Reproducible Structural Connectomics on the Cloud

摘要: This dissertation consists of 3 projects designed to improve the creation, distribution, and analysis of modeling the white matter connectivity as structural connectomes. The primary analysis consists of a novel application of a Canonical Correlation Analysis (CCA) in a healthy population sampled across the lifespan. A novel approach of using the estimated CCA to predict the samples age is described. The model is validated and generalized with an intensive cross-validation of the model is performed. Finally, a novel set of comparisons is described to map the findings from the CCA model to the domains of the brain and behavior as well as the established rich club. The next analysis is an attempt to better capture the anatomy underlying the estimates of a structural connectome during its estimation. A new method of quantifying the evidence of an edge based on the predicted diffusion signal of the tractography is …