AI Detects Early Alzheimer's with 93% Accuracy

Scientists in Massachusetts just cracked a major code in fighting Alzheimer's disease: catching it before most people even know something's wrong.

Researchers at Worcester Polytechnic Institute developed an AI system that predicts Alzheimer's with 93% accuracy by analyzing brain scans for subtle changes invisible to the human eye. At the same time, Mass General Brigham scientists created AI tools that scan doctor visit notes for early warning signs hiding in plain sight.

The timing couldn't be better. Right now, 90% of people in the earliest stage of Alzheimer's go undiagnosed in the United States, missing the narrow window when newer medications can help slow the disease's progression.

"The biggest opportunity to improve Alzheimer's care isn't in a new drug — it's in noticing the earliest signs sooner," said Dr. Lidia Moura, who directs population health in neurology at Mass General Brigham.

The WPI team trained their AI to spot volume loss in three specific brain regions: the hippocampus, amygdala, and entorhinal cortex. These areas shrink as Alzheimer's damages neurons, but the changes are too small for doctors to reliably catch during routine scans.

Sean Terwilliger knows firsthand why early detection matters. The 62-year-old from Holyoke spent five years getting misdiagnosed after a ministroke in 2018 before finally getting his Alzheimer's diagnosis in 2024.

"I'm probably the perfect candidate for being able to use predictive AI to find something," said Terwilliger, who now advocates for the Alzheimer's Association. "It would have saved me loads of grief."

After his diagnosis, Terwilliger started treatment with Leqembi, one of two drugs approved since 2023 that modestly slow Alzheimer's progression in early stages. He completed 18 months of infusions and says he feels sharper than before treatment.

The Ripple Effect

About 7 million Americans currently live with Alzheimer's, the most common form of dementia. Catching the disease earlier doesn't just mean better treatment outcomes. It gives patients more time to participate in clinical trials, manage symptoms longer, and make important decisions about their future while they still can.

The Mass General Brigham team analyzed 3,300 clinical notes from 200 patients and found that warning signs already exist in medical records. AI can spot patterns that busy doctors miss: forgotten appointments, repeated questions, subtle word-finding difficulties mentioned in passing.

Terwilliger now fills his days with brain exercises like Wordle and Connections, activities he wishes he'd started sooner. He writes a blog about living with Alzheimer's, turning his journey into hope for others facing the same diagnosis.

These AI breakthroughs mean future patients won't have to wait five years like Terwilliger did. They'll get answers when treatments work best and time matters most.