Towards Occlusion-Aware Pose Estimation of Surgical Suturing Threads.

Area of Science:

• Computer Vision
• Robotics
• Medical Imaging

Background:

• Robust suture detection is crucial for surgical skill evaluation and robotic-assisted surgery.
• Suture pose estimation faces challenges due to foreground and background occlusions in complex surgical environments.

Purpose of the Study:

• To develop an automated algorithm for accurate suture detection and pose estimation, specifically addressing occlusion issues.
• To enhance the capabilities of robotic surgery through improved suture visualization and tracking.

Main Methods:

• Proposed an Occlusion-Aware Spatial Propagation model to handle suture self-intersection and background occlusion.
• Utilized region connectivity modeling to resolve self-intersecting threads.
• Employed a guided spatial propagation mechanism for context-aware occlusion handling.

Main Results:

• The proposed method demonstrated superior accuracy in suture pose estimation compared to baseline approaches on phantom and ex-vivo datasets.
• Validation confirmed the effectiveness of occlusion-aware connectivity and spatial propagation mechanisms.

Conclusions:

• The developed Occlusion-Aware Spatial Propagation model offers a general, end-to-end framework for suture thread pose estimation.
• The fully automated algorithm effectively addresses surgical occlusions, paving the way for increased autonomy in robotic surgery.