Wenqing Wei: The role of cortico-thalamic feedback for visual information processing
| When |
Jul 04, 2023
from 05:15 PM to 05:45 PM |
|---|---|
| Where | Bernstein Center, Lecture Hall, ground floor, Hansastr. 9a |
| Contact Name | Fiona Siegfried |
| Contact Phone | 0761 203 9549 |
| Add event to calendar |
 vCal iCal
|
Abstract
In the mammalian visual system, light-induced signals from retinal ganglion cells (RGC) are propagated to the primary visual cortex (V1) through relay cells in the dorsal lateral geniculate nucleus (dLGN). In addition, neurons in dLGN receive numerous direct feedback inputs from cortex and indirect feedback inputs through GABAergic cells in thalamic reticular nucleus (TRN), which are also driven by cortical inputs. Feature selectivity of cortical cells is thought to emerge in the feedforward pathway, while the corticothalamic (CT) feedback is hypothesized to modulate the feedforward inputs. Experimental and theoretical work in the past was mainly focused on the influences of CT feedback on the activity of dLGN neurons. However, it is still mysterious what is the exact role of feedback for feature selectivity of V1 neurons.
We developed a computational model of the thalamo-cortico-thalamic network, where the orientation selectivity of V1 neurons emerges in the feedforward pathway. In this model, we are able to turn the CT feedback on and off and record neuronal responses at all levels. We could demonstrate three related phenomena in V1 neurons: (i) the relation between the preferred orientation of input and output is better preserved; (ii) orientation selectivity of cortical responses is stronger for all contrasts; (iii) decoding the neuronal responses of V1 neurons on the population level represents the stimulus orientation with higher confidence. Moreover, our analysis showed that CT feedback generally improves the signal-to-noise ratio (SNR) of the responses of single dLGN neurons, and it increases the SNR of the compound thalamic inputs to V1 neurons. We conclude that the CT feedback may play an important role for improving the feature selectivity of V1 neurons at low contrasts by modulating the activity of dLGN neurons.