Edinburgh Imaging

Menu

28 Oct 20. Featured Paper

DiCyc: GAN-based deformation invariant cross-domain information fusion for medical image synthesis.

Link to paper in Information Fusion

 

Authors

Chengjia Wang, Guang Yang, Giorgos Papanastasiou, Sotirios A. Tsaftaris, David E.Newby, Calum Gray, Gillian Macnaught, Tom J. MacGillivray

 

Abstract

Cycle-consistent generative adversarial network (CycleGAN) has been widely used for cross-domain medical image synthesis tasks particularly due to its ability to deal with unpaired data.

However, most CycleGAN-based synthesis methods can not achieve good alignment between the synthesized images & data from the source domain, even with additional image alignment losses.

This is because the CycleGAN generator network can encode the relative deformations & noises associated to different domains.

This can be detrimental for the downstream applications that rely on the synthesized images, such as generating pseudo-CT for PET-MR attenuation correction.

In this paper, we present a deformation invariant cycle-consistency model that can filter out these domain-specific deformation.

The deformation is globally parameterized by thin-plate-spline (TPS), & locally learned by modified deformable convolutional layers.

Robustness to domain-specific deformations has been evaluated through experiments on multi-sequence brain MR data & multi-modality abdominal CTMR data.

Experiment results demonstrated that our method can achieve better alignment between the source & target data while maintaining superior image quality of signal compared to several state-of-the-art CycleGAN-based methods.

 

Keywords
  • GAN

  • Image synthesis

  • Information fusion

 

 

Social media tags & titles

Featured paper: DiCyc: GAN-based deformation invariant cross-domain information fusion for medical image synthesis. @EdinUniCVS @SchoolOfEng_UoE #ImageSynthesis