A new artificial intelligence model measures how quickly the brain of a patient is outdated and a powerful new tool can be for preventing and treatment of cognitive decline and dementia, according to USC researchers.
The first-in-in-natted tool can follow the pace of brain changes non-invasive by analyzing magnetic resonance image formation (MRI) scans. Snellere hersenveroudering correleert nauw met een hoger risico op cognitieve stoornissen, zei Andrei Irimia, universitair hoofddocent gerontologie, biomedische engineering, kwantitatieve en computationele biologie en neurowetenschappen aan de USC Leonard Davis School of Gerontology and Visiting Associate Professor of Psychological Medicine at King’s College London at King Lond.
“This is a new measurement that can change the way we follow brain health, both in the research lab and in the clinic,” he said. “Knowing how quickly your brain is aging can be powerful.”
Irimia is the senior author of the study describing the new model and his predictive power; The study was published on February 24, 2025 in Proceedings of the National Academy of Sciences.
Biological brain age versus chronological era
The biological age distinguishes itself from the chronological age of an individual, Irimia said. Two people who are the same age based on their date of birth can have very different biological ages because of how well their bodies functions and how “old” seem to be the tissues of the body on a cellular level.
Some common measures of biological age use blood samples to measure epigenetic aging and DNA methylation that influence the role of genes in the cell. Measuring the biological age of blood samples, however, is a bad strategy for measuring the age of the brain, Irimia explained. The barrier between the brain and the bloodstream prevents blood cells from crossing the brain, so that a blood sample of a person’s arm is not immediately a reflection of methylation and other aging processes in the brain. Conversely, it is a much more invasive procedure directly from the brain of a patient, making it unattainable to measure DNA methylation and other aspects of brain aging directly from living human brain cells.
Previous research by Irimia and colleagues emphasized the potential of MRI scans to measure the biological age of the brain non-invasive. The earlier model used AI analysis to compare the brain anatomy of a patient with data collected from the MRI scans of thousands of people of different ages and cognitive health results.
However, the cross-section of analyzing one MRI scan to estimate the age of the brain had great limitations, he said. Although the previous model could, for example, tell whether the brain of a patient was “older” for ten years than their calendar era, it could not provide information about whether this extra aging took place earlier or later in their lives, nor could it indicate whether the brains it Aging was accelerated.
A more accurate picture of brain aging
A newly developed three-dimensional conventional neural network (3D-CNN) offers a more precise way to measure how the brain gets older over time. Made in collaboration with Paul Bogdan, assistant professor of electric and computer technology and holder of the Jack Munushian Early Career Chair at the USC Viterbi School of Engineering, the model was trained and validated on more than 3,000 MRI scans of cognitive normal adults.
In contrast to traditional cross-sectional approaches, which estimate the age of the brain of one scan at a single time, this longitudinal method compares the basic line and follow-up MRI scans of the same person. As a result, the more accurate neuroanatomic changes connected to accelerated or delayed aging. The 3D-CNN also generates interpretable “saliency cards”, which indicate that the specific brain areas that are most important for determining the pace of aging, Bogdan said.
When applied to a group of 104 cognitively healthy adults and 140 patients with Alzheimer’s disease, the calculations of the new model of brain aging of the new model closely correlated with changes in cognitive function tests at both times.
“The coordination of these measures with cognitive test results indicates that the framework can serve as an early biomarker of neurocognitive decline,” Bogdan said. “Moreover, it shows its applicability for both cognitive normal individuals and people with cognitive impairments.”
He added that the model has the potential to better characterize both healthy aging and disease processes, and its predictive power could be applied one day to assess which treatments would be more effective based on individual characteristics.
Speeds of brain aging are considerably correlated with changes in the cognitive function. So if you have a high speed of brain aging, you previously have a high degree of demolition in the cognitive function, including memory, executive speed, executive function and processing speed. It is not just an anatomical measure; The changes we see in anatomy are associated with changes we see in the cognition of these individuals. “
Andrei Irimia, Association Lecturer Gerontology, Biomedical Engineering, Quantitative and Computational Biology and Neurosciences, USC Leonard Davis School of Gerontology
Look forward
In the study, Irimia and CO authors also notice how the new model was able to distinguish different aging speeds in different brain areas. In these differences – including how they vary based on genetics, environment and lifestyle factors – can provide insight into how different pathologies develop in the brain, Irimia said.
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 have different risks for neurodegenerative disorders, including Alzheimer’s, he added.
Irimia said that he is also enthusiastic about the potential for the new model to identify people with faster than normal brain aging before they show symptoms of cognitive disorders. Although new medicines aimed at Alzheimer’s have been introduced, their efficacy has been hoped for less than researchers and doctors, possibly because patients may not start the medicine until much of Alzheimer’s pathology is already present in the brain, he explained.
“One thing that my lab is very interested is to estimate the risk for Alzheimer’s; we want to say one day:” At the moment it seems that this person has a 30% risk for Alzheimer’s. “We are not there yet, but we are working on it,” Irimia said. “I think that these types of measures will be very useful to produce variables that are prognostic and can help predict the risk of Alzheimer’s. That would be really powerful, especially if we start developing potential drugs for prevention.”
Source:
Journal Reference:
Yin, C., et al .. (2025) Learning deeply to quantify the pace of brain aging in relation to neurocognitive changes. Pnas. doi.org/10.1073/pnas.2413442122.