Surface-Based Fiber Tracking and Modeling Techniques for Mapping the Superficial White Matter Connectome with Diffusion MRI

点击翻译按钮获取中文摘要

Abstract The superficial white matter (SWM) lies directly beneath the cortex and contains the short association fibers, or U-fibers, connecting neighboring gyri. The SWM contains around twice as many fiber connections as the deep white matter (DWM) and plays a crucial role in brain development, aging, and various brain disorders. Existing connectome imaging research based on diffusion MRI (dMRI), however, mostly focuses on the connections of long fiber bundles in the DWM even though tremendous advances have been made in human connectome imaging with much improved spatial and angular resolution. In this proposed renewal of our R01 project (NIBIB R01EB022744), we will conduct systematic development of novel computational tools to fill major technical gaps in current SWM research. Our project will provide fundamentally novel solutions to many of the current challenges in SWM connectome research by developing surface-based tools for fiber tracking, atlas construction, and personalized analysis. We will also develop novel personalized dMRI harmonization methods with a particular focus on accounting for the variable cortical anatomy. These developments will for the first time provide dedicated tools for modeling SWM connectome with greatly improved robustness and accuracy. There are three specific aims in our project: 1. Development of novel surface-based fiber tracking and filtering algorithms for the modeling of superficial white matter connectivity. 2. Development of surface-based U-fiber atlases and personalized SWM connectivity analysis. 3. Development of personalized diffusion MRI harmonization tools with improved consistency in cortical anatomy. Rigorous validations of our novel surface-based U-fiber tracking and modeling methods will be performed on high-resolution MRI of post-mortem brains, in vivo intracranial neural recordings from surgically implanted electrodes in patients with epilepsy, and their application in multiple large- scale connectome imaging datasets (n>5000). All software tools and atlases developed in this project will be publicly shared, which will allow brain imaging researchers to augment their current connectome models with U- fibers in SWM and more completely map human brain connectomes for the detection of their alterations in various brain disorders.