{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T23:08:32Z","timestamp":1790032112566,"version":"4.0.1"},"reference-count":43,"publisher":"Georg Thieme Verlag KG","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Rofo"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Artificial intelligence (AI) is increasingly shaping radiology, by performing tasks ranging from image reconstruction and workflow optimization to image analysis, reporting support, and multimodal clinical decision support. This narrative review provides a structured overview of the methodological foundations and clinical evolution of AI in radiology. We first outline the historical development from early rule-based expert systems and classical feature engineering to data-driven machine learning. We then explain the core principles of neural networks, including network depth, nonlinear activation functions, loss-based optimization, and backpropagation, before discussing convolutional neural networks (CNNs) as the dominant architecture for image-based learning in radiology. Particular emphasis is placed on inductive biases such as locality, parameter sharing, and translation equivariance, as well as on encoder\u2013decoder architectures and U-Net-based approaches for segmentation. Subsequently, we discuss attention mechanisms, Transformer architectures, and their role in modeling long-range dependencies in medical imaging. We distinguish Vision Transformers, which process images as token sequences, from vision\u2013language models, which align imaging data with textual information and enable multimodal tasks such as report generation, image\u2013text retrieval, and visual question answering. Recent radiology-specific foundation models and large language models are considered in the context of workflow support, structured reporting, knowledge integration, and emerging multimodal image interpretation. Finally, we address key challenges for clinical translation, including data requirements, bias, generalizability, hallucinations, interpretability, validation, federated learning, and regulatory aspects such as the EU AI Act and the Medical Device Regulation. A basic understanding of these methodological principles is essential for radiologists to critically evaluate current AI systems and to contribute to their safe, transparent, and clinically meaningful implementation.<\/jats:p>","DOI":"10.1055\/a-2944-0073","type":"journal-article","created":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T22:45:33Z","timestamp":1790030733000},"source":"Crossref","is-referenced-by-count":0,"title":["Artificial Intelligence in Radiology: Methodological Foundations, Clinical Applications, and Emerging Multimodal Models"],"prefix":"10.1055","author":[{"given":"Christoph Alexander","family":"Stueckle","sequence":"additional","affiliation":[{"name":"University Witten Herdecke Faculty of Health, Germany","place":["Witten"],"department":["Radiology"]},{"name":"Marienhospital Osnabruck, Germany","place":["Osnabruck"],"department":["Radiology"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Matthias","family":"Limbrock","sequence":"additional","affiliation":[{"name":"University Witten Herdecke Faculty of Health, Germany","place":["Witten"],"department":["Radiology"]},{"name":"Marienhospital Osnabruck, Germany","place":["Osnabruck"],"department":["Radiology"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Patrick","family":"Haage","sequence":"additional","affiliation":[{"name":"HELIOS Universitatsklinikum Wuppertal, Germany","place":["Wuppertal"],"department":["Diagnostic and Interventional Radiology"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"194","published-online":{"date-parts":[[2026,9,21]]},"reference":[{"key":"ref1","volume-title":"Computer-Based Medical Consultations: MYCIN","author":"EH Shortliffe","year":"1976"},{"key":"ref2","doi-asserted-by":"crossref","first-page":"610","DOI":"10.1109\/TSMC.1973.4309314","article-title":"Textural features for image classification","volume":"SMC-3","author":"RM Haralick","year":"1973","journal-title":"IEEE Transactions on Systems, Man, and Cybernetics"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"5799","DOI":"10.1118\/1.3013555","article-title":"Anniversary paper: History and status of CAD and quantitative image analysis: the role of Medical Physics and AAPM","volume":"35","author":"ML Giger","year":"2008","journal-title":"Med Phys"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"386","DOI":"10.1037\/h0042519","article-title":"The perceptron: a probabilistic model for information storage and organization in the brain","volume":"65","author":"F Rosenblatt","year":"1958","journal-title":"Psychol Rev"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"533","DOI":"10.1038\/323533a0","article-title":"Learning