Retirement Living takes an unbiased approach to our reviews. We may earn money when you click a partner link. Learn More

› Retirement News

A new AI model may improve early Alzheimer’s disease detection

ousa-chea-unsplash

New treatments make early detection increasingly important

Avatar photo
By:
Avatar photo

Mark Huffman
Reporter

Mark Huffman has covered consumer issues since 2004, with a special focus on protecting and empowering older adults. A former news correspondent for the Associated Press Radio Network, Westwood One, and ConsumerAffairs, Mark brings decades of journalistic experience to RetirementLiving.com. He is passionate about giving consumers a voice and helping them make informed decisions. Mark has also created educational content, including Senior Scam Alert, a video resource designed to help seniors recognize and avoid fraud. He holds a B.A. from the University of Kentucky and lives with his family in Richmond, Virginia.

Updated:

key insights:

  • Researchers at West Virginia University have identified 21 metabolic biomarkers linked to Alzheimer’s disease and used them to train an artificial intelligence model aimed at predicting the disease in its early stages.
  • The deep-learning system was developed using data from the Alzheimer’s Disease Neuroimaging Initiative and achieved high accuracy in distinguishing Alzheimer’s patients from cognitively healthy individuals.
  • Scientists say earlier detection could improve treatment outcomes, accelerate drug development, and help physicians intervene before symptoms become severe.

Researchers at West Virginia University say they have developed an artificial intelligence-based approach that could improve the early detection of Alzheimer’s disease by identifying metabolic biomarkers associated with the condition. 

The study, published in the Journal of the Neurological Sciences, focused on identifying biological markers that may signal the onset of Alzheimer’s years before clinical symptoms appear. Researchers then used those biomarkers to train a deep-learning AI model designed to predict whether a person has — or may develop — the disease. 

Kesheng Wang, a professor in the WVU School of Nursing who led the study, said deep-learning methods are particularly effective for analyzing the complex biological data associated with neurodegenerative disorders.

“The deep learning method using artificial neural networks … has reached unprecedented prediction performance for complex tasks,” Wang said in the university release.

About the study

The research team analyzed data from 177 participants in the Alzheimer’s Disease Neuroimaging Initiative, including 78 people diagnosed with Alzheimer’s disease and 99 individuals with normal cognitive function. Participants ranged in age from 75 to 82. 

Using statistical software known as LASSO, researchers evaluated 150 metabolic biomarkers, ultimately narrowing the field to the 21 that were considered the most relevant to Alzheimer’s disease. The selected biomarkers were tied to glucose, amino acid, and lipid metabolism, areas increasingly linked to neurodegenerative changes in the brain. 

Some of the biomarkers also correlated with established indicators of Alzheimer’s disease, including amyloid plaques, cognitive decline, and shrinkage of the hippocampus — a brain region associated with memory and often damaged early in the disease process.

Researchers then tested multiple AI models before selecting the one with the highest predictive accuracy. Wang said the findings highlight the growing role artificial intelligence could play in diagnosing Alzheimer’s disease earlier and more precisely.

The power of AI

The study adds to a broader wave of research exploring AI-driven approaches to Alzheimer’s diagnosis and prognosis. Other recent studies have shown that machine-learning tools can outperform some traditional clinical assessments in predicting disease progression and identifying early cognitive decline. 

Scientists say early diagnosis is increasingly important as new treatments aimed at slowing Alzheimer’s progression become available. Experts also believe AI could eventually help reduce reliance on invasive and expensive diagnostic procedures such as PET scans and spinal taps. 

Wang cautioned that AI-based Alzheimer’s detection is still in its early stages and requires additional research before becoming part of routine clinical practice. His team is continuing work to integrate protein and metabolic data into future predictive models.