Related Experiment Video
Updated: Jul 22, 2025
10:31
Intraoperative Gastroscopy for Tumor Localization in Laparoscopic Surgery for Gastric Adenocarcinoma
Published on: August 9, 2016
12.8K
Non-endoscopic Applications of Machine Learning in Gastric Cancer: A Systematic Review
Marianne Linley L Sy-Janairo1, Jose Isagani B Janairo2
1Institute of Digestive and Liver Diseases, St. Luke's Medical Center-Global City, Taguig, Philippines.
Journal of Gastrointestinal Cancer
|July 21, 2023
Summary
Machine learning (ML) offers promising non-endoscopic applications for gastric cancer diagnostics, therapy response prediction, and prognosis. While early-stage, these methods show potential for less invasive gastric cancer detection and management.
Area of Science:
- Oncology
- Medical Informatics
- Artificial Intelligence
Background:
- Gastric cancer presents a significant global health challenge with high incidence and mortality rates.
- Current diagnosis primarily relies on upper gastrointestinal endoscopy and biopsy, a field where most machine learning (ML) tools are concentrated.
- This review explores ML applications beyond endoscopic imaging for gastric cancer.
Purpose of the Study:
- To systematically review machine learning (ML) applications in gastric cancer that do not involve endoscopic image recognition.
- To identify and categorize the diverse non-endoscopic uses of ML in gastric cancer research.
- To assess the current landscape and future potential of ML in gastric cancer management.
Main Methods:
- A systematic literature review was conducted across two databases.
- Studies were independently evaluated by two authors.
- Data extracted included publication year, ML algorithm, performance metrics, specimen type, and clinical application.
Main Results:
- 63 studies were included from 791 screened articles.
- Non-endoscopic ML applications fall into three main categories: diagnostics, predicting therapy response, and prognosis prediction.
- Key applications include histopathologic slide analysis for diagnosis and risk scoring systems for patient survival and prognostic variable identification.
Conclusions:
- Numerous non-endoscopic ML applications for gastric cancer show promise, utilizing various specimens, including non-conventional ones.
- These techniques offer potential for developing less invasive diagnostic and prognostic tools.
- Further research is needed to advance data curation, model interpretability, and clinical validation before widespread adoption.
