AUTHOR=Zhang Bojiang , Zhang Wei , Yao Hongjuan , Qiao Jinggui , Zhang Haimiao , Song Ying 

TITLE=A study on the improvement in the ability of endoscopists to diagnose gastric neoplasms using an artificial intelligence system

JOURNAL=Frontiers in Medicine

VOLUME=Volume 11 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2024.1323516

DOI=10.3389/fmed.2024.1323516

ISSN=2296-858X

ABSTRACT=Background: Artificial intelligence-assisted gastroscopy (AIAG) based on deep learning has been validated in various scenarios, but there was a lack of studies on diagnosing neoplasm under white light endoscopy. This study explored the potential role of AIAG systems in enhancing endoscopists' ability to diagnose gastric tumor lesions under white light.

Methods: A total of 251 patients with complete pathological information of electronic gastroscopy, biopsy, or ESD surgery in Xi'an GaoXin Hospital were retrospectively collected, which included 64 patients with neoplasm lesions (excluding advanced cancer) and 187 patients with non-neoplasm lesions. The diagnosis competence of middle-experienced endoscopists and experts was compared for gastric neoplasms with or without the assistance of AIAG, which was developed based on Resnet-50. 

Results: For the 251 patients with difficult clinical diagnoses included in the study. Compared with intermediated experienced endoscopists, AIAG's diagnostic competence was much higher, with a sensitivity of 79.69% which was higher than that of intermediated experienced endoscopists (79.69% vs. 72.50%, P = 0.012), and a specificity of 73.26% which was higher than that of intermediated experienced endoscopists either (73.26% vs. 52.62%, P < 0.001). With the help of AIAG, the intermediate-experienced endoscopists (< 8 years) demonstrated a significantly higher sensitivity(75.00% vs.72.50%, P<0.001).  Experts ( ≥ 8 years) had similar results with or without AI assistance (with AI vs. without AI, sensitivities: 70.31% vs. 67.81%, P=0.358; specificities: 83.85% vs. 85.88%, P=0.116)
Conclusion: With the assistance of artificial intelligence (AI) systems, the ability of intermediate-experienced endoscopists to diagnose gastric neoplasms is significantly improved. And it has little effect on experts.