% CI: 0.83, 0.90) in Norwich, and 0.85 (95% CI: 0.81, 0.89) in Yeovil. In five zero-shot classification tasks-acute stroke, multiple sclerosis, intracranial hemorrhage, meningioma, and hydrocephalus-the framework achieved a mean AUC of 0.89 (range, 0.77-0.93). For visual-semantic image retrieval, mean precision was 0.84 among the top 15 images across seven pathologies. Conclusion The self-supervised text-vision framework accurately detected brain MRI abnormalities without expert-labeled datasets. Clinical trial registration no. NCT04368481 <b>Keywords:</b> Head and Neck, Unsupervised Learning, Convolutional Neural Network (CNN), Neuroradiology © The Author(s) 2025. Published by the Radiological Society of North America under a CC BY 4.0 license. <i>Supplemental material is available for this article.</i> See also commentary by Ghodasara in this issue.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e240619"},"PeriodicalIF":13.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13019336/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145606735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"2025 Manuscript Reviewers: A Note of Thanks.","authors":"Jeffrey S Klein, Charles E Kahn","doi":"10.1148/ryai.260176","DOIUrl":"https://doi.org/10.1148/ryai.260176","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"8 2","pages":"e260176"},"PeriodicalIF":13.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147436177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The RSNA Lumbar Degenerative Imaging Spine Classification (LumbarDISC) Dataset.","authors":"Tyler J Richards, Adam E Flanders, Errol Colak, Luciano M Prevedello, Robyn L Ball, Felipe Kitamura, John Mongan, Maryam Vazirabad, Hui-Ming Lin, Anne Kendell, Thanat Kanthawang, Salita Angkurawaranon, Emre Altinmakas, Hakan Dogan, Paulo Eduardo de Aguiar Kuriki, Arjuna Somasundaram, Christopher Rushton, Deniz Bulja, Naida Spahović, Jennifer Sommer, Sirui Jiang, Eduardo Moreno Júdice de Mattos Farina, Eduardo Caminha Nunes, Michael Brassil, Megan McNamara, Johanna Ortiz, Jacob Peoples, Vinson L Uytana, Anthony Kam, Venkata N S Dola, Daniel Murphy, David Vu, Arsany Hakim, Jason F Talbott","doi":"10.1148/ryai.250480","DOIUrl":"10.1148/ryai.250480","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e250480"},"PeriodicalIF":13.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145967166","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep Learning for Coronary Stenosis Detection in Heavily Calcified Plaques at Coronary CT Angiography: A Stepwise, Multicenter Study.","authors":"Rui Wang, Siwen Wang, LiBo Zhang, U Joseph Schoepf, Fandong Zhang, Wei Chen, Zhen Zhou, Zhe Fang, Bin Hu, Yizhou Yu, Jiayin Zhang, Ximing Wang, Longjiang Zhang, Lei Xu","doi":"10.1148/ryai.250109","DOIUrl":"10.1148/ryai.250109","url":null,"abstract":"<p><p>Purpose To develop and validate a deep learning (DL) model for automated assessment of coronary stenosis in vessels with heavily calcified plaques at coronary CT angiography (CCTA), using quantitative coronary angiography as the reference standard. Materials and Methods A total of 10 101 CCTA examinations (June 2017-December 2020) from three tertiary hospitals in China were retrospectively collected for DL model development. External testing dataset 1 included 442 CCTA examinations (Agatston score > 300) from two independent hospitals (January 2021-May 2022) for performance evaluation. The separate external testing dataset 2 of 120 CCTA examinations was used for a reader study assessing whether DL assistance improved diagnostic accuracy among junior, attending, and senior radiologists. External testing dataset 3 included 150 prospectively collected CCTA examinations (June-July 2023) that were analyzed to compare model performance against clinical reports, simulating real-world deployment. Model diagnostic performance was assessed using receiver operating characteristic analysis, with quantitative coronary angiography as the reference. Results In external testing dataset 1, specificities for detecting 50% or more stenosis were 78%, 72%, and 48% and the areas under the receiver operating characteristic curve (AUC) were 0.89, 0.90, and 0.87 at the segment, vessel, and patient levels, respectively. In external testing dataset 2, DL assistance improved radiologist specificity by 7%-11% (<i>P</i> < .001) with improving AUC and increased interreader agreement (Δκ = 0.155-0.228; <i>P</i> < .05). In external testing dataset 3, the model demonstrated 53% specificity and a higher AUC versus clinical reports (0.91 vs 0.76; <i>P</i> < .001). Conclusion The proposed DL model accurately detected coronary stenosis of heavily calcified plaques at CCTA and improved diagnostic performance of radiologists. <b>Keywords:</b> CT Angiography, Cardiac, Heart, Arteriosclerosis, Calcifications, Calculi, Quantification, Diagnosis <i>Supplemental material is available for this article.</i> © The Author(s) 2025. Published by the Radiological Society of North America under a CC BY 4.0 license. See also commentary by Maiter and Alabed in this issue.