A Review on Explainable Artificial Intelligence for Gastrointestinal Cancer using Deep Learning

Artificial intelligence is set to take on multiple roles usually controlled by humans, and the use of AI-based medicine in gastroenterology is expected to increase in the near future. Artificial intelligence (AI) is expected to have an immediate impact on clinical image-based diagnoses, such as pathology, radiography, and endoscopy. Recent studies indicate that convolutional neural network (CNN) and deep learning (DL) methodology featuring multilayer perceptrons and designed to operate with minimal preprocessing, holds significant promise in the field of medicine. The requirement for explainability rises along with the usage of deep learning-based techniques, particularly in industries where critical decisions are made, such the interpretation of medical images. The review paper provides an overview of explainable AI used in DL-based medical image analysis. Approximately two million people around the world die as a result of gastrointestinal infections. One of the most cutting-edge clinical imaging techniques for diagnosing and treating gastrointestinal illnesses like bleeding, stomach ulcers, and polyps is video endoscopy. Doctors must spend a lot of time evaluating all of the images produced by medical video endoscopy since there are so many of them. This makes human diagnosis challenging and has sparked research into computer-aided ways for efficiently and accurately identify all generated images. This structured review outlines research findings in gastric cancer and deep-based strategies for the characterization and prognosis of gastrointestinal cancer as well as limitations and future opportunities for AI in gastric cancer.