Emerging trends in Alzheimer's disease diagnosis and prediction using artificial intelligence: A bibliometric...

AD is a significant public health challenge, and AI technologies, including deep learning and machine learning, offer the potential to dramatically improve diagnostic and predictive accuracy. These technologies are widely applied in AD research. However, comprehensive literature summaries of this field remain limited. This study uses bibliometric analysis to examine research hotspots, trends, future development potential, and limitations in AI-based AD diagnosis and prediction.

We conducted a bibliometric analysis of 100 top cited studies on AI-based diagnosis and prediction of AD, using data from the WoSCC. We performed the analysis using CiteSpace 6.3.R2, VOSviewer 1.6.19, Scimago Graphica 1.0.39, Microsoft Excel 2021, and R package Bibliometrix running in RStudio, visualizing the results through graphical representations.

It was found that the top cited 100 articles came from 51 journals and 31 countries. The journal with both the highest number of published articles and the greatest citation frequency was NEUROIMAGE, while PROTEIN ENGINEERING DESIGN & SELECTION boasted the highest average citation rate. The country with the largest volume of published articles was the United States, followed by China and the United Kingdom. In terms of institutions, the University of North Carolina had produced the most publications. The keywords identified fall into 9 categories. The most frequently occurring keywords are "Alzheimers disease", "mild cognitive impairment", "classification", "MRI", "deep learning", "diagnosis", "dementia", "biomarkers", "brain atrophy", "machine learning", "voxel based morphometry", "prediction", and "patterns".

AI-based technologies for AD diagnosis and prediction are becoming indispensable clinical tools. Future research should leverage AI to identify novel AD biomarkers, enabling precision diagnosis and treatment. However, our bibliometric analysis has limitations: language and citation biases may skew interpretation of emerging AI-AD trends.