VS-DDPM: Efficient Low-Cost Diffusion Model for Medical Modality Translation

2026-04-24

AI Summary

This work addresses the low inference efficiency of diffusion models in 3D medical image generation, which hinders their clinical deployment. To overcome this limitation, the authors propose a 3D Variable-Step Denoising Diffusion Probabilistic Model (VS-DDPM) that incorporates an adaptive step-size sampling mechanism to significantly accelerate inference while preserving high fidelity. The method is versatile across multiple cross-modality synthesis tasks, including missing MRI completion, tumor removal, and MRI/CBCT-to-sCT synthesis. Evaluated on the BraTS2025 and SynthRAD2025 challenges, VS-DDPM achieves state-of-the-art performance: for missing MRI completion, it attains Dice scores of 0.80–0.88 and SSIM of 0.95; for tumor removal, it yields an RMSE as low as 0.053 and a PSNR of 26.77.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Diffusion models produce high-quality synthetic data but suffer from slow inference. We propose 3D Variable-Step Denoising Diffusion Probabilistic Model (VS-DDPM) a framework engineered to maintain generative quality while accelerating inference by several factors. We tested our approach on four tasks (missing MRI, tumor removal, MRI-to-sCT, and CBCT-to-sCT) within the BraTS2025 and SynthRAD2025 challenges. Designed for high efficiency under hardware and time constrains imposed by both challenges. VS-DDPM achieved state-of-the-art (SOTA) performance in missing MRI synthesis, yielding Dice scores of 0.80, 0.83, and 0.88 for the enhancing tumor, tumor core, and whole tumor regions, respectively, alongside a structural similarity index (SSIM) of 0.95. For MRI tumor removal, the model attained a root mean squared error (RMSE) of 0.053, a peak signal-to-noise ratio (PSNR) of 26.77, and an SSIM of 0.918. While the framework demonstrated competitive performance in MRI-to-sCT and CBCT-to-sCT tasks, it did not reach SOTA benchmarks, potentially due to sensitivities in data pre and post-processing pipelines or specific loss function configurations. These results demonstrate that VS-DDPM provides a robust and tunable solution for high-fidelity 3D medical image synthesis. The code is available in https://github.com/andre-fs-ferreira/SynthRAD_by_Faking_it.

Problem

• diffusion model
• medical image synthesis
• modality translation
• inference acceleration
• 3D medical imaging

Innovation

• Variable-Step Diffusion
• Medical Image Synthesis
• Modality Translation
• Efficient Inference
• 3D Diffusion Model