{"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/pyro-nn-python-reconstruction-operators-in","title":"PYRO-NN: Python Reconstruction Operators in Neural Networks","arxiv_id":"1904.13342","date":"2019-04-30","proceeding":null,"authors":["Christopher Syben","Markus Michen","Bernhard Stimpel","Stephan Seitz","Stefan Ploner","Andreas K. Maier"],"abstract":"Purpose: Recently, several attempts were conducted to transfer deep learning\nto medical image reconstruction. An increasingly number of publications follow\nthe concept of embedding the CT reconstruction as a known operator into a\nneural network. However, most of the approaches presented lack an efficient CT\nreconstruction framework fully integrated into deep learning environments. As a\nresult, many approaches are forced to use workarounds for mathematically\nunambiguously solvable problems. Methods: PYRO-NN is a generalized framework to\nembed known operators into the prevalent deep learning framework Tensorflow.\nThe current status includes state-of-the-art parallel-, fan- and cone-beam\nprojectors and back-projectors accelerated with CUDA provided as Tensorflow\nlayers. On top, the framework provides a high level Python API to conduct FBP\nand iterative reconstruction experiments with data from real CT systems.\nResults: The framework provides all necessary algorithms and tools to design\nend-to-end neural network pipelines with integrated CT reconstruction\nalgorithms. The high level Python API allows a simple use of the layers as\nknown from Tensorflow. To demonstrate the capabilities of the layers, the\nframework comes with three baseline experiments showing a cone-beam short scan\nFDK reconstruction, a CT reconstruction filter learning setup, and a TV\nregularized iterative reconstruction. All algorithms and tools are referenced\nto a scientific publication and are compared to existing non deep learning\nreconstruction frameworks. The framework is available as open-source software\nat \\url{https://github.com/csyben/PYRO-NN}. Conclusions: PYRO-NN comes with the\nprevalent deep learning framework Tensorflow and allows to setup end-to-end\ntrainable neural networks in the medical image reconstruction context. We\nbelieve that the framework will be a step towards reproducible research","url_abs":"http://arxiv.org/abs/1904.13342v1","url_pdf":"http://arxiv.org/pdf/1904.13342v1.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":"pyro-nn-python-reconstruction-operators-in","repo_url":"https://github.com/csyben/PYRO-NN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"pyro-nn-python-reconstruction-operators-in","repo_url":"https://github.com/csyben/PYRO-NN-Layers","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"ct-reconstruction","task_name":"CT Reconstruction"},{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"image-reconstruction","task_name":"Image Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.13342","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}