A Variational Bayesian Approach to Network Modularity Applied to the Structural Connectome of the Human Brain

Keywords

able accuracy adjusted adult adults advantage alternative although applications applied apply assigned assignment assignments available belonging better biases brain circumvent code comparing computer connected connectivity consensus consistency consistently construction correct correction correlated cortex cortical date decomposition defined degree derived detection diffusion eddy edges evaluated excite female fields finally find five framework generalization graph healthy helpful human identified identify identifying implementation improved included increasingly index individual infer inference institute interestingly international introducing iteratively known literature loosely male maximizing measured model modular modularity module modules motion network networks node nodes noise numerous offers optimal optimization optimize original output outputs pairwise particular partition partitioning phys pipeline poorly poses precision preprocessing previously probabilistic problem processed properties published quantify quantity rand randomized randomizing recent reconstruction related relatively removing reporting researchers resolution robustness scaled scanner science selection smith social source stat statistical structural structure subjects suffer swapping theoretic toolbox uncertain underlying unknown variational version volunteers widely years