Fig. 2: Workflow of MicroSyn-X. | Nature Machine Intelligence

Fig. 2: Workflow of MicroSyn-X.

From: Synthetic X‑ray‑driven tracking and control of miniature medical devices

Fig. 2: Workflow of MicroSyn-X.The alternative text for this image may have been generated using AI.

a, The clinical challenges of MMD perception under real-time X-ray imaging. b, The controlled tissue generation process. Stable diffusion creates high-fidelity tissue images from user-defined masks and prompts. c, The integration of medical devices with tissues. Captured or generated MMD images are seamlessly integrated into the background with flexible parameters, ensuring pixel-accurate labelling. d, A schematic of neural network training for MMD tracking. MMD–tissue images are subdivided to improve precision and efficiency in tiny object localization. The model, trained on synthetic data, is then deployed for real-world MMD tracking.

Back to article page