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e250109"},"PeriodicalIF":13.2,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145769353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging Natural Language Processing to Refine Normative Pediatric Renal US Values.","authors":"Irvine Sihlahla","doi":"10.1148/ryai.250979","DOIUrl":"10.1148/ryai.250979","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"8 1","pages":"e250979"},"PeriodicalIF":13.2,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145913239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reference Trajectories of Extra-Axial Cerebrospinal Fluid during Childhood and Adolescence Defined in a Clinically Acquired MRI Dataset.","authors":"Ayan S Mandal, Lena Dorfschmidt, Jenna M Schabdach, Margaret Gardner, Benjamin E Yerys, Richard A I Bethlehem, Susan Sotardi, M Katherine Henry, Joanne N Wood, Barbara H Chaiyachati, Aaron Alexander-Bloch, Jakob Seidlitz","doi":"10.1148/ryai.250123","DOIUrl":"10.1148/ryai.250123","url":null,"abstract":"<p><p>Purpose To build extra-axial cerebrospinal fluid (eaCSF) growth charts that define key diagnostic criteria for benign enlargement of the subarachnoid space (BESS) by providing an age-related reference benchmark to aid in assessing atypical eaCSF development. Materials and Methods In this retrospective study, T1-weighted MRI scans from patients who underwent imaging at a pediatric health care system between January 2004 and December 2023 were accessed to form a clinical control group. Nine scans from patients diagnosed with BESS by a board-certified pediatric neuroradiologist were also reviewed. T1-weighted scans were segmented into various tissue types, including eaCSF. Growth charts of eaCSF were modeled using the clinical control group. The results of patients with confirmed BESS were then benchmarked against these charts to test the performance of the eaCSF growth charts. Generalized additive models of location, scale, and shape were used. Results The eaCSF measurements were obtained for 1205 patients (619 female; age range, 0.19-19.6 years). Measurements show that eaCSF evolved dynamically with age, steadily decreasing from birth to 2 years, then trending upward in childhood. Seven of the nine patients with a clinical diagnosis of BESS had eaCSF measurements above the 97.5th percentile for at least one measurement. Percentile scores distinguished patients with BESS from controls with areas under the receiver operating characteristic curve of greater than 0.95. Conclusion MRI-derived eaCSF measurements evolved dynamically throughout early life. Patients with atypical CSF development could be differentiated from clinical controls using computational measurements paired with normative modeling. <b>Keywords:</b> MRI, Brain/Brain Stem, Pediatrics, Benign Enlargement of Subarachnoid Space <i>Supplemental material is available for this article.</i> © The Author(s) 2025. Published by the Radiological Society of North America under a CC BY-NC-ND license.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e250123"},"PeriodicalIF":13.2,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12856361/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145716001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mapping the AI \"Mind\": What the AI-STREAM Trial Reveals About Cancers Detected and Missed.","authors":"Synho Do, Manisha Bahl","doi":"10.1148/ryai.251002","DOIUrl":"10.1148/ryai.251002","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"8 1","pages":"e251002"},"PeriodicalIF":13.2,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145769505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Visualizing Radiologic Connections: An Explainable Coarse-to-Fine Foundation Model with Multiview Mammograms and Associated Reports.","authors":"Yuan Gao, Hong-Yu Zhou, Xin Wang, Antonio Portaluri, Tianyu Zhang, Regina Beets-Tan, Luyi Han, Chunyao Lu, Laura Estacio, Anna D'Angelo, Stephan Ursprung, Yizhou Yu, Jonas Teuwen, Tao Tan, Ritse Mann","doi":"10.1148/ryai.240646","DOIUrl":"10.1148/ryai.240646","url":null,"abstract":"<p><p>Purpose To develop a foundational pretraining method for digital mammography that extracts fine-grained visual-language representations from images and reports in label-limited settings. Materials and Methods A multiview mammogram-report pretraining framework for automated breast cancer analysis was developed using retrospectively collected data from January 2010 to December 2020. This framework provides visual explanations of the model's learning, allowing researchers to \"visualize what you learn.\" The abnormality-aware technique was tailored to mammogram characteristics of dense fibroglandular tissue. The proposed framework was evaluated on downstream tasks from four external medical centers, involving label-efficient abnormality recognition in mammograms, including malignancy classification, segmentation, and localization. Statistical analyses were performed using the DeLong test and paired <i>t</i> test for area under the receiver operating characteristic curve and Dice scores, respectively. Results The visualization results, including abnormality-enhanced mammograms and abnormality-awareness maps, could explain that the developed model successfully captures relationships between multiview mammograms and corresponding reports. This reduces the false positives for breast cancer by 37% and enables zero-shot abnormality segmentation. Furthermore, the developed model consistently outperformed existing approaches in fine-tuning for both malignancy classification (area under the receiver operating characteristic curve, INbreast: 0.90 vs 0.78 [<i>P</i> < .001]; Curated Breast Imaging Subset of Digital Database for Screening Mammography [CBIS-DDSM]: 0.85 vs 0.79 [<i>P</i> < .01]; Chinese Mammography Database: 0.85 vs 0.78 [<i>P</i> < .001]; and Cohort of Screen-age Women-Case Control: 0.86 vs 0.77 [<i>P</i> < .001]) and segmentation and localization (Dice score, INbreast: 0.75 vs 0.63 [<i>P</i> < .001]; CBIS-DDSM: 0.76 vs 0.61 [<i>P</i> < .001]). Conclusion The proposed framework enhances interpretability and fine-grained multimodal foundational learning for multiview mammograms and reports. <b>Keywords:</b> Mammography, Breast, Segmentation, Feature Detection, Quantification, Diagnosis, Translation, Transfer Learning, Unsupervised Learning, Breast Cancer, Representation Learning, Visual-Language Foundation Model, Explainable AI <i>Supplemental material is available for this article.</i> © RSNA, 2025.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e240646"},"PeriodicalIF":13.2,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145769463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}