{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T20:50:03Z","timestamp":1792788603557,"version":"4.5.0"},"reference-count":38,"publisher":"Public Library of Science (PLoS)","issue":"7","license":[{"start":{"date-parts":[[2023,7,6]],"date-time":"2023-07-06T00:00:00Z","timestamp":1688601600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Novartis Pharma AG"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS ONE"],"abstract":"<jats:p>Clinical data sharing can facilitate data-driven scientific research, allowing a broader range of questions to be addressed and thereby leading to greater understanding and innovation. However, sharing biomedical data can put sensitive personal information at risk. This is usually addressed by data anonymization, which is a slow and expensive process. An alternative to anonymization is construction of a synthetic dataset that behaves similar to the real clinical data but preserves patient privacy. As part of a collaboration between Novartis and the Oxford Big Data Institute, a synthetic dataset was generated based on images from COSENTYX\u00ae (secukinumab) ankylosing spondylitis (AS) clinical studies. An auxiliary classifier Generative Adversarial Network (ac-GAN) was trained to generate synthetic magnetic resonance images (MRIs) of vertebral units (VUs), conditioned on the VU location (cervical, thoracic and lumbar). Here, we present a method for generating a synthetic dataset and conduct an in-depth analysis on its properties along three key metrics: image fidelity, sample diversity and dataset privacy.<\/jats:p>","DOI":"10.1371\/journal.pone.0280316","type":"journal-article","created":{"date-parts":[[2023,7,6]],"date-time":"2023-07-06T13:36:08Z","timestamp":1688650568000},"page":"e0280316","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":21,"title":["A deep learning approach to private data sharing of medical images using conditional generative adversarial networks (GANs)"],"prefix":"10.1371","volume":"18","author":[{"given":"Hanxi","family":"Sun","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jason","family":"Plawinski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sajanth","family":"Subramaniam","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Amir","family":"Jamaludin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Timor","family":"Kadir","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aimee","family":"Readie","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gregory","family":"Ligozio","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David","family":"Ohlssen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark","family":"Baillie","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7662-8724","authenticated-orcid":true,"given":"Thibaud","family":"Coroller","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2023,7,6]]},"reference":[{"issue":"9","key":"pone.0280316.ref001","doi-asserted-by":"crossref","first-page":"1342","DOI":"10.1038\/s41591-018-0107-6","article-title":"Clinically applicable deep learning for diagnosis and referral in retinal disease","volume":"24","author":"J De Fauw","year":"2018","journal-title":"Nat. Med"},{"issue":"1","key":"pone.0280316.ref002","doi-asserted-by":"crossref","first-page":"3877","DOI":"10.1038\/s41467-020-17678-4","article-title":"A deep learning model to predict RNA-Seq expression of tumours from whole slide images.","volume":"11","author":"B Schmauch","year":"2020","journal-title":"Nat. Commun"},{"issue":"11","key":"pone.0280316.ref003","doi-asserted-by":"crossref","first-page":"e1002711","DOI":"10.1371\/journal.pmed.1002711","article-title":"Deep learning for lung cancer prognostication: A retrospective multi-cohort radiomics study","volume":"15","author":"A Hosny","year":"2018","journal-title":"PLoS Med"},{"issue":"11","key":"pone.0280316.ref004","doi-asserted-by":"crossref","first-page":"3266","DOI":"10.1158\/1078-0432.CCR-18-2495","article-title":"Deep learning predicts lung cancer treatment response from serial medical imaging","volume":"25","author":"Y Xu","year":"2019","journal-title":"Clin. Cancer Res"},{"key":"pone.0280316.ref005","author":"N. C. Institute","year":"2020","journal-title":"PubMed. Definitions"},{"key":"pone.0280316.ref006","doi-asserted-by":"crossref","first-page":"1018","DOI":"10.1016\/j.procs.2021.01.281","article-title":"The