{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T16:59:57Z","timestamp":1791478797209,"version":"4.3.4"},"reference-count":19,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T00:00:00Z","timestamp":1774828800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,3,31]],"date-time":"2026-03-31T00:00:00Z","timestamp":1774915200000},"content-version":"vor","delay-in-days":1,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    The need for clear, trustworthy explanations of deep learning model predictions is essential for high-criticality fields, such as medicine and biometric identification. Class Activation Maps (CAMs) are an increasingly popular category of visual explanation methods for Convolutional Neural Networks (CNNs). However, the performance of individual CAMs depends largely on experimental parameters such as the selected image, target class, and model. Here, we propose MetaCAM, an ensemble-based method for combining multiple existing CAM methods based on the consensus of the top-\n                    <jats:italic>k<\/jats:italic>\n                    % most highly activated pixels across component CAMs. We perform experiments to quantifiably determine the optimal combination of 11 CAMs for a given MetaCAM experiment. A new method denoted Cumulative Residual Effect (CRE) is proposed to summarize large-scale ensemble-based experiments. We also present adaptive thresholding and demonstrate how it can be applied to individual CAMs to improve their performance, measured using pixel perturbation method Remove and Debias (ROAD). Lastly, we show that MetaCAM outperforms existing CAMs and refines the most salient regions of images used for model predictions. In a specific example, MetaCAM improved ROAD performance to 0.393 compared to 11 individual CAMs with ranges from -0.101-0.172, demonstrating the importance of combining CAMs through an ensembling method and adaptive thresholding.\n                  <\/jats:p>","DOI":"10.1038\/s41598-026-42879-0","type":"journal-article","created":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T12:40:04Z","timestamp":1774874404000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["MetaCAM as an ensemble-based class activation mapping framework improves model explainability"],"prefix":"10.1038","volume":"16","author":[{"given":"Kevin","family":"Dick","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Emily","family":"Kaczmarek","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Olivier X.","family":"Miguel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexa C.","family":"Bowie","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robin","family":"Ducharme","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alysha L. J.","family":"Dingwall-Harvey","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Steven","family":"Hawken","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christine M.","family":"Armour","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark C.","family":"Walker","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,3,30]]},"reference":[{"key":"42879_CR1","doi-asserted-by":"publisher","first-page":"425","DOI":"10.1007\/s13218-021-00736-4","volume":"35","author":"T Schmid","year":"2021","unstructured":"Schmid, T., Hildesheim, W., Holoyad, T. & Schumacher, K. The AI methods, capabilities and criticality grid: A three-dimensional classification scheme for artificial intelligence applications. KI-K\u00fcnstliche Intell. 35, 425\u2013440 (2021).","journal-title":"KI-K\u00fcnstliche Intell."},{"key":"42879_CR2","doi-asserted-by":"crossref","unstructured":"Zhou, B., Khosla, A., Lapedriza, A., Oliva, A. & Torralba, A. Learning deep features for discriminative localization. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2921\u20132929 (2016).","DOI":"10.1109\/CVPR.2016.319"},{"key":"42879_CR3","doi-asserted-by":"crossref","unstructured":"Chattopadhay, A., Sarkar, A., Howlader, P. & Balasubramanian, V. N. Grad-cam++: Generalized gradient-based visual explanations for deep convolutional networks. In 2018 IEEE Winter Conference on Applications of Computer Vision (WACV). 839\u2013847 (IEEE, 2018).","DOI":"10.1109\/WACV.2018.00097"},{"key":"42879_CR4","unstructured":"Ramaswamy, H. G. et al. Ablation-cam: Visual explanations for deep convolutional network via gradient-free localization. In Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision. 983\u2013991 (2020)."},{"key":"42879_CR5","unstructured":"Draelos, R. L. & Carin, L. Use hirescam instead of grad-cam for faithful explanations of convolutional neural networks (2020)."},{"key":"42879_CR6","doi-asserted-by":"crossref","unstructured":"Fu, R. et al. Axiom-based grad-cam: Towards accurate visualization and explanation of cnns (2020).","DOI":"10.5244\/C.34.146"},{"key":"42879_CR7","unstructured":"Gildenblat, J. et al. Pytorch Library for CAM Methods. https:\/\/github.com\/jacobgil\/pytorch-grad-cam (2021)."},{"key":"42879_CR8","doi-asserted-by":"publisher","first-page":"5875","DOI":"10.1109\/TIP.2021.3089943","volume":"30","author":"P-T Jiang","year":"2021","unstructured":"Jiang, P.-T., Zhang, C.-B., Hou, Q., Cheng, M.-M. & Wei, Y. Layercam: Exploring hierarchical class activation maps for localization. IEEE Trans. Image Process. 30, 5875\u20135888 (2021).","journal-title":"IEEE Trans. Image Process."},{"key":"42879_CR9","doi-asserted-by":"crossref","unstructured":"Muhammad, M. B. & Yeasin, M. Eigen-cam: Class activation map using principal components. In 2020 International Joint Conference on Neural Networks (IJCNN). 1\u20137 (IEEE, 2020).","DOI":"10.1109\/IJCNN48605.2020.9206626"},{"key":"42879_CR10","doi-asserted-by":"crossref","unstructured":"Selvaraju, R. R. et al. Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE International Conference on Computer Vision. 618\u2013626 (2017).","DOI":"10.1109\/ICCV.2017.74"},{"key":"42879_CR11","doi-asserted-by":"crossref","unstructured":"Wang, H. et al. Score-cam: Score-weighted visual explanations for convolutional neural networks. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops. 24\u201325 (2020).","DOI":"10.1109\/CVPRW50498.2020.00020"},{"key":"42879_CR12","doi-asserted-by":"publisher","first-page":"e0296985","DOI":"10.1371\/journal.pone.0296985","volume":"19","author":"E Kaczmarek","year":"2024","unstructured":"Kaczmarek, E. et al. Camanim: Animating end-to-end network activation maps. PloS one 19, e0296985 (2024).","journal-title":"PloS one"},{"key":"42879_CR13","doi-asserted-by":"crossref","unstructured":"Ornek, A. H. & Ceylan, M. Codcam: A new ensemble visual explanation for classification of medical thermal images. Quant. InfraRed Thermogr. J. 1\u201325 (2023).","DOI":"10.1080\/17686733.2023.2167459"},{"key":"42879_CR14","unstructured":"Srinivas, S. & Fleuret, F. Full-Gradient Representation for Neural Network Visualization. Vol. 32 (2019)."},{"key":"42879_CR15","unstructured":"Rong, Y., Leemann, T., Borisov, V., Kasneci, G. & Kasneci, E. A consistent and efficient evaluation strategy for attribution methods. arXiv:2202.00449 (2022)."},{"key":"42879_CR16","doi-asserted-by":"crossref","unstructured":"Deng, J. et al. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition. 248\u2013255 (IEEE, 2009).","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"42879_CR17","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 770\u2013778 (2016).","DOI":"10.1109\/CVPR.2016.90"},{"key":"42879_CR18","doi-asserted-by":"crossref","unstructured":"Huang, G., Liu, Z., Van Der Maaten, L. & Weinberger, K. Q. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 4700\u20134708 (2017).","DOI":"10.1109\/CVPR.2017.243"},{"key":"42879_CR19","unstructured":"Lundberg, S. M. & Lee, S.-I. A unified approach to interpreting model predictions. Adv. Neural Inf. Process. Syst. 30 (2017)."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-42879-0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-42879-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-42879-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,31]],"date-time":"2026-03-31T09:50:56Z","timestamp":1774950656000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-42879-0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,30]]},"references-count":19,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["42879"],"URL":"https:\/\/doi.org\/10.1038\/s41598-026-42879-0","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,30]]},"assertion":[{"value":"18 July 2024","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"27 February 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 March 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"10613"}}