Flip Distribution Alignment VAE for Multi-Phase MRI Synthesis

Xiaoyan Kui, Qianmu Xiao, Qqinsong Li, Zexin Ji, JIelin Zhang +1 more
10/3/2025
cs.CV

Official code: QianMuXiao/FDA-VAE (1 star, MIT, last commit 2025-06-11).

Abstract

Separating shared and independent features is crucial for multi-phase contrast-enhanced (CE) MRI synthesis. However, existing methods use deep autoencoder generators with low parameter efficiency and lack interpretable training strategies. In this paper, we propose Flip Distribution Alignment Variational Autoencoder (FDA-VAE), a lightweight feature-decoupled VAE model for multi-phase CE MRI synthesis. Our method encodes input and target images into two latent distributions that are symmetric concerning a standard normal distribution, effectively separating shared and independent features. The Y-shaped bidirectional training strategy further enhances the interpretability of feature separation. Experimental results show that compared to existing deep autoencoder-based end-to-end synthesis methods, FDA-VAE significantly reduces model parameters and inference time while effectively improving synthesis quality. The source code is publicly available at https://github.com/QianMuXiao/FDA-VAE.

View on arXivView PDF

Code Implementations(1)

QianMuXiao/FDA-VAEOfficial100%
MIT

Source code for Flip Distribution Alignment VAE for Multi-Phase MRI Synthesis

10PythonJun 10, 20251 years agoMIT

Similar Papers

McCaD: Multi-Contrast MRI Conditioned, Adaptive Adversarial Diffusion Model for High-Fidelity MRI Synthesis

84% match1 repo102024

ViewFusion: Towards Multi-View Consistency via Interpolated Denoising

80% match1 repo542026

Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation

80% match1 repo3672024

Pyramid Hierarchical Masked Diffusion Model for Imaging Synthesis

79% match2025

Generic 3D Diffusion Adapter Using Controlled Multi-View Editing

79% match1 repo3462024

EscherNet: A Generative Model for Scalable View Synthesis

79% match1 repo3772026

Cite this paper

@article{kui2025flip,
  title  = {Flip Distribution Alignment VAE for Multi-Phase MRI Synthesis},
  author = {Xiaoyan Kui and Qianmu Xiao and Qqinsong Li and Zexin Ji and JIelin Zhang and Beiji Zou},
  year   = {2025},
  eprint = {2510.02970},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2510.02970v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2510.02970.svg)](https://codeofpaper.com/paper/2510.02970)

Discussion