representations by back-propagating errors","volume":"323","author":"DE Rumelhart","year":"1986","journal-title":"Nature"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"97","DOI":"10.1016\/S0720-048X(99)00016-9","article-title":"Computer-aided diagnosis in radiology: potential and pitfalls","volume":"31","author":"K Doi","year":"1999","journal-title":"Eur J Radiol"},{"key":"ref8","doi-asserted-by":"crossref","DOI":"10.1093\/oso\/9780198538493.001.0001","volume-title":"Neural Networks for Pattern Recognition","author":"CM Bishop","year":"1995"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"303","DOI":"10.1007\/BF02551274","article-title":"Approximation by superpositions of a sigmoidal function","volume":"2","author":"G Cybenko","year":"1989","journal-title":"Math Control Signals Systems"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1412.6980","article-title":"Adam: A Method for Stochastic Optimization","author":"DP Kingma","year":"2015","journal-title":"International Conference on Learning Representations (ICLR)"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"Y LeCun","year":"1998","journal-title":"Proceedings of the IEEE"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"Y LeCun","year":"2015","journal-title":"Nature"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","article-title":"A survey on deep learning in medical image analysis","volume":"42","author":"G Litjens","year":"2017","journal-title":"Med Image Anal"},{"key":"ref14","first-page":"4898","article-title":"Understanding the Effective Receptive Field in Deep Convolutional Neural Networks","volume":"29","author":"W Luo","year":"2016","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"e21","DOI":"10.23915\/distill.00021","article-title":"Computing Receptive Fields of Convolutional Neural Networks","volume":"4","author":"A Araujo","year":"2019","journal-title":"Distill"},{"key":"ref16","first-page":"1097","article-title":"ImageNet classification with deep convolutional neural networks","volume":"25","author":"A Krizhevsky","year":"2012","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1409.1556","article-title":"Very deep convolutional networks for large-scale image recognition","author":"K Simonyan","year":"2015","journal-title":"International Conference on Learning Representations (ICLR)"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"770","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","volume":"2016","author":"K He","journal-title":"IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"504","DOI":"10.1126\/science.1127647","article-title":"Reducing the dimensionality of data with neural networks","volume":"313","author":"GE Hinton","year":"2006","journal-title":"Science"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"203","DOI":"10.1038\/s41592-020-01008-z","article-title":"nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation","volume":"18","author":"F Isensee","year":"2021","journal-title":"Nat Methods"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1409.0473","article-title":"Neural machine translation by jointly learning to align and translate","author":"D Bahdanau","year":"2015","journal-title":"International Conference on Learning Representations (ICLR)"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.03762","article-title":"Attention Is All You Need.","volume":"30","author":"A Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2010.11929","article-title":"An image is worth 16\u00d716 words: Transformers for image recognition at scale","author":"A Dosovitskiy","year":"2021","journal-title":"International Conference on Learning Representations (ICLR)"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2102.04306","article-title":"TransUNet: Transformers make strong encoders for medical image segmentation","author":"J Chen","year":"2021","journal-title":"arXiv"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2401.12208","article-title":"A Vision-Language Foundation Model to Enhance Efficiency of Chest X-ray Interpretation","author":"Z Chen","year":"2024","journal-title":"arXiv"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"7866","DOI":"10.1038\/s41467-025-62385-7","article-title":"Towards generalist foundation model for radiology by leveraging web-scale 2D and 3D medical data","volume":"16","author":"C Wu","year":"2025","journal-title":"Nat Commun"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"152","DOI":"10.1002\/mrm.28420","article-title":"Accelerating cardiac cine MRI using a deep learning-based ESPIRiT reconstruction","volume":"85","author":"CM Sandino","year":"2021","journal-title":"Magn Reson Med"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"539","DOI":"10.1148\/radiol.2021202624","article-title":"Free-breathing Accelerated Cardiac