Research Trends Using Artificial Intelligence in the MRI from 1989 to 2023: Analysis Using Text Mining

Abstract

Purpose: Although the research areas applying artificial intelligence in the field of magnetic resonance imaging (MRI) have been expanding rapidly in recent years, the means to comprehensively understand these research areas have been limited. The purpose of this study was to visualize the research areas related to artificial intelligence in the field of MRI, and to understand the trend of research. Methods: Using PubMed database, we extracted article titles applying artificial intelligence in the MRI field from January 1, 1989 to December 31, 2023, created an extracted word list, graphs showing the relative frequency of occurrences of words, and drew a co-occurrence network diagram to investigate the frequency of appearance of words and changes in frequency and characteristic words over time. Results: The number of extracted titles was 2870. The most frequently appearing word was “deep learning” (1170 times from 2019 to 2023). Furthermore, deep learning was the word with the strongest co-occurrence (Jaccard coefficient 0.48 from 2019 to 2023). Regarding words related to organs, there was an increasing trend in the appearance frequency of the brain, prostate, and breast. Conclusion: In recent years, the research area related to artificial intelligence in the field of MRI has become a thriving area involving deep learning. In addition, there were many studies in the diagnostic area throughout the period.