|
Neural substrates of odor-guided decisions in the rat| old_uid | 7767 |
|---|
| title | Neural substrates of odor-guided decisions in the rat |
|---|
| start_date | 2009/12/04 |
|---|
| schedule | 11h-12h30 |
|---|
| online | no |
|---|
| location_info | salle 3304 |
|---|
| summary | The mammalian olfactory system has properties that make it unusal amongst other sensory systems. First, stimulus dynamics are under almost complete control by active sampling through respiration or sniffing. Second, stimulus encoding is massively distributed across hundreds or thousands of channels. In this talk I will show that these features have important implications for the neural representation of odors in the olfactory cortex. In these studies, we recorded from ensembles of neurons in the anterior piriform cortex of rats performing an odor mixture categorization task in which rats make a left/right spatial choice based on the dominant component of a binary odor mixture. In this task, rats reach their maximum behavioral accuracy in just one sniff or around 200 ms odor sampling time. We found that odor stimulation evoked rapid and transient bursts of spikes tightly locked to inhalation onset. Thus, olfactory information was chunked into packets locked to the theta cycle. Population decoding analysis showed that the information conveyed by the spikes in the first sniff cycle was more than sufficient to account for the behavioral accuracy. The addition of spikes from subsequent sniffs failed to increase decoding performance. TIhe efficiency of anterior piriform odor representations was due in part to the fact that olfactory cortex neurons, even near neighbors, showed statistically independent odor responses, both with respect to odor tuning and so-called “noise correlations” (correlated fluctuations in response to a given odor). These observations contrast with what has been typically reported in sensory neocortex, where noise correlations are substantial and tuning dependent. Interestingly, noise correlations and trial-to-trial variability were not static, but dynamically shaped by active sampling, being minimal during theta-frequency sniffing, even before stimulus onset. Thus, olfactory system dynamics and architecture cooperate to produce efficient odor processing and help to account for extremely rapid odor-guided decisions. |
|---|
| oncancel | chgt de lieu |
|---|
| responsibles | Gervais |
|---|
| |
|