{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,3]],"date-time":"2026-06-03T13:53:10Z","timestamp":1780494790941,"version":"3.54.1"},"reference-count":35,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2025,10,14]],"date-time":"2025-10-14T00:00:00Z","timestamp":1760400000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Institutes of Health\/National Cancer Institute","award":["P30CA016672"],"award-info":[{"award-number":["P30CA016672"]}]},{"name":"National Institutes of Health\/National Cancer Institute","award":["R01CA235564"],"award-info":[{"award-number":["R01CA235564"]}]},{"name":"National Institutes of Health\/National Cancer Institute","award":["R01CA221971"],"award-info":[{"award-number":["R01CA221971"]}]},{"name":"MD Anderson Strategic Initiative Development Program"},{"name":"Apache corporation"},{"name":"Helen Black Image Guided Fund"},{"name":"Image Guided Cancer Therapy Research Program at The University of Texas MD Anderson Cancer Center through a generous gift from the Apache corporation"},{"name":"Diagnostic Imaging \u2013 Summer Training and Experiences Partnership"},{"name":"APC"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Bioengineering"],"abstract":"<jats:p>This work aims to leverage generative modeling-based anomaly detection to enhance interpretability in AI failure detection systems and to aid data curation for large repositories. For failure detection interpretability, this retrospective study utilized 3339 CT scans (525 patients), divided patient-wise into training, baseline test, and anomaly (having failure-causing attributes\u2014e.g., needles, ascites) test datasets. For data curation, 112,120 ChestX-ray14 radiographs were used for training and 2036 radiographs from the Medical Imaging and Data Resource Center for testing, categorized as baseline or anomalous based on attribute alignment with ChestX-ray14. StyleGAN2 networks modeled the training distributions. Test images were reconstructed with backpropagation and scored using mean squared error (MSE) and Wasserstein distance (WD). Scores should be high for anomalous images, as StyleGAN2 cannot model unseen attributes. Area under the receiver operating characteristic curve (AUROC) evaluated anomaly detection, i.e., baseline and anomaly dataset differentiation. The proportion of highest-scoring patches containing needles or ascites assessed anomaly localization. Permutation tests determined statistical significance. StyleGAN2 did not reconstruct anomalous attributes (e.g., needles, ascites), enabling the unsupervised detection of these attributes: mean (\u00b1standard deviation) AUROCs were 0.86 (\u00b10.13) for failure detection and 0.82 (\u00b10.11) for data curation. 81% (\u00b113%) of the needles and ascites were localized. WD outperformed MSE on CT (p &lt; 0.001), while MSE outperformed WD on radiography (p &lt; 0.001). Generative models detected anomalous image attributes, demonstrating promise for model failure detection interpretability and large-scale data curation.<\/jats:p>","DOI":"10.3390\/bioengineering12101106","type":"journal-article","created":{"date-parts":[[2025,10,15]],"date-time":"2025-10-15T06:06:32Z","timestamp":1760508392000},"page":"1106","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["Generative Modeling for Interpretable Anomaly Detection in Medical Imaging: Applications in Failure Detection and Data Curation"],"prefix":"10.3390","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0995-3695","authenticated-orcid":false,"given":"McKell E.","family":"Woodland","sequence":"first","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"},{"name":"Departments of Computer Science and Electrical and Computer Engineering, Rice University, Houston, TX 77005, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mais","family":"Altaie","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-2871-6692","authenticated-orcid":false,"given":"Caleb S.","family":"O\u2019Connor","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-3015-3570","authenticated-orcid":false,"given":"Austin H.","family":"Castelo","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Olubunmi C.","family":"Lebimoyo","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"},{"name":"Department of Biology, Texas Southern University, Houston, TX 77004, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aashish C.","family":"Gupta","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3752-5997","authenticated-orcid":false,"given":"Joshua P.","family":"Yung","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Paul E.","family":"Kinahan","sequence":"additional","affiliation":[{"name":"Departments of Radiology and Bioengineering, University of Washington, Seattle, WA 98195, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5264-3994","authenticated-orcid":false,"given":"Clifton D.","family":"Fuller","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7675-3461","authenticated-orcid":false,"given":"Eugene J.","family":"Koay","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1771-8233","authenticated-orcid":false,"given":"Bruno C.","family":"Odisio","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ankit B.","family":"Patel","sequence":"additional","affiliation":[{"name":"Departments of Computer Science and Electrical and Computer Engineering, Rice University, Houston, TX 77005, USA"},{"name":"Department