Artificial Intelligence System Approaching Neuroradiologist-level Differential Diagnosis Accuracy at Brain MRI
Open accessThe DOI matches its Crossref record; 35 of the 35 references it cites were found in a registry or an index.
Abstract
Background Although artificial intelligence (AI) shows promise across many aspects of radiology, the use of AI to create differential diagnoses for rare and common diseases at brain MRI has not been demonstrated. Purpose To evaluate an AI system for generation of differential diagnoses at brain MRI compared with radiologists. Materials and Methods This retrospective study tested performance of an AI system for probabilistic diagnosis in patients with 19 common and rare diagnoses at brain MRI acquired between January 2008 and January 2018. The AI system combines data-driven and domain-expertise methodologies, including deep learning and Bayesian networks. First, lesions were detected by using deep learning. Then, 18 quantitative imaging features were extracted by using atlas-based coregistration and segmentation. Third, these image features were combined with five clinical features by using Bayesian inference to develop probability-ranked differential diagnoses. Quantitative feature extraction algorithms and conditional probabilities were fine-tuned on a training set of 86 patients (mean age, 49 years ± 16 [standard deviation]; 53 women). Accuracy was compared with radiology residents, general radiologists, neuroradiology fellows, and academic neuroradiologists by using accuracy of top one, top two, and top three differential diagnoses in 92 independent test set patients (mean age, 47 years ± 18; 52 women). Results For accuracy of top three differential diagnoses, the AI system (91% correct) performed similarly to academic neuroradiologists (86% correct; P = .20), and better than radiology residents (56%; P < .001), general radiologists (57%; P < .001), and neuroradiology fellows (77%; P = .003). The performance of the AI system was not affected by disease prevalence (93% accuracy for common vs 85% for rare diseases; P = .26). Radiologists were more accurate at diagnosing common versus rare diagnoses (78% vs 47% across all radiologists; P < .001). Conclusion An artificial intelligence system for brain MRI approached overall top one, top two, and top three differential diagnoses accuracy of neuroradiologists and exceeded that of less-specialized radiologists. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Zaharchuk in this issue.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“New multimodal AI models have made significant progress in describing video content, such as surgeries, making video data a promising method for AI decision support with further research.68”
From Review of Artificial Intelligence for Clinical Use in Alzheimer's Disease and Related Dementias (2026) · cites this paper as [68]
Citations
Cites
Yushkevich, P. A., Piven, J., Hazlett, H. C., Smith, R. G., Ho, S., Gee, J. C., & Gerig, G. (2006) · NeuroImage · doi:10.1016/j.neuroimage.2006.01.015 · Cited by 8,868
Self-citationDeep learning in medical image analysis
Shen, D., Wu, G., & Suk, H.-I. (2017) · Annual Review of Biomedical Engineering · doi:10.1146/annurev-bioeng-071516-044442 · Cited by 4,968
Subjective probability: A judgment of representativeness
Kahneman, D., & Tversky, A. (1972) · Cognitive Psychology · doi:10.1016/0010-0285(72)90016-3 · Cited by 4,851
A reproducible evaluation of ANTs similarity metric performance in brain image registration
Avants, B. B., Tustison, N. J., Song, G., Cook, P. A., Klein, A., & Gee, J. C. (2011) · NeuroImage · doi:10.1016/j.neuroimage.2010.09.025 · Cited by 4,796 · cited 1×
“com/ANTsX/ANTs) (12,13) (see details in Appendix E1 [online]).”
