Artificial intelligence (AI) has been used to develop brain scans that may enable earlier diagnosis and deeper understanding of dementia, according to researchers in the United States. The new AI-powered imaging approach maps how different regions of the brain age, revealing patterns of neurodegeneration linked to cognitive decline.

The study, published in the Proceedings of the National Academy of Sciences, involved training a deep-learning AI model on magnetic resonance imaging (MRI) data from nearly 15,000 cognitively healthy individuals. This provided a baseline to assess the "local brain age"—or how old various brain regions appear compared to typical aging patterns for people of the same chronological age.

Unlike previous methods that assign a single overall "brain age," this technique produces detailed maps showing the varied rates at which distinct areas of the brain deteriorate. Lead researcher Professor Andrei Irimia noted that some brain regions demonstrate resilience against aging, while others show greater vulnerability to neurodegeneration. The AI model identified accelerated aging in specific brain regions commonly implicated in early stages of mild cognitive impairment and Alzheimer's disease.

Dementia currently affects approximately 982,000 people in the United Kingdom, with projections estimating this number will increase to 1.4 million by 2040. However, more than one-third of those living with dementia remain undiagnosed, underscoring the need for improved detection methods.

Professor Irimia emphasized that providing a more nuanced picture of brain aging holds potential to facilitate earlier diagnosis, enhance understanding of risk factors, and guide the development of new therapeutic approaches. The AI-driven insights allow researchers to move beyond a single, generalized measure and may help pinpoint where neurodegenerative processes begin and how they progress.

The research team believes this advancement represents an important step in neuroscience, offering a sophisticated tool to explore the complexities of brain aging and its relationship to cognitive function decline. Further studies will be needed to validate the clinical applications of this approach and to explore how it might be integrated into routine assessment of individuals at risk of dementia.