Technology is advancing fast.
A recent study reveals that an AI program has surpassed doctors in predicting the progression from early dementia to Alzheimer’s disease, according to findings published in the journal eClinical Medicine.
The AI accurately predicted whether early dementia would remain stable or advance to Alzheimer’s in 4 out of 5 cases, a significant improvement over traditional diagnostic methods that include brain scans, cognitive tests, and analysis of spinal fluid for proteins like tau and amyloid.
Senior researcher Zoe Kourtzi, a professor at the University of Cambridge, highlighted the AI’s sensitivity in utilizing only cognitive test data and MRI scans to forecast progression rates. The study involved developing an AI model using cognitive tests and brain scans from 400 participants in a U.S. research project.
Testing of the AI model involved real-world data from 600 U.S. patients and 900 individuals treated at memory clinics in the U.K. and Singapore. The AI accurately predicted the development of Alzheimer’s within three years in 82% of cases and identified cases of stable mild dementia in 81% of instances.
Furthermore, the AI categorized patients into risk groups, revealing that approximately 50% had stable dementia, 35% progressed slowly to Alzheimer’s, and 15% progressed rapidly.
Researchers suggest that implementing this AI tool could lead to earlier treatment and closer monitoring for those at high risk of Alzheimer’s. It could also redirect attention for the 50% of individuals with stable symptoms, potentially identifying other underlying causes such as mood disorders rather than dementia.
Dr. Ben Underwood, an assistant professor at the University of Cambridge, noted the significant impact of reducing uncertainty around early dementia symptoms, offering potential relief to patients and families while aiding clinicians in providing timely and targeted care.
Looking ahead, the research team aims to expand the AI model to encompass other forms of dementia and incorporate additional data sources such as blood tests. They envision scaling up the AI tool to optimize patient allocation to clinical trials, which could accelerate the development of new treatments for modifying disease progression.
Discussion about this post