Self-citationBest practices for convolutional neural networks applied to visual document analysis
Simard, P. Y., Steinkraus, D., & Platt, J. C. (2005) · Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings. · doi:10.1109/icdar.2003.1227801 · Cited by 2,741
Marcus, D. S., Wang, T. H., Parker, J., Csernansky, J. G., Morris, J. C., & Buckner, R. L. (2007) · Journal of Cognitive Neuroscience · doi:10.1162/jocn.2007.19.9.1498 · Cited by 2,163
Statistical Methods in Diagnostic Medicine
Zhou, X., Obuchowski, N. A., & McClish, D. K. (2011) · Wiley Series in Probability and Statistics · doi:10.1002/9780470906514 · Cited by 1,784
Zech, J. R., Badgeley, M. A., Liu, M., Costa, A. B., Titano, J. J., & Oermann, E. K. (2018) · PLOS Medicine · doi:10.1371/journal.pmed.1002683 · Cited by 1,719 · cited 1×
“Finally, the retrospective work described here was performed at a single health system, potentially limiting generalization (28).”
MethodDeep learning algorithms for detection of critical findings in head CT scans: a retrospective study
Chilamkurthy, S., Ghosh, R., Tanamala, S., Biviji, M., Campeau, N. G., Venugopal, V. K., Mahajan, V., Rao, P., & Warier, P. (2018) · The Lancet · doi:10.1016/s0140-6736(18)31645-3 · Cited by 1,020
Bruno, M. A., Walker, E. A., & Abujudeh, H. H. (2015) · RadioGraphics · doi:10.1148/rg.2015150023 · Cited by 573
Research works printed but not linked: 1
- Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. Cham, Switzerland: Springer International, 2015; 234–241.
Cited by
Items 101–149 of 149
Man Versus Machine? Radiologists and Artificial Intelligence Work Better Together
Weisberg, E. M., Fishman, E. K., Chu, L. C. H., & Catmull, E. (2021) · Journal of the American College of Radiology · doi:10.1016/j.jacr.2020.12.017 · Cited by 5
The clinical value of second-opinion reporting by subspecialist musculoskeletal radiologists
Patel, A., & Isaac, A. (2025) · European Journal of Radiology · doi:10.1016/j.ejrad.2025.112262 · Cited by 2
Early detection of Alzheimer’s disease via multimodal MRI and machine learning
Yuan, J., Liu, Y., Jin, J., Li, S., & Zhang, Y. (2026) · Frontiers in Aging Neuroscience · doi:10.3389/fnagi.2026.1794982 · Cited by 3
Huang, Z. (2021) · International Conference on Computing Methodologies and Communication (ICCMC) · doi:10.1109/iccmc51019.2021.9418460 · Cited by 2
Dash, S., Thakur, R. A., Tripathy, S., & Jindal, S. K. (2022) · Elsevier eBooks · doi:10.1016/b978-0-323-91171-9.00009-0 · Cited by 3
The Role of Artificial Intelligence in Dental Diagnosis
Arefin, S. (2024) · International Journal of Scientific Research and Management (IJSRM) · doi:10.18535/ijsrm/v12i07.mp05 · Cited by 2
“As abnormalities are spotted on imaging or during intraoral inspection, AI can generate an immediate differential diagnoses list ranked by probability (Rauschecker et al., 2020).”
BackgroundCue relevance drives early quitting in visual search
Moher, J., Reyes, A. C. D., & Drew, T. (2024) · Cognitive Research Principles and Implications · doi:10.1186/s41235-024-00587-1 · Cited by 3
“…and highly motivated to find relevant information in medical images, and yet errors in image reading are consistently observed in experimental and clinical settings (e.g., Bruno et al., 2015; Rauschecker et al., 2020; Waite et al., 2020), leading to sometimes fatal outcomes for patients.”
BackgroundIncidental findings in research brain MRI : Definition, prevalence and ethical implications
de Jong, K. J., Poon, E., Foo, M., Maingard, J. T., Kok, H. K., Barras, C. D., Yazdabadi, A., Shaygi, B., Fitt, G. J., Egan, G. F., et al. (2024) · Journal of Medical Imaging and Radiation Oncology · doi:10.1111/1754-9485.13744 · Cited by 3
“46 47 – 49 Such models can ensure that a range of differential diagnoses are considered and help prevent missed aneurysms, particularly those that are incidental or small.”
