The model can predict the chronological ages of cognitively normal participants with an average absolute error of 2.3 years, which is about 1 year more accurate than an existing model.
A new artificial intelligence (AI) tool that analyzes MRI brain scans could accurately capture cognitive decline associated with neurodegenerative diseases, according to researchers at the University of Southern California.
Research has shown that a person’s brain age is a more useful and accurate predictor of health risks and future diseases than their date of birth and is considered a reliable biomarker for neurodegenerative disease risk. These risks increase when an individual’s brain exhibits features that appear “older” than expected for someone of that age.
By using a novel AI model to analyze brain scans, the researchers found that they could detect subtle brain anatomy markers that are otherwise very difficult to detect and that correlate with cognitive decline.
“Our study harnesses the power of deep learning to identify areas of the brain that are aging in ways that reflect a cognitive decline that may lead to Alzheimer’s,” said corresponding author Andrei Irimia, PhD, in a press release. “People age at different rates, and so do tissue types in the body. We know this colloquially when we say, ‘So-and-so is 40, but looks 30.’ The same idea applies to the brain.”
In the study, investigators collated the brain MRIs of 4681 cognitively normal participants, some of whom went on to develop cognitive decline or Alzheimer disease later in life. Using these data, they created an AI model called a neural network to predict participants’ ages from their brain MRIs.
First, the team trained the network to produce detailed anatomic brain maps that reveal subject-specific patterns of aging. They then compared the biological brain ages with the chronological ages of study participants. The greater the difference between the 2, the worse the participants’ cognitive scores, which reflect Alzheimer disease risk.
The results show that the model can predict the true (chronological) ages of cognitively normal participants with an average absolute error of 2.3 years, which is about 1 year more accurate than an existing, award-winning model for brain age estimation that used a different neural network architecture.
“Interpretable [artificial intelligence] can become a powerful tool for assessing the risk for Alzheimer’s and other neurocognitive diseases,” Irimia said in the press release. “The earlier we can identify people at high risk for Alzheimer disease, the earlier clinicians can intervene with treatment options, monitoring, and disease management.”
The study also found sex-specific differences in how aging varies across brain regions. Certain parts of the brain age faster in males than in females, and vice versa, according to the study results.
Men, who have a higher risk of motor impairment due to Parkinson disease, experience faster aging in the brain’s motor cortex, which is responsible for motor function. Similarly, the findings also showed that among women, typical aging may be relatively slowed in the right hemisphere of the brain.
Potential applications for these findings extend far beyond disease risk assessment, Irimia noted. The novel deep learning methods developed by the researchers are used to help patients understand how fast they are aging in general.
“One of the most important applications of our work is its potential to pave the way for tailored interventions that address the unique aging patterns of every individual,” Irimia said in the press release. “Many people would be interested in knowing their true rate of aging. The information could give us hints about different lifestyle changes or interventions that a person could adopt to improve their overall health and well-being. Our methods could be used to design patient-centered treatment plans and personalized maps of brain aging that may be of interest to people with different health needs and goals.”
How old is your brain, really? Artificial intelligence knows. News release. EurekAlert; January 6, 2023. Accessed January 16, 2023. https://www.eurekalert.org/news-releases/975942