New tools seek to boost early detection

Key Points:

• Researchers at Worcester Polytechnic Institute (WPI) developed an AI method analyzing brain scans that predicts Alzheimer’s disease with nearly 93% accuracy by detecting volume loss in specific brain regions like the hippocampus, amygdala, and entorhinal cortex.
• A Massachusetts General Brigham (MGB) team created an AI tool that reviews electronic medical records from routine doctor visits to identify subtle early signs of cognitive impairment, achieving 88% accuracy in detecting potential Alzheimer’s risk.
• Early diagnosis enabled by AI is critical as newly approved drugs like Leqembi and Kisunla can modestly slow Alzheimer’s progression but must be administered during mild cognitive impairment or early dementia stages, which are often missed or mistaken for normal aging.
• Patient advocates emphasize that earlier diagnosis through AI could improve treatment outcomes, symptom management, and planning, as illustrated by Sean Terwilliger, who was diagnosed late despite early symptoms and now benefits from Leqembi treatment.
• Experts caution that AI tools must be carefully validated to avoid false positives, as cognitive decline can result from various causes beyond Alzheimer’s, underscoring the need for high sensitivity and specificity in these predictive systems.