{"id":18759799,"url":"https://github.com/mrphys/tensorflow-mri","last_synced_at":"2025-04-09T10:09:13.639Z","repository":{"id":37230399,"uuid":"388094708","full_name":"mrphys/tensorflow-mri","owner":"mrphys","description":"A Library of TensorFlow Operators for Computational MRI","archived":false,"fork":false,"pushed_at":"2025-03-19T16:07:05.000Z","size":199404,"stargazers_count":40,"open_issues_count":8,"forks_count":3,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-04-02T08:11:21.942Z","etag":null,"topics":["machine-learning","magnetic-resonance-imaging","ml","mri","python","tensorflow"],"latest_commit_sha":null,"homepage":"https://mrphys.github.io/tensorflow-mri/","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/mrphys.png","metadata":{"files":{"readme":"README.rst","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2021-07-21T11:28:50.000Z","updated_at":"2025-03-19T16:07:15.000Z","dependencies_parsed_at":"2024-12-17T20:05:54.307Z","dependency_job_id":"aef9b1d3-1f43-48ec-ba71-495d30c667db","html_url":"https://github.com/mrphys/tensorflow-mri","commit_stats":{"total_commits":383,"total_committers":1,"mean_commits":383.0,"dds":0.0,"last_synced_commit":"cfd8930ee5281e7f6dceb17c4a5acaf625fd3243"},"previous_names":[],"tags_count":35,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mrphys%2Ftensorflow-mri","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mrphys%2Ftensorflow-mri/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mrphys%2Ftensorflow-mri/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mrphys%2Ftensorflow-mri/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/mrphys","download_url":"https://codeload.github.com/mrphys/tensorflow-mri/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248018060,"owners_count":21034048,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["machine-learning","magnetic-resonance-imaging","ml","mri","python","tensorflow"],"created_at":"2024-11-07T18:07:47.661Z","updated_at":"2025-04-09T10:09:08.628Z","avatar_url":"https://github.com/mrphys.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"TensorFlow MRI\n==============\n\n|pypi| |build| |docs| |doi|\n\n.. |pypi| image:: https://badge.fury.io/py/tensorflow-mri.svg\n    :target: https://badge.fury.io/py/tensorflow-mri\n.. |build| image:: https://github.com/mrphys/tensorflow-mri/actions/workflows/build-package.yml/badge.svg\n    :target: https://github.com/mrphys/tensorflow-mri/actions/workflows/build-package.yml\n.. |docs| image:: https://img.shields.io/badge/api-reference-blue.svg\n    :target: https://mrphys.github.io/tensorflow-mri/\n.. |doi| image:: https://zenodo.org/badge/388094708.svg\n    :target: https://zenodo.org/badge/latestdoi/388094708\n\n.. start-intro\n\nTensorFlow MRI is a library of TensorFlow operators for computational MRI.\nThe library has a Python interface and is mostly written in Python. However,\ncomputations are efficiently performed by the TensorFlow backend (implemented in\nC++/CUDA), which brings together the ease of use and fast prototyping of Python\nwith the speed and efficiency of optimized lower-level implementations.\n\nBeing an extension of TensorFlow, TensorFlow MRI integrates seamlessly in ML\napplications. No additional interfacing is needed to include a SENSE operator\nwithin a neural network, or to use a trained prior as part of an iterative\nreconstruction. Therefore, the gap between ML and non-ML components of image\nprocessing pipelines is eliminated.\n\nWhether an application involves ML or not, TensorFlow MRI operators can take\nfull advantage of the TensorFlow framework, with capabilities including\nautomatic differentiation, multi-device support (CPUs and GPUs), automatic\ndevice placement and copying of tensor data, and conversion to fast,\nserializable graphs.\n\nTensorFlow MRI contains operators for:\n\n* Multicoil arrays\n  (`tfmri.coils \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/coils\u003e`_):\n  coil combination, coil compression and estimation of coil sensitivity\n  maps.\n* Convex optimization\n  (`tfmri.convex \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/convex\u003e`_):\n  convex functions (quadratic, L1, L2, Tikhonov, total variation, etc.) and\n  optimizers (ADMM).\n* Keras initializers\n  (`tfmri.initializers \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/initializers\u003e`_):\n  neural network initializers, including support for complex-valued weights.\n* I/O (`tfmri.io \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/io\u003e`_):\n  additional I/O functions potentially useful when working with MRI data.\n* Keras layers\n  (`tfmri.layers \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/layers\u003e`_):\n  layers and building blocks for neural networks, including support for\n  complex-valued weights, inputs and outputs.\n* Linear algebra\n  (`tfmri.linalg \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/linalg\u003e`_):\n  linear operators specialized for image processing and MRI.\n* Loss functions\n  (`tfmri.losses \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/losses\u003e`_):\n  for classification, segmentation and image restoration.\n* Metrics\n  (`tfmri.metrics \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/metrics\u003e`_):\n  for classification, segmentation and image restoration.