New method improves MRI scans by translating between types
S2T-Unet: A Structure-to-Style Framework for Inter-Modality MRI Translation
Computer Vision and Pattern Recognition
Summary
MRI scans come in different types, and sometimes doctors need a type they didn’t scan. The authors designed a new way to create missing types by separating what parts of the image show structure from what parts show style or appearance. Their method, called S2T-Unet, uses special coding to keep important structure details while changing the look to match the target MRI type. Tests showed it works as well or better than existing methods.
What this means in practice
- •For medical imaging teams: Generate missing MRI types reliably to reduce patient scan times and costs using explicit structure and style separation.
- •For radiology software developers: Integrate a model that maintains important anatomy while translating MRI images between modalities for better diagnostic tools.
Authors
Yichao Liu
Abstract
Inter-modality MRI translation aims to synthesize missing MRI modalities from available acquisitions, reducing the need for additional scanning while preserving clinically relevant anatomical information. However, existing image translation methods often learn intensity mappings without explicitly separating modality-invariant structural information from modality-specific appearance, which may lead to structural information loss or unrealistic image details. In this work, we propose S2T-Unet, a structure-to-style framework that explicitly models these two aspects. Specifically, vector quantization is introduced at the lower-level bottleneck to encode modality-invariant structural information using a learned discrete codebook. At higher levels, a modality transformation module uses decoder features to condition and transform encoder representations toward the target modality, thereby recovering modality-specific intensity and contrast information. Experiments on the IXI multi-contrast MRI dataset across four translation tasks demonstrate that S2T-Unet is comparable or outperform with state-of-art method.