Please use this identifier to cite or link to this item:
|Title:||Sensitivity enhancement mechanisms at the periphery of the olfactory pathway|
|Authors:||Mackenzie, Josephine Ann|
|Presented at:||University of Leicester|
|Abstract:||Massive convergence of input from olfactory receptor neurons (ORNs) with identical tunings leads to spatial integration of sensory signals, thereby boosting sensitivity to sensory cues. The consequent reduction in detection thresholds is assumed to derive from the pooling of elevated firing rates across the ORN population. By comparing detection thresholds at the first two stages of the olfactory pathway in an olfactory specialist, the moth, allowed for the quantification of the sensitivity boost achieved during early sensory processing. This boost was found to be at least 3 orders of magnitude, which was shown to exceed that achieved by a theoretical model of spike train integration. The sensitivity enhancement achieved by this system therefore goes beyond straightforward spatial summation of receptor firing rates, suggesting subtler coding and readout mechanisms. To discount the possibility of ORNs employing a temporal encoding scheme, an inves tigation into spike patterns at the periphery was performed. While no temporal patterns were evident, a temporal encoding scheme remains a possibility. Despite the inconclusive result found here, the analysis demonstrates the need for an investigation of the stationar- ity of spike trains, where a statistical basis underlies the analysis method, before drawing conclusions. Regardless of the encoding scheme employed at the periphery, due to the multitude of possible synaptic connections within a glomerulus, it seems unlikely that this site of conver gence of receptor input would be passive. A simple, but biologically plausible computational model was developed, where specific zones of the dendritic tree of an output neuron form individual subunits capable of performing a nonlinear threshold function on ORN inputs. This nonlinear model consistently outperformed a comparable linear model when assessing the stimulus detection performance of the output neuron.|
|Rights:||Copyright © the author. All rights reserved.|
|Appears in Collections:||Theses, Dept. of Engineering|
Items in LRA are protected by copyright, with all rights reserved, unless otherwise indicated.