{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T00:26:37Z","timestamp":1771547197648,"version":"3.50.1"},"reference-count":46,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,5,7]],"date-time":"2025-05-07T00:00:00Z","timestamp":1746576000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,5,7]],"date-time":"2025-05-07T00:00:00Z","timestamp":1746576000000},"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":["npj Digit. Med."],"abstract":"<jats:title>Abstract<\/jats:title>\n          <jats:p>This study proposes a semi-weakly supervised learning approach for pulmonary embolism (PE) detection on CT pulmonary angiography (CTPA) to alleviate the resource-intensive burden of exhaustive medical image annotation. Attention-based CNN-RNN models were trained on the RSNA pulmonary embolism CT dataset and externally validated on a pooled dataset (Aida and FUMPE). Three configurations included weak (examination-level labels only), strong (all examination and slice-level labels), and semi-weak (examination-level labels plus a limited subset of slice-level labels). The proportion of slice-level labels varying from 0 to 100%. Notably, semi-weakly supervised models using approximately one-quarter of the total slice-level labels achieved an AUC of 0.928, closely matching the strongly supervised model\u2019s AUC of 0.932. External validation yielded AUCs of 0.999 for the semi-weak and 1.000 for the strong model. By reducing labeling requirements without sacrificing diagnostic accuracy, this method streamlines model development, accelerates the integration of models into clinical practice, and enhances patient care.<\/jats:p>","DOI":"10.1038\/s41746-025-01594-2","type":"journal-article","created":{"date-parts":[[2025,5,7]],"date-time":"2025-05-07T05:30:06Z","timestamp":1746595806000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["High performance with fewer labels using semi-weakly supervised learning for pulmonary embolism diagnosis"],"prefix":"10.1038","volume":"8","author":[{"given":"Zixuan","family":"Hu","sequence":"first","affiliation":[]},{"given":"Hui Ming","family":"Lin","sequence":"additional","affiliation":[]},{"given":"Shobhit","family":"Mathur","sequence":"additional","affiliation":[]},{"given":"Robert","family":"Moreland","sequence":"additional","affiliation":[]},{"given":"Christopher D.","family":"Witiw","sequence":"additional","affiliation":[]},{"given":"Laura","family":"Jimenez-Juan","sequence":"additional","affiliation":[]},{"given":"Matias F.","family":"Callejas","sequence":"additional","affiliation":[]},{"given":"Djeven P.","family":"Deva","sequence":"additional","affiliation":[]},{"given":"Ervin","family":"Sejdi\u0107","sequence":"additional","affiliation":[]},{"given":"Errol","family":"Colak","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2025,5,7]]},"reference":[{"key":"1594_CR1","doi-asserted-by":"publisher","first-page":"e230860","DOI":"10.1148\/radiol.230860","volume":"309","author":"S Bennani","year":"2023","unstructured":"Bennani, S., et al. Using AI to improve radiologist performance in detection of abnormalities on chest radiographs. Radiology 309, e230860 (2023).","journal-title":"Radiology"},{"key":"1594_CR2","doi-asserted-by":"publisher","first-page":"505","DOI":"10.3348\/kjr.2021.0476","volume":"23","author":"JH Lee","year":"2022","unstructured":"Lee, J. H. et al. Improving the performance of radiologists using artificial intelligence-based detection support software for mammography: a multi-reader study. Korean J. Radiol. 23, 505\u2013516 (2022).","journal-title":"Korean J. Radiol."},{"key":"1594_CR3","doi-asserted-by":"publisher","first-page":"692","DOI":"10.1148\/radiol.2021204021","volume":"301","author":"DK Eng","year":"2021","unstructured":"Eng, D. K. et al. Artificial intelligence algorithm improves radiologist performance in skeletal age assessment: a prospective multicenter randomized controlled trial. Radiology 301, 692\u2013699 (2021).","journal-title":"Radiology"},{"key":"1594_CR4","doi-asserted-by":"publisher","first-page":"607","DOI":"10.1016\/j.crad.2021.03.021","volume":"76","author":"MDBS Tam","year":"2021","unstructured":"Tam, M. D. B. S. et al. Augmenting lung cancer diagnosis on chest radiographs: positioning artificial intelligence to improve radiologist performance. Clin. Radiol. 