ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis
Researchers release ERCPMP-Gx, a 160-image endoscopic dataset linking colorectal polyposis imagery with histopathology and germline findings for AI development.
The ERCPMP-GX dataset provides 160 endoscopic images and video clips captured mainly with the Olympus EVIS X1 system in white-light, NBI, and magnifying NBI modes. Roughly 80% of cases are confirmed hereditary polyposis syndromes such as FAP, PJS, JPS, and ganglioneuroma syndrome, with the remainder being non-hereditary mimics for differential classification. Records are linked to standardized endoscopic annotations, histopathology, and germline findings, and the dataset is publicly available on Mendeley.
- 160 endoscopic images plus video clips supporting AI recognition of colorectal polyposis syndromes
- About 80% of cases are genetically or clinically confirmed hereditary syndromes (FAP, PJS, JPS, GNS)
- Patient-level links between endoscopy, histopathology, and germline findings, unlike polyp-centric datasets
- Publicly released on Mendeley for AI model training and evaluation
- Vendors
- Olympus
- Products
- Olympus EVIS X1
From the original article
1 min read at arxiv.orgHereditary polyposis syndromes can be precursor lesions to colorectal cancer and are associated with a broad spectrum of extracolonic tumors. Early identification and accurate classification of these syndromes are essential for timely diagnosis, individualized patient management, and targeted surveillance strategies for affected families.
By Zahra Ghaffari · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.