New AI-powered diagnostic tools promise earlier, faster detection of Alzheimer's disease

New diagnostic tools, including AI models and blood tests, are transforming early Alzheimer's detection. A blood test called pTau217 detects brain changes decades before symptoms appear with 90% accuracy. Meanwhile, AI systems like CortexNet_V1 analyze MRI scans with 98% accuracy, and machine-learning models estimate brain age from sleep EEG signals, where a 10-year increase in brain age correlates with a 40% higher dementia risk. These advancements aim to address global diagnostic delays and high healthcare costs.

AI and biomarker-based Alzheimer's diagnosis

• ▪The pTau217 blood test works by detecting phosphorylated tau 217, a biomarker for amyloid plaque buildup in the brain.
• ▪A machine-learning model developed by researchers at UC San Francisco and Beth Israel Deaconess Medical Center estimates brain age from sleep EEG signals.
• ▪Researchers at Universidad Carlos III de Madrid and Hospital Universitario Severo Ochoa developed an AI program that analyzes sleep to detect Alzheimer's.
• ▪A new blood test called pTau217 can detect changes in the brain decades before Alzheimer's symptoms develop, with results in 18 minutes.

REM sleep deficiency and brain aging effects

• ▪A lack of REM sleep prevents the clearance of metabolic waste from the brain, causing cell death and accelerating neurological diseases like dementia.
• ▪For every 10-year increase between a person's estimated sleep-based brain age and chronological age, the likelihood of developing dementia rises by nearly 40%.
• ▪Specific EEG patterns during sleep, such as delta waves and sleep spindles, are known to support memory and cognitive health.

Electroencephalogram and imaging study results

• ▪CortexNet_V1 requires only 0.059 GFLOPs of computation per brain scan, making it suitable for modest hardware with limited resources.
• ▪A study of 7,000 participants over 3.5 to 17 years found that sudden spikes in EEG signals, called kurtosis, were associated with lower dementia risk.
• ▪An AI model named CortexNet_V1 classified brain MRI scans with over 98% accuracy across two, three, and four-class diagnostic tasks.
• ▪A sleep EEG study of 42 Alzheimer's patients and 58 healthy controls successfully differentiated the two groups using AI-analyzed brain signals.

Home-based and clinical diagnostic testing

• ▪Researchers aim to validate sleep EEG tests for home use using domestic encephalograms, which could enable low-cost, wide-scale screening.
• ▪Traditional sleep metrics, such as total sleep time or sleep efficiency, show no meaningful association with dementia risk.
• ▪The Asia-Pacific Centre of Excellence for Alzheimer's Disease Diagnosis in Sydney aims to enable earlier detection through blood-based and spinal fluid biomarker testing.

Global Alzheimer's detection challenges

• ▪In Spain, Alzheimer's affects 800,000 people, with over 50% of mild cases undiagnosed and an average diagnostic delay exceeding two years.
• ▪Dementia, including Alzheimer's, caused over 17,500 deaths in Australia in 2024, making it the nation's leading cause of death.
• ▪The annual economic cost of dementia in Australia exceeds $4.7 billion, placing immense pressure on clinicians and the healthcare system.