Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

EndoGSLAM: Real-Time Dense Reconstruction and Tracking in Endoscopic Surgeries using Gaussian Splatting

Kailing Wang, Chen Yang, Yuehao Wang, Sikuang Li, Yan Wang, Qi Dou, Xiaokang Yang, Wei Shen

TL;DR

EndoGSLAM tackles the need for online, dense reconstruction and tracking in endoscopic surgery by introducing a streamlined, explicit 3D Gaussian representation coupled with differentiable rasterization for fast, gradient-based optimization. The approach simplifies Gaussian parameters to isotropic Gaussians with per-Gaussian color, enabling efficient initialization, tracking, and real-time rendering, while Gaussian expanding and a keyframe-guided partial refining strategy maintain reconstruction completeness and fidelity. Quantitative results on the C3VD dataset show EndoGSLAM achieves superior online tracking, reconstruction quality, and rendering speed compared with traditional and neural SLAM methods, with EndoGSLAM-R providing real-time performance and EndoGSLAM-H prioritizing quality. The work demonstrates substantial potential for intraoperative navigation by delivering robust, real-time dense visualization alongside precise pose estimation, with future directions including depth-free operation, deformation handling, and deeper integration into surgical workflows.

Abstract

Precise camera tracking, high-fidelity 3D tissue reconstruction, and real-time online visualization are critical for intrabody medical imaging devices such as endoscopes and capsule robots. However, existing SLAM (Simultaneous Localization and Mapping) methods often struggle to achieve both complete high-quality surgical field reconstruction and efficient computation, restricting their intraoperative applications among endoscopic surgeries. In this paper, we introduce EndoGSLAM, an efficient SLAM approach for endoscopic surgeries, which integrates streamlined Gaussian representation and differentiable rasterization to facilitate over 100 fps rendering speed during online camera tracking and tissue reconstructing. Extensive experiments show that EndoGSLAM achieves a better trade-off between intraoperative availability and reconstruction quality than traditional or neural SLAM approaches, showing tremendous potential for endoscopic surgeries. The project page is at https://EndoGSLAM.loping151.com

EndoGSLAM: Real-Time Dense Reconstruction and Tracking in Endoscopic Surgeries using Gaussian Splatting

TL;DR

EndoGSLAM tackles the need for online, dense reconstruction and tracking in endoscopic surgery by introducing a streamlined, explicit 3D Gaussian representation coupled with differentiable rasterization for fast, gradient-based optimization. The approach simplifies Gaussian parameters to isotropic Gaussians with per-Gaussian color, enabling efficient initialization, tracking, and real-time rendering, while Gaussian expanding and a keyframe-guided partial refining strategy maintain reconstruction completeness and fidelity. Quantitative results on the C3VD dataset show EndoGSLAM achieves superior online tracking, reconstruction quality, and rendering speed compared with traditional and neural SLAM methods, with EndoGSLAM-R providing real-time performance and EndoGSLAM-H prioritizing quality. The work demonstrates substantial potential for intraoperative navigation by delivering robust, real-time dense visualization alongside precise pose estimation, with future directions including depth-free operation, deformation handling, and deeper integration into surgical workflows.

Abstract

Precise camera tracking, high-fidelity 3D tissue reconstruction, and real-time online visualization are critical for intrabody medical imaging devices such as endoscopes and capsule robots. However, existing SLAM (Simultaneous Localization and Mapping) methods often struggle to achieve both complete high-quality surgical field reconstruction and efficient computation, restricting their intraoperative applications among endoscopic surgeries. In this paper, we introduce EndoGSLAM, an efficient SLAM approach for endoscopic surgeries, which integrates streamlined Gaussian representation and differentiable rasterization to facilitate over 100 fps rendering speed during online camera tracking and tissue reconstructing. Extensive experiments show that EndoGSLAM achieves a better trade-off between intraoperative availability and reconstruction quality than traditional or neural SLAM approaches, showing tremendous potential for endoscopic surgeries. The project page is at https://EndoGSLAM.loping151.com
Paper Structure (12 sections, 6 equations, 3 figures, 2 tables)

This paper contains 12 sections, 6 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Preliminaries and Initialization
  4. Camera Tracking
  5. Gaussian Expanding
  6. Partial Refining
  7. Experiments
  8. Dataset and Evaluation Metrics
  9. Implementation Details
  10. Evaluation
  11. Ablation Study
  12. Conclusion and Future Work

Figures (3)

  • Figure 1: Comparative Visualization of Novel View Synthesis. From left to right, we show the holistic rendering from EndoGSLAM, the ground truth of one given viewpoint, renderings of EndoGSLAM, NICESLAM niceslam and Endo-Depth endodepth. These comparisons highlight EndoGSLAM's superior fidelity.
  • Figure 2: Overview. EndoGSLAM aims to track the camera and reconstruct tissues among endoscopic surgeries while enabling online visualization.
  • Figure 3: Qualitative results on sequence cecum_t2_b and sigmoid_t2_a.