{"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/framing-u-net-via-deep-convolutional","title":"Framing U-Net via Deep Convolutional Framelets: Application to Sparse-view CT","arxiv_id":"1708.08333","date":"2017-08-28","proceeding":null,"authors":["Yoseob Han","Jong Chul Ye"],"abstract":"X-ray computed tomography (CT) using sparse projection views is a recent\napproach to reduce the radiation dose. However, due to the insufficient\nprojection views, an analytic reconstruction approach using the filtered back\nprojection (FBP) produces severe streaking artifacts. Recently, deep learning\napproaches using large receptive field neural networks such as U-Net have\ndemonstrated impressive performance for sparse- view CT reconstruction.\nHowever, theoretical justification is still lacking. Inspired by the recent\ntheory of deep convolutional framelets, the main goal of this paper is,\ntherefore, to reveal the limitation of U-Net and propose new multi-resolution\ndeep learning schemes. In particular, we show that the alternative U- Net\nvariants such as dual frame and the tight frame U-Nets satisfy the so-called\nframe condition which make them better for effective recovery of high frequency\nedges in sparse view- CT. Using extensive experiments with real patient data\nset, we demonstrate that the new network architectures provide better\nreconstruction performance.","url_abs":"http://arxiv.org/abs/1708.08333v3","url_pdf":"http://arxiv.org/pdf/1708.08333v3.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":"framing-u-net-via-deep-convolutional","repo_url":"https://github.com/hanyoseob/framing-u-net","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"framing-u-net-via-deep-convolutional","repo_url":"https://github.com/hjahan58/framing-u-net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"framing-u-net-via-deep-convolutional","repo_url":"https://github.com/jongcye/FramingUNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"ct-reconstruction","task_name":"CT Reconstruction"},{"task_slug":"computed-tomography-ct","task_name":"Computed Tomography (CT)"}],"methods":[{"method_slug":"concatenated-skip-connection","method_name":"Concatenated Skip Connection"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"u-net","method_name":"U-Net"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.08333","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}