Minor fixes broken urls, dockerfile, CITATION.cff, and adds paper preprints

Minor updates since v1.0.0

• Fixes broken urls on docs/, README.md, and notebooks under tutorials/.
• Fixes Dockerfile.
• Adds paper as listed:
  • ArXiv PrePrint,
  • SSRN PrePrint,
  • MIDL 2024 Short Paper.
• Updates CITATION.cff to SSRN citation.

ATOMMIC v1.0.0 Release Notes

This is the initial first release of ATOMMIC. We’ve decided to directly publish v1.0.0 instead of versions <1 since the project has been developed for a long time and reached a stable version.

Highlights

The Advanced Toolbox for Multitask Medical Imaging Consistency (ATOMMIC) s a toolbox for applying AI methods for accelerated MRI reconstruction (REC), MRI segmentation (SEG), quantitative MR imaging (qMRI), as well as multitask learning (MTL), i.e., performing multiple tasks simultaneously, such as reconstruction and segmentation. Each task is implemented in a separate collection consisting of data loaders, transformations, models, metrics, and losses.

Pretrained models are available on our HuggingFace account and can be downloaded and used for inference.

Collections

ATOMMIC is organized into the following collections in v1.0.0, each of which implements a specific task and various models as listed:

MultiTask Learning (MTL)

1. End-to-End Recurrent Attention Network (SERANet), 2. Image domain Deep Structured Low-Rank Network (IDSLR), 3. Image domain Deep Structured Low-Rank UNet (IDSLRUNet), 4. Multi-Task Learning for MRI Reconstruction and Segmentation (MTLRS), 5. Reconstruction Segmentation method using UNet (RecSegUNet), 6. Segmentation Network MRI (SegNet).

Quantitative MR Imaging (qMRI)

1. Quantitative Recurrent Inference Machines (qRIMBlock), 2. Quantitative End-to-End Variational Network (qVarNet), 3. Quantitative Cascades of Independently Recurrent Inference Machines (qCIRIM).

MRI Reconstruction (REC)

1. Cascades of Independently Recurrent Inference Machines (CIRIM), 2. Convolutional Recurrent Neural Networks (CRNNet), 3. Deep Cascade of Convolutional Neural Networks (CascadeNet), 4. Down-Up Net (DUNet), 5. End-to-End Variational Network (VarNet), 6. Independently Recurrent Inference Machines (RIMBlock), 7. Joint Deep Model-Based MR Image and Coil Sensitivity Reconstruction Network (JointICNet), 8. KIKINet, 9. Learned Primal-Dual Net (LPDNet), 10. Model-based Deep Learning Reconstruction (MoDL), 11. MultiDomainNet, 12. ProximalGradient, 13. Recurrent Inference Machines (RIMBlock), 14. Recurrent Variational Network (RecurrentVarNet), 15. UNet, 16. Variable Splitting Network (VSNet), 17. XPDNet, 18. Zero-Filled reconstruction (ZF).

MRI Segmentation (SEG)

1. SegmentationAttentionUNet, 2. SegmentationDYNUNet, 3. SegmentationLambdaUNet, 4. SegmentationUNet, 5. Segmentation3DUNet, 6. SegmentationUNetR, 7. SegmentationVNet.

MRI Datasets

The following public datasets are supported natively in v1.0.0:

• AHEAD: Supports the (qMRI) and (REC) tasks.
• BraTS 2023 Adult Glioma: Supports the (SEG) task.
• CC359: Supports the (REC) task.
• fastMRI Brains Multicoil: Supports the (REC) task.
• fastMRI Knees Multicoil: Supports the (REC) task.
• fastMRI Knees Singlecoil: Supports the (REC) task.
• ISLES 2022 Sub Acute Stroke: Supports the (SEG) task.
• SKM-TEA: Supports the (REC), (SEG), and (MTL) tasks.
• Stanford Knees: Supports the (REC) task.