Robust Compressed Sensing MRI with Deep Generative Priors
arXiv:2108.01368
Abstract
The CSGM framework (Bora-Jalal-Price-Dimakis'17) has shown that deep generative priors can be powerful tools for solving inverse problems. However, to date this framework has been empirically successful only on certain datasets (for example, human faces and MNIST digits), and it is known to perform poorly on out-of-distribution samples. In this paper, we present the first successful application of the CSGM framework on clinical MRI data. We train a generative prior on brain scans from the fastMRI dataset, and show that posterior sampling via Langevin dynamics achieves high quality reconstructions. Furthermore, our experiments and theory show that posterior sampling is robust to changes in the ground-truth distribution and measurement process. Our code and models are available at: \url{https://github.com/utcsilab/csgm-mri-langevin}.
References in corpus (21)
- Auto-Encoding Variational Bayes
- Deep Convolutional Neural Network for Inverse Problems in Imaging
- Simultaneous analysis of Lasso and Dantzig selector
- Generative Modeling by Estimating Gradients of the Data Distribution
- On instabilities of deep learning in image reconstruction - Does AI come at a cost?
- Results of the 2020 fastMRI Challenge for Machine Learning MR Image Reconstruction
- Compressed Sensing using Generative Models
- Improved Techniques for Training Score-Based Generative Models
- Deep Generative Adversarial Networks for Compressed Sensing Automates MRI
- Measuring Robustness in Deep Learning Based Compressive Sensing
- Plug-in Estimation in High-Dimensional Linear Inverse Problems: A Rigorous Analysis
- Modeling Sparse Deviations for Compressed Sensing using Generative Models
- Invertible generative models for inverse problems: mitigating representation error and dataset bias
- Simultaneous Multi-Slice MRI using Cartesian and Radial FLASH and Regularized Nonlinear Inversion: SMS-NLINV
- Task-Aware Compressed Sensing with Generative Adversarial Networks
- Global Guarantees for Blind Demodulation with Generative Priors
- Lower Bounds for Compressed Sensing with Generative Models
- Deep Learning Techniques for Inverse Problems in Imaging
- Robust Compressed Sensing using Generative Models
- Prior Image-Constrained Reconstruction using Style-Based Generative Models
- Instance-Optimal Compressed Sensing via Posterior Sampling