VDSB-GWSyn: Diffusion Schrödinger Bridge for Controllable and Anatomically Feasible Guidewire Synthesis in Coronary Angiography

2026-05-26

AI Summary

This work addresses the challenge of guidewire tip localization in coronary angiography, which is hindered by the scarcity of annotated data and the limited adaptability of existing synthetic image generation methods. To overcome this, the authors propose a diffusion Schrödinger bridge (DSB)-based approach for controllable guidewire image synthesis. The method integrates vessel segmentation mask constraints, a SPADE conditional mechanism, and shape priors to ensure anatomically plausible synthetic images with preserved background structures and precise endpoint control. This paradigm significantly enhances the clinical utility of generated data and is readily transferable to other interventional device perception tasks. Experimental results demonstrate superior performance on ROI-FID, ROI-KID, and IPR metrics; when used for pretraining, the synthetic images reduce the mean positional error (MPE) of guidewire tip localization to 7.71 pixels and achieve a 3-pixel PCK of 86.27%.

Abstract

Coronary guidewire endpoint localization is a fundamental capability for computer-assisted PCI, and its importance increases as robot-assisted PCI is progressively adopted to reduce operator radiation exposure. However, the scarcity of annotated CAG images with guidewires and the limited adaptability of existing guidewire synthesis models remain key bottlenecks for guidewire endpoint localization. To address this issue, we propose VDSB-GWSyn, a Diffusion Schrödinger Bridge (DSB) model-based framework, enabling synthesis of controllable, high-fidelity guidewire samples under complex anatomical backgrounds. VDSB-GWSyn first uses our shape prior algorithm to learn the basic guidewire geometry. It then generates guidewire masks under constraints imposed by the vessel segmentation masks and outputs the corresponding endpoint coordinates. Finally, it synthesizes realistic guidewire samples on real CAG images using DSB conditioned with SPADE. Experimental results show that the guidewire samples synthesized by VDSB-GWSyn achieve favorable ROI-FID and ROI-KID, as well as high IPR scores. In addition, incorporating our synthesized data for synthetic pre-training followed by real fine-tuning substantially improves downstream guidewire endpoint localization, reducing MPE from 16.01~px to 7.71~px and increasing PCK at 3~px from 52.63\% to 86.27\%, leading to more clinically reliable deployment of robot-assisted guidewire delivery systems. Moreover, the core design philosophy of controllable device synthesis with strict background preservation and anatomical feasibility constraints has the potential to transfer to other interventional device perception tasks where annotated data are scarce.

Problem

• guidewire endpoint localization
• coronary angiography
• data scarcity
• anatomical feasibility
• interventional device synthesis

Innovation

• Diffusion Schrödinger Bridge
• Guidewire Synthesis
• Anatomical Feasibility
• Controllable Generation
• SPADE Conditioning