UF AI Tool Diagnoses Dementia Types with 99% Accuracy

University of Florida researchers created an artificial intelligence tool that distinguishes Alzheimer's disease from dementia with Lewy bodies with near-perfect accuracy, addressing a diagnostic problem that causes misdiagnosis in up to 50% of dementia with Lewy bodies cases.

The tool, called Automated Imaging Differentiation for Dementia (AIDD), combines specialized MRI brain scans with AI to map water-movement patterns caused by brain cell damage and inflammation. Researchers published their findings in Neurology this month.

Misdiagnosis carries serious consequences. Dementia with Lewy bodies typically begins with attention, alertness and movement issues, while Alzheimer's patients show memory problems first. The diseases require different treatments, and incorrect diagnosis can worsen cognitive and motor functions.

Researchers analyzed 519 brain scans from patients with Alzheimer's, dementia with Lewy bodies and healthy controls collected between January 2007 and March 2022 at multiple research centers. They used 387 scans to train and test the AI model, splitting the data 80-20 for training and testing.

In a final validation, researchers applied AIDD to 13 patients whose diagnoses doctors confirmed through autopsy after death. The tool correctly identified all 13 cases.

"Developing precision biomarkers will offer better outcomes for patients since therapies for these diseases differ," said David Vaillancourt, a distinguished professor at UF's Department of Applied Physiology and Kinesiology.

Alzheimer's disease and related dementias are expected to more than double by 2060, according to projections.