Breakthrough AI Achieves 88% Accuracy in Identifying Nine Types of Dementia

Mayo Clinic researchers have introduced a groundbreaking AI tool called StateViewer that detects nine types of dementia, including Alzheimer’s, using a single brain scan. The tool achieved an 88% accuracy rate in identifying dementia types, while also doubling interpretation speed and tripling diagnostic accuracy compared to standard methods. Trained on over 3,600 brain scans, StateViewer analyzes fluorodeoxyglucose positron emission tomography (FDG-PET) scans, which reveal brain glucose usage, and compares them against a comprehensive database. This allows the AI to recognize distinct brain activity patterns correlated with specific dementia types, essential for timely and precise diagnoses. Given the rising global dementia burden, with over 55 million affected individuals, StateViewer aims to provide diagnostic support where specialized care is lacking. The researchers plan to evaluate its performance across various clinical settings to enhance early interventions and improve patient outcomes. The study was published in the journal Neurology.

Source link