Learning-based Hierarchical Tracheal Anatomy Understanding from Sparse Surgical Demonstration Annotations for Ultrasound Robots

2026-07-24

AI Summary

This study addresses the challenges of subjective palpation and experience-dependent ultrasound guidance in tracheostomy procedures, which often lead to inaccurate incision placement. To overcome these limitations, the authors propose a two-stage perception framework: an initial coarse localization of the trachea is achieved using a lightweight YOLOv8n model, followed by high-fidelity segmentation via a sparsely prompted, optimized SAM2. Innovatively decoupling the localization module from a vision foundation model enables precise, efficient, and generalizable anatomical understanding of the trachea with only sparse annotations. By integrating a hybrid training strategy that combines laboratory-collected and real surgical data, the method achieves a mean Dice coefficient of 0.777 in both controlled and generalized settings—significantly outperforming U-Net baselines (≤0.494)—and attains an inference speed of 6.92 FPS, satisfying the latency requirements for robotic closed-loop operation.

Abstract

Tracheostomy requires precise localization of the tracheal incision site; however, conventional manual palpation is subjective and often unreliable, while ultrasound utility remains operator-dependent. This work presents a learning-based framework for hierarchical tracheal anatomy understanding, designed specifically for ultrasound-guided robotic systems. We propose a two-stage perception pipeline integrating a YOLOv8n localization backbone with a sparse, prompt-optimized SAM2 decoder to achieve high-fidelity segmentation from sparse surgical annotations. Our hybrid training strategy, bridging curated laboratory data with unconstrained sequences, ensures clinical robustness. Experimental benchmarks demonstrate that this decoupled architecture effectively balances generalization, precision, and efficiency. The YOLOv8n and SAM2 framework achieves a consistent Mean Dice Similarity Coefficient (DSC) of 0.777 across both controlled and generalized domains. This significantly outperforms U-Net baselines, which often suffer from anatomical fragmentation and performance degradation (Generalization DSC $\le$ 0.494). By constraining mask decoding to targeted, sparse regions of interest, our model achieves a throughput of 6.92 FPS, which is vital for closed-loop robotic teleoperation. This study confirms that a robust hierarchical understanding of tracheal anatomy can be derived by coupling lightweight localization with foundation-scale visual models. Our framework establishes a scalable foundation for standardized, autonomous surgical assistance, effectively navigating the variability of real-world ultrasound to enhance the safety and precision of robotic-assisted tracheostomy.

Problem

• tracheostomy
• ultrasound guidance
• anatomical localization
• robotic surgery
• sparse annotations

Innovation

• hierarchical anatomy understanding
• sparse annotation learning
• YOLOv8n-SAM2 fusion
• ultrasound-guided robotics
• real-time surgical segmentation