3D Nephrographic Image Synthesis in CT Urography with the Diffusion Model and Swin Transformer

2025-02-26

AI Summary

To address the increased radiation dose associated with missing nephrographic-phase images in CT urography (CTU), this work proposes an end-to-end 3D medical image synthesis method integrating diffusion models with Swin Transformers, enabling high-fidelity reconstruction of nephrographic-phase volumetric images solely from low-dose unenhanced and excretory-phase scans. This is the first application of Swin Transformers within a 3D diffusion framework for multiphase CTU synthesis. The method incorporates affine registration and 3D voxel-wise reconstruction, and is validated via quantitative metrics—PSNR (26.3 ± 4.4 dB), SSIM (0.84 ± 0.069), Fréchet Volume Distance (FVD)—and blinded radiologist assessment (3.4/5.0, *p* = 0.5). Compared to standard three-phase CTU, the proposed approach reduces radiation exposure by 33.3% while yielding diagnostically acceptable synthetic images with no statistically significant difference from clinical ground truth.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Security and Privacy: Data transparency and provenance

Abstract

Purpose: This study aims to develop and validate a method for synthesizing 3D nephrographic phase images in CT urography (CTU) examinations using a diffusion model integrated with a Swin Transformer-based deep learning approach. Materials and Methods: This retrospective study was approved by the local Institutional Review Board. A dataset comprising 327 patients who underwent three-phase CTU (mean $pm$ SD age, 63 $pm$ 15 years; 174 males, 153 females) was curated for deep learning model development. The three phases for each patient were aligned with an affine registration algorithm. A custom deep learning model coined dsSNICT (diffusion model with a Swin transformer for synthetic nephrographic phase images in CT) was developed and implemented to synthesize the nephrographic images. Performance was assessed using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Mean Absolute Error (MAE), and Fr'{e}chet Video Distance (FVD). Qualitative evaluation by two fellowship-trained abdominal radiologists was performed. Results: The synthetic nephrographic images generated by our proposed approach achieved high PSNR (26.3 $pm$ 4.4 dB), SSIM (0.84 $pm$ 0.069), MAE (12.74 $pm$ 5.22 HU), and FVD (1323). Two radiologists provided average scores of 3.5 for real images and 3.4 for synthetic images (P-value = 0.5) on a Likert scale of 1-5, indicating that our synthetic images closely resemble real images. Conclusion: The proposed approach effectively synthesizes high-quality 3D nephrographic phase images. This model can be used to reduce radiation dose in CTU by 33.3% without compromising image quality, which thereby enhances the safety and diagnostic utility of CT urography.

Problem

• Synthesizes 3D nephrographic images in CT urography.
• Uses diffusion model with Swin Transformer for image synthesis.
• Reduces radiation dose in CTU by 33.3%.

Innovation

• Diffusion model for 3D synthesis
• Swin Transformer enhances deep learning
• Reduces CTU radiation dose safely