ColonSplat: Reconstruction of Peristaltic Motion in Colonoscopy with Dynamic Gaussian Splatting

Weronika Smolak-Dyżewska^1,*, Joanna Kaleta^2,3,*, Diego Dall'Alba^4,2, Przemyslaw Spurek^1,5

¹ Jagiellonian University, ² SANO - Centre for Computational Personalised Medicine

³ Warsaw University of Technology, ⁴ University of Verona, ⁵ IDEAS Research Institute, Poland

^* Equal contribution

Abstract

Accurate 3D reconstruction of colonoscopy data, accounting for complex peristaltic movements, is crucial for advanced surgical navigation and retrospective diagnostics. While recent novel view synthesis and 3D reconstruction methods have demonstrated remarkable success in general endoscopic scenarios, they struggle in the highly constrained environment of the colon. Due to the limited field of view of a camera moving through an actively deforming tubular structure, existing endoscopic methods reconstruct the colon appearance only for initial camera trajectory. However, the underlying anatomy remains largely static; instead of updating Gaussians' spatial coordinates (xyz), these methods encode deformation through either rotation, scale or opacity adjustments. In this paper, we first present a benchmark analysis of state-of-the-art dynamic endoscopic methods for realistic colonoscopic scenes, showing that they fail to model true anatomical motion. To enable rigorous evaluation of global reconstruction quality, we introduce DynamicColon, a synthetic dataset with ground-truth point clouds at every timestep. Building on these insights, we propose ColonSplat, a dynamic Gaussian Splatting framework that captures peristaltic-like motion while preserving global geometric consistency, achieving superior geometric fidelity on C3VDv2 and DynamicColon datasets.

Method overview

Qualitative comparison with existing methods

ColonSplat captures true colon-wall motion significantly better than existing baselines, where motion is mostly encoded through rotation and scale while Gaussian centers do not follow the underlying anatomical deformation.

Ablation Study

Acknowledgments

We thank the authors of publicly available repositories:
ENDO-4DGX, EndoPlanar, SurgicalGS, Endo-4DGS, Deform3DGS, RADE-GS, and C3VDv2.

If you find this work useful, please consider citing as:

@article{colonsplat_2026, title = {ColonSplat: Reconstruction of Peristaltic Motion in Colonoscopy with Dynamic Gaussian Splatting}, author = {Smolak-Dyzewska, Weronika and Kaleta, Joanna and Dall'Alba, Diego and Spurek, Przemyslaw}, journal = {arXiv}, year = {2026}, note = {Citation info placeholder (arXiv link coming soon)} }

For our DynamicColon dataset we used meshes provided by authors of C3VDv2 dataset. Please cite also:

@misc{golhar2025c3vdv2colonoscopy3d, title={C3VDv2 -- Colonoscopy 3D video dataset with enhanced realism}, author={Mayank V. Golhar and Lucas Sebastian Galeano Fretes and Loren Ayers and Venkata S. Akshintala and Taylor L. Bobrow and Nicholas J. Durr}, year={2025}, eprint={2506.24074}, archivePrefix={arXiv}, primaryClass={eess.IV}, url={https://arxiv.org/abs/2506.24074} }