Artificial intelligence in upper gastrointestinal endoscopy: what detection rates cannot tell us

Abstract

Impact of real-time artificial intelligence integration on detection of gastric lesions: an exploratory single-center before-and-after study using low-definition routine endoscopyPost hoc analysis indicates that real-time AI modestly improves detection of subcentimeter (≤0.5 cm) gastric lesionsdespite lesion size not being predefined-suggesting AI may help mitigate visual blind spots in routine endoscopy.Example images of the detection results.(A-F) Gastric lesions detected by endoscopy using AI so ware.LDR, lesion detection rate; BDR, benign detection rate; MDR, malignancy detection rate.Lesion detected (by size) ≤0.5 cm 269 166 >0.5 cm, ≤1.0 cm 316 153 >1.0 cm, ≤1.5 cm 9 6 >1.5 cm 76 36 Overall LDR 1.15 1.20 0.035 LDR (by size) ≤0.5 cm 0.18 0.20 <0.05 >0.5 cm, ≤1.0 cm 0.21 0.18 0.11 >1.0 cm, ≤1.5 cm 0.006 0.007 0.73 >1.5 cm 0.051 0.043 0.51 Biopsy performed (n, %) 242 (13.8) 81 (8.5) <0.05 Biopsy result Benign (BDR) 213 (0.124) 59 (0.058) <0.05 Malignant (MDR) 26 (0.015) 10 (0.01) 0.32