Files
Abstract
A connectome is a comprehensive map of neural connections in the brain. It plays a critical role in implementing brain functions such as memory, decision making, emotion, and language, and is believed to correlate with mental disorders such as autism and schizophrenia. To study brain connectome, researchers need to investigate from different views such as structural connectivity, functional connectivity, molecular regulators, development progress, and plasticity property. Since brain connectome is a multi-scale concept and a finest neuron wiring map of the human brain is not feasible due to the technique limits, current studies usually focus on single view in a specific resolution scale using certain imaging modalities and animal models. Despite many novel findings achieved in these studies, a comprehensive map of brain network is still missing. To achieve such map, in my study, I propose to jointly analysis brain in different scales and fuse brain connectome derived from different image modalities, different animal models, and different resolution scales. In micro-scale, I developed a set of software to automatically reconstruct neuron morphologies. In meso-scale, I computed whole brain connectome derived from neuron tracing experiments and employed it to evaluate the result of diffusion tensor image which is in macro-scale. And in macro-scale, I developed a set of brain landmarks to study group-wise inter-regional connectivity. Then I jointly analyzed brain wiring pattern and folding pattern across primate species and adopted machine learning algorithms to fuse brain functional connectome and structural connectome. Those studies involve different imaging modalities such as confocal microscopy imaging, neuron tracers, and structural/functional/diffusion magnetic resonance imaging. A wide range of study subjects has also been included to enable analysis in different resolution scales. In these studies, I have identified interesting brain connectome patterns that preserved or altered across species, modalities, populations, and between healthy and diseased human brains. Moreover, the newly developed computational frameworks will be further applied in other studies and shed light on the understanding of brain architectures and development mechanisms.