Spectral Consistent Flow for One-step 3D Medical Image Translation

2026-07-12

AI Summary

This work addresses the challenges of modality entanglement, excessive smoothing, and artifacts in cross-modal translation of 3D medical images, which stem from implicit low-pass modulation. To overcome these issues, the authors propose a generative framework based on a single-step stochastic Brownian bridge in latent space. The method directly models the mapping from source to target domains by predicting a support-regularized mean velocity field and incorporates a learnable frequency-domain gain modulation mechanism. This mechanism explicitly links spatial textures with spectral energy flow, enabling dynamic correction of power spectral density evolution. Extensive experiments across four datasets demonstrate that the proposed approach significantly outperforms existing methods in various translation tasks, achieving superior structural consistency and robustness while preserving high-fidelity anatomical details.

Abstract

We present Spectral Consistent Flow (SC-Flow), a 3D medical image translation framework with a single function evaluation (1-NFE) in the latent space. This approach reformulates medical image translation as a stochastic Brownian bridge process that directly constructs a mapping between source and target modalities by predicting the support regularized mean velocity field. To mitigate modality entanglement, over-smoothing, and artifacts induced by the implicit low-pass modulation of the latent average velocity, we introduce a Spectral Consistency Corrector that dynamically regularizes the evolution of the power spectral density via learnable frequency-domain gain modulation. This mechanism establishes an explicit bridge between spatial textures and spectral energy flow, enabling the model to recover fine-grained anatomical fidelity while maintaining global structural coherence. Extensive experiments on four datasets demonstrate that SC-Flow delivers significantly more accurate, consistent, and robust performance across various translation scenarios.

Problem

• 3D medical image translation
• modality entanglement
• over-smoothing
• artifacts
• spectral consistency

Innovation

• Spectral Consistent Flow
• Brownian bridge
• frequency-domain gain modulation
• power spectral density regularization
• one-step 3D image translation