Frontiers and hotspots evolution of machine learning in gout from 2004-2023: A bibliometric analysis

This study aimed to explore the evolution of frontiers and hotspots in machine learning for gout from 2004 to 2023.

The database of publications was sourced from the Science Citation Index-Expanded within the Web of Science Core Collection. Employing CiteSpace 6.2.R5 software, a visual analysis of the core information within these publications was conducted. This retrieval spanned from 2004 to 2023 and was completed within the confines of a single day, specifically on October 20, 2023.

In total, 219 articles were eventually included in the analysis. Publication activity remained limited before 2018 but increased markedly thereafter; 47 articles were published in 2023 by the search date of October 20. In terms of citation impact, Breiman L holds the most significant influence. China and the United States of America, along with their affiliated academic institutions, exert substantial academic influence in this field, surpassing other countries and institutions. "Ascorbic acid" is the keyword with the longest burst duration, and "prediction" holds the highest strength. The cited journals encompass various disciplines, including internal medicine, cardiology, analytical chemistry, and endocrinology and metabolism. The prevailing developmental pattern in the field of machine learning in gout primarily revolves around interdisciplinary integration.

Machine learning has indeed become a frontier and hotspot in aiding the prediction and diagnosis of gout over the past 5 years. The acceptance of predictive diagnostic models and the determination of the most suitable algorithmic models remain the current and future research focal points in this field.