MRI Using Deep Learning: Validation in Children and Young Adults","volume":"300","author":"EJ Zucker","year":"2021","journal-title":"Radiology"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3571730","article-title":"Survey of hallucination in natural language generation","volume":"55","author":"Z Ji","year":"2023","journal-title":"ACM Comput Surv"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"800","DOI":"10.1148\/radiol.2017171920","article-title":"Methodologic Guide for Evaluating Clinical Performance and Effect of Artificial Intelligence Technology for Medical Diagnosis and Prediction","volume":"286","author":"SH Park","year":"2018","journal-title":"Radiology"},{"key":"ref37","doi-asserted-by":"publisher","first-page":"e271","DOI":"10.1016\/S2589-7500(19)30123-2","article-title":"A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis","volume":"1","author":"X Liu","year":"2019","journal-title":"Lancet Digit Health"},{"key":"ref38","doi-asserted-by":"publisher","first-page":"65","DOI":"10.1038\/s41746-021-00438-z","article-title":"Diagnostic accuracy of deep learning in medical imaging: a systematic review and meta-analysis","volume":"4","author":"R Aggarwal","year":"2021","journal-title":"NPJ Digit Med"},{"key":"ref39","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1038\/s41586-019-1799-6","article-title":"International evaluation of an AI system for breast cancer screening","volume":"577","author":"SM McKinney","year":"2020","journal-title":"Nature"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"936","DOI":"10.1016\/S1470-2045(23)00298-X","article-title":"Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study","volume":"24","author":"K L\u00e5ng","year":"2023","journal-title":"Lancet Oncol"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"505","DOI":"10.1016\/S0140-6736(25)02464-X","article-title":"Interval cancer, sensitivity, and specificity comparing AI-supported mammography screening with standard double reading without AI in the MASAI study: a randomised, controlled, non-inferiority, single-blinded, population-based, screening-accuracy trial","volume":"407","author":"J Gommers","year":"2026","journal-title":"Lancet"},{"key":"ref42","first-page":"1877","article-title":"Language models are few-shot learners","volume":"33","author":"TB Brown","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref43","doi-asserted-by":"crossref","first-page":"23716","DOI":"10.52202\/068431-1723","article-title":"Flamingo: a visual language model for few-shot learning","volume":"35","author":"J-B Alayrac","year":"2022","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref44","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","article-title":"Large language models encode clinical knowledge","volume":"620","author":"K Singhal","year":"2023","journal-title":"Nature"},{"key":"ref45","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1038\/s41586-023-05881-4","article-title":"Foundation models for generalist medical artificial intelligence","volume":"616","author":"M Moor","year":"2023","journal-title":"Nature"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1056\/AIoa2300138","article-title":"Towards Generalist Biomedical AI","volume":"1","author":"T Tu","year":"2024","journal-title":"NEJM AI"},{"key":"ref47","first-page":"9459","article-title":"Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks","volume":"33","author":"P Lewis","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1259\/bjr.20220890","article-title":"Federated learning for medical imaging radiology","volume":"96","author":"MHU Rehman","year":"2023","journal-title":"Br J Radiol"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2404.18416","article-title":"Capabilities of Gemini Models in Medicine","author":"K Saab","year":"2024","journal-title":"arXiv"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2405.03162","article-title":"Advancing Multimodal Medical Capabilities of Gemini","author":"L Yang","year":"2024","journal-title":"arXiv"}],"container-title":["R\u00f6Fo - Fortschritte auf dem Gebiet der R\u00f6ntgenstrahlen und der bildgebenden Verfahren"],"original-title":[],"language":"de","link":[{"URL":"http:\/\/www.thieme-connect.de\/products\/ejournals\/pdf\/10.1055\/a-2944-0073.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T22:45:40Z","timestamp":1790030740000},"score":1,"resource":{"primary":{"URL":"http:\/\/www.thieme-connect.de\/DOI\/DOI?10.1055\/a-2944-0073"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,21]]},"references-count":43,"URL":"https:\/\/doi.org\/10.1055\/a-2944-0073","archive":["Portico","CLOCKSS"],"relation":{},"ISSN":["1438-9029","1438-9010"],"issn-type":[{"value":"1438-9029","type":"print"},{"value":"1438-9010","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,21]]}}}