Deep Learning in neuroimaging for neurodegenerative diseases: State-of-the art, Challenges, and Opportunities

Author: Akan T, Akan S, Alp S, Ledbetter CR, Tafti AP, Arevalo O, Bhuiyan MA.

Editor's Choice
Journal of the Neurological Sciences | Deep Learning in neuroimaging for neurodegenerative diseases: State-of-the art, Challenges, and Opportunities

Article Link: https://www.jns-journal.com/article/S0022-510X(25)00355-7 ⧉

A new article explores how deep learning (DL) is transforming neuroimaging for the early detection and monitoring of neurodegenerative diseases (NDDs). While DL models have shown impressive diagnostic accuracy, their clinical use remains limited by real-world variability across scanners, sites, and patient demographics.

This contribution broadens the research focus beyond Alzheimer's and Parkinson's to include Frontotemporal Dementia, Lewy Body Dementia, Huntington's Disease, Amyotrophic Lateral Sclerosis, and Multiple Sclerosis. It highlights how recent models are leveraging spatial and temporal brain imaging data, while addressing persistent challenges such as data imbalance, limited annotations, and scanner inconsistencies.

Read the full article for a deeper look at how AI-driven neuroimaging is advancing NDD diagnosis and what obstacles remain before full clinical integration.