The Importance of Artificial Intelligence in Upper Gastrointestinal Endoscopy

Abstract

Recently, there has been a growing interest in the application of artificial intelligence (AI) in medicine, especially in specialties where visualization methods are applied. AI is defined as a computer’s ability to achieve human cognitive performance, which is accomplished through enabling computer “learning”. This can be conducted in two ways, as machine learning and deep learning. Deep learning is a complex learning system involving the application of artificial neural networks, whose algorithms imitate the human form of learning. Upper gastrointestinal endoscopy allows examination of the esophagus, stomach and duodenum. In addition to the quality of endoscopic equipment and patient preparation, the performance of upper endoscopy depends on the experience and knowledge of the endoscopist. The application of artificial intelligence in endoscopy refers to computer-aided detection and the more complex computer-aided diagnosis. The application of AI in upper endoscopy is aimed at improving the detection of premalignant and malignant lesions, with special attention on the early detection of dysplasia in Barrett’s esophagus, the early detection of esophageal and stomach cancer and the detection of H. pylori infection. Artificial intelligence reduces the workload of endoscopists, is not influenced by human factors and increases the diagnostic accuracy and quality of endoscopic methods.

Integrating bioinformatics and multiple machine learning to identify mitophagy-related targets for the diagnosis and treatment of diabetic foot ulcers: evidence from transcriptome analysis and drug docking

“…Certainly, the performance of DMGs for DFU still needs both single-gene as well as multi-gene tests, including ROC and ANN ( Mandair et al, 2023 ). ANN, as a form of artificial intelligence, has found extensive application in clinical medicine ( Kufel et al, 2023 ; Mandair et al, 2023 ; Popovic et al, 2023 ). However, the current clinical methods for diagnosing DFU have limitations, particularly in terms of molecular-level diagnosis, which is often reliant on the clinical experience ( Voelker, 2023 ).…”