Online 3D reconstruction and dense tracking in endoscopic videos

TL;DR

This paper introduces an online framework for dense 3D reconstruction and tracking in endoscopic videos, improving surgical scene understanding and assisting interventions with real-time capabilities.

Contribution

It presents a novel online method using Gaussian splatting and tissue deformation modeling for real-time 3D reconstruction and tracking in endoscopic videos.

Findings

Outperforms state-of-the-art tracking methods

Achieves comparable results to offline reconstruction

Enables real-time surgical scene analysis

Abstract

3D scene reconstruction from stereo endoscopic video data is crucial for advancing surgical interventions. In this work, we present an online framework for online, dense 3D scene reconstruction and tracking, aimed at enhancing surgical scene understanding and assisting interventions. Our method dynamically extends a canonical scene representation using Gaussian splatting, while modeling tissue deformations through a sparse set of control points. We introduce an efficient online fitting algorithm that optimizes the scene parameters, enabling consistent tracking and accurate reconstruction. Through experiments on the StereoMIS dataset, we demonstrate the effectiveness of our approach, outperforming state-of-the-art tracking methods and achieving comparable performance to offline reconstruction techniques. Our work enables various downstream applications thus contributing to advancing the capabilities of surgical assistance systems.