quantitative (#82)

* implemented quantitative imaging 
* added  qRIM, qCIRIM & qE2EVN models and ahead dataset preprocessing scripts

README.md

@@ -12,6 +12,7 @@
 is provided for MRI data analysis. Primarily it focuses on the following tasks:**
 
 ### **Reconstruction**:
+The following models are implemented for accelerated MRI reconstruction:
 1.[Cascades of Independently Recurrent Inference Machines (CIRIM)](https://iopscience.iop.org/article/10.1088/1361-6560/ac6cc2),
 2.[Compressed Sensing (CS)](https://ieeexplore.ieee.org/document/4472246),
 3.[Convolutional Recurrent Neural Networks (CRNN)](https://ieeexplore.ieee.org/document/8425639),
@@ -30,17 +31,22 @@ is provided for MRI data analysis. Primarily it focuses on the following tasks:*
 16.[XPDNet](https://arxiv.org/abs/2010.07290),
 17.and Zero-Filled reconstruction (ZF).
 
+### **Quantitative Imaging**:
+The following models are implemented for quantitative imaging:
+1.[quantitative Cascades of Independently Recurrent Inference Machines (qCIRIM)](https://iopscience.iop.org/article/10.1088/1361-6560/ac6cc2),
+2.[quantitative End-to-End Variational Network (qE2EVN)](https://link.springer.com/chapter/10.1007/978-3-030-59713-9_7),
+3.[quantitative Independently Recurrent Inference Machines (qIRIM)](http://arxiv.org/abs/2012.07819),
+4.[quantitative Recurrent Inference Machines (qRIM)](https://www.sciencedirect.com/science/article/abs/pii/S1361841518306078?via%3Dihub),
+
+_Note: Currently only the above models are implemented. More models can be added by extending the reconstruction models
+for quantitative imaging. If you wish to extend the toolbox, please open an issue._
+
 ### **Segmentation**:
 _Coming soon..._
 
-### **Acknowledgements**
-
-MRIDC is based on the [NeMo](https://github.com/NVIDIA/NeMo) framework, using PyTorch Lightning for feasible
-high-performance multi-GPU/multi-node mixed-precision training.
+## Usage
 
-For the reconstruction methods:
-- the implementations of 6 and 14 are thanks to and based on the [fastMRI repo](https://github.com/facebookresearch/fastMRI).
-- The implementations of 7, 9, 10, 11, 13, and 16 are thanks to and based on the [DIRECT repo](https://github.com/NKI-AI/direct).
+Check the [projects](https://github.com/wdika/mridc/blob/main/projects/README.md) page for more information of how to use **mridc**.
 
 ## Installation
 
@@ -66,17 +72,6 @@ cd mridc
 ./reinstall.sh
 ```
 
-## Usage
-
-Check the [projects](https://github.com/wdika/mridc/blob/main/projects/README.md) page for more information of how to use **mridc**.
-
-### Datasets
-
-Recommended public datasets to use with this repo:
-
-- [fastMRI](http://arxiv.org/abs/1811.08839),
-- [Fully Sampled Knees](http://old.mridata.org/fullysampled/knees/).
-
 ## API Documentation
 
 [![Documentation Status](https://readthedocs.org/projects/mridc/badge/?version=latest)](https://mridc.readthedocs.io/en/latest/?badge=latest)
@@ -87,15 +82,24 @@ Access the API Documentation [here](https://mridc.readthedocs.io/en/latest/modul
 
 [![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
 
+## Acknowledgements
+
+MRIDC is based on the [NeMo](https://github.com/NVIDIA/NeMo) framework, using PyTorch Lightning for feasible
+high-performance multi-GPU/multi-node mixed-precision training.
+
+For the reconstruction methods:
+- the implementations of 6 and 14 are thanks to and based on the [fastMRI repo](https://github.com/facebookresearch/fastMRI).
+- The implementations of 7, 9, 10, 11, 13, and 16 are thanks to and based on the [DIRECT repo](https://github.com/NKI-AI/direct).
+
 ## Citation
 
 Please cite MRIDC using the "_Cite this repository_" button or as
 
