{"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/efficient-multi-class-fetal-brain","title":"Efficient multi-class fetal brain segmentation in high resolution MRI reconstructions with noisy labels","arxiv_id":"2009.06275","date":"2020-09-14","proceeding":null,"authors":[],"abstract":"Segmentation of the developing fetal brain is an important step in\nquantitative analyses. However, manual segmentation is a very time-consuming\ntask which is prone to error and must be completed by highly specialized\nindi-viduals. Super-resolution reconstruction of fetal MRI has become standard\nfor processing such data as it improves image quality and resolution. However,\ndif-ferent pipelines result in slightly different outputs, further complicating\nthe gen-eralization of segmentation methods aiming to segment super-resolution\ndata. Therefore, we propose using transfer learning with noisy multi-class\nlabels to automatically segment high resolution fetal brain MRIs using a single\nset of seg-mentations created with one reconstruction method and tested for\ngeneralizability across other reconstruction methods. Our results show that the\nnetwork can auto-matically segment fetal brain reconstructions into 7 different\ntissue types, regard-less of reconstruction method used. Transfer learning\noffers some advantages when compared to training without pre-initialized\nweights, but the network trained on clean labels had more accurate\nsegmentations overall. No additional manual segmentations were required.\nTherefore, the proposed network has the potential to eliminate the need for\nmanual segmentations needed in quantitative analyses of the fetal brain\nindependent of reconstruction method used, offering an unbiased way to quantify\nnormal and pathological neurodevelopment.","url_abs":"http://arxiv.org/abs/2009.06275v1","url_pdf":"http://arxiv.org/pdf/2009.06275v1.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":"efficient-multi-class-fetal-brain","repo_url":"https://github.com/FNNDSC/pl-surfaces-fetus","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"brain-segmentation","task_name":"Brain Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"super-resolution","task_name":"Super-Resolution"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}