UNSTRUCTURED
Abstract
Background: A growing number of studies have reported artificial intelligence (AI) has been developed for diagnosis and outcome prediction in clinical practice. Furthermore, AI in digestive endoscopy has attracted much attention, which has shown promising and stimulating results. Our study aimed to visualize the articles to determine the trends and hotspots of AI in digestive endoscopy.
Methods: Publications on AI in digestive endoscopy research were retrieved from the Web of Science Core Collection (WoSCC) on March 14, 2021. Microsoft Excel 2016, VOSviewer 1.6.11.0, and CiteSpace V were used to assess and plot the research output.
Results: The analytic research was based on original articles and reviews. A total of 121 records of AI research in digestive endoscopy published from 2017 to 2021 were retrieved. The citation number for these articles ranged from 0 to 142. The number of published articles increased 68-fold just from 2017 to 2020. All publications were distributed among 31 countries and 296 institutions. Asian countries had the most publications in this field (80.17%). Among the 31 countries, China and Japan were consistently the leading driving force and contributed mostly (31.40% and 28.93%, respectively), with a strong academic reputation in this area. Tada Tomohiro distributed the most related articles (13.22%) and was cited the most frequently. Gastrointestinal endoscopy published the largest number of publications (14.88%), and 4 of the top 10 cited references were in this leading journal. “Barrett’s esophagus” was the leading research hotspot. The keywords “classification,” “polyps,” “risk,” “histology,” and “resection” appeared most recently as research frontiers.
Conclusions: Our study provides a systematic elaboration for researchers to obtain a good comprehension of AI development in digestive endoscopy.