A bibliometric analysis of 12,564 papers (2014–2024) mapping the rapid growth, key countries/institutions/sources, and dominant themes (deep learning, MRI) including applications to diagnosis of Parkinson's, Alzheimer's, and MCI.

Helps researchers navigate the neuroimaging–AI landscape to find datasets, methods, and collaborators for imaging-based biomarkers or diagnostic ML for Parkinson's, but offers little mechanistic or therapeutic actionable insight.

This paper presents a bibliometric analysis of the fast-growing area of deep learning in neuroimaging. Using data from the Scopus database, we analyzed 12564 peer-reviewed publications originating from 102 countries, published in 2259 sources over the period from 2014 to 2024. The field demonstrated a compound average annual growth rate of 51.7%. We found that China emerged as the most productive contributor, accounting for 22.9% of the total publications and 18% of total citations. The Chinese Academy of Sciences was identified as the most productive research institution with 149 publications and 1557 citations, while Lecture Notes in Computer Science was noted as the most highly cited source in this domain. High usage of deep learning, brain, and magnetic resonance imaging identified the most prominent research themes. Also, our analysis noted strong research emphasis on the application of various deep learning architectures for the diagnosis and study of important neurological disorders like Parkinson's Disease, Alzheimer's Disease, and Mild Cognitive Impairment. The article would be useful in understanding the current state-of-the-art deep learning for neuroimaging by identifying key research trends, influential institutions, and prominent research themes. In this way, it will contribute to helping future researchers go further in this fast-growing field.