A new artificial intelligence model measures how fast a patient’s brain is aging and could be a powerful new tool for understanding, preventing and treating cognitive decline and dementia, according to USC researchers. The first of its kind, it can track the pace of brain changes non-invasively by analyzing magnetic resonance imaging (MRI) scans.
Faster brain aging correlates closely with a higher risk of cognitive impairment, according to Andrei Irimia, associate professor of gerontology, biomedical engineering, quantitative and computational biology and neuroscience at the USC Leonard Davis School of Gerontology and visiting professor of psychological medicine at King’s College London. “This is a novel measurement that could change the way we track brain health both in the research lab and in the clinic,” he said. Knowing how fast the brain ages can be very revealing. Irimia is the lead author of the study describing the new model and its predictive power. The study was published in the Proceedings of the National Academy of Sciences.
Biological Brain Age and Chronological Age May Differ
Biological age is different from a person’s chronological age. Two people who are the same age based on their date of birth may have very different biological ages based on the functioning of their bodies and the “age” that the body tissues appear to have at the cellular level.
Some common methods of measuring biological age use blood samples to measure epigenetic aging and DNA methylation, which affect the role of genes in the cell. However, measuring biological age using blood samples is a poor strategy for measuring the age of the brain, Irimia explained. The barrier between the brain and the bloodstream prevents blood cells from entering the brain, so a blood sample from the arm does not directly reflect methylation and other age-related processes in the brain. Conversely, taking a sample directly from a patient’s brain is a much more invasive procedure, so it is not possible to measure DNA methylation and other aspects of brain aging directly from living human brain cells.
Previous research by Irimia and colleagues has shown the potential of MRI scans to non-invasively measure the biological age of the brain. The earlier model used AI analyses to compare a patient’s brain anatomy with data compiled from MRI scans of thousands of people of different ages and with different cognitive health outcomes. However, the cross-sectional analysis of an MRI scan to estimate brain age has significant limitations, according to the researchers. For example, while the previous model could detect whether a patient’s brain was ten years “older” than their calendar age, it could not provide information on whether this additional ageing had occurred earlier or later in their life, nor could it indicate whether the ageing of the brain was accelerated.
New Model Provides More Accurate Picture of How the Brain Ages
A newly developed three-dimensional convolutional neural network (3D-CNN) provides a more accurate way to measure how the brain ages over time. The model was developed in collaboration with Paul Bogdan, associate professor of electrical and computer engineering and holder of the Jack Munushian Early Career Chair at the USC Viterbi School of Engineering, and trained and validated using more than 3,000 MRI scans of cognitively normal adults.
In contrast to conventional cross-sectional approaches, where brain age is estimated from a single scan at a specific point in time, this longitudinal method compares MRI scans of the same person at the beginning and at the end. This allows neuroanatomical changes associated with accelerated or slowed aging to be more accurately determined. The 3D-CNN also generates interpretable “salience maps” that show the specific brain regions that are most important for determining the pace of aging, Bogdan said. When applied to a group of 104 cognitively healthy adults and 140 Alzheimer’s patients, the new model’s calculations of the rate of brain aging correlated closely with changes in cognitive function tests performed at both time points.
According to Bogdan, the agreement between these measurements and the results of the cognitive tests indicates that the model can serve as an early biomarker for neurocognitive decline. It also shows its applicability to both cognitively normal individuals and those with cognitive impairment. He added that the model has the potential to better characterize both healthy aging and disease progression, and that its predictive power could one day be used to assess which treatments would be more effective based on individual characteristics.
“The rate of brain aging correlates significantly with changes in cognitive function,” says Irimia. “So if brain aging is rapid, cognitive function, including memory, execution speed, executive function and processing speed, is also more likely to decline rapidly. It’s not just an anatomical measurement; the changes we see in the anatomy are related to changes we see in the cognition of these individuals.”
Gender Differences and Prognostic Possibilities
In the study, Irimia and his co-authors also note how the new model was able to distinguish different rates of aging in different regions of the brain. Examining these differences – including how they vary according to genetics, environment and lifestyle – could shed light on how different pathologies develop in the brain. The study also showed that the pace of brain aging in certain regions differed between the sexes, which could shed light on why men and women are at different risks for neurodegenerative diseases, including Alzheimer’s.
Irimia said he is also excited about the potential of the new model to identify people with faster-than-normal brain aging before they show symptoms of cognitive impairment. While new drugs for Alzheimer’s have been introduced, their effectiveness has been less than researchers and doctors had hoped, possibly because patients don’t take the drug until a large amount of Alzheimer’s pathology is already present in the brain. The researchers hope to create prognostic variables in the future that can help predict Alzheimer’s risk. This would be extremely helpful, especially when it comes to developing potential drugs for prevention.