Bayesian Inference of Whole-Brain Directed Networks Using Neuroimaging Data

This research project will develop new statistical models and computationally efficient algorithms to analyze functional magnetic resonance imaging (fMRI) data. The advent of fMRI offers unprecedented opportunities for scientists to study the functional organization of the human brain because fMRI provides non-invasive measurements of the entire human brain's activity with a high spatial resolution. However, the massive size and large noise of fMRI data and the complex functional organizations of human brains pose challenges to scientists. This project will develop efficient fMRI data analysis methods to address these challenges. The new methods will be applied to investigate human brains' functional organizations, functional organization changes during brain development, and the relationship between the human brain and behavior in the population. The project results will contribute to the critical knowledge of the development of both healthy brain functions and risks for mental health challenges. Open-source software that implements the new statistical tools will be developed and made publicly available. The project will provide educational and training opportunities for undergraduate and graduate students. This project will build Bayesian models for whole-brain networks of many subjects based on their fMRI data. The new models will specifically characterize functionally specialized modules of brain regions and connections within and between the modules in whole-brain networks. In addition, the new models offer model flexibility, improved estimation efficiency, and robustness to model error and data noise in characterizing many subjects' whole-brain networks. Efficient computational algorithms will be developed to address the computational challenge in analyzing massive fMRI data and to map many subjects' whole-brain networks simultaneously. The project also will propose scalar-on-network regressions that feature heterogeneous relationships between behavior and brain networks in the population. The investigator will apply the new methods to neuroimaging and behavioral data of many subjects and examine the variation and distribution of the human brain's functional organization and its relationship to human behavior in the population. The project's findings will enhance understanding of the brain, human behavior, the risk for mental health challenges, and the role of the brain in overall health and well-being. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.