Federated Learning-Based Virtual Dual-Energy CT Generation from Single-Energy CT for Gout Detection

Digital Health, 2025
Yufang Dong1,2*, Min Liu1,2*, Jiajun Feng1,3,4*, Yuezhe Yang2*, Yong Dai5, Zhe Jin2†
*Equal Contribution †Corresponding author
1Nankai University, 2Anhui University, 3Guangzhou First People's Hospital, 4South China University of Technology, 5Anhui University of Science and Technology
Paper

Abstract

Objective: This study aims to develop and validate OneGout, a federated learning (FL)-based framework for early and accurate gout diagnosis to address the limitations of current diagnostic methods, specifically the invasiveness of joint aspiration and the accessibility, cost, and radiation exposure associated with advanced imaging techniques like dual-energy computed tomography (DECT).

Methods: We introduce OneGout, which pioneers a deep learning-based method for generating virtual DECT images. This approach offers a low-cost and low-radiation alternative for gout diagnosis. Furthermore, OneGout integrates federated learning (OneGout-FL) to enable collaborative model training across multiple medical institutions while ensuring patient data privacy is preserved.

Results: Experiments demonstrate that our method successfully generates high-quality virtual DECT images. The framework based on U-Net achieves a PSNR of 22.44 dB and an SSIM of 0.92 for the generation of 140kV from 80kV images. It also shows strong diagnostic performance, with an IoU of 46.66 and a Dice score of 63.20, indicating promising accuracy comparable to diagnoses made with real DECT scans.

Conclusion: OneGout presents an efficient, scalable, and privacy-preserving diagnostic solution for gout, particularly beneficial for resource-limited medical institutions. This framework has the potential to significantly enhance global gout management by providing a more accessible and safer diagnostic alternative.

Highlights

  • Generate virtual DECT images from single-energy CT scans.
  • Enable privacy-preserving multi-institution training with FL.
  • Support low-cost and low-radiation gout diagnosis.

Method

OneGout overview

Overview of the proposed OneGout framework.

Federated learning workflow

Federated learning workflow of OneGout-FL.

Results

Quantitative Results

Task Metric Result
140kV generation from 80kV images PSNR 22.44 dB
140kV generation from 80kV images SSIM 0.92
Gout detection IoU 46.66
Gout detection Dice 63.20

Quantitative performance reported by OneGout.

Qualitative Results

Virtual DECT generation results

Virtual DECT generation examples.

Gout detection results

Gout detection visualization.

Comparative analysis

Comparative analysis of generated and reference images.

Citation

@article{dong2025federated,
  title={Federated learning-based virtual dual-energy CT generation from single-energy CT for gout detection},
  author={Dong, Yufang and Liu, Min and Feng, Jiajun and Yang, Yuezhe and Dai, Yong and Jin, Zhe},
  journal={Digital Health},
  volume={11},
  pages={20552076251375570},
  year={2025},
  publisher={SAGE Publications Sage UK: London, England}
}