{"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/on-the-compactness-efficiency-and","title":"On the Compactness, Efficiency, and Representation of 3D Convolutional Networks: Brain Parcellation as a Pretext Task","arxiv_id":"1707.01992","date":"2017-07-06","proceeding":null,"authors":["Wenqi Li","Guotai Wang","Lucas Fidon","Sebastien Ourselin","M. Jorge Cardoso","Tom Vercauteren"],"abstract":"Deep convolutional neural networks are powerful tools for learning visual\nrepresentations from images. However, designing efficient deep architectures to\nanalyse volumetric medical images remains challenging. This work investigates\nefficient and flexible elements of modern convolutional networks such as\ndilated convolution and residual connection. With these essential building\nblocks, we propose a high-resolution, compact convolutional network for\nvolumetric image segmentation. To illustrate its efficiency of learning 3D\nrepresentation from large-scale image data, the proposed network is validated\nwith the challenging task of parcellating 155 neuroanatomical structures from\nbrain MR images. Our experiments show that the proposed network architecture\ncompares favourably with state-of-the-art volumetric segmentation networks\nwhile being an order of magnitude more compact. We consider the brain\nparcellation task as a pretext task for volumetric image segmentation; our\ntrained network potentially provides a good starting point for transfer\nlearning. Additionally, we show the feasibility of voxel-level uncertainty\nestimation using a sampling approximation through dropout.","url_abs":"http://arxiv.org/abs/1707.01992v1","url_pdf":"http://arxiv.org/pdf/1707.01992v1.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":"on-the-compactness-efficiency-and","repo_url":"https://github.com/gift-surg/HighRes3DNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"on-the-compactness-efficiency-and","repo_url":"https://github.com/black0017/MedicalZooPytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"on-the-compactness-efficiency-and","repo_url":"https://github.com/fepegar/highresnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"on-the-compactness-efficiency-and","repo_url":"https://github.com/khanlab/hippunfold","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"3d-medical-imaging-segmentation","task_name":"3D Medical Imaging Segmentation"},{"task_slug":"image-segmentation","task_name":"Image Segmentation"},{"task_slug":"medical-image-segmentation","task_name":"Medical Image Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"},{"task_slug":"volumetric-medical-image-segmentation","task_name":"Volumetric Medical Image Segmentation"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1707.01992","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}