3D-LLDM: Label-Guided 3D Latent Diffusion Model for Improving High-Resolution Synthetic MR Imaging in Hepatic Structure Segmentation

2026-03-24

AI Summary

High-quality annotated data are scarce in medical imaging, limiting the application of generative models for liver structure segmentation. This work proposes a label-guided 3D latent diffusion model based on ControlNet, introducing structural label guidance into 3D medical image generation for the first time. Leveraging hepatobiliary-phase Gd-EOB-DTPA-enhanced MR images and their corresponding anatomical masks, the model simultaneously synthesizes high-fidelity MR volumes and accurate segmentation labels. The method achieves a Fréchet Inception Distance (FID) of 28.31, representing improvements of 70.9% and 26.7% over conventional GANs and existing diffusion models, respectively. When employed for data augmentation, it boosts the Dice score for liver tumor segmentation by up to 11.153%, substantially enhancing downstream task performance.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Deep learning and generative models are advancing rapidly, with synthetic data increasingly being integrated into training pipelines for downstream analysis tasks. However, in medical imaging, their adoption remains constrained by the scarcity of reliable annotated datasets. To address this limitation, we propose 3D-LLDM, a label-guided 3D latent diffusion model that generates high-quality synthetic magnetic resonance (MR) volumes with corresponding anatomical segmentation masks. Our approach uses hepatobiliary phase MR images enhanced with the Gd-EOB-DTPA contrast agent to derive structural masks for the liver, portal vein, hepatic vein, and hepatocellular carcinoma, which then guide volumetric synthesis through a ControlNet-based architecture. Trained on 720 real clinical hepatobiliary phase MR scans from Samsung Medical Center, 3D-LLDM achieves a Fréchet Inception Distance (FID) of 28.31, improving over GANs by 70.9% and over state-of-the-art diffusion baselines by 26.7%. When used for data augmentation, the synthetic volumes improve hepatocellular carcinoma segmentation by up to 11.153% Dice score across five CNN architectures.

Problem

• synthetic medical imaging
• data scarcity
• hepatic segmentation
• annotated datasets
• MR image synthesis

Innovation

• label-guided diffusion
• 3D latent diffusion model
• synthetic MR imaging
• ControlNet
• medical image segmentation