Artificial intelligence (AI) can accurately classify gastric lesions, but its clinician-level impact in real-world practice remains uncertain. We compared endoscopists’ diagnostic performance with vs. without AI assistance using both still-image (M1) and video (M2) datasets. We analyzed 1,584 cases (226 cancer, 282 dysplasia, 90 non-neoplastic lesions NNL, 649 intestinal metaplasia IM, and 337 gastritis/normal). One representative still image per case was extracted for M1; edited five-second video clips formed M2. Six in-training endoscopists (< 3 years’ experience) independently read M1 and M2 with and without AI after a one-week washout. As a stand-alone model, AI achieved 91.31% (M1) and 92.51% (M2) accuracy for focal lesions (sensitivities 91.02% and 91.91%; specificities 95.50% and 96.12%). For IM, accuracy was 91.83% (M1) and 92.45% (M2). With AI assistance, overall reader accuracy increased from 74.92% to 86.66% in M1 (AUC 0.742 to 0.860) and likewise from 74.92% to 86.81% in M2 (AUC 0.796 to 0.900); all p < 0.05. By subtype (videos, M2), accuracy improved 80.01% to 89.85% for cancer (+ 9.84%), % 67.16%to 81.08% for dysplasia (+ 13.92%), 77.59% to 89.50% for NNL (+ 11.91%), and 68.95% to 85.34% for IM (+ 16.39%). Still-image results showed similar gains (e.g., dysplasia 67.16% to 81.32%, IM 68.95% to 79.22%, both p < 0.05). AI assistance significantly enhances endoscopists’ diagnostic accuracy across lesion types and modalities, with the largest benefits for dysplasia and IM—conditions prone to clinician-level variability. These findings suggest that AI assistance may help improve reliability and support earlier recognition of clinically significant lesions.
Lee et al. (Fri,) studied this question.