End2Reg: Learning Task-Specific Segmentation for Markerless Registration in Spine Surgery

2025-12-15

AI Summary

Intraoperative markerless spinal navigation demands millimeter-accurate, minimally invasive registration, yet existing radiographic or bone-anchored marker-based approaches entail high radiation exposure and significant procedural disruption; mainstream RGB-D markerless methods rely on weakly supervised segmentation labels, propagating errors. This paper proposes End2Reg, an end-to-end deep learning framework that jointly optimizes segmentation and registration for the first time. It introduces a task-driven, unsupervised segmentation mask learning mechanism wherein the segmentation module is entirely driven by registration loss—requiring no segmentation annotations—and integrates RGB-D perception, implicit shape modeling, and a differentiable registration architecture. Evaluated on both ex vivo and in vivo benchmarks, End2Reg achieves state-of-the-art performance: median target registration error of 1.83 mm (−32%) and RMSE of 3.95 mm (−45%).

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Security and Privacy: Browser and end-point security
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Purpose: Intraoperative navigation in spine surgery demands millimeter-level accuracy. Current systems based on intraoperative radiographic imaging and bone-anchored markers are invasive, radiation-intensive and workflow disruptive. Recent markerless RGB-D registration methods offer a promising alternative, but existing approaches rely on weak segmentation labels to isolate relevant anatomical structures, which can propagate errors throughout registration. Methods: We present End2Reg an end-to-end deep learning framework that jointly optimizes segmentation and registration, eliminating the need for weak segmentation labels and manual steps. The network learns segmentation masks specifically optimized for registration, guided solely by the registration objective without direct segmentation supervision. Results: The proposed framework achieves state-of-the-art performance on ex- and in-vivo benchmarks, reducing median Target Registration Error by 32% to 1.83mm and mean Root Mean Square Error by 45% to 3.95mm, respectively. An ablation study confirms that end-to-end optimization significantly improves registration accuracy. Conclusion: The presented end-to-end RGB-D registration pipeline removes dependency on weak labels and manual steps, advancing towards fully automatic, markerless intraoperative navigation. Code and interactive visualizations are available at: https://lorenzopettinari.github.io/end-2-reg/.

Problem

• Develops a markerless registration method for spine surgery navigation
• Eliminates reliance on weak segmentation labels and manual steps
• Achieves high accuracy by jointly optimizing segmentation and registration

Innovation

• Jointly optimizes segmentation and registration end-to-end
• Learns segmentation masks guided solely by registration objective
• Eliminates need for weak segmentation labels and manual steps