| Literature DB >> 27543118 |
Peter Bratby1, James Sneyd2, John Montgomery3.
Abstract
Here, we introduce a novel mechanism for temporal recoding by the cerebellar granular layer based on three key properties: the granule cell-Golgi cell inhibitory feedback loop, bursting behaviour of granule cells and the large ratio of granule cells to Golgi cells. We propose that mutual inhibition of granule cells, mediated by Golgi cell feedback inhibition, prevents simultaneous activation. Granule cells are differentiated by firing threshold, resulting in sequential bursts of spikes. We demonstrate the plausibility of the mechanism through a computational simulation of a firing rate model, and further examine its robustness by developing a spiking model incorporating realistic postsynaptic potentials.Keywords: Adaptive filter; Cerebellum; Computational simulation; Granular layer; Neural network; Sequential winner-take-all
Mesh:
Year: 2017 PMID: 27543118 DOI: 10.1007/s12311-016-0820-y
Source DB: PubMed Journal: Cerebellum ISSN: 1473-4222 Impact factor: 3.847