Artificial Intelligence applications in healthcare: A bibliometric and topic model-based analysis

• Published in: Intelligent systems with applications 2024-03, Vol.21, p.200299, Article 200299
• Main Authors: Alhashmi, Saadat M., Hashem, Ibrahim Abaker Targio, Al-Qudah, Islam
• Format: Article
• Language: English
• Subjects: Analysis; Artificial Intelligence; Bibliometric; Healthcare
• ISSN: 2667-3053; 2667-3053
• DOI: 10.1016/j.iswa.2023.200299

•We provide a bibliometric analysis of current AI.•We present research trends that evaluate highly academic articles and topic modelling.•We highlight the most popular counties.•We analysis trends in machine learning, deep learning, and cloud computing. Artificial Intelligence (AI) has emerged as a leading technology that can significantly enhance healthcare systems., including diagnosis and treatment recommendations, patient engagement and adherence, and health predictions, because of recent developments in digitized data acquisition, cloud computing, IoT, and Machine learning. In this study, we conducted a bibliometric analysis to evaluate the trend of healthcare applications' research assessment publications indexed in Scopus from 1991 to 2022. A biblioshiny program was used for data visualization to produce distance- and graph-based maps. Moreover, the study presented a unique set of topics and terms that correlate with certain areas related to AI. using the popular Latent Dirichlet Allocation technique. A Corpus of 2,335 articles from 8,536 authors were analysed. The top 20 journals have been extracted to provide the recent trends in healthcare applications concerning AI Results reveal shifting trends in AI and its applications in healthcare. Certain areas of machine learning and deep learning are gaining momentum while others are diminishing. Artificial intelligence (AI) has transformed modern healthcare since its 1950s inception. AI, particularly machine learning, has enriched disease prediction, diagnosis, and treatment, benefiting patients and healthcare providers. This paper presents a comprehensive analysis of AI's current healthcare research landscape. Employing bibliometric analytics, it explores document trends, top sources, influential countries, dynamic keywords, and emerging research topics. The study highlights the United States as a dominant force in AI healthcare research, with over 5,000 citations. Keyword analysis reveals the shift from fuzzy logic to deep learning, signifying its increasing relevance. Deep learning research surged, reaching 616 publications in 2021. The analysis identifies common keywords in AI healthcare articles. Moreover, using the popular Latent Dirichlet Allocation technique, the study presented a unique set of topics and terms that correlate with certain areas related to AI. A Corpus of 2,335 articles from 8,536 authors were analysed. While limitations exist, such as the need for broader databases like the Web