Fostering Creative sports talents with transformer models for inclusive financial
Gao, Y., Wang, W., & Li, Y. (2025) · Egyptian Informatics Journal · doi:10.1016/j.eij.2025.100838 · Cited by 3
Chen, X., Xu, H., Huang, Y., Zhu, F., Tang, J., Li, H., Min, X., Hu, L., & Lu, L. J. (2026) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2026.106346 · Cited by 1
Kumar, P., Alnaimi, N. A. A., Soman, S., Suansing, L., Arriola, D. R., & Jamea, L. A. (2026) · Sci · doi:10.3390/sci8040073 · Cited by 2
“…[39] Patients-100 60 65 Homayounieh 2021 [21] Patients-100 80 86 86 Luo 2021 [43] Patients-150 59 51 Michael Gottlieb 2024 [18] Patients-71 69 68 Rauschecker 2020 [15] Patients-86 78 73 48 Tamai 2023 [23] Patients-42 34 28 Choi 2020 [33] Patients-1198 1174 1198 910 Wang 2021 [22] Patients-80 71…”
InfluentialBasics of MR imaging for the radiation oncologist
Das, I. J., Sagreiya, H., Yadav, P., & Allen, B. D. (2022) · Advances in magnetic resonance technology and applications · doi:10.1016/b978-0-323-91689-9.00002-9 · Cited by 1
Ferles, A., & Barkhof, F. (2023) · European Radiology · doi:10.1007/s00330-023-09863-y · Cited by 1
Hasan, H. H. (2025) · medRxiv · doi:10.1101/2025.09.10.25335333 · Cited by 4
Li, M., & Chang, C. Y. (2021) · Radiology Artificial Intelligence · doi:10.1148/ryai.2021210022
“To order printed copies, contact reprints@rsna.org diagnoses on brain MRI studies (7,8).”
BackgroundCRISE TIREOTÓXICA: UM DESAFIO NO DIAGNÓSTICO NA SALA DE EMERGÊNCIA
Roque, C. D. F., Silveira, A. P. D. O., & Teixeira, E. B. (2021) · doi:10.22533/at.ed.6162108067 · Cited by 24
de Almeida, S. A., Rodrigues, T. Q., Teodoro, J. S. T. S., Campos, L. P., Lima, E. F., Cabral, M. D., Machado, N. F., Filho, C. R. D. S., Lemos, P. C. O., Soerger, M., et al. (2021) · doi:10.22533/at.ed.6162108062
ANSIEDADE E CÂNCER DE MAMA: INFLUÊNCIA DA ATIVIDADE FÍSICA
Aguiar, R. R. R., da Silva, C. E. S., Rodrigues, N. H. G., Lima, C. A. G., de Jesus, Y. N. O. N., de Paula, J. G. S., Xavier, M. D., Bicalho, B. L. R., Souto, S. V. D., Bauman, J. M., et al. (2021) · doi:10.22533/at.ed.6162108063
van Staalduinen, E. K. (2022) · Journal of Magnetic Resonance Imaging · doi:10.1002/jmri.28119
Cook, T. S. (2021) · Springer eBooks · doi:10.1007/978-1-0716-1756-4_10
Self-citationWood, D. A., Wood, D. A., Kafiabadi, S., Busaidi, A. A., Guilhem, E., Montvila, A., Agarwal, S., Lynch, J., Townend, M., Barker, G. J., et al. (2022) · doi:10.22541/au.166785452.22212589/v1 · Cited by 2
Neuroradiology: Current Status and Future Prospects
Ferraciolli, S. F., Mota, A. L., Ayres, A. S., Polsin, L. L. M., Kitamura, F., & Leite, C. D. C. (2022) · Integrated science · doi:10.1007/978-3-031-11199-0_4
Application of Artificial Intelligence System in Chamber Music Teaching
