The Locust Olfactory System has three main stages of Odor Representation: detection by the Olfactory Receptor Neurons (ORNs), spatiotemporal coding by the Antennal Lobe (AL), and reliable odor identity representation in the Mushroom Body (MB). Previous theoretical studies have proposed a model of odor representation via activation of sub-networks of \textit{all} AL neurons which leads to spatiotemporal patterning. This pattern of activity then converges on the Mushroom Body giving us a identity readout. We propose an alternate idea in which the network structure of the AL gives rise to a dynamical system with a large number of attractors and the activity of a sub-network of PNs acts as a director of the activity on odor onset. We also suggest that the fixed point representation of odors in the trajectories of the activity space may reflect this in the activity. We try to build a scaled-down model of the complete locust olfactory system and study the dynamics of the odor representations with realistic ORN activity, a physiological model of AL and dynamic MB activity as a readout.
assisilab/DAMN
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|