{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T23:22:17Z","timestamp":1780356137522,"version":"3.54.1"},"reference-count":38,"publisher":"Springer Science and Business Media LLC","issue":"2","license":[{"start":{"date-parts":[[2025,6,2]],"date-time":"2025-06-02T00:00:00Z","timestamp":1748822400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,6,2]],"date-time":"2025-06-02T00:00:00Z","timestamp":1748822400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Digit Imaging. Inform. med."],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>This study addresses the critical challenge of detecting out-of-distribution (OOD) chest X-rays, where subtle view differences between lateral and frontal radiographs can lead to diagnostic errors. We develop a GAN-based framework that learns the inherent feature distribution of frontal views from the MIMIC-CXR dataset through latent space optimization and Kolmogorov\u2013Smirnov statistical testing. Our approach generates similarity scores to reliably identify OOD cases, achieving exceptional performance with 100% precision, and 97.5% accuracy in detecting lateral views. The method demonstrates consistent reliability across operating conditions, maintaining accuracy above 92.5% and precision exceeding 93% under varying detection thresholds. These results provide both theoretical insights and practical solutions for OOD detection in medical imaging, demonstrating how GANs can establish feature representations for identifying distributional shifts. By significantly improving model reliability when encountering view-based anomalies, our framework enhances the clinical applicability of deep learning systems, ultimately contributing to improved diagnostic safety and patient outcomes.<\/jats:p>","DOI":"10.1007\/s10278-025-01559-7","type":"journal-article","created":{"date-parts":[[2025,6,2]],"date-time":"2025-06-02T13:11:15Z","timestamp":1748869875000},"page":"1774-1782","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Robust Detection of Out-of-Distribution Shifts in Chest X-ray Imaging"],"prefix":"10.1007","volume":"39","author":[{"given":"Fatemeh","family":"Karimi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Farzan","family":"Farnia","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kyongtae Tyler","family":"Bae","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,6,2]]},"reference":[{"key":"1559_CR1","unstructured":"Subbaswamy A, Schulam P, Saria S, editors. Preventing failures due to dataset shift: Learning predictive models that transport. The 22nd International Conference on Artificial Intelligence and Statistics: PMLR; 2019; p. 3118\u20133127."},{"key":"1559_CR2","unstructured":"Ziou D, Tabbone S. Edge detection techniques-an overview. Pac\u043fo\u0437\u043da\u0432a\u043d\u0438e o\u0431pa\u0437o\u0432 \u0438 a\u043da\u043b\u0438\u0437 \u0438\u0437o\u0431pa\u0436e\u043d\/Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications. 1998;8(4):537-559."},{"key":"1559_CR3","doi-asserted-by":"crossref","unstructured":"Pass G, Zabih R, editors. Histogram refinement for content-based image retrieval. Proceedings third IEEE workshop on applications of computer vision WACV\u201996: IEEE; 1996; p. 96\u2013102.","DOI":"10.1109\/ACV.1996.572008"},{"key":"1559_CR4","unstructured":"Hendrycks D, Gimpel K. A baseline for detecting misclassified and out-of-distribution examples in neural networks. arXiv preprint arXiv:161002136. 2016."},{"key":"1559_CR5","unstructured":"Liang S, Li Y, Srikant R. Enhancing the reliability of out-of-distribution image detection in neural networks. arXiv preprint arXiv:170602690. 2017."},{"key":"1559_CR6","unstructured":"Hendrycks D, Mazeika M, Dietterich T. Deep anomaly detection with outlier exposure. arXiv preprint arXiv:181204606. 2018."},{"key":"1559_CR7","unstructured":"Lee K, Lee H, Lee K, Shin J. Training confidence-calibrated classifiers for detecting out-of-distribution samples. arXiv preprint arXiv:171109325. 2017."},{"key":"1559_CR8","unstructured":"DeVries T, Taylor GW. Learning confidence for out-of-distribution detection in neural networks. arXiv preprint arXiv:180204865. 2018."},{"key":"1559_CR9","unstructured":"Ruff L, Vandermeulen R, Goernitz N, Deecke L, Siddiqui SA, Binder A, et al., editors. Deep one-class classification. International conference on machine learning: PMLR; 2018; p. 4393\u20134402."},{"key":"1559_CR10","unstructured":"Agustsson E, Theis L. Universally quantized neural compression. Advances in neural information processing systems. 