of Neuroscience, Baylor College of Medicine, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9364-5040","authenticated-orcid":false,"given":"Kristy K.","family":"Brock","sequence":"additional","affiliation":[{"name":"Departments of GI Radiation Oncology, Imaging Physics, Interventional Radiology, and Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2025,10,14]]},"reference":[{"key":"ref_1","first-page":"142","article-title":"Evaluating the performance of StyleGAN2-ADA on medical images","volume":"Volume 13570","author":"Zhao","year":"2022","journal-title":"Proceedings of the SASHIMI 2022, Singapore, 18 September 2022"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"5635","DOI":"10.1007\/s11263-024-02117-4","article-title":"Generalized out-of-distribution detection: A survey","volume":"132","author":"Yang","year":"2024","journal-title":"Int. J. Comput. Vis."},{"key":"ref_3","first-page":"e141","article-title":"Deep learning for medical anomaly detection\u2014A survey","volume":"54","author":"Fernando","year":"2021","journal-title":"ACM Comput. Surv."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"497","DOI":"10.1016\/j.neucom.2021.12.093","article-title":"GAN-based anomaly detection: A review","volume":"493","author":"Xia","year":"2022","journal-title":"Neurocomputing"},{"key":"ref_5","unstructured":"Burgos, N., and Svoboda, D. (2022). Unsupervised abnormality detection in medical images with deep generative methods. Biomedical Image Synthesis and Simulation: Methods and Applications, Academic Press."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Zech, J.R., Badgeley, M.A., Liu, M., Costa, A.B., Titano, J.J., and Oermann, E.K. (2018). Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study. PLoS Med., 15.","DOI":"10.1371\/journal.pmed.1002683"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"4","DOI":"10.1148\/radiol.2020192224","article-title":"Preparing medical imaging data for machine learning","volume":"295","author":"Willemink","year":"2020","journal-title":"Radiology"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"100464","DOI":"10.1016\/j.adro.2020.04.023","article-title":"Automated contouring of contrast and noncontrast computed tomography liver images with fully convolutional networks","volume":"6","author":"Anderson","year":"2021","journal-title":"Adv. Radiat. Oncol."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"2006","DOI":"10.59275\/j.melba.2024-g93a","article-title":"Dimensionality reduction and nearest neighbors for improving out-of-distribution detection in medical image segmentation","volume":"2","author":"Woodland","year":"2024","journal-title":"J. Mach. Learn. Biomed. Imaging"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"7","DOI":"10.1016\/j.crad.2019.04.002","article-title":"Open access image repositories: High-quality data to enable machine learning research","volume":"75","author":"Prior","year":"2020","journal-title":"Clin. Radiol."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Johnston, L.R., Curty, R., Braxton, S.M., Carlson, J., Hadley, H., Lafferty-Hess, S., Luong, H., Petters, J.L., and Kozlowski, W.A. (2024). Understanding the value of curation: A survey of US data repository curation practices and perceptions. PLoS ONE, 19.","DOI":"10.1371\/journal.pone.0301171"},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Wang, X., Peng, Y., Lu, L., Lu, Z., Bagheri, M., and Summers, R.M. (2017, January 21\u201326). ChestX-ray8: Hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. Proceedings of the CVPR 2017, Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.369"},{"key":"ref_13","first-page":"442","article-title":"Software-based versus visual assessment of the minimal ablative margin in patients with liver tumours undergoing percutaneous thermal ablation (COVERALL): A randomized phase 2 trial","volume":"10","author":"Odisio","year":"2025","journal-title":"Lancet"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"e064501","DOI":"10.1117\/1.JMI.10.6.064501","article-title":"Sequestration of imaging studies in MIDRC: A multi-institutional data commons","volume":"10","author":"Baughan","year":"2023","journal-title":"J. Med. Imaging"},{"key":"ref_15","unstructured":"Ghahramani, Z., Welling, M., Cortes, C., Lawrence, N.D., and Weinberger, K.Q. (2014). Generative Adversarial Nets. Advances in Neural Information Processing Systems 27, Proceedings of NIPS 2014, Montreal, QC, Canada, 8\u201313 December 2014, Curran Associates, Inc."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Karras, T., Laine, S., Aittala, M., Hellsten, J., Lehtinen, J., and Aila, T. (2020, January 13\u201319). Analyzing and improving the image quality of StyleGAN. Proceedings of the CVPR 2020, Seattle, WA, USA.","DOI":"10.1109\/CVPR42600.2020.00813"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Karras, T., Laine, S., and Aila, T. (2019, January 16\u201320). A style-based generator architecture for generative adversarial networks. Proceedings of the CVPR 2019, Long Beach, CA, USA.","DOI":"10.1109\/CVPR.2019.00453"},{"key":"ref_18","unstructured":"Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M.F., and Lin, H. (2020). Training generative adversarial networks with limited data. Advances in Neural Information Processing Systems 33, Proceedings of NeurIPS 2020, Online, 6\u201312 December 2020, Curran Associates, Inc."