Answers are grounded in this paper and cite their sources.
Background: Esophageal and stomach cancers pose a serious threat to patients due to their nonspecific symptoms and aggressive nature. By the time an accurate diagnosis is obtained, the disease has frequently progressed to an advanced stage, resulting in a significantly limited range of treatment options. Recently, computer-aided detection and computer-aided diagnosis systems have been developed to identify and characterize subtle gastric mucosal lesions, precancerous lesions, and early-stage cancers. Methodology: A literature review of the PubMed and Scopus databases was conducted in search of studies published from 2015 to 2025. Included publications were analyzed to assess the potential of artificial intelligence-based models in the clinical diagnosis of upper gastrointestinal tract lesions. Results: Artificial intelligence provides automated, accurate and consistent diagnostic performance for the detection and characterization of Barrett’s esophagus, esophageal squamous cell carcinoma and early gastric cancer. It significantly reduces blind spots and is capable of determining invasion depth and tumor margins. Artificial intelligence models can greatly improve the diagnostic accuracy of endoscopists enhancing success rate of lesion detection and minimizing the incidence of misdiagnosis. Conclusions: The assistance of artificial intelligence-based models holds particular potential for the diagnosis of upper gastrointestinal tract lesions. However, current evidence is limited by idealized datasets and lack of real-world validation, highlighting the need for larger, diverse, and clinically representative studies.
No takes yet. Share an insight, caveat, or question.
Plizga et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: