EndoMapper dataset of complete calibrated endoscopy procedures

Pablo Azagra,Carlos Sostres,Ángel Ferrandez,Luis Riazuelo,Clara Tomasini,Oscar León Barbed,Javier Morlana,David Recasens,Victor M. Batlle,Juan J. Gómez-Rodríguez,Richard Elvira,Julia López,Cristina Oriol,Javier Civera,Juan D. Tardós,Ana Cristina Murillo,Angel Lanas,José M.M. Montiel

DOI: https://doi.org/10.1038/s41597-023-02564-7

2023-10-10

Abstract:Computer-assisted systems are becoming broadly used in medicine. In endoscopy, most research focuses on the automatic detection of polyps or other pathologies, but localization and navigation of the endoscope are completely performed manually by physicians. To broaden this research and bring spatial Artificial Intelligence to endoscopies, data from complete procedures is needed. This paper introduces the Endomapper dataset, the first collection of complete endoscopy sequences acquired during regular medical practice, making secondary use of medical data. Its main purpose is to facilitate the development and evaluation of Visual Simultaneous Localization and Mapping (VSLAM) methods in real endoscopy data. The dataset contains more than 24 hours of video. It is the first endoscopic dataset that includes endoscope calibration as well as the original calibration videos. Meta-data and annotations associated with the dataset vary from the anatomical landmarks, procedure labeling, segmentations, reconstructions, simulated sequences with ground truth and same patient procedures. The software used in this paper is publicly available.

Image and Video Processing,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The main purpose of this paper is to introduce the Endomapper dataset, the first dataset containing complete endoscopic examination sequences, aimed at facilitating the development and evaluation of Visual Simultaneous Localization and Mapping (VSLAM) methods in real endoscopic data. Specifically, the paper attempts to address the following issues:

1. **Insufficiency of existing datasets**: Currently, most endoscopic research focuses on the automatic detection of polyps or other pathological changes, but endoscopic localization and navigation still rely entirely on manual operation by doctors. Existing datasets do not meet the research needs of VSLAM algorithms in the endoscopic field.

2. **Introduction of spatial AI capabilities**: By introducing spatial AI technology, particularly VSLAM, the functionality of endoscopic examinations can be enhanced, including augmented reality insertion, blind spot detection, polyp measurement, and guidance to previously discovered polyp locations.

3. **Need for high-quality calibrated videos**: To achieve the above goals, a large amount of high-quality and calibrated endoscopic video data is required. The Endomapper dataset provides over 24 hours of high-definition video, collected during routine medical practice, and includes geometric and photometric calibration parameters of the endoscope.

4. **Support for challenging tasks**: The dataset not only includes easily processed video segments but also some challenging segments to test the limits of existing algorithms and indicate directions for future research.

In summary, this paper attempts to promote the development of VSLAM technology in the endoscopic field through the Endomapper dataset, addressing the deficiencies in quality and diversity of existing datasets, thereby supporting broader spatial AI applications.