Artificial Intelligence-Driven Approaches to Endoscopic Gastric Cancer Detection: Current Progress and Future Directions

Sep 24, 2026

Gastric cancer is a common and highly lethal malignancy worldwide, where survival depends heavily on early detection. Endoscopy is the main diagnostic approach for GC and early gastric cancer, yet accuracy varies with endoscopist skill and is limited by misrecognition and operator-related errors. Deep learning and other AI systems, including convolutional neural networks, can improve detection, reduce missed lesions, and standardize assessments. Remaining barriers include data quality, false results, regional bias, and regulatory hurdles, while ongoing research and more diverse datasets support future clinical adoption.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages3

Reading Time8 min

AccessFree to download

Introduction

Gastric cancer burden and prognosis

Endoscopy diagnostic limitations

Advanced endoscopic imaging and need for adjunct tools

Artificial intelligence in medical imaging

AI impact in gastroenterology and oncology

Review began 03/12/2025 Review ended 03/25/2025 Published 03/25/2025 © Copyright 2025 Hasan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License CC-BY 4.0. , which permits unrestricted use , distribution , and reproduction in any medium , provided the original author and source are credited. DOI: 10.7759/cureus.81194 Open Access Editorial Artificial Intelligence-Driven Approaches to Endoscopic Gastric Cancer Detection: Current Progress and Future Directions Abdul Haseeb Hasan 1 , Muhammad Ali Abid 1, Nida Waheed 1, Taimoor Wajid 1 1. Medicine, Mayo Hospital, Lahore, PAK Corresponding author: Abdul Haseeb Hasan, [abdulhaseebhassan60@gmail.com](abdulhaseebhassan60@gmail.com) Abstract Gastric cancer (GC) is one of the most prevalent and lethal cancers globally, with early detection being crucial for improving patient survival rates. Endoscopy remains the primary diagnostic tool for gastric cancer, but its accuracy is highly dependent on the endoscopist 's skill. Misdiagnosis rates, caused by factors such as incomplete lesion recognition and operator errors, remain a significant challenge. Artificial intelligence (AI), particularly deep learning models like convolutional neural networks (CNNs), has shown substantial potential in enhancing the detection of gastric cancer during endoscopic procedures. AI-driven systems have demonstrated the ability to improve diagnostic accuracy, reduce missed lesions, and standardize assessments, regardless of endoscopist experience. Studies have highlighted AI 's effectiveness, with some models achieving diagnostic accuracy comparable to senior practitioners. However, challenges such as data quality, false positives/negatives, geographical biases, and regulatory barriers need to be addressed for broader clinical implementation. Despite these limitations, ongoing advancements in AI technology, coupled with multinational research and improved dataset diversity, hold promise for improving the early detection of gastric cancer, enhancing patient outcomes, and optimizing clinical workflows. Categories: Gastroenterology, Oncology, Healthcare Technology Keywords: ai-assisted detection, artificial intelligence (ai), cancer, convolutional neural networks (cnns), deep learning, early gastric cancer (egc), endoscopic imaging, endoscopy, gastric cancer, machine learning (ml) Editorial Introduction Gastric cancer (GC) remains one of the most prevalent and lethal malignancies worldwide, ranking among the top five cancers in incidence and as the third leading cause of cancer-related mortality [1,2] . In 2020, gastric cancer was diagnosed in approximately 1.1 million people worldwide, with 720,000 cases in males and 370,000 in females [3] . The disease also led to 770,000 deaths that year [3] . If current incidence rates persist, the number of new cases is expected to rise by 62%, reaching 1.8 million by 2040, while mortality could increase to 1.3 million [3] . Its development is influenced by a combination of dietary, environmental, and genetic factors and is characterized by aggressive and heterogeneous biological behaviour [4] . The prognosis of GC is largely dependent on the stage at diagnosis, with patients in advanced stages facing a median survival of less than a year and a five-year survival rate below 20%[1,4] . In contrast, early gastric cancer (EGC), defined as tumor involvement confined to the mucosa and submucosa regardless of lymph node metastasis, has a significantly improved prognosis [4] . When detected and treated in its early stages, EGC has a five-year survival rate exceeding 90%[1,2,4]. Consequently, early diagnosis and timely intervention are crucial in reducing GC-related mortality [1,2,4]. Endoscopy remains the primary modality for diagnosing both GC and EGC, yet its diagnostic accuracy is highly dependent on the experience and proficiency of the endoscopist [1,4] . Studies have highlighted notable

FAQs

Why is early detection critical for gastric cancer outcomes?

Survival strongly depends on the stage at diagnosis. Advanced-stage patients have markedly worse median survival and five-year survival, while early gastric cancer confined to mucosa and submucosa shows a substantially higher five-year survival rate.

What limits the accuracy of endoscopy for diagnosing gastric cancer?

Diagnostic performance depends on endoscopist experience, with misdiagnosis linked to incomplete lesion recognition, inadequate mapping of gastric anatomy, and biopsy sampling errors.

How does AI improve endoscopic gastric cancer detection?

AI, particularly deep learning models such as convolutional neural networks, can enhance diagnostic accuracy, reduce missed lesions, and help standardize assessment independent of endoscopist experience.

Sep 24, 2026

Uploaded by นรินทร์