Deep Multi-Modal Detection of Early Alzheimer’s Disease (2025-2026)

This project team investigated how advanced neuroimaging, electrophysiological and proteomic methods can be used to identify novel biomarkers for Alzheimer’s disease and improve the early detection and prediction of disease progression. With Alzheimer’s disease affecting nearly 7 million Americans and cases expected to rise sharply in coming decades, the team sought to better understand the subtle brain and molecular changes that occur before significant cognitive decline. Drawing on large datasets and expertise from neuroscience, statistics and computer science, the team explored how MRI, EEG and proteomic data can reveal patterns not captured by traditional Alzheimer’s biomarkers alone.

Over the past year, team members applied computational topology, machine learning and network analysis techniques to diffusion MRI, functional MRI and EEG data to identify features associated with Alzheimer’s pathology, cognitive performance and disease stage. The team used persistent homology and graph-based methods to characterize changes in brain connectivity and organization, while also examining how high-dimensional proteomic profiles relate to functional disruptions in the brain. Their work demonstrated the value of integrating multiple data sources to capture the biological complexity and heterogeneity of Alzheimer’s disease.

The project advanced new approaches for modeling disease progression and identifying early indicators of neurological decline. By combining neuroimaging, EEG and proteomic analyses within predictive frameworks, the team generated insights into how molecular changes translate into alterations in brain function and network structure. Their findings help lay the foundation for more accurate, personalized diagnostic tools and could ultimately support earlier interventions and improved treatment strategies for individuals at risk of Alzheimer’s disease.

Timing

Summer 2025 – Spring 2026

Team Outputs

Deep Multi-Modal Detection of Early Alzheimer’s Disease (Team profile)

Deep Multi-Modal Detection of Early Alzheimer’s Disease (Poster presentation at the Fortin Foundation Bass Connections Showcase, April 15, 2026)

Manuscripts in progress

See related Data+ summer project, Deep Multimodal Detection of Early Alzheimer’s Disease (2025), and related team, Analyzing Alzheimer's Biomarkers Through Dynamic Brain Topology (2024-2025).