Background
Artificial Intelligence (AI) is an increasingly popular research focus for multiple areas of science. The trend of using AI-based clinical research in different fields of medicine and defining the shortcomings of those trials will guide researchers and future studies.
Methods
We systematically reviewed trials registered in ClinicalTrials.gov that apply AI in clinical research. We explored the trend of AI-applied clinical research and described the design and conduct of such trials. Also, we considered high-quality trials to represent their enrollees’ and other characteristics.
Results
Our search yielded 839 trials involving a direct application of AI, among which 330 (39.3 %) trials were interventional, and the rest were observational (60.7 %). Most of the studies aimed to improve diagnosis (70.2 %); in less than a quarter of trials, management was targeted (22.8 %), and AI was implemented in an acute setting (13 %). Gastrointestinal, cardiovascular, and neurology were the significant fields of medicine with the application of AI in their research. High-quality published AI trials showed good generalizability in terms of their enrollees’ characteristics, with an average age of 52.46 years old and 50.28 % female participants.
Conclusion
The incorporation of AI in different fields of medicine needs to be more balanced, and attempts should be made to broaden the spectrum of AI-based clinical research and to improve its deployment in real-world practice.
