Neurology and Neurosurgery

March 24, 2023

Chronological age can differ from "brain age," which incorporates age-related brain changes. As described in the September 2022 issue of The Lancet, Mayo Clinic researchers are using a deep learning model to better understand brain age and neurodegenerative disease.

The researchers created their model using a large collection of brain MRI and fluorodeoxyglucose positron emission tomography (FDG-PET) scans. The structural and metabolic data were used to train deep neural networks to predict individuals' ages.

The research involved 4,127 scans from 2,349 individuals. Among those individuals, 1,085 were healthy, 480 had mild cognitive impairment, 215 had Alzheimer's disease, 86 had Lewy body dementia and 45 had frontotemporal dementia. The development of that model was described in a study published in the May 2022 issue of Nature Aging.

The difference between the model's prediction of an individual's age and that individual's chronological age is known as brain age gap.

Key findings

• Higher brain age gap was found in individuals with neurodegenerative diseases compared with healthy individuals.
• Brain age gap was associated with pathologic tau protein deposition in individuals with Alzheimer's disease.
• Brain age gap was associated with longitudinal cognitive decline across all the neurodegenerative disorders studied.

The researchers note that the deep learning model produces a biomarker of neurodegenerative disease states.

"It can index severity and aging pathology in a general way, and has emerging applications for diagnosis, disease staging and monitoring of age-related diseases in the brain," says David T. Jones, M.D., a neurologist at Mayo Clinic in Rochester, Minnesota.

For more information

Jones DT, et al. Digitising brain age. The Lancet. 2022;400:988.

Lee J, et al. Deep learning-based brain age prediction in normal aging and dementia. Nature Aging. 2022;2:412.

Refer a patient to Mayo Clinic.