{"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/one-shot-domain-adaptation-in-multiple","title":"One-shot domain adaptation in multiple sclerosis lesion segmentation using convolutional neural networks","arxiv_id":"1805.12415","date":"2018-05-31","proceeding":null,"authors":["Sergi Valverde","Mostafa Salem","Mariano Cabezas","Deborah Pareto","Joan C. Vilanova","Lluís Ramió-Torrentà","Àlex Rovira","Joaquim Salvi","Arnau Oliver","Xavier Lladó"],"abstract":"In recent years, several convolutional neural network (CNN) methods have been\nproposed for the automated white matter lesion segmentation of multiple\nsclerosis (MS) patient images, due to their superior performance compared with\nthose of other state-of-the-art methods. However, the accuracies of CNN methods\ntend to decrease significantly when evaluated on different image domains\ncompared with those used for training, which demonstrates the lack of\nadaptability of CNNs to unseen imaging data. In this study, we analyzed the\neffect of intensity domain adaptation on our recently proposed CNN-based MS\nlesion segmentation method. Given a source model trained on two public MS\ndatasets, we investigated the transferability of the CNN model when applied to\nother MRI scanners and protocols, evaluating the minimum number of annotated\nimages needed from the new domain and the minimum number of layers needed to\nre-train to obtain comparable accuracy. Our analysis comprised MS patient data\nfrom both a clinical center and the public ISBI2015 challenge database, which\npermitted us to compare the domain adaptation capability of our model to that\nof other state-of-the-art methods. For the ISBI2015 challenge, our one-shot\ndomain adaptation model trained using only a single image showed a performance\nsimilar to that of other CNN methods that were fully trained using the entire\navailable training set, yielding a comparable human expert rater performance.\nWe believe that our experiments will encourage the MS community to incorporate\nits use in different clinical settings with reduced amounts of annotated data.\nThis approach could be meaningful not only in terms of the accuracy in\ndelineating MS lesions but also in the related reductions in time and economic\ncosts derived from manual lesion labeling.","url_abs":"http://arxiv.org/abs/1805.12415v1","url_pdf":"http://arxiv.org/pdf/1805.12415v1.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":"one-shot-domain-adaptation-in-multiple","repo_url":"https://github.com/NIC-VICOROB/nicmslesions","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"one-shot-domain-adaptation-in-multiple","repo_url":"https://github.com/giorgiberiani/nicmslesions","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"one-shot-domain-adaptation-in-multiple","repo_url":"https://github.com/sergivalverde/nicMSlesions","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"lesion-segmentation","task_name":"Lesion Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}