Medical News Observer

• Researchers developed a machine-learning model trained on FLAIR MRI imaging to predict brain age and identify early signs of Alzheimer’s.
• The BrainAGE metric, showing the gap between predicted and actual brain age, effectively distinguishes between stable and converting mild cognitive impairment.
• This approach provides a significant accuracy in predicting Alzheimer’s up to four years before onset, correlating with established biomarkers.

Researchers have made a significant breakthrough in the early detection of Alzheimer’s disease, leveraging the power of brain imaging and machine learning. In a recent study, scientists developed a novel approach using brain age prediction models derived from FLAIR MRI imaging. This method primarily focuses on identifying individuals at the prodromal stage of Alzheimer’s, specifically targeting those with mild cognitive impairment (MCI). The distinction between patients who will progress to Alzheimer’s (converting mild cognitive impairment) and those who will remain stable (stable mild cognitive impairment) is crucial for early intervention strategies.

The core of this research lies in a sophisticated machine-learning model trained on data from 3239 FLAIR MRI volumes. This model predicts the brain age of normal control subjects based on volume, intensity, and texture features. The brain age gap estimation (BrainAGE) metric, computed as the difference between the predicted and actual age, emerges as a vital biomarker. Results showed that the BrainAGE metric differentiated between stable and converting MCI groups but did so with significant accuracy up to four years before the onset of Alzheimer’s disease. 

This early detection capability is a game-changer, offering a window for potential intervention. Additionally, this BrainAGE metric correlates significantly with established Alzheimer’s disease biomarkers, bolstering its reliability. This research paves the way for a new era in Alzheimer’s detection, where a single, explainable value can indicate whether an individual is experiencing normal aging or veering towards high Alzheimer’s risk, revolutionizing how we approach this debilitating disease.

Ref: Crystal, Owen, Pejman J. Maralani, Sandra Black, Corinne Fischer, Alan R. Moody, and April Khademi. 2023. “Brain Age Estimation on a Dementia Cohort Using FLAIR MRI Biomarkers.” AJNR. American Journal of Neuroradiology, November. https://doi.org/10.3174/ajnr.A8059.