See More, Change Less: Anatomy-Aware Diffusion for Contrast Enhancement

TL;DR

This paper introduces SMILE, an anatomy-aware diffusion model for contrast enhancement in medical imaging that improves image quality and diagnostic accuracy by focusing on clinically relevant regions while respecting anatomical structures.

Contribution

The paper presents a novel anatomy-aware diffusion model with structure-aware supervision and registration-free learning, enhancing medical images without over-editing or distorting anatomy.

Findings

Outperforms existing methods in image quality metrics (SSIM, PSNR, FID)

Improves cancer detection accuracy from non-contrast CT scans

Provides fast, consistent enhancement across contrast phases

Abstract

Image enhancement improves visual quality and helps reveal details that are hard to see in the original image. In medical imaging, it can support clinical decision-making, but current models often over-edit. This can distort organs, create false findings, and miss small tumors because these models do not understand anatomy or contrast dynamics. We propose SMILE, an anatomy-aware diffusion model that learns how organs are shaped and how they take up contrast. It enhances only clinically relevant regions while leaving all other areas unchanged. SMILE introduces three key ideas: (1) structure-aware supervision that follows true organ boundaries and contrast patterns; (2) registration-free learning that works directly with unaligned multi-phase CT scans; (3) unified inference that provides fast and consistent enhancement across all contrast phases. Across six external datasets, SMILE outperforms existing methods in image quality (14.2% higher SSIM, 20.6% higher PSNR, 50% better FID) and in clinical usefulness by producing anatomically accurate and diagnostically meaningful images. SMILE also improves cancer detection from non-contrast CT, raising the F1 score by up to 10 percent.