Reviewed by Frances BriggsJun 30 2025
Mayo Clinic researchers have developed an AI tool, StateViewer, that accurately identifies nine types of dementia from a single brain scan. This tool significantly speeds up diagnosis and improves accuracy even in non-specialist settings.
Dr. David Jones reviews brain scans on a computer at the Mayo Clinic. Image Credit: Mayo Clinic
According to a recent study in Neurology, the tool can even identify Alzheimer’s disease, using a single, widely accessible scan. It marks a major advance for early, accurate diagnosis.
Using StateViewer, researchers correctly identified the dementia type present in 88 % of instances. Additionally, the AI tool allowed doctors to interpret brain scans nearly twice as quickly and with up to three times the accuracy of traditional processes. Researchers trained and tested the AI on more than 3,600 brain scans, drawing from patients with dementia and individuals without cognitive issues.
Detecting dementia early and precisely is one of the biggest challenges in dementia care, especially when patient symptoms stem from multiple overlapping diseases. Early prognosis is key when matching patients with therapies, improving the efficacy of treatments. Using tools like StateViewer may be a way to extend expert-level diagnostic support to clinics with limited neurology resources.
The Rising Toll of Dementia
Dementia affects around 55 million individuals globally, with approximately ten million new cases diagnosed each year. Alzheimer’s disease, the most prevalent type, is currently the fifth largest cause of mortality worldwide.
Diagnosing dementia involves cognitive tests, imaging, blood work, interviews, and specialist evaluations. Even then, distinguishing among types like Alzheimer’s, Lewy body dementia, or frontotemporal dementia can be difficult, even for experienced neurologists.
StateViewer was developed by David Jones, M.D., a Mayo Clinic neurologist and director of the Mayo Clinic Neurology Artificial Intelligence Program.
Every patient who walks into my clinic carries a unique story shaped by the brain's complexity. That complexity drew me to neurology and continues to drive my commitment to clearer answers. StateViewer reflects that commitment—a step toward earlier understanding, more precise treatment, and, one day, changing the course of these diseases.
David Jones, Neurologist, Mayo Clinic
Dr. Jones collaborated with data scientist Leland Barnard, Ph.D., who led the engineering of the AI StateViewer.
As we were designing StateViewer, we never lost sight of the fact that behind every data point and brain scan was a person facing a difficult diagnosis and urgent questions. Seeing how this tool could assist physicians with real-time, precise insights and guidance highlights the potential of machine learning for clinical medicine.
Leland Barnard, Data Scientist, Mayo Clinic
Turning Brain Patterns into Clinical Insight
The tool examines an FDG-PET scan, which shows how the brain uses glucose for energy. It then compares the scan to a large database of scans from patients with proven dementia diagnoses, identifying patterns that correspond to specific types or combinations of dementia.
Different types of dementia affect the brain differently: Alzheimer’s disease primarily impacts memory and processing centers. Lewy body dementia targets regions involved in attention and movement, while frontotemporal dementia affects the parts of the brain that control language, behavior, and social interaction.
The StateViewer tool shows these patterns as color-coded brain maps that emphasize significant regions of brain activity. These maps provide a visual explanation of the scans and help doctors, including those without neurology expertise, make an accurate diagnosis. Mayo Clinic researchers intend to broaden the tool's use and continue to assess its efficacy in a range of clinical contexts.
Journal Reference:
Barnard, L., et al. (2025) An FDG-PET–Based Machine Learning Framework to Support Neurologic Decision-Making in Alzheimer's Disease and Related Disorders. Neurology. doi.org/10.1212/wnl.0000000000213831.