Frequency-Supervised MR-to-CT Image Synthesis

TL;DR

This paper introduces a frequency-supervised deep learning method to improve high-frequency detail reconstruction in MR-to-CT image synthesis, addressing limitations of existing CNN-based approaches.

Contribution

It proposes a novel frequency decomposition layer and a high-frequency adversarial refinement module to enhance detail preservation in synthetic CT images.

Findings

Improved high-frequency detail in synthetic CT images.

Effective on a new brain MR-CT dataset with 45 pairs.

Code available for reproducibility.

Abstract

This paper strives to generate a synthetic computed tomography (CT) image from a magnetic resonance (MR) image. The synthetic CT image is valuable for radiotherapy planning when only an MR image is available. Recent approaches have made large strides in solving this challenging synthesis problem with convolutional neural networks that learn a mapping from MR inputs to CT outputs. In this paper, we find that all existing approaches share a common limitation: reconstruction breaks down in and around the high-frequency parts of CT images. To address this common limitation, we introduce frequency-supervised deep networks to explicitly enhance high-frequency MR-to-CT image reconstruction. We propose a frequency decomposition layer that learns to decompose predicted CT outputs into low- and high-frequency components, and we introduce a refinement module to improve high-frequency reconstruction through high-frequency adversarial learning. Experimental results on a new dataset with 45 pairs of 3D MR-CT brain images show the effectiveness and potential of the proposed approach. Code is available at \url{https://github.com/shizenglin/Frequency-Supervised-MR-to-CT-Image-Synthesis}.