Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

On the Benefit of Dual-domain Denoising in a Self-supervised Low-dose CT Setting

Fabian Wagner, Mareike Thies, Laura Pfaff, Oliver Aust, Sabrina Pechmann, Daniela Weidner, Noah Maul, Maximilian Rohleder, Mingxuan Gu, Jonas Utz, Felix Denzinger, Andreas Maier

TL;DR

This work presents an end-to-end trainable CT reconstruction pipeline that contains denoising operators in both the projection and the image domain and that are optimized simultaneously without requiring ground-truth high-dose CT data.

Abstract

Computed tomography (CT) is routinely used for three-dimensional non-invasive imaging. Numerous data-driven image denoising algorithms were proposed to restore image quality in low-dose acquisitions. However, considerably less research investigates methods already intervening in the raw detector data due to limited access to suitable projection data or correct reconstruction algorithms. In this work, we present an end-to-end trainable CT reconstruction pipeline that contains denoising operators in both the projection and the image domain and that are optimized simultaneously without requiring ground-truth high-dose CT data. Our experiments demonstrate that including an additional projection denoising operator improved the overall denoising performance by 82.4-94.1%/12.5-41.7% (PSNR/SSIM) on abdomen CT and 1.5-2.9%/0.4-0.5% (PSNR/SSIM) on XRM data relative to the low-dose baseline. We make our entire helical CT reconstruction framework publicly available that contains a raw projection rebinning step to render helical projection data suitable for differentiable fan-beam reconstruction operators and end-to-end learning.

On the Benefit of Dual-domain Denoising in a Self-supervised Low-dose CT Setting

TL;DR

This work presents an end-to-end trainable CT reconstruction pipeline that contains denoising operators in both the projection and the image domain and that are optimized simultaneously without requiring ground-truth high-dose CT data.

Abstract

Computed tomography (CT) is routinely used for three-dimensional non-invasive imaging. Numerous data-driven image denoising algorithms were proposed to restore image quality in low-dose acquisitions. However, considerably less research investigates methods already intervening in the raw detector data due to limited access to suitable projection data or correct reconstruction algorithms. In this work, we present an end-to-end trainable CT reconstruction pipeline that contains denoising operators in both the projection and the image domain and that are optimized simultaneously without requiring ground-truth high-dose CT data. Our experiments demonstrate that including an additional projection denoising operator improved the overall denoising performance by 82.4-94.1%/12.5-41.7% (PSNR/SSIM) on abdomen CT and 1.5-2.9%/0.4-0.5% (PSNR/SSIM) on XRM data relative to the low-dose baseline. We make our entire helical CT reconstruction framework publicly available that contains a raw projection rebinning step to render helical projection data suitable for differentiable fan-beam reconstruction operators and end-to-end learning.
Paper Structure (10 sections, 6 equations, 3 figures, 2 tables)

This paper contains 10 sections, 6 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methods
  3. End-to-end CT denoising
  4. Self-supervised training
  5. Projection rebinning
  6. Experiments
  7. Results and discussion
  8. Conclusion
  9. Compliance with ethical standards
  10. Acknowledgments

Figures (3)

  • Figure 1: Overview of the proposed trainable dual-domain and self-supervised CT denoising pipeline.
  • Figure 2: Illustration of the proposed end-to-end self-supervised denoising pipeline. Projections are split in two stacks and reconstructed separately. The loss calculated between denoised prediction and reconstructed target stack is backpropagated through $\mathbf{D}_{\text{img}}^w$ and $\mathbf{R}$ to $\mathbf{D}_{\text{proj}}^w$ to optimize all denoising operators. The gradient flow from Eq. \ref{['eq:gradient_flow']} is indicated by dashed lines.
  • Figure 3: Examplary predictions of an abdomen CT slice (left) and the cross sections of a mouse tibia bone (XRM) close to the knee area (right). Below high-dose overview images, ROIs (red squares) of (LD) low-dose, (HD) high-dose, (1a) dual U-Nets (self-supervised), (1b) dual U-Nets (supervised), (1c) reco U-Net (self-supervised), (2a) dual BFs (self-supervised), (2b) dual BFs (supervised), and (2c) reco BF (self-supervised) predictions are presented. Windows are $[-150, 250]\,\text{HU}$ (abdomen) and $[0.05, 0.32]\,\text{arb.}\,\text{unit}$ (XRM).