University of Florida Researchers Develop AI Tool to Distinguish Between Alzheimer's and Lewy Body Dementia
Three University of Florida researchers have developed an AI system that identifies two common forms of dementia with near-perfect accuracy, a step toward earlier diagnosis as cases are expected to more than double by 2060.
The tool, called Automated Imaging Differentiation for Dementia (AIDD), combines brain scans with machine learning to distinguish between Alzheimer's disease dementia and dementia with Lewy bodies. The study was published in Neurology.
Why the distinction matters
The two conditions require different treatments. Dementia with Lewy bodies typically begins with attention, alertness, and movement problems, while Alzheimer's typically presents with memory loss.
Yet up to 50% of patients with dementia with Lewy bodies are misdiagnosed with Alzheimer's. Incorrect diagnosis can lead to treatments that worsen cognitive and motor function.
Current diagnosis methods rely on a combination of evaluations, testing, and brain scans rather than a single definitive test.
How the system works
Researchers analyzed 519 brain scans collected between January 2007 and March 2022 from multiple research centers. They used 387 scans-129 from Alzheimer's patients, 129 from dementia with Lewy bodies patients, and 129 from controls-to train and test the AI model.
The scans used a specialized MRI technique that measures fluid movement in the brain, which can signal cell damage and inflammation. The AI identified subtle water-movement patterns to distinguish between the two diseases.
The researchers trained the model on 80% of the scans and tested it on the remaining 20%. To verify accuracy, they applied the tool to 13 patients whose diagnoses were later confirmed through autopsy. The system correctly identified all 13 cases.
David Vaillancourt, a distinguished professor in the UF Department of Applied Physiology & Kinesiology, said the results "suggest it could be a promising future tool for clinicians."
Next steps
The researchers performed validation tests using data from multiple scanners and imaging centers to ensure reliability across different equipment and facilities.
"Since the therapies for Alzheimer's disease and dementia with Lewy bodies differ, developing precision biomarkers will offer better outcomes for patients," Vaillancourt said.
Professionals working in research and healthcare who want to understand AI applications in medical diagnostics can explore AI Research Courses and AI for Healthcare learning paths.
Your membership also unlocks:
