{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T16:05:32Z","timestamp":1784563532402,"version":"3.55.0"},"reference-count":33,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T00:00:00Z","timestamp":1784505600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T00:00:00Z","timestamp":1784505600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/100019559","name":"Carl von Ossietzky Universit\u00e4t Oldenburg","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100019559","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Artificial intelligence (AI) systems for chest X-ray (CXR) interpretation can achieve high standalone diagnostic accuracy, yet limited transparency raises concerns about subgroup disparities. We audited sex-stratified performance and attention patterns in CNN-based pneumonia detection using the RSNA Pneumonia Detection Challenge dataset (26,684 frontal adult CXRs). ImageNet-pretrained InceptionV3 models were trained across five prespecified random seeds and combined as a seed-ensemble. Analyses used a prespecified, fixed, patient-level sex- and label-balanced test set (n\u2009=\u20091,000; 250 pneumonia-positive and 250 pneumonia-negative per sex) and a single validation-derived operating point held constant across sex. Baseline performance was robust (AUROC 0.859 [0.836, 0.881], Brier 0.155 [0.141, 0.169]). Males showed higher discrimination than females (AUROC 0.882 vs 0.837, gap\u2009+\u20090.045 [+\u20090.002,\u2009+\u20090.091]) and a lower false-positive rate (FPR gap\u2009\u2212\u20090.073 [\u2212\u20090.145,\u2009\u2212\u20090.006]). To contextualize disparities, we quantified Grad-CAM attention relative to lung segmentation masks; in-lung activation was lower in females (60.8% vs 64.7%, M\u2009\u2212\u2009F\u2009+\u20093.9\u00a0pp [+\u20091.6,\u2009+\u20096.4]). A combined mitigation (group-balanced sampling\u2009+\u2009adversarial debiasing) preserved performance (AUROC 0.864 [0.843, 0.885]) while reducing AUROC\/specificity gaps and attenuating the false-positive disparity. Quantitative explainability can complement fairness auditing and mitigation in medical imaging.<\/jats:p>","DOI":"10.1038\/s41598-026-62761-3","type":"journal-article","created":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T15:05:22Z","timestamp":1784559922000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Explainable AI for chest radiographs: sex-stratified fairness auditing in CNN-based pneumonia detection"],"prefix":"10.1038","volume":"16","author":[{"given":"Matteo","family":"Haupt","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Elena","family":"Jolkver","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anne","family":"Schwerk","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,7,20]]},"reference":[{"key":"62761_CR1","doi-asserted-by":"publisher","DOI":"10.3390\/jcm13164659","volume":"13","author":"J Kufel","year":"2024","unstructured":"Kufel, J. et al. Measurement of cardiothoracic ratio on chest X-rays using artificial intelligence\u2014A systematic review and meta-analysis. J. Clin. Med. 13, 4659 (2024).","journal-title":"J. Clin. Med."},{"key":"62761_CR2","doi-asserted-by":"publisher","DOI":"10.1016\/j.asoc.2022.109319","volume":"126","author":"D Meedeniya","year":"2022","unstructured":"Meedeniya, D. et al. Chest X-ray analysis empowered with deep learning: A systematic review. Appl. Soft Comput. 126, 109319 (2022).","journal-title":"Appl. Soft Comput."