# Regularization-Agnostic Compressed Sensing MRI Reconstruction with Hypernetworks

arXiv: https://arxiv.org/abs/2101.02194 · HTML: https://mlsanity.com/paper/2101.02194/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 33 references that are in the map.
- First posted on arXiv 6 Jan 2021.

## Builds on (most cited first, 17 of 17)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [VoxelMorph: A Learning Framework for Deformable Medical Image Registration](https://mlsanity.com/paper/1809.05231/) (2018)
- [SMASH: One-Shot Model Architecture Search through HyperNetworks](https://mlsanity.com/paper/1708.05344/) (2017)
- [Inference Suboptimality in Variational Autoencoders](https://mlsanity.com/paper/1801.03558/) (2018)
- [Graph HyperNetworks for Neural Architecture Search](https://mlsanity.com/paper/1810.05749/) (2018)
- [Iterative Amortized Inference](https://mlsanity.com/paper/1807.09356/) (2018)
- [HyperMorph: Amortized Hyperparameter Learning for Image Registration](https://mlsanity.com/paper/2101.01035/) (2021)
- [Stochastic Hyperparameter Optimization through Hypernetworks](https://mlsanity.com/paper/1802.09419/) (2018)
- [Bayesian Hypernetworks](https://mlsanity.com/paper/1710.04759/) (2017)
- [Hypernetwork functional image representation](https://mlsanity.com/paper/1902.10404/) (2019)
- [Image Reconstruction: From Sparsity to Data-adaptive Methods and Machine Learning](https://mlsanity.com/paper/1904.02816/) (2019)
- [Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentation](https://mlsanity.com/paper/1903.03148/) (2019)
- [Amortized Inference Regularization](https://mlsanity.com/paper/1805.08913/) (2018)
- [Neural Network-based Reconstruction in Compressed Sensing MRI Without Fully-sampled Training Data](https://mlsanity.com/paper/2007.14979/) (2020)
- [HyperST-Net: Hypernetworks for Spatio-Temporal Forecasting](https://mlsanity.com/paper/1809.10889/) (2018)
- [Optimization Methods for Inverse Problems](https://mlsanity.com/paper/1712.00154/) (2017)
- [Meta Networks for Neural Style Transfer](https://mlsanity.com/paper/1709.04111/) (2017)

## Built on by (most cited first, 9 of 9)

- [Magnitude Invariant Parametrizations Improve Hypernetwork Learning](https://mlsanity.com/paper/2304.07645/) (2023)
- [A review and experimental evaluation of deep learning methods for MRI reconstruction](https://mlsanity.com/paper/2109.08618/) (2021)
- [Hyper-Convolution Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/2105.10559/) (2021)
- [Combiner and HyperCombiner Networks: Rules to Combine Multimodality MR Images for Prostate Cancer Localisation](https://mlsanity.com/paper/2307.08279/) (2023)
- [HyperLips: Hyper Control Lips with High Resolution Decoder for Talking Face Generation](https://mlsanity.com/paper/2310.05720/) (2023)
- [Hyper-Convolutions via Implicit Kernels for Medical Imaging](https://mlsanity.com/paper/2202.02701/) (2022)
- [Generalizing Supervised Deep Learning MRI Reconstruction to Multiple and Unseen Contrasts using Meta-Learning Hypernetworks](https://mlsanity.com/paper/2307.06771/) (2023)
- [HyperCoil-Recon: A Hypernetwork-based Adaptive Coil Configuration Task Switching Network for MRI Reconstruction](https://mlsanity.com/paper/2308.04821/) (2023)
- [RBF Weighted Hyper-Involution for RGB-D Object Detection](https://mlsanity.com/paper/2310.00342/) (2023)

## Abstract (arXiv)

Reconstructing under-sampled k-space measurements in Compressed Sensing MRI (CS-MRI) is classically solved with regularized least-squares. Recently, deep learning has been used to amortize this optimization by training reconstruction networks on a dataset of under-sampled measurements. Here, a crucial design choice is the regularization function(s) and corresponding weight(s). In this paper, we explore a novel strategy of using a hypernetwork to generate the parameters of a separate reconstruction network as a function of the regularization weight(s), resulting in a regularization-agnostic reconstruction model. At test time, for a given under-sampled image, our model can rapidly compute reconstructions with different amounts of regularization. We analyze the variability of these reconstructions, especially in situations when the overall quality is similar. Finally, we propose and empirically demonstrate an efficient and data-driven way of maximizing reconstruction performance given limited hypernetwork capacity. Our code is publicly available at https://github.com/alanqrwang/RegAgnosticCSMRI.