Semantically Conditioned Diffusion Models for Cerebral DSA Synthesis
This work addresses data scarcity for medical imaging researchers and clinicians, enabling synthetic DSA generation for algorithm development and training, though it is incremental as it applies existing diffusion models to a new medical domain.
The authors tackled the problem of limited digital subtraction angiography (DSA) data for cerebrovascular disease by developing a semantically conditioned latent diffusion model to synthesize arterial-phase cerebral DSA frames, achieving realistic images with Likert scores of 3.1-3.3 from medical experts and a low FID of 15.27.
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 Fréchet 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.