AI Tool Predicts Markers of Alzheimer’s Disease

While some blood tests can detect signs of the disease, they cannot pinpoint the exact brain regions affected. Researchers have now developed an artificial intelligence (AI) solution that uses widely available and less costly tests to predict the presence and location of these proteins.

This AI tool, developed by scientists at Boston University Chobanian & Avedisian School of Medicine (Boston, MA, USA), uses brain scans, memory assessments, genetic information, health records, and medical history to detect Alzheimer’s biomarkers. The model was trained on data from seven research cohorts totaling 12,185 participants, learning to identify patterns linked to protein buildup. It was also designed to operate effectively even when some patient data was missing.

The team tested the AI on an independent group of participants not included in the training phase. The results, published in Nature Communications, show that the tool could accurately identify individuals with high amyloid or tau levels. According to the study, this AI tool could streamline patient selection for clinical trials and new drug treatments, enabling faster and more affordable diagnoses.

Additionally, by identifying disease earlier, it could support personalized intervention plans, such as targeted diets or exercise programs to slow disease progression. Researchers believe it may one day be applied to related conditions like frontotemporal dementia and chronic traumatic encephalopathy.

“The tool can help doctors quickly pick people for treatment with new drugs or to participate in research studies, thus saving time and money while reaching more patients who might not have access to costly and complicated tests,” said Vijaya B. Kolachalama, PhD, FAHA, associate professor of medicine and computer science, Boston University. “For the public, this means faster diagnoses, fewer unnecessary exams, and hope for treatments that slow the disease, improving daily life for those affected and their loved ones.”

Related Links:
Chobanian & Avedisian School of Medicine