76, 607\u2013614 (2021).","journal-title":"Clin. Radiol."},{"key":"1594_CR5","doi-asserted-by":"publisher","first-page":"e220163","DOI":"10.1148\/ryct.220163","volume":"5","author":"L Topff","year":"2023","unstructured":"Topff, L., et al. Artificial intelligence tool for detection and worklist prioritization reduces time to diagnosis of incidental pulmonary embolism at CT. Radiol. Cardiothorac. Imaging 5, e220163 (2023).","journal-title":"Radiol. Cardiothorac. Imaging"},{"key":"1594_CR6","doi-asserted-by":"publisher","first-page":"923","DOI":"10.1148\/radiol.2017162664","volume":"285","author":"LM Prevedello","year":"2017","unstructured":"Prevedello, L. M. et al. Automated critical test findings identification and online notification system using artificial intelligence in imaging. Radiology 285, 923\u2013931 (2017).","journal-title":"Radiology"},{"key":"1594_CR7","doi-asserted-by":"publisher","first-page":"196","DOI":"10.1148\/radiol.2018180921","volume":"291","author":"M Annarumma","year":"2019","unstructured":"Annarumma, M. et al. Automated triaging of adult chest radiographs with deep artificial neural networks. Radiology 291, 196\u2013202 (2019).","journal-title":"Radiology"},{"key":"1594_CR8","doi-asserted-by":"publisher","first-page":"55","DOI":"10.1097\/RLI.0000000000000509","volume":"54","author":"DJ Winkel","year":"2019","unstructured":"Winkel, D. J., Heye, T., Weikert, T. J., Boll, D. T. & Stieltjes, B. Evaluation of an AI-based detection software for acute findings in abdominal computed tomography scans: toward an automated work list prioritization of routine CT examinations. Invest. Radiol. 54, 55\u201359 (2019).","journal-title":"Invest. Radiol."},{"key":"1594_CR9","doi-asserted-by":"publisher","first-page":"1046","DOI":"10.3348\/kjr.2023.0839","volume":"24","author":"F Meng","year":"2023","unstructured":"Meng, F., Zhan, L., Liu, S. & Zhang, H. The growing problem of radiologist shortage: China\u2019s perspective. Korean J. Radiol. 24, 1046\u20131048 (2023).","journal-title":"Korean J. Radiol."},{"key":"1594_CR10","doi-asserted-by":"publisher","first-page":"1044","DOI":"10.3390\/children9071044","volume":"9","author":"CKC Ng","year":"2022","unstructured":"Ng, C. K. C. Artificial intelligence for radiation dose optimization in pediatric radiology: a systematic review. Children 9, 1044 (2022).","journal-title":"Children"},{"key":"1594_CR11","doi-asserted-by":"publisher","first-page":"e220246","DOI":"10.1148\/ryai.220246","volume":"5","author":"Y-RJ Wang","year":"2023","unstructured":"Wang, Y. -R. J. et al. AI transformers for radiation dose reduction in serial whole-body PET scans. Radiol. Artif. Intell. 5, e220246 (2023).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR12","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1186\/s12880-023-01081-8","volume":"23","author":"GD Jo","year":"2023","unstructured":"Jo, G. D. et al. 75% radiation dose reduction using deep learning reconstruction on low-dose chest CT. BMC Med. Imaging 23, 121 (2023).","journal-title":"BMC Med. Imaging"},{"key":"1594_CR13","doi-asserted-by":"publisher","unstructured":"Tamada, D. Review: noise and artifact reduction for MRI using deep learning. Preprint at https:\/\/doi.org\/10.48550\/ARXIV.2002.12889 (2020).","DOI":"10.48550\/ARXIV.2002.12889"},{"key":"1594_CR14","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1148\/radiol.2020192224","volume":"295","author":"MJ Willemink","year":"2020","unstructured":"Willemink, M. J. et al. Preparing medical imaging data for machine learning. Radiology 295, 4\u201315 (2020).","journal-title":"Radiology"},{"key":"1594_CR15","doi-asserted-by":"publisher","first-page":"e230227","DOI":"10.1148\/ryai.230227","volume":"6","author":"FC Kitamura","year":"2024","unstructured":"Kitamura, F. C. et al. Lessons learned in building expertly annotated multi-institution datasets and hosting the RSNA AI challenges. Radiol. Artif. Intell. 