Semantically Conditioned Diffusion Models for Cerebral DSA Synthesis

2026-02-12

AI Summary

This study addresses the scarcity of digital subtraction angiography (DSA) data—caused by its invasive nature and high cost—which hinders cerebrovascular disease research and algorithm development. To overcome this limitation, the authors propose the first semantic-conditioned latent diffusion model (LDM) that explicitly encodes anatomical regions and C-arm imaging angles via text embeddings, enabling precise control over both anatomical structure and geometric perspective in synthesized arterial-phase cerebral DSA images. Trained on a large-scale single-center DSA dataset, the model generates images achieving Likert scores of 3.1–3.3 and a Fréchet Inception Distance (FID) of 15.27. Expert evaluations confirm high clinical realism and strong inter-rater reliability (ICC = 0.80–0.87), significantly enhancing the clinical utility of synthetic medical imagery.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Large language models for search
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Digital subtraction angiography (DSA) plays a central role in the diagnosis and treatment of cerebrovascular disease, yet its invasive nature and high acquisition cost severely limit large-scale data collection and public data sharing. Therefore, we developed a semantically conditioned latent diffusion model (LDM) that synthesizes arterial-phase cerebral DSA frames under explicit control of anatomical circulation (anterior vs.\ posterior) and canonical C-arm positions. We curated a large single-centre DSA dataset of 99,349 frames and trained a conditional LDM using text embeddings that encoded anatomy and acquisition geometry. To assess clinical realism, four medical experts, including two neuroradiologists, one neurosurgeon, and one internal medicine expert, systematically rated 400 synthetic DSA images using a 5-grade Likert scale for evaluating proximal large, medium, and small peripheral vessels. The generated images achieved image-wise overall Likert scores ranging from 3.1 to 3.3, with high inter-rater reliability (ICC(2,k) = 0.80--0.87). Distributional similarity to real DSA frames was supported by a low median Fr\'echet inception distance (FID) of 15.27. Our results indicate that semantically controlled LDMs can produce realistic synthetic DSAs suitable for downstream algorithm development, research, and training.

Problem

• Digital Subtraction Angiography
• Cerebrovascular Disease
• Data Scarcity
• Medical Image Synthesis
• Invasive Imaging

Innovation

• Semantically Conditioned Diffusion
• Latent Diffusion Model
• Cerebral DSA Synthesis
• Text-to-Image Generation
• Medical Image Synthesis