SCIENCE & TECH: New AI tool predicts brain decline before symptoms appear

Who needs a crystal ball when you’ve got artificial intelligence?

Researchers at Mass General Brigham have developed a potentially game-changing AI tool capable of detecting subtle changes in brain wave patterns that could predict cognitive decline years before symptoms appear.

“This could completely change how we approach dementia prevention,” said Dr. Shahab Haghayegh, lead author of the study and an instructor at Massachusetts General Hospital and Harvard Medical School.

Unlocking the secrets of brain waves

The AI tool works its magic by analyzing brain wave activity recorded during sleep using an electroencephalography (EEG) — a non-invasive technique that tracks electrical activity in the brain through sensors on the scalp. 

The AI tool was developed using data from a separate trial on fracture risk, where women aged 65 and older took cognitive tests around the same time they participated in a sleep study that included an overnight EEG.

Haghayegh and his team honed in on 281 participants who had normal cognitive function at the time of the initial sleep study, then repeated the same cognitive tests five years later.  By the second round of assessments, 96 of the women had developed cognitive impairment.

Researchers applied advanced information theory techniques to extract brainwave patterns from the EEG data collected during the sleep study. They then used AI to detect subtle changes in brain wave activity — especially in gamma band frequencies during deep sleep — among individuals who later developed cognitive impairment, even before symptoms surfaced.

The AI tool correctly flagged 85% of individuals who eventually experienced cognitive decline, with an overall accuracy of 77%.

“Using novel sophisticated analyses, advanced information theory tools and AI, we can detect subtle changes in brain wave patterns during sleep that signal future cognitive impairment, offering a window of opportunity for intervention years before symptoms appear,” Haghayegh said. 

The promise of early intervention

A little cognitive decline is par for the course as we age, with most people experiencing slower processing speed, memory lapses and trouble multitasking over time. But when it begins to interfere with daily life, it can be a red flag that something more serious might be at play. 

Research suggests that one in 10 US adults aged 65 and older have dementia, while another 22% have mild cognitive impairment. Experts predict these numbers will rise significantly in the coming years, largely driven by longer life expectancies and an aging population.

Thanks to the AI tool, early detection could give people valuable time to make lifestyle changes — such as exercising, staying mentally active and eating a balanced diet — that may help preserve cognitive health before the effects of impairment set in.

Early identification could also open the door for therapies and medications that slow disease progression or alleviate symptoms.

“The new, FDA-approved treatments for Alzheimer’s disease are effective at the earlier stages of dementia, but not the more advanced stages,” said Dr. Kun Hu, senior author and physiologist at Brigham and Women’s Hospital and Harvard Medical School. 

“Interventions that are performed even earlier — before someone shows signs of cognitive decline — are likely to be even more effective,” Hu added. 

The research also paves the way for an exciting new possibility: manipulating brain electrical activity to reduce the risk of cognitive decline.

Haghayegh and Hu are currently designing a clinical study to test whether electrical stimulation can alter EEG patterns during sleep.