2020;33:12367-12376."},{"key":"1559_CR11","unstructured":"Gal Y, Ghahramani Z, editors. Dropout as a bayesian approximation: Representing model uncertainty in deep learning. international conference on machine learning: PMLR; 2016; p. 1050\u20131059."},{"key":"1559_CR12","unstructured":"Kingma DP, Ba J. Adam: A method for stochastic optimization. arXiv preprint arXiv:14126980. 2014."},{"key":"1559_CR13","unstructured":"Lee K, Lee K, Lee H, Shin J. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. Advances in neural information processing systems. 2018;31."},{"key":"1559_CR14","unstructured":"Sun Y, Ming Y, Zhu X, Li Y, editors. Out-of-distribution detection with deep nearest neighbors. International Conference on Machine Learning: PMLR; 2022; p. 20827\u201320840."},{"key":"1559_CR15","doi-asserted-by":"crossref","unstructured":"Chen W, Feng C, Ke A, Xie X, Zhou SK, editors. Out-of-distribution detection for learning-based chest x-ray diagnosis. ICASSP 2024\u20132024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP): IEEE; 2024; p. 1816\u20131820.","DOI":"10.1109\/ICASSP48485.2024.10447414"},{"key":"1559_CR16","doi-asserted-by":"crossref","unstructured":"Zamzmi G, Venkatesh K, Nelson B, Prathapan S, Yi P, Sahiner B, Delfino JG. Out-of-Distribution Detection and Radiological Data Monitoring Using Statistical Process Control. Journal of Imaging Informatics in Medicine. 2024:1\u201319.","DOI":"10.1007\/s10278-024-01212-9"},{"key":"1559_CR17","doi-asserted-by":"crossref","unstructured":"Wollek A, Willem T, Ingrisch M, Sabel B, Lasser T. Out\u2010of\u2010distribution detection with in\u2010distribution voting using the medical example of chest x\u2010ray classification. Medical Physics. 2024;51(4):2721-2732.","DOI":"10.1002\/mp.16790"},{"key":"1559_CR18","unstructured":"\u00c7all\u0131 E, Murphy K, Sogancioglu E, Van Ginneken B. FRODO: Free rejection of out-of-distribution samples: application to chest x-ray analysis. arXiv preprint arXiv:190701253. 2019."},{"key":"1559_CR19","doi-asserted-by":"crossref","unstructured":"Zhang Q, Chen X, He Z, Wang K, Wu L, Shen H, Sun J. U2AD: Uncertainty-based Unsupervised Anomaly Detection Framework for Detecting T2 Hyperintensity in MRI Spinal Cord. arXiv preprint arXiv:250313400. 2025.","DOI":"10.1016\/j.media.2026.103939"},{"key":"1559_CR20","doi-asserted-by":"crossref","unstructured":"Xu J, Zhu A, Lin J, Ke Q, Chen C. Skeleton-OOD: An End-to-End Skeleton-Based Model for Robust Out-of-Distribution Human Action Detection. arXiv preprint arXiv:240520633. 2024.","DOI":"10.1016\/j.neucom.2024.129158"},{"key":"1559_CR21","doi-asserted-by":"crossref","unstructured":"Dionelis N, Yaghoobi M, Tsaftaris SA, editors. Tail of distribution GAN (TailGAN): GenerativeAdversarial-network-based boundary formation. 2020 sensor signal processing for defence conference (SSPD): IEEE; 2020; p. 1\u20135.","DOI":"10.1109\/SSPD47486.2020.9272100"},{"key":"1559_CR22","doi-asserted-by":"crossref","unstructured":"Marek P, Naik VI, Auvray V, Goyal A. Oodgan: Generative adversarial network for out-of-domain data generation. arXiv preprint arXiv:210402484. 2021.","DOI":"10.18653\/v1\/2021.naacl-industry.30"},{"key":"1559_CR23","doi-asserted-by":"crossref","unstructured":"Azizmalayeri M, Soltani Moakhar A, Zarei A, Zohrabi R, Manzuri M, Rohban MH. Your out-of-distribution detection method is not robust! Advances in Neural Information Processing Systems. 2022;35:4887-4901.","DOI":"10.52202\/068431-0353"},{"key":"1559_CR24","doi-asserted-by":"crossref","unstructured":"Kwon H-J, Shin D-H, Chung K. PGGAN-based anomaly classification on chest x-ray using weighted multi-scale similarity. IEEE Access. 2021;9:113315-113325.","DOI":"10.1109\/ACCESS.2021.3102954"},{"key":"1559_CR25","unstructured":"Woodland M, Wood J, O\u2019Connor C, Patel AB, Brock KK. StyleGAN2-based Out-of-Distribution Detection for Medical Imaging. arXiv preprint arXiv:230710193. 2023."},{"key":"1559_CR26","doi-asserted-by":"crossref","unstructured":"Caetano F, Viviers C, Filatova L, de With PH, van der Sommen F. AdverX-Ray: Ensuring X-Ray Integrity Through Frequency-Sensitive Adversarial VAEs. arXiv preprint arXiv:250216610. 2025.","DOI":"10.1117\/12.3045893"},{"key":"1559_CR27","doi-asserted-by":"crossref","unstructured":"Massey Jr FJ. The Kolmogorov-Smirnov test for goodness of fit. Journal of the American statistical Association. 