Modeling Longitudinal Brain Atrophy to Optimize MRI Timing for Early Alzheimer’s Detection

Overview

Alzheimer’s disease (AD) affects 55 million people worldwide and is a major contributor to the global burden on healthcare systems, families, and caregivers. It has been found that detectable structural changes in the brain begin years before symptoms appear, and earlier detection could improve treatment. I investigated whether longitudinal brain magnetic resonance imaging (MRI) scans improve diagnosis beyond baseline and how to optimally schedule MRI scans. To do so, I modeled MRI data from 740 NACC participants. Regression models (linear, mixed-effects, logistic, and regularized) were used to characterize atrophy, evaluate the predictive values of longitudinal MRI, and determine scan frequency needed for accurate measurement of brain changes. I found that longitudinal MRI improved AUC from 80% to 86%, and an MRI every 2 years was reasonable. Furthermore, I developed a mathematical framework that enables personalized MRI scheduling by predicting how an individual’s disease risk evolves over time.

Competition history

• CSEF 2026 Mathematical Sciences (Senior Division) · Entry S-14-10

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Mathematics projects
• CSEF Mathematics projects
• Every CSEF 2026 project