6, e230227 (2024).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR16","doi-asserted-by":"publisher","first-page":"101693","DOI":"10.1016\/j.media.2020.101693","volume":"63","author":"N Tajbakhsh","year":"2020","unstructured":"Tajbakhsh, N. et al. Embracing imperfect datasets: a review of deep learning solutions for medical image segmentation. Med. Image Anal. 63, 101693 (2020).","journal-title":"Med. Image Anal."},{"key":"1594_CR17","doi-asserted-by":"publisher","first-page":"171","DOI":"10.1177\/08465371231184425","volume":"75","author":"RC Adams-McGavin","year":"2024","unstructured":"Adams-McGavin, R. C. et al. Interrater agreement of CT grading of blunt splenic injuries: does the AAST grading need to be reimagined?. Can. Assoc. Radiol. J. 75, 171\u2013177 (2024).","journal-title":"Can. Assoc. Radiol. J."},{"key":"1594_CR18","doi-asserted-by":"publisher","first-page":"725","DOI":"10.1007\/s00330-021-08132-0","volume":"32","author":"DA Wood","year":"2022","unstructured":"Wood, D. A. et al. Deep learning to automate the labelling of head MRI datasets for computer vision applications. Eur. Radiol. 32, 725\u2013736 (2022).","journal-title":"Eur. Radiol."},{"key":"1594_CR19","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.230034","volume":"5","author":"HM Lin","year":"2023","unstructured":"Lin, H. M. et al. The RSNA cervical spine fracture CT dataset. Radiol. Artif. Intell. 5, e230034 (2023).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR20","unstructured":"Flanders, A. et al. RSNA 2022 cervical spine fracture detection. Kaggle. https:\/\/kaggle.com\/competitions\/rsna-2022-cervical-spine-fracture-detection (2020)."},{"key":"1594_CR21","doi-asserted-by":"publisher","first-page":"e230159","DOI":"10.1148\/ryai.230159","volume":"6","author":"J Teneggi","year":"2024","unstructured":"Teneggi, J., Yi, P. H. & Sulam, J. Examination-level supervision for deep learning\u2013based intracranial hemorrhage detection on head CT scans. Radiol. Artif. Intell. 6, e230159 (2024).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR22","doi-asserted-by":"publisher","first-page":"102105","DOI":"10.1016\/j.media.2021.102105","volume":"72","author":"P Chikontwe","year":"2021","unstructured":"Chikontwe, P. et al. Dual attention multiple instance learning with unsupervised complementary loss for COVID-19 screening. Med. Image Anal. 72, 102105 (2021).","journal-title":"Med. Image Anal."},{"key":"1594_CR23","doi-asserted-by":"publisher","first-page":"2584","DOI":"10.1109\/TMI.2020.2996256","volume":"39","author":"Z Han","year":"2020","unstructured":"Han, Z. et al. Accurate screening of COVID-19 using attention-based deep 3D multiple instance learning. IEEE Trans. Med. Imaging 39, 2584\u20132594 (2020).","journal-title":"IEEE Trans. Med. Imaging"},{"key":"1594_CR24","doi-asserted-by":"publisher","first-page":"2615","DOI":"10.1109\/TMI.2020.2995965","volume":"39","author":"X Wang","year":"2020","unstructured":"Wang, X. et al. A weakly-supervised framework for COVID-19 classification and lesion localization from chest CT. IEEE Trans. Med. Imaging 39, 2615\u20132625 (2020).","journal-title":"IEEE Trans. Med. Imaging"},{"key":"1594_CR25","first-page":"129","volume":"18","author":"J B\u0115lohl\u00e1vek","year":"2013","unstructured":"B\u0115lohl\u00e1vek, J., Dytrych, V. & Linhart, A. Pulmonary embolism, part I: epidemiology, risk factors and risk stratification, pathophysiology, clinical presentation, diagnosis and nonthrombotic pulmonary embolism. Exp. Clin. Cardiol. 18, 129\u2013138 (2013).","journal-title":"Exp. Clin. Cardiol."},{"key":"1594_CR26","doi-asserted-by":"publisher","first-page":"127","DOI":"10.1007\/s40119-022-00290-6","volume":"12","author":"M Imura","year":"2023","unstructured":"Imura, M., Yamamoto, T. & Hiasa, K. -I. Pulmonary thromboembolism developed during hospitalization: a nationwide retrospective observational study using claims data. Cardiol. Ther. 12, 127\u2013141 (2023).","journal-title":"Cardiol. Ther."