Dynamics and orientation selectivity in a cortical model of rodent V1 with excess bidirectional connections

Abstract

Recent experiments have revealed fine structure in cortical microcircuitry. In particular, bidirectional connections are more prevalent than expected by chance. Whether this fine structure affects cortical dynamics and function has not yet been studied. Here we investigate the effects of excess bidirectionality in a strongly recurrent network model of rodent V1. We show that reciprocal connections have only a very weak effect on orientation selectivity. We find that excess reciprocity between inhibitory neurons slows down the dynamics and strongly increases the Fano factor, while for reciprocal connections between excitatory and inhibitory neurons it has the opposite effect. In contrast, excess bidirectionality within the excitatory population has a minor effect on the neuronal dynamics. These results can be explained by an effective delayed neuronal self-coupling which stems from the fine structure. Our work suggests that excess bidirectionality between inhibitory neurons decreases the efficiency of feature encoding in cortex by reducing the signal to noise ratio. On the other hand it implies that the experimentally observed strong reciprocity between excitatory and inhibitory neurons improves the feature encoding.

Figure 1

Activity in the network without excess bidirectionality. Other parameters are given in Methods section. (a) Sample voltage trace of cells (firing rates: top, E: 5.36 Hz, bottom, I: 8.7 Hz). (b) Population averaged tuning curves for both populations (E: black; I: Red). The tuning curves are normalized to the peak rate. (c) Distribution of orientation selectivity index (OSI) for excitatory (black) and inhibitory (red) neurons. Unlike in van Vreeswijk and Hansel, the average number of feedforward inputs from layer 4 in excitatory and inhibitory neurons are different: $K_{ff}^E=100$, $K_{ff}^I=800$. Inhibitory neurons receive more but weaker feedforward inputs leading to less selectivity in their response.

Figure 2

Bidirectionality in E-to-E has negligible effect on spiking irregularity. (a) Population averaged autocorrelation functions for excitatory and inhibitory populations for different values of p. (b) Fano factor distributions different values of p. (c) Distribution of CV. (d) Distribution of CV₂ (see Methods). In all subfigures the top panel is for the excitatory population and the bottom one is for the inhibitory population.

Figure 3

Bidirectionality in I-to-I slows down fluctuations and increases response variability. (a) Example voltage trace of an inhibitory cell for p = 0.8 (firing rate: 8.8 Hz). Dependence on p of the Fano factor (b), population averaged autocorrelation functions (c), CV and CV₂ in (d). Top panels: Excitatory population. Bottom panesl: Inhibitory population. (e) Decorrelation time (see Methods) of the network activity as a function of p.

Figure 4

Bidirectionality in E-to-I connections leads to rapid decorelation and reduced response variability. (a) Population averaged autocorrelation functions for different values of p. (b) Average Fano factor decreases with p. The distributions of CV (c) and CV₂ (d) have negiligible dependance on p. Top panels: Excitatory neurons. Bottom panels: Inhibitory neurons.

Figure 5

Bidirectionality has a weak effect on feature selectivity. (a and c) Excess bidirectionality within the excitatory population (a) and between excitatory and inhibitory populations (c) have no effect on the selectivity of excitatory (top) and inhibitory (bottom) neurons. Excess bidirectionality between the inhibitory neurons slightly decreases the selectivity of excitatory neurons (b, top) while it slightly increases the selectivity in inhibitory population (b, bottom).