Advanced Artificial Intelligence for Structural Brain Connectomics of Neurodegenerative Diseases

Understanding how the brain changes at different stages of neurodegenerative diseases can give important information for improving their early detection and diagnosis.

A promising tool for that purpose is brain connectomics, whose aim is to study the connectivity among different areas of the brain.

Structural connectivity estimates the degree of connectivity among regions in the cortex by analyzing the distribution of neural fibers in the white matter (WM) computed from diffusion MRI (dMRI) data. The current tools for connectomic analyses are not accurate enough for clinical use.

In this project, we aim at closing this gap by proposing new advanced artificial intelligence (AI)-based tools for performing brain connectivity analysis. We will test their effectiveness in Alzheimer’s disease (AD) and multiple sclerosis (MS). Applying AI to dMRI data is challenging and requires advanced AI methods.

The specific objectives of the project are to:Develop AI-based tractography and tractogram filtering methods that are reliable for performing structural connectivity analysisAdapt generative AI architectures for creating synthetic dMRI data to improve tractography and brain connectivity analyses.Combine mechanical properties of brain tissue and microstructure parameters of WM with tractography for improving structural connectivity analyses.