Sun, P. (2023) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-981-19-9373-2_35
Rozzanigo, U., Casagranda, G., Mochen, M., & FERRARI, M. (2023) · Journal of Radiological Review · doi:10.23736/s2723-9284.23.00244-9
Nielsen, E. E. (2024) · Cronfa (Swansea University) · doi:10.23889/suthesis.66964
Contrasting the performance of mainstream Large Language Models in Radiology Board ExaminationsPreprint
Wei, B., Zhang, X., Shao, Y., Sun, X., & Chen, L. (2024) · Research Square · doi:10.21203/rs.3.rs-4573702/v1
Torik, I. A., Gharib, M. H., Roshandel, G., Zare, A., & Amanian, D. (2025) · The Royal College of Radiologists Open · doi:10.1016/j.rcro.2025.100271
Pneumonia and brain tumors diagnosis using machine learning algorithms
Filipović, S., & Kadry, S. (2025) · Elsevier eBooks · doi:10.1016/b978-0-443-23979-3.00011-7
Tan, R., Wang, M., Kang, B., Yu, X., Zhao, G., Zhang, S., & Wang, X. (2025) · Research Square · doi:10.21203/rs.3.rs-6439190/v1
Differential diagnosis making with large language models and probabilistic logic program
Galitsky, B. A. (2025) · Elsevier eBooks · doi:10.1016/b978-0-443-30046-2.00010-7
An Overview of Artificial Intelligence in Neurology
Parker, T., & Brush, B. (2025) · Seminars in Neurology · doi:10.1055/a-2693-0547 · Cited by 2
“32 This review aims to provide neurologists with the foundational principles of modern AI algorithms, equipping them to critically assess inputs and interpret outputs across various clinical applications.”
Rao, H., Wei, S., Jiang, S., Wang, M., Ding, W., & Huang, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05182-0_65
Uppalapati, K., Yimenicioglu, B., Abdulkareem, S., Eftekhari, A., Uppalapati, B., & Kamath, V. (2025) · doi:10.1109/cises66934.2025.11265234
Han, X., Luo, N., Xu, L., Cao, J., Guo, N., He, Y., Hong, M., Jia, X., Wang, Z., & Yang, Z. (2022) · Figshare · doi:10.6084/m9.figshare.19194200
Additional file 1 of Cue relevance drives early quitting in visual search
Moher, J., Delos Reyes, A., & Drew, T. (2024) · Figshare · doi:10.6084/m9.figshare.26829035
Bedrikovetski, S., Dudi-Venkata, N. N., Kroon, H. M., Seow, W., Vather, R., Carneiro, G., Moore, J. W., & Sammour, T. (2021) · Figshare · doi:10.6084/m9.figshare.16682841
Beyond the Hype: Risks, Realism, and Responsibilities in Digital Twins for Alzheimer’s Disease
Guttmann-Flury, E., Zhao, J., & Sawan, M. (2026) · Artificial intelligence · doi:10.5772/intechopen.1014167
Association of Rare Disease Diagnosis and Its Treatment Using Artificial Intelligence – A Review
Kamboj, A., Singh, A. K., Kumar, S., & Garg, V. K. (2025) · Global AI Summit - International Conference on Artificial Intelligence and Emerging Technology (AI Summit) · doi:10.1109/aisummit66170.2025.11411235
Advanced neuroimaging for central nervous system infections: when, where, how, and why?
