Background

Current Alzheimer’s disease (AD) biomarkers recommended by the National Institute on Aging and the Alzheimer’s Association have notable diagnostic limitations. Positron emission tomography (PET) — a procedure that measures metabolic activity in cells — can be effective at identifying biomarkers, but the procedure presents challenges due to its high costs and potential radiation risks. Similarly, invasive cerebrospinal fluid assays, while informative, lack standardization and carry infection risks. Thus, there remains a pressing need for noninvasive, cost-effective methods for routine AD biomarker detection, especially with a focus on heightened sensitivity to detect early dysfunction.

Alzheimer’s disease profoundly affects the brain’s connectivity, extending disruptions beyond local regions to impact the neural connectivity of the entire brain. Understanding these mechanisms requires a shift toward capturing higher-order relations. Paired with imaging like MRI and diffusion tensor imaging, persistent homology (PH) — a method for computing topological features at different spatial resolutions — offers a mathematically rigorous perspective on high-order network attributes. However, conventional PH encounters challenges in analyzing large, highly interconnected brain networks, especially during the prolonged progression of AD.

Project Description

This project team will investigate intricate alterations in the topology of structural and functional MRI-based brain networks from a cognitively normal state through mild cognitive impairment to Alzheimer’s disease (AD). The overall objective is a comprehensive, longitudinal exploration that employs a novel spatiotemporal topological approach to characterize the collective dynamics of brain networks over time, thereby enhancing researchers’ and clinicians’ understanding of pathological shifts across the lifespan.

Team members will leverage recent advancements in PH to analyze large-scale data without approximation, potentially transforming Alzheimer’s research. This new topological framework will facilitate an unprecedented exploration of large and noisy whole brain connectivities, allowing for the extraction of intricate topological patterns and for topological integration of heterogeneous data. 

The team will design a coherent statistical solution to provide inference, enabling potential multimodal approaches alongside existing measures beyond network-based biomarkers. Success could mean earlier detection of network topology changes preceding clinical symptoms and a high-resolution spatioregional atlas of brain topological changes, offering detailed insights into localized dynamics.

Anticipated Outputs

Development of a whole-brain connectivity dataset; novel algorithm solution for understanding AD; high-resolution spatioregional atlas; poster presentation; publication submission

Student Opportunities

Ideally, this project team will include 2 graduate students and 6 undergraduate students with interests in computer science, mathematics, neuroscience, biostatistics, bioinformatics and engineering. 

Team members will contribute to improving AD biomarkers and gain skills in team-based problem-solving, project management, organization, and verbal and written communication skills. Team members will utilize preprocessing tools to generate a robust, reproducible brain connectivities dataset from neuroimaging and learn to apply data analytics tools to analyze human brain connectivities datasets, employing various computational, statistical, and mathematical techniques. Team members will also have the chance to contribute to open-source projects and learn from industry experts to proficiently engage in pair programming, use git for code version control and conduct code reviews.

Graduate students will contribute to guiding the project’s research direction and actively acquire essential leadership skills.

See the related Data+ project for Summer 2024; there is a separate application process for students who are interested in this optional component.

Timing

Summer 2024 – Spring 2025

• Summer 2024 (optional): Data+ students: Complete project orientation; complete literature review;  contribute to dataset preprocessing; compile report; set up derived dataset repository
• Fall 2024: Conduct exploratory data analysis; define goals and roles; implement and execute algorithms on the dataset; obtain preliminary results; engage in visualization, interpretation and results presentation
• Spring 2025: Develop a statistical solution for early detection statistics in AD; generate a high-resolution spatioregional atlas of brain topological changes; compile research outcomes into comprehensive documentation; present research findings; potentially prepare and submit manuscripts for publication

Crediting

Academic credit available for fall and spring semesters; summer funding available

See related Data+ summer project, Analyzing Alzheimer’s Biomarkers through Dynamic Brain Topology (2024).

Image: Brain showing hallmarks of Alzheimer’s disease (plaques in blue), by Alvin Gogineni, Genentech, NIH Image Gallery, licensed under CC BY-NC 2.0