\n* Image processing\n  (`tfmri.image \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/image\u003e`_):\n  filtering, gradients, phantoms, image quality assessment, etc.\n* Image reconstruction\n  (`tfmri.recon \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/recon\u003e`_):\n  Cartesian/non-Cartesian, 2D/3D, parallel imaging, compressed sensing.\n* *k*-space sampling\n  (`tfmri.sampling \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/sampling\u003e`_):\n  Cartesian masks, non-Cartesian trajectories, sampling density compensation,\n  etc.\n* Signal processing\n  (`tfmri.signal \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/signal\u003e`_):\n  N-dimensional fast Fourier transform (FFT), non-uniform FFT (NUFFT)\n  (see also `TensorFlow NUFFT \u003chttps://github.com/mrphys/tensorflow-nufft\u003e`_),\n  discrete wavelet transform (DWT), *k*-space filtering, etc.\n* Unconstrained optimization\n  (`tfmri.optimize \u003chttps://mrphys.github.io/tensorflow-mri/api_docs/tfmri/optimize\u003e`_):\n  gradient descent, L-BFGS.\n* And more, e.g., supporting array manipulation and math tasks.\n\n.. end-intro\n\nInstallation\n------------\n\n.. start-install\n\nYou can install TensorFlow MRI with ``pip``:\n\n.. code-block:: console\n\n    $ pip install tensorflow-mri\n\nNote that only Linux is currently supported.\n\nTensorFlow Compatibility\n^^^^^^^^^^^^^^^^^^^^^^^^\n\nEach TensorFlow MRI release is compiled against a specific version of\nTensorFlow. To ensure compatibility, it is recommended to install matching\nversions of TensorFlow and TensorFlow MRI according to the table below.\n\n.. start-compatibility-table\n\n======================  ========================  ============\nTensorFlow MRI Version  TensorFlow Compatibility  Release Date\n======================  ========================  ============\nv0.23.0                 v2.10.x                   Jan 28, 2025 \nv0.22.0                 v2.10.x                   Sep 26, 2022\nv0.21.0                 v2.9.x                    Jul 24, 2022\nv0.20.0                 v2.9.x                    Jun 18, 2022\nv0.19.0                 v2.9.x                    Jun 1, 2022\nv0.18.0                 v2.8.x                    May 6, 2022\nv0.17.0                 v2.8.x                    Apr 22, 2022\nv0.16.0                 v2.8.x                    Apr 13, 2022\nv0.15.0                 v2.8.x                    Apr 1, 2022\nv0.14.0                 v2.8.x                    Mar 29, 2022\nv0.13.0                 v2.8.x                    Mar 15, 2022\nv0.12.0                 v2.8.x                    Mar 14, 2022\nv0.11.0                 v2.8.x                    Mar 10, 2022\nv0.10.0                 v2.8.x                    Mar 3, 2022\nv0.9.0                  v2.7.x                    Dec 3, 2021\nv0.8.0                  v2.7.x                    Nov 11, 2021\nv0.7.0                  v2.6.x                    Nov 3, 2021\nv0.6.2                  v2.6.x                    Oct 13, 2021\nv0.6.1                  v2.6.x                    Sep 30, 2021\nv0.6.0                  v2.6.x                    Sep 28, 2021\nv0.5.0                  v2.6.x                    Aug 29, 2021\nv0.4.0                  v2.6.x                    Aug 18, 2021\n======================  ========================  ============\n\n.. end-compatibility-table\n\n.. end-install\n\nDocumentation\n-------------\n\nVisit the `docs \u003chttps://mrphys.github.io/tensorflow-mri/\u003e`_ for guides,\ntutorials and the API reference.\n\nVideo Tutorial\n---------------------\n\n\nHere is a video tutorial demonstrating how TensorFlow MRI can be use (including a specific example problem for creating fully sampled k-space data from undersampled raw data with partial-fourier, as well as creating coil-combined 'ground truth' images from this, and paired undersampled radial multi-coild complex data, which is used to train a 3D Unet. I also show how to do a CS recosntruction of the same raw-data)\n\n[![Watch the video](tools/assets/thumb.png)](https://vimeo.com/1054518675/e19c8abad3)\n\n\nIssues\n-------------\n\nIf you use this package and something does not work as you expected, please\n`file an issue \u003chttps://github.com/mrphys/tensorflow-mri/issues/new\u003e`_\ndescribing your problem. We're here to help!\n\nCredits\n-------------\n\nIf you like this software, star the repository! |stars|\n\n.. |stars| image:: https://img.shields.io/github/stars/mrphys/tensorflow-mri?style=social\n    :target: https://github.com/mrphys/tensorflow-mri/stargazers\n\nIf you find this software useful in your research, you can cite TensorFlow MRI\nusing its `Zenodo record \u003chttps://doi.org/10.5281/zenodo.5151590\u003e`_.\n\nIn the above link, scroll down to the \"Export\" section and select your favorite\nexport format to get an up-to-date citation.\n\nContributions\n-------------\n\nContributions of any kind are welcome! Open an issue or pull request to begin.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmrphys%2Ftensorflow-mri","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmrphys%2Ftensorflow-mri","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmrphys%2Ftensorflow-mri/lists"}