},{"key":"62761_CR3","doi-asserted-by":"publisher","first-page":"545","DOI":"10.1378\/chest.10-1302","volume":"141","author":"S Raoof","year":"2012","unstructured":"Raoof, S. et al. Interpretation of plain chest roentgenogram. Chest 141, 545\u2013558 (2012).","journal-title":"Chest"},{"key":"62761_CR4","doi-asserted-by":"publisher","first-page":"650","DOI":"10.1016\/j.chest.2022.10.039","volume":"163","author":"WB Gefter","year":"2023","unstructured":"Gefter, W. B., Post, B. A. & Hatabu, H. Commonly missed findings on chest radiographs: causes and consequences. Chest 163, 650\u2013661 (2023).","journal-title":"Chest"},{"key":"62761_CR5","first-page":"31","volume":"2","author":"A Kareem","year":"2022","unstructured":"Kareem, A., Liu, H. & Sant, P. Review on Pneumonia Image Detection: A Machine Learning Approach. Hum.-Centric Intell. Syst. 2, 31\u201343 (2022).","journal-title":"Syst."},{"key":"62761_CR6","doi-asserted-by":"publisher","DOI":"10.3390\/jimaging10100239","volume":"10","author":"D Bhati","year":"2024","unstructured":"Bhati, D., Neha, F. & Amiruzzaman, M. A survey on explainable artificial intelligence (XAI) techniques for visualizing deep learning models in medical imaging. J. Imaging 10, 239 (2024).","journal-title":"J. Imaging"},{"key":"62761_CR7","doi-asserted-by":"publisher","DOI":"10.1117\/1.JMI.9.6.061102","volume":"9","author":"EAM Stanley","year":"2022","unstructured":"Stanley, E. A. M., Wilms, M., Mouches, P. & Forkert, N. D. Fairness-related performance and explainability effects in deep learning models for brain image analysis. J. Med. Imaging 9, 061102 (2022).","journal-title":"J. Med. Imaging"},{"key":"62761_CR8","doi-asserted-by":"publisher","first-page":"4431","DOI":"10.3390\/jcm10194431","volume":"10","author":"CY Yang","year":"2021","unstructured":"Yang, C. Y. et al. Using Deep Neural Networks for Predicting Age and Sex in Healthy Adult Chest Radiographs. J. Clin. Med. 10, 4431 (2021).","journal-title":"J. Clin. Med."},{"key":"62761_CR9","doi-asserted-by":"publisher","DOI":"10.1016\/j.ejrad.2023.110786","volume":"162","author":"K Borys","year":"2023","unstructured":"Borys, K. et al. Explainable AI in medical imaging: An overview for clinical practitioners - Beyond saliency-based XAI approaches. Eur. J. Radiol. 162, 110786 (2023).","journal-title":"Eur. J. Radiol."},{"key":"62761_CR10","doi-asserted-by":"publisher","first-page":"12","DOI":"10.3390\/make7010012","volume":"7","author":"M Ennab","year":"2025","unstructured":"Ennab, M. & Mcheick, H. (2025) Advancing AI Interpretability in Medical Imaging: A Comparative Analysis of Pixel-Level Interpretability and Grad-CAM Models. Mach. Learn. Knowl. Extr. 7, 12 (2025).","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"62761_CR11","doi-asserted-by":"publisher","DOI":"10.3390\/s25134166","volume":"25","author":"Z Cheng","year":"2025","unstructured":"Cheng, Z., Wu, Y., Li, Y., Cai, L. & Ihnaini, B. A comprehensive review of explainable artificial intelligence (XAI) in computer vision. Sensors 25, 4166 (2025).","journal-title":"Sensors"},{"key":"62761_CR12","doi-asserted-by":"publisher","first-page":"1456486","DOI":"10.3389\/frai.2024.1456486","volume":"7","author":"J Kim","year":"2024","unstructured":"Kim, J., Maathuis, H. & Sent, D. Human-centered evaluation of explainable AI applications: a systematic review. Front. Artif. Intell. 7, 1456486 (2024).","journal-title":"Front. Artif. Intell."},{"key":"62761_CR13","doi-asserted-by":"publisher","DOI":"10.1145\/3194770.3194776","author":"S Verma","year":"2018","unstructured":"Verma, S. & Rubin, J. Fairness definitions explained. Proc. - Int. Conf. Softw. Eng. https:\/\/doi.org\/10.1145\/3194770.3194776 (2018).","journal-title":"Proc. - Int. Conf. Softw. Eng."