Adventist Health Glendale’s James Tabibian, MD, Medical Director of Adventist Health Glendale’s Digestive Diseases Center recently served as a co-investigator (Tabibian et al, 2025) in a systematic review and meta-analysis evaluating how artificial intelligence (AI)-assisted endoscopy can improve the detection and depth assessment of early gastric cancer (EGC) — a form of stomach cancer that is highly curable when diagnosed early.
James Tabibian, MD
Why This Research Matters
Stomach cancer is the sixth most commonly diagnosed cancer worldwide and the fourth leading cause of cancer death. In early stages, when cancer is limited to the stomach’s mucosa or submucosa, survival rates can exceed 90%. However, conventional endoscopy — the current standard for detection — can miss subtle signs of EGC, especially in less-experienced hands. Accurate assessment of tumor depth is also challenging. Yet, it is crucial in determining whether a patient can be treated with minimally invasive procedures such as endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD).
The study reviewed 19 high-quality research papers and found that AI-assisted endoscopy, particularly using convolutional neural network (CNN) technology, achieved a pooled sensitivity of 84% and specificity of 92% for detecting EGC. It also showed strong accuracy in assessing invasion depth, with an area under the curve (AUC) of 0.89. These findings suggest AI tools could significantly enhance diagnostic precision, supporting faster and more effective treatment decisions.
Potential Impact for Local Patients
While much of the research on AI-assisted endoscopy has been conducted in Asia, where gastric cancer is more prevalent, Dr. Tabibian is helping to explore how these methods could benefit patients in ethnically diverse communities of Southern California, where early detection can be particularly challenging due to varied genetic and environmental risk factors.
“AI has the potential to add the expertise and accuracy of the most experienced endoscopists to every patient encounter,” Dr. Tabibian notes. “For local patients, that could mean earlier diagnosis, less invasive treatment, and better outcomes.”
Citation:
Tabibian, J. H., Agarwal, S., Singh Rajput, M., Pandey, S., Ramai, D., Ofosu, A., Barakat, M. T., Girotra, M., & Jagannath Mahapatra, S. (2025). Evaluation of the use of convoluted neural network for detecting early gastric cancer and predicting its invasion depth: A systematic review and meta-analysis. Digestive and Liver Disease, 57(10). https://doi.org/10.1016/j.dld.2025.05.030.
Source:
ScienceDirect meta-analysis
Purpose: Review the role of artificial intelligence (AI) assisted endoscopy in the diagnosis and depth assessment of EGC.