Researchers at the University of Florida have developed an artificial intelligence tool that separates two commonly confused forms of dementia with what they describe as near-perfect accuracy. Reported in the journal Neurology, the system is called Automated Imaging Differentiation for Dementia, or AIDD, and it pairs brain scans with AI analysis to distinguish Alzheimer's disease dementia from dementia with Lewy bodies.
That distinction carries real weight in the clinic. The two conditions can look similar as they progress, but they tend to start differently: dementia with Lewy bodies frequently announces itself through problems with attention, alertness and movement, while Alzheimer's typically begins with memory loss. They also call for different treatment, so a misread diagnosis is not a technicality - it shapes what care a patient receives.
David Vaillancourt, a distinguished professor and Orchid Endowed Chair in UF's Department of Applied Physiology and Kinesiology, said combining AI with advanced imaging holds considerable promise for uncovering the patterns of brain degeneration behind dementia. The appeal is that scans are already part of many diagnostic workups; a tool that extracts more information from them could sharpen decisions clinicians are making anyway.
AIDD is not yet in routine practice, and the UF team frames it as a promising future tool rather than a finished product. With Alzheimer's disease and related dementias projected to more than double by 2060, methods that push toward earlier and more precise diagnosis are drawing increasing attention from researchers.