RARE: Image Reconstruction using Deep Priors Learned without Ground Truth

TL;DR

RARE introduces a novel image reconstruction method that learns artifact-removal priors directly from undersampled measurements, enabling high-quality reconstructions without requiring fully-sampled ground truth data.

Contribution

It extends RED by using artifact-removal networks trained on undersampled data, broadening applicability to practical scenarios with limited ground truth.

Findings

Effective reconstruction of 3D MRIs from heavily undersampled data.

Demonstrates potential of learned regularizers on real and simulated data.

Achieves high-quality images without fully-sampled groundtruth.

Abstract

Regularization by denoising (RED) is an image reconstruction framework that uses an image denoiser as a prior. Recent work has shown the state-of-the-art performance of RED with learned denoisers corresponding to pre-trained convolutional neural nets (CNNs). In this work, we propose to broaden the current denoiser-centric view of RED by considering priors corresponding to networks trained for more general artifact-removal. The key benefit of the proposed family of algorithms, called regularization by artifact-removal (RARE), is that it can leverage priors learned on datasets containing only undersampled measurements. This makes RARE applicable to problems where it is practically impossible to have fully-sampled groundtruth data for training. We validate RARE on both simulated and experimentally collected data by reconstructing a free-breathing whole-body 3D MRIs into ten respiratory phases from heavily undersampled k-space measurements. Our results corroborate the potential of learning regularizers for iterative inversion directly on undersampled and noisy measurements.