🛰️ Daily AI Frontier
‹ back to 2026-09-18

ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis

arXiv cs.CV Medical/Healthcare AI Zahra Ghaffari, Massih Bahar, Mojgan Forootan, Ali Darvishi, Hamidreza Bolhasani 2026-09-17
Representative image for ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis

TL;DR - ERCPMP-Gx is a public, patient-level dataset linking colorectal polyposis endoscopy images and videos with histopathology and germline findings. It addresses a gap in multimodal data for AI-assisted recognition and classification of hereditary polyposis syndromes.

  • Contains 160 images plus video clips captured with white-light and multiple narrow-band imaging modes.
  • Approximately 80% of cases involve clinically or genetically confirmed hereditary syndromes; 20% are non-hereditary polyps or mimicking lesions for differential classification.
  • Covers familial adenomatous, Peutz-Jeghers, juvenile, and ganglioneuroma polyposis syndromes.
  • Records include standardized endoscopic annotations and, where available, representative histopathology and clinically reported germline findings.

view merged work →