Data-Efficient Limited-Angle CT Using Deep Priors and Regularization

TL;DR

This paper introduces a data-efficient deep prior-based method for limited-angle CT reconstruction, effectively reducing artifacts with minimal training data by combining multiple regularization techniques and gradient-based optimization.

Contribution

It presents a novel approach that reconstructs images from severely limited-angle sinograms using very few data points and multiple regularizations, outperforming prior methods in data efficiency.

Findings

Achieved high-quality reconstructions with only 12 data points.

Comparable results to state-of-the-art approaches with minimal data.

Effective use of deep priors and regularization in ill-posed inverse problems.

Abstract

Reconstructing an image from its Radon transform is a fundamental computed tomography (CT) task arising in applications such as X-ray scans. In many practical scenarios, a full 180-degree scan is not feasible, or there is a desire to reduce radiation exposure. In these limited-angle settings, the problem becomes ill-posed, and methods designed for full-view data often leave significant artifacts. We propose a very low-data approach to reconstruct the original image from its Radon transform under severe angle limitations. Because the inverse problem is ill-posed, we combine multiple regularization methods, including Total Variation, a sinogram filter, Deep Image Prior, and a patch-level autoencoder. We use a differentiable implementation of the Radon transform, which allows us to use gradient-based techniques to solve the inverse problem. Our method is evaluated on a dataset from the Helsinki Tomography Challenge 2022, where the goal is to reconstruct a binary disk from its limited-angle sinogram. We only use a total of 12 data points--eight for learning a prior and four for hyperparameter selection--and achieve results comparable to the best synthetic data-driven approaches.