Framing U-Net via Deep Convolutional Framelets: Application to Sparse-view CT

TL;DR

This paper analyzes the limitations of U-Net in sparse-view CT reconstruction and proposes new multi-resolution deep learning architectures based on deep convolutional framelets, demonstrating improved performance on real patient data.

Contribution

It introduces U-Net variants satisfying frame conditions, enhancing high-frequency edge recovery in sparse-view CT, supported by theoretical insights and extensive experiments.

Findings

U-Net variants with frame conditions outperform standard U-Net.

New architectures better recover high-frequency details.

Experimental results show improved reconstruction quality.

Abstract

X-ray computed tomography (CT) using sparse projection views is a recent approach to reduce the radiation dose. However, due to the insufficient projection views, an analytic reconstruction approach using the filtered back projection (FBP) produces severe streaking artifacts. Recently, deep learning approaches using large receptive field neural networks such as U-Net have demonstrated impressive performance for sparse- view CT reconstruction. However, theoretical justification is still lacking. Inspired by the recent theory of deep convolutional framelets, the main goal of this paper is, therefore, to reveal the limitation of U-Net and propose new multi-resolution deep learning schemes. In particular, we show that the alternative U- Net variants such as dual frame and the tight frame U-Nets satisfy the so-called frame condition which make them better for effective recovery of high frequency edges in sparse view- CT. Using extensive experiments with real patient data set, we demonstrate that the new network architectures provide better reconstruction performance.