‘Local Brain Age’ AI Mapping Tool Reveals Patterns of Cognitive Impairment

Conventional brain age measures often compress complex, region-specific changes into a single value, limiting clinical insight. Variability in how different parts of the brain age can mask patterns tied to cognitive decline. To help address this challenge, researchers have developed an artificial intelligence approach that creates detailed maps of local brain aging from MRI

Developed at the University of Southern California (USC; Los Angeles, CA, USA), the deep-learning method produces anatomically detailed maps showing how old specific brain regions appear relative to a person’s chronological age. Unlike single-number estimates, the model measures local brain age at the voxel level, capturing regional vulnerability and resilience. This framing enables comparisons across brain structures that are directly relevant to cognition.

The team trained the neural network on MRI scans from 14,748 cognitively healthy adults ages 19 to 100 drawn from six large public datasets, including the UK Biobank, the Human Connectome Project, and the Alzheimer’s Disease Neuroimaging Initiative. These data established a normative baseline against which the model estimates local brain age. In healthy adults, the model consistently indicated that frontal and temporal lobes appeared biologically older than parietal and occipital regions, and it found slightly more advanced aging in the right hemisphere than the left regardless of handedness.

The approach was then tested using MRI from more than 1,900 additional participants in the Alzheimer’s Disease Neuroimaging Initiative, spanning cognitively normal adults, people with mild cognitive impairment, and people with Alzheimer’s disease. In impaired groups, the maps showed accelerated local aging in structures affected early in neurodegeneration, including the hippocampus, amygdala, and several deep brain regions involved in memory and cognitive processing. Older local brain age was associated with poorer performance on cognitive assessments, with the strongest relationships observed in Alzheimer’s disease.

Findings were published in Proceedings of the National Academy of Sciences on August 3, 2026. The authors note the method remains a research tool because it was trained primarily on research-quality MRI and relies largely on cross-sectional data. Additional validation on more diverse clinical datasets and longitudinal studies will be needed to evaluate progression and potential responses to experimental therapies.

“This more nuanced understanding of how the brain ages could pave the way for earlier identification of dementia, a better understanding of what factors affect risk and new ideas for treatment approaches,” said Andrei Irimia, Associate Professor at the USC Leonard Davis School of Gerontology.

Related Links
USC Leonard Davis School of Gerontology