{"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/networks-for-joint-affine-and-non-parametric","title":"Networks for Joint Affine and Non-parametric Image Registration","arxiv_id":"1903.08811","date":"2019-03-21","proceeding":"CVPR 2019 6","authors":["Zhengyang Shen","Xu Han","Zhenlin Xu","Marc Niethammer"],"abstract":"We introduce an end-to-end deep-learning framework for 3D medical image\nregistration. In contrast to existing approaches, our framework combines two\nregistration methods: an affine registration and a vector\nmomentum-parameterized stationary velocity field (vSVF) model. Specifically, it\nconsists of three stages. In the first stage, a multi-step affine network\npredicts affine transform parameters. In the second stage, we use a Unet-like\nnetwork to generate a momentum, from which a velocity field can be computed via\nsmoothing. Finally, in the third stage, we employ a self-iterable map-based\nvSVF component to provide a non-parametric refinement based on the current\nestimate of the transformation map. Once the model is trained, a registration\nis completed in one forward pass. To evaluate the performance, we conducted\nlongitudinal and cross-subject experiments on 3D magnetic resonance images\n(MRI) of the knee of the Osteoarthritis Initiative (OAI) dataset. Results show\nthat our framework achieves comparable performance to state-of-the-art medical\nimage registration approaches, but it is much faster, with a better control of\ntransformation regularity including the ability to produce approximately\nsymmetric transformations, and combining affine and non-parametric\nregistration.","url_abs":"http://arxiv.org/abs/1903.08811v1","url_pdf":"http://arxiv.org/pdf/1903.08811v1.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":"networks-for-joint-affine-and-non-parametric","repo_url":"https://github.com/uncbiag/OAI_analysis","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"networks-for-joint-affine-and-non-parametric","repo_url":"https://github.com/uncbiag/registration","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"networks-for-joint-affine-and-non-parametric","repo_url":"https://github.com/MindSpore-scientific/code-12/tree/main/Fourier-Features-Let-Networks-Learn-High-Frequency","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"image-registration","task_name":"Image Registration"},{"task_slug":"medical-image-registration","task_name":"Medical Image Registration"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-registration-on-osteoarthritis","task":"Image Registration","dataset":"Osteoarthritis Initiative","model":"vSVF-net [shen2019networks]","rank_in_archive_order":2,"of":2,"metrics":{"Dice":"67.59"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1903.08811","atlas_url":"https://app.syntology.ai/?focus=1903.08811","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}