},{"key":"ref_19","unstructured":"Guyon, I., Von Luxburg, U., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., and Garnett, R. (2017). GANs trained by a two time-scale update rule converge to a local Nash equilibrium. Advances in Neural Information Processing Systems 30, Proceedings of NIPS 2017, Long Beach, CA, USA, 4\u20139 December 2017, Curran Associates, Inc."},{"key":"ref_20","unstructured":"Morozov, S., Voynov, A., and Babenko, A. (2021, January 3\u20137). On self-supervised image representations for GAN evaluation. Proceedings of the ICLR 2021, Online."},{"key":"ref_21","unstructured":"Linguraru, M.G., Dou, Q., Feragen, A., Giannarou, S., Glocker, B., Lekadir, K., and Schnabel, J.A. (2024). Towards learning contrast kinetics with multi-condition latent diffusion models. LNCS, Vol. 15005, Proceedings of MICCAI 2024, Marrakesh, MA, Morocco, 6\u201310 October 2024, Springer."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1016\/j.cviu.2018.10.009","article-title":"Pros and cons of GAN evaluation measures","volume":"179","author":"Borji","year":"2018","journal-title":"Comput. Vis. Image Underst."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., and Fei-Fei, L. (2009, January 20\u201325). ImageNet: A large-scale hierarchical image database. Proceedings of the CVPR 2009, Miami Beach, FL, USA.","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2015, January 7\u201312). Going deeper with convolutions. Proceedings of the CVPR 2015, Boston, MA, USA.","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"ref_25","unstructured":"Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M.F., and Lin, H. (2020). Unsupervised learning of visual features by contrasting cluster assignments. Advances in Neural Information Processing Systems 33, Proceedings of NeurIPS 2020, Online, 6\u201312 December 2020, Curran Associates, Inc."},{"key":"ref_26","unstructured":"Linguraru, M.G., Dou, Q., Feragen, A., Giannarou, S., Glocker, B., Lekadir, K., and Schnabel, J.A. (2024). Feature extraction for generative medical imaging evaluation: New evidence against an evolving trend. LNCS, Vol. 15012, Proceedings of MICCAI 2024, Marrakesh, Morocco, 6\u201310 October 2024, Springer."},{"key":"ref_27","unstructured":"Niethammer, M., Styner, M., Aylward, S., Zhu, H., Oguz, I., Yap, P.-T., and Shen, D. (2017). Unsupervised anomaly detection with generative adversarial networks to guide marker discovery. LNCS, Vol. 10265, Proceedings of IPMI 2017, Boone, NC, USA, 25\u201330 June 2017, Springer."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Gong, D., Liu, L., Le, V., Saha, B., Mansour, M.R., Venkatesh, S., and van den Hengel, A. (November, January 27). Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detection. Proceedings of the ICCV 2019, Seoul, Korea.","DOI":"10.1109\/ICCV.2019.00179"},{"key":"ref_29","unstructured":"Woodland, M., and Brock, K. (2025, February 20). Generative Modeling for Interpretable Failure Detection in Liver CT Segmentation and Scalable Data Curation of Chest Radiographs (Version v1). Zenodo 2025. Available online: https:\/\/zenodo.org\/records\/14901472."},{"key":"ref_30","first-page":"1050","article-title":"Dropout as a Bayesian approximation: Representing model uncertainty in deep learning","volume":"Volume 48","author":"Balcan","year":"2016","journal-title":"Proceedings of the 33rd International Conference on Machine Learning (ICML 2016), New York, NY, USA, 19\u201324 June 2016"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Nguyen, A., Yosinski, J., and Clune, J. (2015, January 7\u201312). Deep neural networks are easily fooled: High confidence predictions for unrecognizable images. Proceedings of the CVPR 2015, Boston, MA, USA.","DOI":"10.1109\/CVPR.2015.7298640"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"418","DOI":"10.1007\/s10278-020-00413-2","article-title":"Unsupervised deep anomaly detection in chest radiographs","volume":"32","author":"Nakao","year":"2021","journal-title":"J. Digit. Imaging"},{"key":"ref_33","first-page":"372","article-title":"Improving unsupervised defect segmentation by applying structural similarity to autoencoders","volume":"Volume 5","author":"Tremeau","year":"2019","journal-title":"Proceedings of the VISIGRAPP 2019"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Yan, Z., Fang, Q., Lv, W., and Su, Q. (2024). AnomalySD: Few-shot multi-class anomaly detection with stable diffusion model. arXiv.","DOI":"10.2139\/ssrn.5266814"},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Cheng, X., Wen, M., Gao, C., and Wang, Y. (2022). Hyperspectral anomaly detection based on Wasserstein distance and spatial filtering. Remote Sens, 14.","DOI":"10.3390\/rs14122730"}],"container-title":["Bioengineering"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2306-5354\/12\/10\/1106\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,16]],"date-time":"2025-10-16T04:10:49Z","timestamp":1760587849000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2306-5354\/12\/10\/1106"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,14]]},"references-count":35,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2025,10]]}},"alternative-id":["bioengineering12101106"],"URL":"https:\/\/doi.org\/10.3390\/bioengineering12101106","relation":{},"ISSN":["2306-5354"],"issn-type":[{"value":"2306-5354","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,14]]}}}