Skip to main content
Fig. 6 | The Journal of Mathematical Neuroscience

Fig. 6

From: Shifting Spike Times or Adding and Deleting Spikes—How Different Types of Noise Shape Signal Transmission in Neural Populations

Fig. 6

The lower bound on the mutual information rate for the AD population (a, c) and the STS population (b, d) as a function of the independent noise level . The analytical results Eq. (42) and Eq. (47) (black lines) are compared with numerical simulations (circles) of the lower bound on the mutual information rate for various N for a weak sensory signal with (panels a, b) and a stronger sensory signal with (panels c, d). For the lower bound on the mutual information rate is identical for both models and is degraded by independent noise. For both models exhibit SSR, where a non-vanishing level of independent noise increases the lower bound on the mutual information rate. Hereby, the STS population profits significantly more from independent noise than the AD population. For in panel d the theory Eq. (47) fails due to the linearization of the logarithm in Eq. (40). Here we plot Eq. (4), where we inserted the analytically calculated coherence function Eq. (46) and integrated numerically. We note that our analytical theory appropriately describes the increase of the mutual information for weak independent noise levels. The firing rate was and the signal and noise cutoff frequencies were and

Back to article page