Mridc Versions Save

What's in

Version 0.2.0 is a major stable update.

The following tools have been implemented:

• Quantitative MRI data loaders & transforms.
• Quantitative MRI models: qCIRIM, qRIM, qE2EVN.
• AHEAD dataset preprocessing tools for quantitative MRI.

What's Changed

• added quantitative MRI tools by @wdika in https://github.com/wdika/mridc/pull/82
• Bump to v.0.2.0 by @wdika in https://github.com/wdika/mridc/pull/101

Full Changelog: https://github.com/wdika/mridc/compare/v.0.1.1...v.0.2.0

Important

• Python 3.10 cannot be supported yet due to onnxruntime 1.11.1 inconsistency (#69).
• hydra-core>=1.2.0 cannot be supported due to omegaconf > 2.1 inconsistency (#72).

What's in

Version 0.1.1 is a minor stable update.

The following tools have been implemented:

• Noise PreWhitening
• Geometric Decomposition Coil Compression
• The RIM-based models can support multi-slice 2D inputs.

What's Changed

• fetch upstream updates by @wdika in https://github.com/wdika/mridc/pull/65
• Addresses #57, #58, #60 by @wdika in https://github.com/wdika/mridc/pull/66
• Refactor workflows & Add tox by @wdika in https://github.com/wdika/mridc/pull/70
• Bump to v.0.1.1 by @wdika in https://github.com/wdika/mridc/pull/74

Full Changelog: https://github.com/wdika/mridc/compare/v.0.1.0...v.0.1.1

Important

• Python 3.10 cannot be supported yet due to onnxruntime 1.11.1 inconsistency (#69).
• hydra-core>=1.2.0 cannot be supported due to omegaconf > 2.1 inconsistency (#72).

What's in

Version 0.1.0 is a major stable update.

The following reconstruction methods have been added:

• Convolutional Recurrent Neural Networks (CRNN)
• Deep Cascade of Convolutional Neural Networks (CCNN)
• Down-Up Net (DUNET)
• Joint Deep Model-Based MR Image and Coil Sensitivity Reconstruction Network (Joint-ICNet)
• KIKI-Net
• Learned Primal-Dual Net (LPDNet)
• MultiDomainNet
• Recurrent Variational Network (RVN)
• Variable Splitting Network (VSNet)
• XPDNet

PyTorch-Lighting, Hydra, ONXX are now used.

What's Changed

• Rebased - refactored logic by @wdika in https://github.com/wdika/mridc/pull/34
• Updated base recon model inheritance to other models by @wdika in ce648e73691471aefa3b35b1e1b91b4b049dd059
• Update README.md by @wdika in https://github.com/wdika/mridc/pull/39
• fixed trainer, updated PL to 1.6.0 by @wdika in https://github.com/wdika/mridc/commit/4cc91ba0e348e52935de365d10fafc7b66875eaa
• [fixes documentation & codecov uploader @wdika in e2c50f3bec7bfc73fdab9abd82bf05d3f4c4b7fc

Full Changelog: https://github.com/wdika/mridc/compare/v.0.0.1...v.0.1.0

What's in

This initial version includes the implementation of the following reconstruction methods:

• Cascades of Independently Recurrent Inference Machines (CIRIM),
• Independently Recurrent Inference Machines (IRIM),
• End-to-End Variational Network (E2EVN),
• the UNet,
• Compressed Sensing (CS), and
• zero-filled (ZF).

Also, includes coil sensitivity estimation and evaluation or reconstructions.

What's Changed

• Circleci project setup by @wdika in https://github.com/wdika/mridc/pull/2
• Remove assert statement from non-test files by @deepsource-autofix in https://github.com/wdika/mridc/pull/3
• Remove unnecessary parentheses after keyword by @deepsource-autofix in https://github.com/wdika/mridc/pull/4

New Contributors

• @wdika made their first contribution in https://github.com/wdika/mridc/pull/2
• @deepsource-autofix made their first contribution in https://github.com/wdika/mridc/pull/3

Full Changelog: https://github.com/wdika/mridc/commits/v.0.0.1