Longitudinal brain age prediction and cognitive function after stroke.

Area of Science:

• Neuroimaging and Machine Learning
• Neurology and Cognitive Science
• Gerontology

Background:

• Advanced age is a known risk factor for cognitive decline following a stroke.
• Brain age gap (BAG), calculated using machine learning on brain scans, reflects the difference between a patient's brain age and chronological age.
• Previous research has explored BAG in various conditions, but its role in post-stroke neurocognitive disorder (NCD) requires further investigation.

Purpose of the Study:

• To investigate the association between the brain age gap (BAG) and the development of post-stroke neurocognitive disorder (NCD) over time.
• To determine if baseline BAG predicts the future risk of developing NCD after a stroke.

Main Methods:

• Utilized machine learning algorithms to estimate brain age from patient brain scans.
• Calculated the brain age gap (BAG) for each patient.
• Employed survival analysis to assess the relationship between baseline BAG and the incidence of post-stroke NCD at 18 and 36 months.

Main Results:

• A lower BAG (indicating a younger-appearing brain relative to chronological age) was associated with a reduced risk of post-stroke NCD up to 36 months post-stroke.
• This association was observed even in patients without cognitive impairments at 3 months post-hospital admission.
• For patients without baseline NCD, a higher BAG at baseline predicted an increased risk of developing NCD at 18 and 36 months.

Conclusions:

• A younger brain appearance, as indicated by a lower brain age gap, is associated with a decreased risk of developing neurocognitive disorder after a stroke.
• BAG may serve as a valuable imaging biomarker for predicting long-term cognitive outcomes in stroke survivors.
• These findings underscore the potential of neuroimaging biomarkers in understanding and managing post-stroke cognitive decline.