},{"key":"62761_CR14","doi-asserted-by":"publisher","first-page":"e745","DOI":"10.1016\/S2589-7500(21)00208-9","volume":"3","author":"M Ghassemi","year":"2021","unstructured":"Ghassemi, M., Oakden-Rayner, L. & Beam, A. L. The false hope of current approaches to explainable artificial intelligence in health care. Lancet Digit Health 3, e745\u2013e750 (2021).","journal-title":"Lancet Digit Health"},{"key":"62761_CR15","doi-asserted-by":"publisher","first-page":"12592","DOI":"10.1073\/pnas.1919012117","volume":"117","author":"AJ Larrazabal","year":"2020","unstructured":"Larrazabal, A. J., Nieto, N., Peterson, V., Milone, D. H. & Ferrante, E. Gender imbalance in medical imaging datasets produces biased classifiers for computer-aided diagnosis. Proc. Natl. Acad. Sci. U. S. A. 117, 12592\u201312594 (2020).","journal-title":"Proc. Natl. Acad. Sci. U. S. A."},{"key":"62761_CR16","doi-asserted-by":"publisher","first-page":"2176","DOI":"10.1038\/s41591-021-01595-0","volume":"27","author":"L Seyyed-Kalantari","year":"2021","unstructured":"Seyyed-Kalantari, L., Zhang, H., McDermott, M. B. A., Chen, I. Y. & Ghassemi, M. Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations. Nat. Med. 27, 2176\u20132182 (2021).","journal-title":"Nat. Med."},{"key":"62761_CR17","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.230060","volume":"5","author":"B Glocker","year":"2023","unstructured":"Glocker, B., Jones, C., Roschewitz, M. & Winzeck, S. Risk of bias in chest radiography deep learning foundation models. Radiol. Artif. Intell. 5, e230060 (2023).","journal-title":"Radiol. Artif. Intell."},{"key":"62761_CR18","doi-asserted-by":"publisher","first-page":"1575","DOI":"10.1007\/s11739-022-02999-7","volume":"17","author":"B Corica","year":"2022","unstructured":"Corica, B., Tartaglia, F., D\u2019Amico, T., Romiti, G. F. & Cangemi, R. Sex and gender differences in community-acquired pneumonia. Intern. Emerg. Med. 17, 1575\u20131588 (2022).","journal-title":"Intern. Emerg. Med."},{"key":"62761_CR19","doi-asserted-by":"publisher","first-page":"70","DOI":"10.1038\/s41746-020-0273-z","volume":"3","author":"Y-X Tang","year":"2020","unstructured":"Tang, Y.-X. et al. Automated abnormality classification of chest radiographs using deep convolutional neural networks. Npj Digit. Med. 3, 70 (2020).","journal-title":"Npj Digit. Med."},{"key":"62761_CR20","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-019-42294-8","volume":"9","author":"IM Baltruschat","year":"2019","unstructured":"Baltruschat, I. M., Nickisch, H., Grass, M., Knopp, T. & Saalbach, A. Comparison of deep learning approaches for multi-label chest X-ray classification. Sci. Rep. 9, 6381 (2019).","journal-title":"Sci. Rep."},{"key":"62761_CR21","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pmed.1002683","volume":"15","author":"JR Zech","year":"2018","unstructured":"Zech, J. R. et al. Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study. PLoS Med. 15, e1002683 (2018).","journal-title":"PLoS Med."},{"key":"62761_CR22","doi-asserted-by":"publisher","unstructured":"Selvaraju, R. R. et al. Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization. in 2017 IEEE International Conference on Computer Vision (ICCV) 618\u2013626 (2017). https:\/\/doi.org\/10.1109\/ICCV.2017.74.","DOI":"10.1109\/ICCV.2017.74"},{"key":"62761_CR23","doi-asserted-by":"publisher","first-page":"867","DOI":"10.1038\/s42256-022-00536-x","volume":"4","author":"A Saporta","year":"2022","unstructured":"Saporta, A. et al. Benchmarking saliency methods for chest X-ray interpretation. Nat. Mach. Intell. 4, 867\u2013878 (2022).","journal-title":"Nat. Mach. Intell."},{"key":"62761_CR24","unstructured":"Adebayo, J. et al. Sanity Checks for Saliency Maps. https:\/\/arxiv.org\/abs\/1810.03292v3 (2018)."},{"key":"62761_CR25","doi-asserted-by":"crossref","unstructured":"Wang, X. et al. ChestX-ray8: Hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. Proc. - 30th IEEE Conf. Comput. Vis. Pattern Recognit. CVPR 2017 2017-January, 3462\u20133471 (2017).","DOI":"10.1109\/CVPR.2017.369"},{"key":"62761_CR26","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s41073-016-0007-6","volume":"1","author":"S Heidari","year":"2016","unstructured":"Heidari, S., Babor, T. F., Castro, P. D., Tort, S. & Curno, M. Sex and gender equity in research: Rationale for the SAGER guidelines and recommended use. Res. Integr. Peer Rev. 1, 1\u20139 (2016).","journal-title":"Res. Integr. Peer Rev."