# Learning Sampling and Model-Based Signal Recovery for Compressed Sensing MRI

arXiv: https://arxiv.org/abs/2004.10536 · HTML: https://mlsanity.com/paper/2004.10536/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Apr 2020.

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

- [A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learning](https://mlsanity.com/paper/2110.01515/) (2021)
- [On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction](https://mlsanity.com/paper/2203.16392/) (2022)
- [Dynamic Probabilistic Pruning: A general framework for hardware-constrained pruning at different granularities](https://mlsanity.com/paper/2105.12686/) (2021)
- [PUERT: Probabilistic Under-sampling and Explicable Reconstruction Network for CS-MRI](https://mlsanity.com/paper/2204.11189/) (2022)
- [Adaptive Illumination based Depth Sensing using Deep Superpixel and Soft Sampling Approximation](https://mlsanity.com/paper/2103.12297/) (2021)
- [Single-pass Object-adaptive Data Undersampling and Reconstruction for MRI](https://mlsanity.com/paper/2111.09212/) (2021)
- [Artificial Intelligence for Neuro MRI Acquisition: A Review](https://mlsanity.com/paper/2406.05982/) (2024)
- [Gradient-Based Learning of Discrete Structured Measurement Operators for Signal Recovery](https://mlsanity.com/paper/2202.03391/) (2022)
- [Reinforcement Learning for Sampling on Temporal Medical Imaging Sequences](https://mlsanity.com/paper/2308.14946/) (2023)

## Abstract (arXiv)

Compressed sensing (CS) MRI relies on adequate undersampling of the k-space to accelerate the acquisition without compromising image quality. Consequently, the design of optimal sampling patterns for these k-space coefficients has received significant attention, with many CS MRI methods exploiting variable-density probability distributions. Realizing that an optimal sampling pattern may depend on the downstream task (e.g. image reconstruction, segmentation, or classification), we here propose joint learning of both task-adaptive k-space sampling and a subsequent model-based proximal-gradient recovery network. The former is enabled through a probabilistic generative model that leverages the Gumbel-softmax relaxation to sample across trainable beliefs while maintaining differentiability. The proposed combination of a highly flexible sampling model and a model-based (sampling-adaptive) image reconstruction network facilitates exploration and efficient training, yielding improved MR image quality compared to other sampling baselines.