{"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/brain-extraction-from-normal-and-pathological","title":"Brain Extraction from Normal and Pathological Images: A Joint PCA/Image-Reconstruction Approach","arxiv_id":"1711.05702","date":"2017-11-15","proceeding":null,"authors":["Xu Han","Roland Kwitt","Stephen Aylward","Spyridon Bakas","Bjoern Menze","Alexander Asturias","Paul Vespa","John Van Horn","Marc Niethammer"],"abstract":"Brain extraction from images is a common pre-processing step. Many approaches\nexist, but they are frequently only designed to perform brain extraction from\nimages without strong pathologies. Extracting the brain from images with strong\npathologies, for example, the presence of a tumor or of a traumatic brain\ninjury, is challenging. In such cases, tissue appearance may deviate from\nnormal tissue and violates algorithmic assumptions for these approaches; hence,\nthe brain may not be correctly extracted. This paper proposes a brain\nextraction approach which can explicitly account for pathologies by jointly\nmodeling normal tissue and pathologies. Specifically, our model uses a\nthree-part image decomposition: (1) normal tissue appearance is captured by\nprincipal component analysis, (2) pathologies are captured via a total\nvariation term, and (3) non-brain tissue is captured by a sparse term.\nDecomposition and image registration steps are alternated to allow statistical\nmodeling in a fixed atlas space. As a beneficial side effect, the model allows\nfor the identification of potential pathologies and the reconstruction of a\nquasi-normal image in atlas space. We demonstrate the effectiveness of our\nmethod on four datasets: the IBSR and LPBA40 datasets which show normal images,\nthe BRATS dataset containing images with brain tumors and a dataset containing\nclinical TBI images. We compare the performance with other popular models:\nROBEX, BEaST, MASS, BET, BSE and a recently proposed deep learning approach.\nOur model performs better than these competing methods on all four datasets.\nSpecifically, our model achieves the best median (97.11) and mean (96.88) Dice\nscores over all datasets. The two best performing competitors, ROBEX and MASS,\nachieve scores of 96.23/95.62 and 96.67/94.25 respectively. Hence, our approach\nis an effective method for high quality brain extraction on a wide variety of\nimages.","url_abs":"http://arxiv.org/abs/1711.05702v2","url_pdf":"http://arxiv.org/pdf/1711.05702v2.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":"brain-extraction-from-normal-and-pathological","repo_url":"https://github.com/uncbiag/pstrip","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"image-reconstruction","task_name":"Image Reconstruction"},{"task_slug":"image-registration","task_name":"Image Registration"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.05702","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}