Early gastric cancer and Artificial Intelligence: Is it time for population screening?. (June 2021)

• Record Type: Journal Article
• Authors: Arribas Anta, Julia; Dinis-Ribeiro, Mario
• Is Part Of: Best practice & research. Volume 52/53(2021)
• Publication Date: 2021-06
• Subjects: Artificial intelligence -- Gastric cancer -- Screening -- Diagnosis; Gastroenterology -- Periodicals; Gastrointestinal system -- Diseases -- Periodicals; Gastrointestinal Diseases; Gastroenterology; Klinische geneeskunde; Gastro-enterologie; Gastroenterology; Gastrointestinal system -- Diseases; Periodicals; 616.33005
• DOI: 10.1016/j.bpg.2020.101710
• Languages: English
• ISSNs: 1521-6918

Abstract

Abstract: Gastric cancer is a common cause of death worldwide and its early detection is crucial to improve its prognosis. Its incidence varies throughout countries, and screening has been found to be cost-effective at least in high-incidence regions. Identification of individuals harbouring preneoplastic lesions and their surveillance or of those with early gastric cancer are extremely important processes and endoscopy play a key role for this purpose. Unfortunately, also quality and accuracy for endoscopic detection varies among centres and endoscopists. Recent studies about Artificial Intelligence applied to endoscopic imaging show that these technologies perform very well and could be extremely useful for endoscopists to achieve the accuracy needed for gastric cancer screening. Nonetheless, as its introduction in this field is very recent, most studies are carried out offline and its results in clinical practice need to be further validated namely by incorporating all the components/dimensions of endoscopy from pre to post-assessment.