Back
Science

AI System Developed by WPI Predicts Alzheimer's Disease with High Accuracy

Generated on: Last updated: 1 sources
View source

WPI AI Method Predicts Alzheimer's Disease with 93% Accuracy

Worcester Polytechnic Institute (WPI) researchers have developed a method using artificial intelligence (AI) to analyze anatomical brain changes and predict Alzheimer's disease with approximately 93% accuracy. Their findings, published in the journal Neuroscience, indicate that these anatomical changes, specifically brain volume loss, vary by age and sex.

Machine learning technologies can identify subtle changes from scans to accurately predict Alzheimer's and related cognitive states. This development could lead to earlier and more effective diagnosis and treatment.

— Benjamin Nephew, Assistant Research Professor at WPI

Alzheimer's Disease Context

Alzheimer's disease is a neurodegenerative disorder that impairs mental functions, leading to cell death and loss of brain tissue.

  • An estimated 6.9 million Americans aged 65 and older are living with Alzheimer's disease.

Methodology and Findings

The research team, including PhD student Senbao Lu and Bhaavin Jogeshwar, MS '24, analyzed 815 MRI scans from the Alzheimer's Disease Neuroimaging Initiative. These scans included individuals with normal mental function, mild cognitive impairment, and Alzheimer's disease.

Researchers used machine learning to measure brain volumes in 95 regions. An algorithm then used these measurement differences to make predictions. The method achieved 92.87% accuracy in distinguishing Alzheimer's disease from normal brains and those with mild cognitive impairment.

Key Predictors and Sex/Age Differences

Volume loss in specific brain regions served as top predictors:

  • Hippocampus: Responsible for memory and learning.
  • Amygdala: Controls emotions.
  • Entorhinal cortex: A hub for memory, navigation, and perception, often an early impact area for Alzheimer's.

Specific findings regarding age and sex revealed important distinctions. Both males and females aged 69 to 76 (the youngest group studied) showed volume loss in the right hippocampus, suggesting its importance for early diagnosis across sexes in this age bracket.

In females, volume loss was observed in the left middle temporal cortex, which is involved in language, memory, and visual perception. Conversely, in males, volume loss was notable in the right entorhinal cortex. These sex-specific differences may be linked to interactions between Alzheimer's progression and changes in sex hormones.

Future Research

Nephew and WPI students plan to explore deep learning models and examine other factors, such as diabetes, that may impact the brain and Alzheimer's disease. The research emphasizes an interdisciplinary approach, drawing students from various scientific and computational fields.