Mohan, S., Duong, M. T., Curl, B. J., & Chawla, S. (2026) · Current Opinion in Infectious Diseases · doi:10.1097/qco.0000000000001191
Self-citationKumar, K. N., Dhanalakshmi, P., & Baseer, K. K. (2026) · AIP conference proceedings · doi:10.1063/5.0313022
Portable neuroimaging and AI: democratizing brain diagnostics
León-Rojas, J. E. (2026) · Frontiers in Neurology · doi:10.3389/fneur.2026.1786660
Zholshybek, N., Abdurakhmanova, E. G., Bimakhan, A. Z., Jumadilova, D., Baiturlin, Z., Almazan, J., & Bolla, S. R. (2026) · Frontiers in Neuroinformatics · doi:10.3389/fninf.2026.1821249
Cheraya, G., Tejani, A. S., Prakash, M., Lutz, A., & Chhabra, A. B. (2026) · Clinical Radiology · doi:10.1016/j.crad.2026.107377
Research on tunnel construction quality control based on blockchain technology and Bayesian networks
Cai, L.-P., Gong, Q., Jiang, F., Yuan, M., Xiao, Z.-Y., Zhang, X., Zheng, C., & Wu, Y. (2026) · Applied Geophysics · doi:10.1007/s11770-026-1418-x
Sankar, S., Rabbi, K. A., Abdelkhalek, K., Alharbi, A., Okafor, E., Singla, K., Karunanayaka, P. P., & Sharma, P. (2026) · Clinical neuroimaging. · doi:10.1002/neo2.70080
Artificial Intelligence in Neuroradiology: Transforming Diagnosis, Prognostication, and Workflow
Ghosh, T. (2026) · Indian journal of radiology and imaging - new series/Indian journal of radiology and imaging/Indian Journal of Radiology & Imaging · doi:10.1055/s-0046-1829088
Combining artificial intelligence and neuroscience to diagnose and predict neurological disorders
Rahaman, S. A., Babu, M. R., Patel, S., Kumar, R., Löebenberg, R., & Chettupalli, A. K. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-27722-1.00015-8
Training visual language models for MRI-based diagnostics
Gusarova, N., Dobrenko, N., Zubanenko, A., Tomilov, I., & Ulyanov, P. G. (2026) · Scientific and technical journal of information technologies mechanics and optics · doi:10.17586/2226-1494-2026-26-5-1000-1008
Metadata
- Authors
- Rauschecker, A. M., Rudie, J. D., Xie, L., Wang, J., Duong, M. T., Botzolakis, E. J., Kovalovich, A. M., Egan, J., Cook, T. C., Bryan, R. N., Nasrallah, I. M., Mohan, S., & Gee, J. C.
- Affiliations
- University of California, San Francisco (US)
- From the Department of Radiology & Biomedical Imaging, University of California, San Francisco, 513 Parnassus Ave, Room S-261, Box 0628, San Francisco, CA 94143-0628 (A.M.R.); Department of Radiology, Hospital of the University of Pennsylvania, Philadelphia, PA 19104 (A.M.R., J.D.R., L.X., J.W., M.T.D., A.M.K., J.E., T.C.C., I.M.N., S.M., J.C.G.); Mecklenburg Radiology Associates, Charlotte, NC (E.J.B.); and Department of Radiology, University of Texas at Austin, Austin, TX (R.N.B.).
- ORCID
- Venue
- Radiology
- Published
- 2020-06
- Volume
- 295
- Issue
- 3
- Pages
- 626-637
- ISSN
- 0033-8419 (ISSN-L)
- 1527-1315
- PMC
- PMC7263320
- OpenAlex
- W3014301887
- Cited by
- 148
- References
- 34
- Citations by year
- 2026: 14 · 2025: 28 · 2024: 22 · 2023: 20 · 2022: 26 · 2021: 26 · 2020: 12
- FWCI
- 6.18
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Radiology (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- eScholarship (California Digital Library) (submittedVersion)
- Funders
- National Institutes of Health (T32)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Brain Tumor Detection and Classification
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiomics and Machine Learning in Medical Imaging
- Keywords
- artificial intelligence, differential diagnosis, brain MRI, neuroradiology, Bayesian networks, quantitative imaging features, probabilistic diagnosis, radiologist comparison
- MeSH terms
- Adult, Aged, Artificial Intelligence, Brain, Brain Diseases, Diagnosis, Computer-Assisted, Diagnosis, Differential, Female, Humans, Magnetic Resonance Imaging, Male, Middle Aged, Retrospective Studies, Sensitivity and Specificity, Rare Diseases
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.