Alzheimer's disease (AD) is a progressive neurodegenerative disorder for which there is currently no cure, and its incidence is on the rise. Early detection is essential for timely intervention and slowing the progression of the disease. While the brain structures of healthy aging individuals change gradually, the aging trajectories in AD patients deviate significantly. Recent advancements in deep learning have enabled the detection of subtle changes in brain structures using neuroimaging data. By developing brain age prediction models based on data from healthy individuals, it is possible to estimate brain age at various stages of AD progression and assess the Brain Age Gap (BAG)-the difference between predicted brain age and chronological age-which holds promise as an early diagnostic biomarker for AD. In recent years, the use of artificial intelligence, particularly deep learning, for brain age prediction has attracted considerable attention. This systematic review provides a comprehensive summary of the current state of research in this field, focusing on the progress and limitations of machine learning techniques for brain age prediction. We place particular emphasis on deep learning methods, addressing data sources, model development, and interpretability. Additionally, we analyze key challenges in the field, including site effects, bias correction, insufficient data, hardware requirements, model accuracy, and clinical applicability. Finally, we offer insights and recommendations for future research directions to address these challenges and further enhance the potential of BAG as a diagnostic tool for AD.
Keywords: Alzheimer's disease; Brain age gap; Deep learning; Early diagnosis; Neuroimaging.
Copyright © 2025 Elsevier B.V. All rights reserved.