},{"key":"62761_CR27","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. & Wojna, Z. Rethinking the Inception Architecture for Computer Vision. Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit. 2016-December, 2818\u20132826 (2016).","DOI":"10.1109\/CVPR.2016.308"},{"key":"62761_CR28","unstructured":"Kingma, D. P. & Ba, J. L. Adam: A Method for Stochastic Optimization. 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf. Track Proc. https:\/\/arxiv.org\/abs\/1412.6980v9 (2014)."},{"key":"62761_CR29","doi-asserted-by":"publisher","first-page":"458","DOI":"10.1002\/bimj.200410135","volume":"47","author":"R Fluss","year":"2005","unstructured":"Fluss, R., Faraggi, D. & Reiser, B. Estimation of the Youden index and its associated cutoff point. Biom. J. Biom. Z. 47, 458\u2013472 (2005).","journal-title":"Biom. J. Biom. Z."},{"key":"62761_CR30","doi-asserted-by":"publisher","first-page":"419","DOI":"10.1002\/bimj.200710415","volume":"50","author":"MD Ruopp","year":"2008","unstructured":"Ruopp, M. D., Perkins, N. J., Whitcomb, B. W. & Schisterman, E. F. Youden index and optimal cut-point estimated from observations affected by a lower limit of detection. Biom. J. Biom. Z. 50, 419\u2013430 (2008).","journal-title":"Biom. J. Biom. Z."},{"key":"62761_CR31","doi-asserted-by":"publisher","unstructured":"Cohen, J. P. et al. TorchXRayVision: A library of chest X-ray datasets and models. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2111.00595 (2021).","DOI":"10.48550\/arXiv.2111.00595"},{"key":"62761_CR32","doi-asserted-by":"publisher","unstructured":"Lian, J. et al. A Structure-Aware Relation Network for Thoracic Diseases Detection and Segmentation. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2104.10326 (2021).","DOI":"10.48550\/arXiv.2104.10326"},{"key":"62761_CR33","doi-asserted-by":"crossref","unstructured":"Ronneberger, O., Fischer, P. & Brox, T.(2015) U-Net: Convolutional Networks for Biomedical Image Segmentation. Lect. Notes Comput. Sci. Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinforma. 9351, 234\u2013241.","DOI":"10.1007\/978-3-319-24574-4_28"}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-62761-3","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-62761-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-62761-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T15:30:34Z","timestamp":1784561434000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-62761-3"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,20]]},"references-count":33,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["62761"],"URL":"https:\/\/doi.org\/10.1038\/s41598-026-62761-3","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,20]]},"assertion":[{"value":"5 March 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 July 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 July 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study is a secondary analysis of a publicly available, fully de-identified chest radiograph dataset. All experiments were performed using the RSNA Pneumonia Detection Challenge dataset, which was anonymized prior to public release and contains no directly identifiable patient information. No additional institutional review board approval or informed consent was required. The study was conducted in accordance with the Declaration of Helsinki.","order":2,"name":"Ethics","label":"Ethical approval","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"22709"}}