Cross-modality image synthesis from unpaired data using CycleGAN: Effects of gradient consistency loss and training data size

TL;DR

This paper enhances CycleGAN for MR-to-CT image synthesis by incorporating gradient consistency loss, analyzing the impact of training data size, and evaluating segmentation accuracy on synthesized images.

Contribution

The study introduces gradient consistency loss into CycleGAN for improved boundary accuracy in unpaired MR-to-CT synthesis and assesses effects of training data size.

Findings

Gradient consistency loss improves boundary accuracy.

More training data enhances synthesis quality.

Synthesized images enable effective segmentation.

Abstract

CT is commonly used in orthopedic procedures. MRI is used along with CT to identify muscle structures and diagnose osteonecrosis due to its superior soft tissue contrast. However, MRI has poor contrast for bone structures. Clearly, it would be helpful if a corresponding CT were available, as bone boundaries are more clearly seen and CT has standardized (i.e., Hounsfield) units. Therefore, we aim at MR-to-CT synthesis. The CycleGAN was successfully applied to unpaired CT and MR images of the head, these images do not have as much variation of intensity pairs as do images in the pelvic region due to the presence of joints and muscles. In this paper, we extended the CycleGAN approach by adding the gradient consistency loss to improve the accuracy at the boundaries. We conducted two experiments. To evaluate image synthesis, we investigated dependency of image synthesis accuracy on 1) the number of training data and 2) the gradient consistency loss. To demonstrate the applicability of our method, we also investigated a segmentation accuracy on synthesized images.