New AI Tool Helps Clinicians Identify Brain Activity Patterns Linked to Different Types of Dementia

Mayo Clinic researchers have developed a new artificial intelligence (AI) tool called StateViewer that helps clinicians identify brain activity patterns linked to nine types of dementia, including Alzheimer's disease, using a single, widely available scan. This advancement in early and accurate diagnosis was published online on June 27, 2025, in Neurology, the medical journal of the American Academy of Neurology. StateViewer enabled clinicians to identify the dementia type in 88% of cases, interpret brain scans faster, and with greater accuracy compared to standard workflows. The AI was trained and tested on over 3,600 scans, including images from patients with dementia and those without cognitive impairment.

This innovation addresses the challenge of identifying dementia early and precisely, even in the presence of multiple conditions. Timely diagnosis is crucial as new treatments emerge, helping match patients with appropriate care. StateViewer could provide advanced diagnostic support to clinics lacking neurology expertise.

Dementia affects over 55 million people worldwide, with nearly 10 million new cases annually. Alzheimer's disease is the most common form and is the fifth-leading global cause of death. Diagnosing dementia usually involves cognitive tests, blood draws, imaging, clinical interviews, and specialist referrals. Distinguishing between conditions like Alzheimer's, Lewy body dementia, and frontotemporal dementia remains difficult, even for experienced specialists.

StateViewer was developed under the direction of Dr. David Jones, a Mayo Clinic neurologist and director of the Mayo Clinic Neurology Artificial Intelligence Program. Dr. Jones emphasizes the importance of early understanding and precise treatment in dementia care. The tool was created in collaboration with Dr. Leland Barnard, a data scientist leading the AI engineering behind StateViewer.

The tool analyzes fluorodeoxyglucose positron emission tomography (FDG-PET) scans to compare brain glucose usage patterns with a database of scans from confirmed dementia cases. It identifies patterns matching specific types or combinations of dementia, such as Alzheimer's affecting memory and processing regions, Lewy body dementia involving attention and movement areas, and frontotemporal dementia altering language and behavior regions. StateViewer displays these patterns through color-coded brain maps, providing clinicians with visual insights supporting the diagnosis.

Mayo Clinic researchers aim to expand the tool's use and evaluate its performance in various clinical settings.

According to the source: News-Medical.