Real-time intraoperative depth estimation in transsphenoidal surgery using deep learning: A feasibility study
Through the application of DINOv2, we were able to estimate depth in endoscopic imaging from transsphenoidal endonasal surgeries by generating numeric maps and depth colormaps. This illustrates the potential of deep learning-based depth estimations, which in the future could contribute to improving intraoperative orientation. It also highlights the opportunities in using artificial intelligence to augment endoscopic video feeds.
Authors
- Victor E. Staartjes (ORCID: https://orcid.org/0000-0003-1039-2098)
- Raffaele Da Mutten (ORCID: https://orcid.org/0000-0002-2645-0865)
- Matteo Zoli (ORCID: https://orcid.org/0000-0001-8838-5408)
- Olivier Zanier (ORCID: https://orcid.org/0000-0002-3286-3770)
- Aron Alakmeh (ORCID: https://orcid.org/0009-0004-3395-3307)
- Alessandro Carretta
- Luca Regli
- Carlo Serra
- Diego Mazzatenta
Institutions
- University of Lucerne (CH)
- University of Zurich (CH)
- University Hospital of Zurich (CH)
- Istituto delle Scienze Neurologiche di Bologna (IT)
- University of Gothenburg (SE)
Publication Details
- Journal
- Journal of Clinical Neuroscience
- Published
- 2026-02-14
- DOI
- https://doi.org/10.1016/j.jocn.2026.111910
- Primary Topic
- Surgical Simulation and Training
- Type
- article
- Field-Weighted Citation Impact
- 0.00