1951;46(253):68-78.","DOI":"10.1080\/01621459.1951.10500769"},{"key":"1559_CR28","doi-asserted-by":"crossref","unstructured":"Cheng KC, Miller EL, Hughes MC, Aeron S. On matched filtering for statistical change point detection. IEEE Open Journal of Signal Processing. 2020;1:159-176.","DOI":"10.1109\/OJSP.2020.3035070"},{"key":"1559_CR29","unstructured":"Jiang D, Sun S, Yu Y, editors. Revisiting flow generative models for out-of-distribution detection. International conference on learning representations; 2021."},{"key":"1559_CR30","doi-asserted-by":"crossref","unstructured":"Johnson AE, Pollard TJ, Berkowitz SJ, Greenbaum NR, Lungren MP, Deng C-y, et al. MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports. Scientific data. 2019;6(1):317.","DOI":"10.1038\/s41597-019-0322-0"},{"key":"1559_CR31","unstructured":"Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, et al. Generative adversarial nets. Advances in neural information processing systems. 2014;27."},{"key":"1559_CR32","unstructured":"Dumoulin V, Visin F. A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:160307285. 2016."},{"key":"1559_CR33","unstructured":"Ioffe S, Szegedy C, editors. Batch normalization: Accelerating deep network training by reducing internal covariate shift. International conference on machine learning: pmlr; 2015; p. 448\u2013456."},{"key":"1559_CR34","unstructured":"Nair V, Hinton GE, editors. Rectified linear units improve restricted boltzmann machines. Proceedings of the 27th international conference on machine learning (ICML-10); 2010; p. 807\u2013814."},{"key":"1559_CR35","doi-asserted-by":"crossref","unstructured":"LeCun Y, Bengio Y, Hinton G. Deep learning. nature. 2015;521(7553):436-444.","DOI":"10.1038\/nature14539"},{"key":"1559_CR36","unstructured":"Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems. 2012;25."},{"key":"1559_CR37","unstructured":"Maas AL, Hannun AY, Ng AY, editors. Rectifier nonlinearities improve neural network acoustic models. Proc icml: Atlanta, GA; 2013; p. 3."},{"key":"1559_CR38","doi-asserted-by":"crossref","unstructured":"Rumelhart DE, Hinton GE, Williams RJ. Learning representations by back-propagating errors. nature. 1986;323(6088):533-536.","DOI":"10.1038\/323533a0"}],"container-title":["Journal of Imaging Informatics in Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10278-025-01559-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10278-025-01559-7","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10278-025-01559-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,22]],"date-time":"2026-04-22T16:23:18Z","timestamp":1776874998000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10278-025-01559-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,2]]},"references-count":38,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,4]]}},"alternative-id":["1559"],"URL":"https:\/\/doi.org\/10.1007\/s10278-025-01559-7","relation":{},"ISSN":["2948-2933"],"issn-type":[{"value":"2948-2933","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6,2]]},"assertion":[{"value":"13 January 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 May 2025","order":2,"name":"revised","label":"Revised","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 May 2025","order":3,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"2 June 2025","order":4,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study used the MIMIC dataset, which is publicly available and fully de-identified. In accordance with the University of Hong Kong\u2019s guidelines, no ethical approval was required for this research.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics Approval"}},{"value":"Since the study relied on the de-identified and anonymized data from the MIMIC dataset, informed consent was not necessary according to the guidelines governing the secondary use of publicly available datasets.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent to Participate"}},{"value":"The chest radiographs analyzed in this study were sourced from the publicly available and de-identified MIMIC dataset, as cited in [\n                      \n                      ]. In line with the dataset\u2019s usage policy, all data were anonymized to ensure participant confidentiality and adherence to ethical standards.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for Publication"}},{"value":"The authors declare no competing interests.","order":5,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}]}}