PIVM: Diffusion-Based Prior-Integrated Variation Modeling for Anatomically Precise Abdominal CT Synthesis

TL;DR

PIVM introduces a diffusion-based framework that synthesizes anatomically precise abdominal CT images by integrating organ-specific priors, improving realism and detail while addressing data scarcity issues.

Contribution

It proposes a novel diffusion model that predicts intensity variations relative to priors, operating directly in image space for high-fidelity CT synthesis.

Findings

Produces realistic, organ-aligned CT images

Preserves full Hounsfield Unit range and fine textures

Operates efficiently in image space

Abstract

Abdominal CT data are limited by high annotation costs and privacy constraints, which hinder the development of robust segmentation and diagnostic models. We present a Prior-Integrated Variation Modeling (PIVM) framework, a diffusion-based method for anatomically accurate CT image synthesis. Instead of generating full images from noise, PIVM predicts voxel-wise intensity variations relative to organ-specific intensity priors derived from segmentation labels. These priors and labels jointly guide the diffusion process, ensuring spatial alignment and realistic organ boundaries. Unlike latent-space diffusion models, our approach operates directly in image space while preserving the full Hounsfield Unit (HU) range, capturing fine anatomical textures without smoothing. Source code is available at https://github.com/BZNR3/PIVM.