 ```BibTeX
 @misc{mridc,
-    author = {Karkalousos, Dimitrios and Caan, Matthan},
+    author = {Karkalousos Dimitrios, Zhang Chaoping, and Caan Matthan},
     title = {MRIDC: Data Consistency for Magnetic Resonance Imaging},
-    year = {2021},
+    year = {2022},
     url = {https://github.com/wdika/mridc},
 }
 ```
@@ -106,3 +110,5 @@ The following papers use the MRIDC repo:
 
 [1] [Karkalousos, D. et al. (2021) ‘Assessment of Data Consistency through Cascades of Independently Recurrent
 Inference Machines for fast and robust accelerated MRI reconstruction’](https://iopscience.iop.org/article/10.1088/1361-6560/ac6cc2)
+
+[2] Zhang, C. et al. (2022) 'A unified model for reconstruction and R2* mapping of accelerated 7T data using the quantitative Recurrent Inference Machine'

docs/source/mridc.collections.quantitative.data.rst

@@ -0,0 +1,21 @@
+mridc.collections.quantitative.data package
+===========================================
+
+Submodules
+----------
+
+mridc.collections.quantitative.data.qmri\_data module
+-----------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.data.qmri_data
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative.data
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.quantitative.models.qrim.rst

@@ -0,0 +1,29 @@
+mridc.collections.quantitative.models.qrim package
+==================================================
+
+Submodules
+----------
+
+mridc.collections.quantitative.models.qrim.qrim\_block module
+-------------------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.qrim.qrim_block
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+mridc.collections.quantitative.models.qrim.utils module
+-------------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.qrim.utils
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative.models.qrim
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.quantitative.models.qvarnet.rst

@@ -0,0 +1,21 @@
+mridc.collections.quantitative.models.qvarnet package
+=====================================================
+
+Submodules
+----------
+
+mridc.collections.quantitative.models.qvarnet.qvn\_block module
+---------------------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.qvarnet.qvn_block
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative.models.qvarnet
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.quantitative.models.rst

@@ -0,0 +1,46 @@
+mridc.collections.quantitative.models package
+=============================================
+
+Subpackages
+-----------
+
+.. toctree::
+   :maxdepth: 4
+
+   mridc.collections.quantitative.models.qrim
+   mridc.collections.quantitative.models.qvarnet
+
+Submodules
+----------
+
+mridc.collections.quantitative.models.base module
+-------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.base
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+mridc.collections.quantitative.models.qcirim module
+---------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.qcirim
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+mridc.collections.quantitative.models.qvn module
+------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.models.qvn
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative.models
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.quantitative.parts.rst

@@ -0,0 +1,21 @@
+mridc.collections.quantitative.parts package
+============================================
+
+Submodules
+----------
+
+mridc.collections.quantitative.parts.transforms module
+------------------------------------------------------
+
+.. automodule:: mridc.collections.quantitative.parts.transforms
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative.parts
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.quantitative.rst

@@ -0,0 +1,20 @@
+mridc.collections.quantitative package
+======================================
+
+Subpackages
+-----------
+
+.. toctree::
+   :maxdepth: 4
+
+   mridc.collections.quantitative.data
+   mridc.collections.quantitative.models
+   mridc.collections.quantitative.parts
+
+Module contents
+---------------
+
+.. automodule:: mridc.collections.quantitative
+   :members:
+   :undoc-members:
+   :show-inheritance:

docs/source/mridc.collections.rst

@@ -8,6 +8,7 @@ Subpackages
    :maxdepth: 4
 
    mridc.collections.common
+   mridc.collections.quantitative
    mridc.collections.reconstruction
 
 Module contents

mridc/collections/common/losses/ssim.py

@@ -38,9 +38,11 @@ def forward(self, X: torch.Tensor, Y: torch.Tensor, data_range: torch.Tensor):
         -------
         SSIM loss.
         """
-        if not isinstance(self.w, torch.Tensor):
+        if not isinstance(self.w, torch.Tensor):  # type: ignore
             raise AssertionError
 
+        self.w = self.w.to(X)  # type: ignore
+
         data_range = data_range[:, None, None, None]
         C1 = (self.k1 * data_range) ** 2
         C2 = (self.k2 * data_range) ** 2

mridc/collections/quantitative/__init__.py

@@ -0,0 +1,12 @@
+# encoding: utf-8
+
+from mridc.package_info import __version__
+
+# Set collection version equal to MRIDC version.
+__version = __version__
+
+# Authorship.
+__author__ = "Dimitrios Karkalousos, Chaoping Zhang"
+
+# Set collection name.
+__description__ = "Quantitative models collection"

mridc/collections/quantitative/data/__init__.py

@@ -0,0 +1,2 @@
+# encoding: utf-8
+__author__ = "Dimitrios Karkalousos, Chaoping Zhang"