Alzheimer’s disease and related dementias are expected to more than double by 2060. June is Alzheimer’s Disease and Brain Awareness Month, and three University of Florida researchers are working to improve clinicians’ ability to differentiate between these diseases. This is an important step towards earlier diagnosis and better outcomes.
In a recent study published in the journal Neurology, researchers developed a new tool called Automated Imaging Differentiation for Dementia (AIDD). The tool combines brain scans and AI to differentiate between two common dementias: Alzheimer’s disease dementia and Lewy body dementia. The results showed that AIDD identified the two diseases with near-perfect accuracy, suggesting that AIDD could be a promising tool for clinicians in the future.
“The use of AI and advanced imaging techniques holds considerable promise for revealing patterns of brain degeneration in dementia,” said David Vaillancourt, Ph.D., Distinguished Professor and Orchid Endowed Chair in the Department of Applied Physiology and Kinesiology at the University of California Health and Human Performance College.
Both conditions are types of dementia, but the symptoms manifest differently. For example, Lewy body dementia often begins with problems with attention, alertness, and movement, whereas patients with Alzheimer’s disease exhibit memory impairment. Unlike Alzheimer’s disease, Lewy body dementia requires different treatments.
Unfortunately, the two diseases are frequently confused, and up to 50% of patients with Lewy body dementia are misdiagnosed as having Alzheimer’s disease. Today’s diagnostic methods rely on a combination of assessments, tests, and brain scans rather than a single definitive test. In some cases, misdiagnosis can lead to treatments that worsen cognitive and motor function.
To build the tool, researchers analyzed 519 brain scans of Alzheimer’s disease, Lewy body dementia, and disease-free patients (control group) collected at multiple research data centers from January 2007 to March 2022. From this group, a subset of 387 scans (129 Alzheimer’s disease, 129 Lewy body dementia, and 129 controls) were used to train and test the AI model. 80% of the scans were used to train the machine, and the remaining 20% were used for testing.
“To ensure the highest level of reliability, we conducted extensive validation experiments using data collected from multiple scanners and imaging centers,” said Dr. Angelos Burnputis, professor at the Digital World Institute at UF College of the Arts, who worked on the study with Vaillancourt and Dr. Robin Chen, a postdoctoral fellow in the J. Clayton Pruitt Family Department of Biomedical Engineering.
The scan used a special MRI technique that measures excess fluid in the brain, which often indicates brain cell damage or inflammation. By using AI to analyze these subtle water movement patterns in the brain, it has become possible to more accurately identify each disease. Across multiple brain scan comparisons, this tool showed strong performance. To further test the system, the researchers applied the tool to another group of 13 patients whose diagnosis was confirmed by post-mortem autopsy. The tool correctly identified all 13 cases.
“The treatments for Alzheimer’s disease and Lewy body dementia are different, so developing highly accurate biomarkers will lead to better outcomes for patients,” Vaillancourt said.
sauce:
Reference magazines:
DOI: 10.1212/WN9.00000000000000093
