{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/exclusive-autoencoder-xae-for-nucleus","title":"eXclusive Autoencoder (XAE) for Nucleus Detection and Classification on Hematoxylin and Eosin (H&E) Stained Histopathological Images","arxiv_id":"1811.11243","date":"2018-11-27","proceeding":null,"authors":["Chao-Hui Huang","Daniel Racoceanu"],"abstract":"In this paper, we introduced a novel feature extraction approach, named\nexclusive autoencoder (XAE), which is a supervised version of autoencoder (AE),\nable to largely improve the performance of nucleus detection and classification\non hematoxylin and eosin (H&E) histopathological images. The proposed XAE can\nbe used in any AE-based algorithm, as long as the data labels are also provided\nin the feature extraction phase. In the experiments, we evaluated the\nperformance of an approach which is the combination of an XAE and a fully\nconnected neural network (FCN) and compared with some AE-based methods. For a\nnucleus detection problem (considered as a nucleus/non-nucleus classification\nproblem) on breast cancer H&E images, the F-score of the proposed XAE+FCN\napproach achieved 96.64% while the state-of-the-art was at 84.49%. For nucleus\nclassification on colorectal cancer H&E images, with the annotations of four\ncategories of epithelial, inflammatory, fibroblast and miscellaneous nuclei.\nThe F-score of the proposed method reached 70.4%. We also proposed a lymphocyte\nsegmentation method. In the step of lymphocyte detection, we have compared with\ncutting-edge technology and gained improved performance from 90% to 98.67%. We\nalso proposed an algorithm for lymphocyte segmentation based on nucleus\ndetection and classification. The obtained Dice coefficient achieved 88.31%\nwhile the cutting-edge approach was at 74%.","url_abs":"http://arxiv.org/abs/1811.11243v1","url_pdf":"http://arxiv.org/pdf/1811.11243v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"exclusive-autoencoder-xae-for-nucleus","repo_url":"https://github.com/huangch/xae4hne","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"miscellaneous","task_name":"Miscellaneous"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}