Related Experiment Video
Updated: Dec 23, 2025
10:31
Intraoperative Gastroscopy for Tumor Localization in Laparoscopic Surgery for Gastric Adenocarcinoma
Published on: August 9, 2016
13.1K
Deep learning in gastric tissue diseases: a systematic review.
Wanderson Gonçalves E Gonçalves1,2, Marcelo Henrique de Paula Dos Santos3, Fábio Manoel França Lobato4
1Laboratório de Genética Humana e Médica - Instituto de Ciências Biológicas, Universidade Federal do Pará, Belém, Pará, Brazil.
BMJ Open Gastroenterology
|April 28, 2020
Summary
Deep learning shows great potential for analyzing gastric tissue diseases from medical images. However, challenges in evaluation metrics and data availability need addressing for improved reproducibility in gastric cancer and other conditions.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Gastroenterology
Background:
- Deep learning (DL) is increasingly vital in medical image analysis, often matching or exceeding human expert performance.
- Despite DL's rise in gastric disease research, comprehensive reviews remain scarce.
- This study focuses on DL applications for diagnosing gastric tissue diseases.
Purpose of the Study:
- To systematically review deep learning applications in analyzing gastric tissue diseases.
- To identify the potential and limitations of DL in this field.
- To highlight areas for improvement in research methodology and data accessibility.
Main Methods:
- Conducted a systematic review of studies utilizing deep learning for gastric tissue disease analysis.
- Included analyses of digital histology, endoscopy, and radiology images.
- Focused on applications for gastric cancer, ulcers, gastritis, and non-malignant conditions.
Main Results:
- Deep learning demonstrates significant effectiveness in analyzing gastric tissues across various conditions.
- Identified high potential for DL in diagnosing gastric cancer, ulcers, and gastritis.
- Highlighted shortcomings in current DL research, including evaluation metrics and data availability.
Conclusions:
- Deep learning holds substantial promise for advancing gastric tissue disease diagnosis.
- Gaps in evaluation metrics and image data availability impede experimental reproducibility.
- Further standardization and data sharing are crucial for robust DL applications in gastroenterology.