challenge of privacy and security when using technology to track people in times of COVID-19 pandemic.","volume":"181","author":"H. J. Smidt","year":"2021","journal-title":"Procedia Comput. Sci"},{"key":"pone.0280316.ref007","article-title":"Ten Research Challenge Areas in Data Science","author":"J. M. Wing","year":"2020","journal-title":"Harvard Data Sci. Rev"},{"key":"pone.0280316.ref008","doi-asserted-by":"crossref","DOI":"10.4337\/9781788976220","author":"J. Cannataci","year":"2020","journal-title":"Legal Challenges of Big Data. (Edward Elgar Publishing"},{"key":"pone.0280316.ref009","author":"J. L. Grama","year":"2020","journal-title":"Legal and Privacy Issues in Information Security. (Jones & Bartlett Learning"},{"key":"pone.0280316.ref010","volume-title":"Reproducibility and Replicability in Science.","author":"National Academies of Sciences and Medicine, E.","year":"2019"},{"key":"pone.0280316.ref011","doi-asserted-by":"crossref","first-page":"170124","DOI":"10.1038\/sdata.2017.124","article-title":"The public cancer radiology imaging collections of The Cancer Imaging Archive","volume":"4","author":"F Prior","year":"2017","journal-title":"Sci. Data"},{"key":"pone.0280316.ref012","unstructured":"El Emam, K. and Arbuckle, L. Anonymizing health data: case studies and methods to get you started. O\u2019Reilly Media, Inc. https:\/\/www.oreilly.com\/library\/view\/anonymizing-health-data\/9781449363062\/ (2013)."},{"key":"pone.0280316.ref013","unstructured":"Ryffel, T. et al. A generic framework for privacy preserving deep learning."},{"key":"pone.0280316.ref014","article-title":"Generative Adversarial Nets.","author":"I. J. Goodfellow","year":"2014","journal-title":"Proceedings of the 27th International Conference on Neural Information Processing Systems\u2014Volume 2 2672\u20132680"},{"issue":"26","key":"pone.0280316.ref015","doi-asserted-by":"crossref","first-page":"2534","DOI":"10.1056\/NEJMoa1505066","article-title":"Secukinumab, an Interleukin-17A Inhibitor, in Ankylosing Spondylitis","volume":"373","author":"D Baeten","year":"2015","journal-title":"N. Engl. J. Med"},{"key":"pone.0280316.ref016","doi-asserted-by":"crossref","first-page":"63","DOI":"10.1016\/j.media.2017.07.002","article-title":"SpineNet: Automated classification and evidence visualization in spinal MRIs","volume":"41","author":"A Jamaludin","year":"2017","journal-title":"Med. Image Anal"},{"issue":"9906","key":"pone.0280316.ref017","doi-asserted-by":"crossref","first-page":"1705","DOI":"10.1016\/S0140-6736(13)61134-4","article-title":"Anti-interleukin-17A monoclonal antibody secukinumab in treatment of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial","volume":"382","author":"D Baeten","year":"2013","journal-title":"Lancet"},{"key":"pone.0280316.ref018","article-title":"Conditional image synthesis with auxiliary classifier GANs","author":"A. Odena","year":"2017","journal-title":"International conference on machine learning 2642\u20132651"},{"key":"pone.0280316.ref019","article-title":"Adam: A method for stochastic optimization. arXiv Prepr","author":"D. P. Kingma","year":"2014","journal-title":"arXiv1412.6980"},{"key":"pone.0280316.ref020","doi-asserted-by":"crossref","first-page":"171","DOI":"10.1080\/01621459.1987.10478410","article-title":"Better Bootstrap Confidence Intervals","volume":"82","author":"B. Efron","year":"1987","journal-title":"J. Am. Stat. Assoc"},{"key":"pone.0280316.ref021","article-title":"A Style-Based Generator Architecture for Generative Adversarial Networks","author":"T. Karras","year":"2020","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"pone.0280316.ref022","article-title":"UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction.","author":"L. McInnes","year":"2018"},{"issue":"2008","key":"pone.0280316.ref023","article-title":"Visualizing Data using t-SNE.","volume":"9","author":"L. Van Der Maaten","journal-title":"Journal of Machine Learning Research"},{"key":"pone.0280316.ref024","article-title":"Deep residual learning for image recognition","author":"K. He","year":"2016","journal-title":"Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition vols 2016-December 770\u2013778 (IEEE Computer