Efficient Computational Strategies for Simulating Neural Activity on Branched Structures
Mathematical Biosciences Institute (MBI), The Ohio State University
(October 19, 2006 10:30 AM - 11:30 AM)
Ever since Hodgkin and Huxley developed their revolutionary mathematical model of neuronal activity, the field of computational neuroscience has been growing rapidly. Today, computer simulations are used extensively to model systems as small as individual ion channels and as large as networks consisting of 10,000 morphologically accurate model cells.
A drawback of simulations performed on detailed cell models is that the most commonly used numerical schemes are implicit, meaning that the voltage update is global in scope. This approach does not allow focusing of computational effort on those regions of the cell that are most active. The result is an unnecessary slow-down in neural simulations when activity is localized to a small region of the cell.
I will present a predictor-corrector numerical method we developed that decouples all of the branches from each other. This allows the algorithm to detect which regions of the cell are active and to focus computational effort on those regions while saving computations in other regions of the cell that are at rest. As a result, the computational cost of a simulation scales with activity, not with the physical size of the system. I will show several simulations that illustrate this idea, including reproductions of recent experimental results from our lab.