Disentangled Anatomy-Disease Diffusion (DADD) for Controllable Ulcerative Colitis Progression Synthesis

2026-05-03

AI Summary

This work addresses the challenge of synthesizing longitudinal medical images with controllable disease stages, where patient anatomy and pathological textures are highly entangled, making it difficult to simultaneously preserve anatomical fidelity and precisely modulate disease progression. To this end, the authors propose the DADD framework, which achieves controllable generation through disentangled anatomical and pathological representations. Specifically, a Feature Purifier removes pathological information from anatomical features, while a three-pathway cross-attention mechanism and a training-free Delta Steering strategy are integrated within a latent diffusion model. This architecture combines a pretrained encoder with ordinal embeddings to enable high-fidelity, single-forward synthesis of ulcerative colitis endoscopic images. Evaluated on the LIMUC dataset, the method effectively balances class distribution and significantly enhances downstream classification performance.

Abstract

Synthesizing longitudinal medical images at controllable disease stages while preserving patient-specific anatomy is hindered by the entanglement of pathological textures and structural features. We address this challenge for ulcerative colitis (UC) endoscopy, where severity follows a continuous ordinal progression along the Mayo Endoscopic Score (MES). Our framework, Disentangled Anatomy-Disease Diffusion (DADD), conditions a latent diffusion model on two complementary embeddings: a pretrained image encoder for patient anatomy and a separately trained ordinal embedder for cumulative disease severity. Since image embeddings inevitably capture disease information, we introduce a Feature Purifier, a cross-attention-based erasure mechanism that identifies and suppresses disease-correlated channels, yielding purified anatomical representations. These cleaned anatomy tokens and target disease tokens are injected into the denoising network via a Triple-Pathway Cross-Attention mechanism with resolution-dependent routing gates. This architecture leverages the U-Net hierarchy, in which different network depths encode global structure versus fine-grained pathological texture. Furthermore, we introduce Delta Steering, a training-free directional signal derived from the ordinal embeddings that enables explicit, single-pass control over disease transitions at inference without requiring additional forward passes. Validated on the LIMUC dataset, our approach produces high-fidelity images across all severity levels and effectively rebalances skewed class distributions, enhancing performance for downstream classification tasks. The dataset is available at zenodo.org/records/5827695 and the code base at github.com/umutdundar99/progressive-stable-diffusion

Problem

• disentanglement
• ulcerative colitis
• medical image synthesis
• anatomy-disease separation
• controllable progression

Innovation

• Disentangled Representation
• Latent Diffusion Model
• Feature Purifier
• Triple-Pathway Cross-Attention
• Delta Steering