GeoCFNet: Geometry-Aware Confidence Field Network for Robot-Assisted Endoscopic Submucosal Dissection

2026-06-11

AI Summary

This study addresses the instability of confidence field estimation in robot-assisted endoscopic submucosal dissection, where challenges such as smoke, specular highlights, tissue deformation, and weak texture degrade performance. To tackle this, the authors formulate dissection guidance as a geometry-aware confidence field estimation problem and propose GeoCFNet. The method leverages a DINOv2-pretrained backbone for feature extraction, introduces a Token-Differentiated Fusion module to integrate class tokens with dense patch embeddings, and incorporates a SegFormer decoder enhanced by geometry-aware spatial regularization to improve spatial consistency and local geometric continuity. Experimental results demonstrate that the proposed approach achieves an RMSE of 0.0480, PSNR of 27.1995, SSIM of 0.3397, and CC of 0.2466 on the test set, significantly advancing the accuracy and geometric stability of confidence field estimation for thin-layer anatomical structures.

Abstract

Advanced surgical robotics has made robot-assisted endoscopic submucosal dissection (ESD) a promising approach for the en-bloc resection of large lesions, with the potential to reduce recurrence and improve long-term outcomes. However, the technical complexity and risk of complications in ESD demand stable and precise visual guidance to maintain an accurate dissection corridor and a safe tissue margin. Dense confidence fields provide an effective representation for this purpose by describing both the preferred dissection region and its spatial transition to surrounding tissue. However, reliable confidence field estimation remains challenging in dynamic endoscopic scenes due to smoke, specular highlights, tissue deformation, weak texture, and the thin geometric structure of the target region. To address these challenges, we formulate dissection guidance as a geometry-aware confidence field estimation problem and propose GeoCFNet, a geometry-aware confidence field network built on a pretrained DINOv3 backbone. GeoCFNet integrates a Token-Differentiated Fusion module to aggregate class-token context with dense patch representations, a SegFormer decoder for confidence regression, and Geometry-Aware Spatial Regularization (GASR) to preserve spatial coherence and local geometric transitions. Experimental results show that GeoCFNet achieves RMSE 0.0480, PSNR 27.1995, SSIM 0.3397, and CC 0.2466, indicating accurate and geometrically stable confidence field estimation for robot-assisted ESD guidance.

Problem

• robot-assisted ESD
• confidence field estimation
• endoscopic vision
• geometric structure
• surgical guidance

Innovation

• Geometry-Aware Confidence Field
• GeoCFNet
• Token-Differentiated Fusion
• Geometry-Aware Spatial Regularization
• Robot-Assisted ESD