Advancing artificial intelligence applicability in endoscopy through source-agnostic camera signal extraction from endoscopic images
This work introduces an AI-based method that effectively segments the endoscope camera signal from the raw endoscopic data in a source-agnostic way. Utilizing the proposed method as a preprocessing step allows existing AI models to use any endoscopic image, independent of its source, without compromising performance. Additionally, EPIC, a dataset of diverse endoscopic images, is generated. The proposed method, trained AI model weights, and the EPIC dataset are made publicly available.
Authors
- Robert Hüneburg (ORCID: https://orcid.org/0000-0001-9957-8299)
- Alexander Hann (ORCID: https://orcid.org/0000-0001-8035-3559)
- Jacob Nattermann (ORCID: https://orcid.org/0000-0001-5972-0780)
- Nora Martens
- Alexander Meining (ORCID: https://orcid.org/0009-0001-3773-138X)
- Ioannis Kafetzis (ORCID: https://orcid.org/0000-0002-3533-2246)
- Philipp Sodmann (ORCID: https://orcid.org/0000-0002-5598-8533)
- Wolfram G. Zoller
- Daniel R. Englmann
Institutions
- University of Applied Sciences Worms (DE)
- University Hospital Bonn (DE)
- Katharinenhospital (DE)
- National Center for Tumor Diseases (DE)
- Universitätsklinikum Würzburg (DE)
- Technische Universität Dresden (DE)
Publication Details
- Journal
- PLoS ONE
- Published
- 2025-06-11
- DOI
- https://doi.org/10.1371/journal.pone.0325987
- Citations
- 1
- Primary Topic
- Colorectal Cancer Screening and Detection
- Type
- article
- Field-Weighted Citation Impact
- 1.05
Funders
- Baden-Württemberg Stiftung