add bib

.gitignore

@@ -16,6 +16,7 @@
 !*.yml
 !*.yaml
 !*.toml
+!*.bib
 
 
 # even if in sub-dirs

paper/paper.bib

@@ -0,0 +1,114 @@
+@article{CG,
+  author = {{Hestenes}, Magnus Rudolph and {Stiefel}, Eduard and others},
+  title = {{Methods of conjugate gradients for solving linear systems}},
+  journal = {Journal of Research of the National Bureau of Standards},
+  volume = {49},
+  number = {1},
+  year = {1952},
+  publisher = {NBS Washington, DC}
+}
+
+@article{deep_learning,
+  author = {{Wang}, Ge and {Ye}, Jong Chu and {Mueller}, Klaus and {Fessler}, Jeffrey A.},
+  title = {{Image Reconstruction Is a New Frontier of Machine Learning}},
+  journal = {IEEE Transactions on Medical Imaging},
+  volume = {37},
+  number = {6},
+  pages = {1289--1296},
+  year = {2018},
+  month = jun,
+  doi = {10.1109/TMI.2018.2833635}
+}
+
+@article{physics_informed,
+  author = {{Aggarwal}, Hemant K. and {Mani}, Merry P. and {Jacob}, Mathews},
+  title = {{MoDL: Model-Based Deep Learning Architecture for Inverse Problems}},
+  journal = {IEEE Transactions on Medical Imaging},
+  volume = {38},
+  number = {2},
+  pages = {394--405},
+  year = {2019},
+  month = feb,
+  doi = {10.1109/TMI.2018.2865356}
+}
+
+@misc{MIRT,
+  author = {{Fessler}, Jeffrey A.},
+  title = {{Michigan Image Reconstruction Toolbox}},
+  url = {https://github.com/JeffFessler/mirt},
+  note = {Accessed: 2024-01-01},
+  year = {2024}
+}
+
+@inproceedings{SigPy,
+  author = {{Ong}, Frank and {Lustig}, Michael},
+  title = {{SigPy: a python package for high performance iterative reconstruction}},
+  booktitle = {Proceedings of the ISMRM 27th Annual Meeting},
+  volume = {4819},
+  number = {5},
+  year = {2019},
+  address = {Montreal, Quebec, Canada}
+}
+
+@article{FISTA,
+  author = {{Beck}, Amir and {Teboulle}, Marc},
+  title = {{A fast iterative shrinkage-thresholding algorithm for linear inverse problems}},
+  journal = {SIAM Journal on Imaging Sciences},
+  volume = {2},
+  number = {1},
+  pages = {183--202},
+  year = {2009},
+  publisher = {SIAM}
+}
+
+@article{primal_dual,
+  author = {{Chambolle}, Antonin and {Pock}, Thomas},
+  title = {{A first-order primal-dual algorithm for convex problems with applications to imaging}},
+  journal = {Journal of Mathematical Imaging and Vision},
+  volume = {40},
+  pages = {120--145},
+  year = {2011},
+  publisher = {Springer}
+}
+
+@article{proximal,
+  author = {{Parikh}, Neal and {Boyd}, Stephen and others},
+  title = {{Proximal algorithms}},
+  journal = {Foundations and Trends in Optimization},
+  volume = {1},
+  number = {3},
+  pages = {127--239},
+  year = {2014},
+  publisher = {Now Publishers, Inc.}
+}
+
+@article{sampling,
+  author = {{Wang}, Guanhua and {Nielsen}, Jon-Fredrik and {Fessler}, Jeffrey A. and {Noll}, Douglas C.},
+  title = {{Stochastic Optimization of Three-Dimensional Non-Cartesian Sampling Trajectory}},
+  journal = {Magnetic Resonance in Medicine},
+  volume = {90},
+  number = {2},
+  pages = {417--431},
+  year = {2023},
+  doi = {10.1002/mrm.29645}
+}
+
+@inproceedings{BART,
+  author = {{Uecker}, Martin and {Tamir}, Jonathan I. and {Ong}, Frank and {Lustig}, Michael},
+  title = {{The BART toolbox for computational magnetic resonance imaging}},
+  booktitle = {Proceedings of the International Society for Magnetic Resonance in Medicine},
+  volume = {24},
+  pages = {1},
+  year = {2016}
+}
+
+@article{Gadgetron,
+  author = {{Hansen}, Michael Schacht and {S{\o}rensen}, Thomas Sangild},
+  title = {{Gadgetron: an open source framework for medical image reconstruction}},
+  journal = {Magnetic Resonance in Medicine},
+  volume = {69},
+  number = {6},
+  pages = {1768--1776},
+  year = {2013},
+  publisher = {Wiley Online Library}
+}

paper/paper.md

@@ -65,9 +65,9 @@ With native PyTorch support, MIRTorch satisfies the following specific needs:
 
 1. Fast prototyping of model-based deep learning. Many related projects hard-coded the physics-informed components, such as system operators and iterative solvers. This non-modular approach hampers reproducibility and comparability. MIRTorch provides a standardized and modular implementation. It also facilitates transferring algorithms across different imaging modalities. Additionally, the code using MIRTorch better matches the mathematical expressions, facilitating understanding.
 
-2. Optimization of the imaging system. Since the toolbox is fully differentiable, it enables gradient methods for tuning imaging system parameters. For example, [@sampling1] and [@sampling2] use MIRTorch to optimize MRI sampling trajectories via stochastic gradient descent.
+2. Optimization of the imaging system. Since the toolbox is fully differentiable, it enables gradient methods for tuning imaging system parameters. For example, [@sampling] uses MIRTorch to optimize MRI sampling trajectories via stochastic gradient descent.
 
-3. User-friendly fast reconstruction. Many applications, such as fMRI, are vectorized and large-dimensional. CPU-based computation can be very time-consuming for iterative algorithms [@fMRI]. Benefitting from PyTorch's intrinsic multi-GPU support, user-friendly installation, and cross-platform capability, MIRTorch provides researchers with fast reconstruction at a minimal switching cost.
+3. User-friendly fast reconstruction. Many applications, such as functional MRI (fMRI), are vectorized and large-dimensional. CPU-based computation can be very time-consuming for iterative algorithms. Benefitting from PyTorch's intrinsic multi-GPU support, user-friendly installation, and cross-platform capability, MIRTorch provides researchers with fast reconstruction at a minimal switching cost.
 
 # Acknowledgements