Society"},{"issue":"2","key":"pone.0280316.ref025","doi-asserted-by":"crossref","first-page":"36","DOI":"10.1167\/tvst.9.2.36","article-title":"Protecting Data Privacy in the Age of AI-Enabled Ophthalmology.","volume":"9","author":"E Tom","year":"2020","journal-title":"Transl. Vis. Sci. Technol"},{"issue":"2020","key":"pone.0280316.ref026","article-title":"Protecting GANs against privacy attacks by preventing overfitting.","volume":"0","author":"S. Mukherjee","year":"2001","journal-title":"CoRR abs"},{"key":"pone.0280316.ref027","first-page":"1","author":"C. Dwork","year":"2008","journal-title":"Differential privacy: A survey of results"},{"key":"pone.0280316.ref028","article-title":"PATE-GAN: Generating Synthetic Data with Differential Privacy Guarantees.","author":"J. Yoon","year":"2019","journal-title":"International Conference on Learning Representations"},{"issue":"2019","key":"pone.0280316.ref029","article-title":"GAN-Leaks: A Taxonomy of Membership Inference Attacks against GANs","volume":"0","author":"D. Chen","journal-title":"CoRR abs\/1909"},{"key":"pone.0280316.ref030","doi-asserted-by":"crossref","first-page":"e005122","DOI":"10.1161\/CIRCOUTCOMES.118.005122","article-title":"Privacy-Preserving Generative Deep Neural Networks Support Clinical Data Sharing","volume":"12","author":"B. K. Beaulieu-Jones","year":"2019","journal-title":"Circ. Cardiovasc. Qual. Outcomes"},{"key":"pone.0280316.ref031","article-title":"GANs for Medical Image Analysis","volume":"101938","author":"S. Kazeminia","year":"2020","journal-title":"Artif. Intell. Med"},{"key":"pone.0280316.ref032","doi-asserted-by":"crossref","first-page":"101552","DOI":"10.1016\/j.media.2019.101552","article-title":"Generative adversarial network in medical imaging: A review","volume":"58","author":"X. Yi","year":"2019","journal-title":"Med. Image Anal"},{"key":"pone.0280316.ref033","unstructured":"Shin, H.-C. et al. Medical Image Synthesis for Data Augmentation and Anonymization Using Generative Adversarial Networks: Third International Workshop, SASHIMI 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 16, 2018, Proceedings. in Simulation and Synthesis in Medical Imaging (eds. Gooya, A., Goksel, O., Oguz, I. & Burgos, N.) vol. 11037 1\u201311 (Springer International Publishing, 2018)."},{"key":"pone.0280316.ref034","doi-asserted-by":"crossref","DOI":"10.1109\/ISBI.2018.8363678","article-title":"GAN-based synthetic brain MR image generation.","author":"C. Han","year":"2018","journal-title":"2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018)"},{"key":"pone.0280316.ref035","author":"S. Fossen-Romsaas","year":"2020","journal-title":"Synthesizing Skin Lesion Images Using Generative Adversarial Networks"},{"key":"pone.0280316.ref036","doi-asserted-by":"crossref","first-page":"321","DOI":"10.1016\/j.neucom.2018.09.013","article-title":"GAN-based synthetic medical image augmentation for increased CNN performance in liver lesion classification","volume":"321","author":"M. Frid-Adar","year":"2018","journal-title":"Neurocomputing"},{"key":"pone.0280316.ref037","doi-asserted-by":"crossref","DOI":"10.1145\/3424978.3425118","article-title":"Medical Image Synthetic Data Augmentation Using GAN.","author":"H. Zhang","year":"2020","journal-title":"Proceedings of the 4th International Conference on Computer Science and Application Engineering"},{"issue":"4","key":"pone.0280316.ref038","doi-asserted-by":"crossref","first-page":"600","DOI":"10.1109\/TIP.2003.819861","article-title":"Image quality assessment: from error visibility to structural similarity","volume":"13","author":"Z Wang","year":"2004","journal-title":"IEEE Trans. Image Process"}],"container-title":["PLOS ONE"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0280316","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,7,6]],"date-time":"2023-07-06T13:36:47Z","timestamp":1688650607000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0280316"}},"subtitle":[],"editor":[{"given":"Sathishkumar","family":"V. E.","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2023,7,6]]},"references-count":38,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2023,7,6]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0280316","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,7,6]]}}}