},{"key":"1594_CR27","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.2021200254","volume":"3","author":"E Colak","year":"2021","unstructured":"Colak, E. et al. The RSNA pulmonary embolism CT dataset. Radiol. Artif. Intell. 3, e200254 (2021).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR28","doi-asserted-by":"publisher","first-page":"84849","DOI":"10.1109\/ACCESS.2019.2925210","volume":"7","author":"X Yang","year":"2019","unstructured":"Yang, X. et al. A Two-stage convolutional neural network for pulmonary embolism detection from CTPA images. IEEE Access 7, 84849\u201384857 (2019).","journal-title":"IEEE Access"},{"key":"1594_CR29","unstructured":"Shi, L. et al. Automatic Diagnosis of Pulmonary Embolism Using an Attention-guided Framework: A Large-scale Study. In Proc. Third Conference on Medical Imaging with Deep Learning 743\u2013754 (PMLR, 2020)."},{"key":"1594_CR30","doi-asserted-by":"publisher","first-page":"61","DOI":"10.1038\/s41746-020-0266-y","volume":"3","author":"S-C Huang","year":"2020","unstructured":"Huang, S. -C. et al. PENet-a scalable deep-learning model for automated diagnosis of pulmonary embolism using volumetric CT imaging. NPJ Digit. Med. 3, 61 (2020).","journal-title":"NPJ Digit. Med."},{"key":"1594_CR31","unstructured":"Rajan, D., Beymer, D., Abedin, S. & Dehghan, E. Pi-PE: A Pipeline for Pulmonary Embolism Detection using Sparsely Annotated 3D CT Images. In Proc. of the Machine Learning for Health NeurIPS Workshop 220\u2013232 (PMLR, 2020)."},{"key":"1594_CR32","doi-asserted-by":"publisher","unstructured":"Suman, S. et al. Attention Based CNN-LSTM Network for Pulmonary Embolism Prediction on Chest Computed Tomography Pulmonary Angiograms. In Medical Image Computing and Computer Assisted Intervention \u2013 MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 \u2013 October 1, 2021, Proceedings, Part VII 356\u2013366 (Springer-Verlag, Berlin, Heidelberg, 2021). https:\/\/doi.org\/10.1007\/978-3-030-87234-2_34.","DOI":"10.1007\/978-3-030-87234-2_34"},{"key":"1594_CR33","doi-asserted-by":"publisher","first-page":"102988","DOI":"10.1016\/j.media.2023.102988","volume":"91","author":"NU Islam","year":"2024","unstructured":"Islam, N. U., Zhou, Z., Gehlot, S., Gotway, M. B. & Liang, J. Seeking an optimal approach for computer-aided diagnosis of pulmonary embolism. Med. Image Anal. 91, 102988 (2024).","journal-title":"Med. Image Anal."},{"key":"1594_CR34","doi-asserted-by":"publisher","first-page":"227","DOI":"10.7326\/0003-4819-132-3-200002010-00009","volume":"132","author":"SW Rathbun","year":"2000","unstructured":"Rathbun, S. W., Raskob, G. E. & Whitsett, T. L. Sensitivity and specificity of helical computed tomography in the diagnosis of pulmonary embolism: a systematic review. Ann. Intern. Med. 132, 227\u2013232 (2000).","journal-title":"Ann. Intern. Med."},{"key":"1594_CR35","doi-asserted-by":"publisher","first-page":"2317","DOI":"10.1056\/NEJMoa052367","volume":"354","author":"PD Stein","year":"2006","unstructured":"Stein, P. D. et al. Multidetector computed tomography for acute pulmonary embolism. N. Engl. J. Med. 354, 2317\u20132327 (2006).","journal-title":"N. Engl. J. Med."},{"key":"1594_CR36","doi-asserted-by":"publisher","first-page":"165","DOI":"10.1080\/02841850601100859","volume":"48","author":"W Ghanima","year":"2007","unstructured":"Ghanima, W. et al. Multidetector computed tomography (MDCT) in the diagnosis of pulmonary embolism: interobserver agreement among radiologists with varied levels of experience. Acta Radiol 48, 165\u2013170 (2007).","journal-title":"Acta Radiol"},{"key":"1594_CR37","doi-asserted-by":"publisher","first-page":"214","DOI":"10.1177\/08465371211000739","volume":"73","author":"ET Nguyen","year":"2022","unstructured":"Nguyen, E. T. et al. Canadian Society of Thoracic Radiology\/Canadian Association of Radiologists Best Practice Guidance for investigation of acute pulmonary embolism, Part 2: technical issues and interpretation Pitfalls. Can. Assoc. Radiol. J. 73, 214\u2013227 (2022).","journal-title":"Can. Assoc. Radiol. J."