North Florida Gastroenterology/Advance Center for Daycare Surgery/Burjeel Medical City Abu Dhabi UAE Edison, NJ, United States
Abdulhadi Quadri, MD, Makki Fayadh, MRCP, Salim Awad, MRCP, Imran Rasid, MD, Ahmed Hasan, MD North Florida Gastroenterology/Advance Center for Daycare Surgery/Burjeel Medical City Abu Dhabi UAE, Edison, NJ
Introduction: VCE is time consuming (30-120 minutes) and human analysis is imprecise resulting in missed lesions. AI, VCE BOT was developed to study if it will help shorten the reading time and detect any additional findings
Methods: Five hundred and fifty studies were pre read by an experienced endoscopist and labeled as normal, abnormal, or inconclusive (poor prep). The normal studies were then scanned by the VCEBOT. Any discrepancy between the endoscopist read and VCEBOT scanned studies were then analyzed by two separate experience endoscopists. Total time to read a VCEBOT scanned study and additional findings were recorded
Results: Out of five hundred and fifty studies, fifty-five studies were labelled as discrepant (read normal by the endoscopist and abnormal by the VCEBOT). Total time to read the study after scanned by the VCEBOT by either endoscopist ranged between 8-12 minutes. Additional findings detected by the VCEBOT and confirmed by the two endoscopist were in small bowl, polyps 4, extrinsic compression/submucosal lesion 12, ulcerated lesions 2, ulcers/erosions 2, lipoma 2, active bleeding 1, thick folds? Lymphoma 1, worms 2, and in stomach, ulcers 2, erosions irrespective of gastritis 4, submucosal lesions 2.
Discussion: The use of AI significantly decreased the reading time of VCE. It also detected additional findings not seen by the endoscopist
Disclosures: Abdulhadi Quadri indicated no relevant financial relationships. Makki Fayadh indicated no relevant financial relationships. Salim Awad indicated no relevant financial relationships. Imran Rasid indicated no relevant financial relationships. Ahmed Hasan indicated no relevant financial relationships.
Abdulhadi Quadri, MD, Makki Fayadh, MRCP, Salim Awad, MRCP, Imran Rasid, MD, Ahmed Hasan, MD. P1955 - Artificial Intelligence (AI) Shortens the Reading Time and Increases the Yield of Video Capsule Enteroscopy(VCE), ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.
