Detection of Helicobacter Pylori Infection
Description
This study evaluated the diagnostic accuracy of an artificial intelligence (AI) algorithm for predicting Helicobacter pylori infection using endoscopic images, given the absence of established optical diagnosis methods. A meta-analysis of 8 studies showed that AI had a pooled sensitivity of 87%, specificity of 86%, and an area under the curve of 0.92 for detecting H. pylori infection. The AI algorithm demonstrated reliable performance, although further external validation and broader geographical testing are needed to address limitations.
Creator
Hallym University College of Medicine, Chuncheon Sacred Heart Hospital
Information
Pediatrics or Adult
Adult
Speciality
Pathology
Modality
Endoscopic Images
Training
Endoscopic images from 1719 patients (385 patients with H pylori infection vs 1334 controls)
Github
Publication
FDA