ABSTRACT: To date, little is known about the cellular microstructure of the human brain, in particular about synaptic circuits. These intertwined circuits underlie the unparalleled capabilities of the human mind, and when disrupted, likely underlie disorders of brain function. In our ongoing work, we aim to better understand the human brain as a spatial network based on the latest 3D reconstruction of a ~1 mm3 volume, in comparison to similar datasets in the mouse and fruit fly. Specifically, we aim to understand how i) the connection probability of neurons decays as a function of their distance. Moreover, we explore whether 3D geometry dictates brain wiring or merely puts constraints on it. There is increasing evidence that ii) the connectome is a higher dimensional non-Euclidean object, in line with the results of our recent Spatial Connectome Model in C. elegans. Addressing the geometric complexity of the brain, we show that iii) the neuron level brain anatomy shows critical structural features, with universal algebraic correlations across all studied organisms. We believe that an eloquent brain model needs to capture both the critical geometric correlations and the non-metric aspects of brain topology. Our aim is to develop such a spatial brain model, towards a suitable baseline to address species specific or disease related deviations, as well as powerful design strategies for artificial neural networks.
Noyce Conference Room
US Mountain Time
Our campus is closed to the public for this event.