},{"key":"1594_CR38","doi-asserted-by":"publisher","DOI":"10.23698\/AIDA\/CTPA","author":"T Sj\u00f6blom","year":"2019","unstructured":"Sj\u00f6blom, T., Sladoje, N., Kahraman, A. T., Toumpanakis, D. & Fr\u00f6ding, T. Computed tomography pulmonary angiography (CTPA) data. AIDA https:\/\/doi.org\/10.23698\/AIDA\/CTPA (2019).","journal-title":"AIDA"},{"key":"1594_CR39","doi-asserted-by":"publisher","DOI":"10.1038\/sdata.2018.180","volume":"5","author":"M Masoudi","year":"2018","unstructured":"Masoudi, M. et al. A new dataset of computed-tomography angiography images for computer-aided detection of pulmonary embolism. Sci. Data 5, 180180 (2018).","journal-title":"Sci. Data"},{"key":"1594_CR40","doi-asserted-by":"publisher","unstructured":"Chen, Y. et al. SCUNet++: Swin-UNet and CNN Bottleneck Hybrid Architecture with Multi-Fusion Dense Skip Connection for Pulmonary Embolism CT Image Segmentation. In 2024 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV) 7744\u20137752 (IEEE, Waikoloa, HI, USA, 2024). https:\/\/doi.org\/10.1109\/WACV57701.2024.00758.","DOI":"10.1109\/WACV57701.2024.00758"},{"key":"1594_CR41","doi-asserted-by":"publisher","first-page":"e230024","DOI":"10.1148\/ryai.230024","volume":"5","author":"J Wasserthal","year":"2023","unstructured":"Wasserthal, J. et al. TotalSegmentator: robust segmentation of 104 anatomic structures in CT images. Radiol. Artif. Intell. 5, e230024 (2023).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR42","doi-asserted-by":"publisher","first-page":"e210162","DOI":"10.1148\/ryai.210162","volume":"4","author":"LD Hahn","year":"2022","unstructured":"Hahn, L. D., Hall, K., Alebdi, T., Kligerman, S. J. & Hsiao, A. Automated deep learning analysis for quality improvement of CT pulmonary angiography. Radiol. Artif. Intell. 4, e210162 (2022).","journal-title":"Radiol. Artif. Intell."},{"key":"1594_CR43","doi-asserted-by":"publisher","first-page":"125","DOI":"10.3390\/info11020125","volume":"11","author":"A Buslaev","year":"2020","unstructured":"Buslaev, A. et al. Albumentations: fast and flexible image augmentations. Information 11, 125 (2020).","journal-title":"Information"},{"key":"1594_CR44","unstructured":"Dai, Z., Liu, H., Le, Q. V. & Tan, M. CoAtNet: Marrying Convolution and Attention for All Data Sizes. In Advances in Neural Information Processing Systems (eds Ranzato, M., Beygelzimer, A., Dauphin, Y., Liang, P. S. & Vaughan, J. W.) Vol. 34, 3965\u20133977 (Curran Associates, Inc., 2021)."},{"key":"1594_CR45","doi-asserted-by":"publisher","DOI":"10.5281\/ZENODO.4414861","author":"R Wightman","year":"2023","unstructured":"Wightman, R. et al. rwightman\/pytorch-image-models: v0.8.10dev0 Release. Zenodo https:\/\/doi.org\/10.5281\/ZENODO.4414861 (2023).","journal-title":"Zenodo"},{"key":"1594_CR46","doi-asserted-by":"publisher","first-page":"1389","DOI":"10.1109\/LSP.2014.2337313","volume":"21","author":"X Sun","year":"2014","unstructured":"Sun, X. & Xu, W. Fast implementation of DeLong\u2019s algorithm for comparing the areas under correlated receiver operating characteristic curves. IEEE Signal Process. Lett. 21, 1389\u20131393 (2014).","journal-title":"IEEE Signal Process. Lett."}],"container-title":["npj Digital Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41746-025-01594-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-025-01594-2","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-025-01594-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,7]],"date-time":"2025-05-07T05:30:09Z","timestamp":1746595809000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41746-025-01594-2"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,7]]},"references-count":46,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,12]]}},"alternative-id":["1594"],"URL":"https:\/\/doi.org\/10.1038\/s41746-025-01594-2","relation":{},"ISSN":["2398-6352"],"issn-type":[{"value":"2398-6352","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,5,7]]},"assertion":[{"value":"1 October 2024","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"28 March 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"7 May 2025","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","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"254"}}