Search papers, labs, and topics across Lattice.
This paper introduces GeoCFNet, a geometry-aware confidence field network designed to enhance visual guidance in robot-assisted endoscopic submucosal dissection (ESD) by addressing challenges such as dynamic scenes and tissue deformation. By leveraging a pretrained DINOv3 backbone and incorporating a Token-Differentiated Fusion module alongside Geometry-Aware Spatial Regularization, the model effectively estimates confidence fields that maintain spatial coherence and local geometric transitions. Experimental results demonstrate that GeoCFNet achieves significant improvements in accuracy and stability, with RMSE of 0.0480 and PSNR of 27.1995, crucial for ensuring safe and precise surgical dissection.
GeoCFNet achieves unprecedented accuracy in estimating confidence fields for robot-assisted ESD, crucial for minimizing surgical complications.
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.
