WING: A Window-Prior-Based Generative Network with Gated Inception for Cross-Modality CT Synthesis

2026-07-07

AI Summary

This work addresses the challenge of generating full-range CT images directly from MRI or CBCT, where high dynamic range and long-tailed intensity distributions often cause sparse yet critical anatomical structures to be averaged out. To overcome this, the authors propose a window-level prior–guided generative framework that, for the first time, incorporates clinical CT windowing knowledge into cross-modality synthesis. Leveraging the structural determinism and window-level separability of CT intensities, they design a gated Inception generator and a Fuse-and-Refine Transformer. The framework employs multi-window prediction, structure-aware feature fusion, and window-conditioned adversarial training to enable high-fidelity, unified modeling across diverse anatomical regions. Evaluated on both MRI-to-CT and CBCT-to-CT tasks, a single model achieves state-of-the-art performance, significantly enhancing the fidelity of clinically crucial structures.

Abstract

Generating CT volumes from MRI and CBCT can improve treatment planning in adaptive radiotherapy while avoiding additional radiation exposure. However, direct regression of CT intensities is challenged by the inherently high dynamic range and long-tailed distributions, thereby averaging out sparse yet clinically important structures. To alleviate this issue, we reformulate the regression target into multiple windowed representations, leveraging the inductive prior that CT intensities are structure-deterministic and window-separable. These windowed views exhibit smoother distributions and admit structured fusion back to the full-range CT. Building on this reformulation, we introduce WING, a WINdow-prior-based Generative network comprising: 1) a new Gated Inception Generator to produce multi-window predictions, enabling multi-shape kernel interactions to capture cross-modality correspondence; 2) a Fuse-and-Refine Transformer to aggregate the windowed outputs and learn residuals for detail refinement; and 3) a joint adversarial training objective to enhance window-conditioned realism. Extensive experiments demonstrate that our compact WING achieves state-of-the-art performance on the MRI-to-CT and CBCT-to-CT benchmarks, while supporting multi-anatomy synthesis with a single model.

Problem

• CT synthesis
• cross-modality
• intensity regression
• long-tailed distribution
• adaptive radiotherapy

Innovation

• window-prior
• Gated Inception
• Fuse-and-Refine Transformer
• cross-modality synthesis
• CT generation