We define the LFP amplitude σ as the standard deviation of the compound LFP signal ϕ(t)ϕ(t) across time. With increasing population radius R  , more and more cells contribute to the compound signal ϕ(t)ϕ(t). The amplitude σ(R)σ(R) is thus expected to increase with R  . On the other hand, the contribution to the potential from a single neuron decreases with its distance r   from the electrode ( Lindén et al., 2010). Intuitively, one might therefore expect that σ(R)σ(R) approaches a constant value σ∗σ∗ as the population size R   increases. If so, it is natural to define the reach  R∗R∗ of the electrode as the population size at which the signal amplitude selleck kinase inhibitor captures a certain

fraction α of this limit value σ∗σ∗. In the present article, we set α to 95 %. It is, however, a priori not clear that σ(R)σ(R) converges, i.e., that a finite limit value σ∗σ∗ and thus a finite reach R∗R∗ indeed exist. Below we will therefore first consider a simplified model to demonstrate MK 2206 which factors shape the dependence of the LFP amplitude σ(R)σ(R) on the population size R and to illustrate under which conditions the spatial reach is finite. Next, we investigate these factors in detail by means of comprehensive numerical simulations of the LFP generated by cortical populations consisting of thousands of neurons with realistic dendritic morphologies. This idea suggests

that the amplitude σ generated by a population of neuronal sources surrounding the electrode is essentially controlled by three factors: • The attenuation f(r) of the contribution to the LFP signal from a single neuron with increasing distance r ( Thymidine kinase Figure 1B), The distance-dependent attenuation f(r) of the extracellular

potential around a neuron is determined by the distribution of the underlying transmembrane current density ( Pettersen and Einevoll, 2008 and Lindén et al., 2010). The potential generated by a pure current dipole source, for example, typically decreases in amplitude as 1/r2 with distance r (blue curve in Figure 1B). A hypothetical point source, in contrast, would generate a potential which decays in amplitude as 1/r (red curve in Figure 1B). Assuming a constant area density of neuronal sources, the decrease in amplitude is to some extent compensated by the increase in the number of neurons with increasing distance from the electrode. In this article, we consider populations of neurons symmetrically distributed around the electrode on a 2D plane with a constant density ρ. The number N(r)Δr=2πrρΔrN(r)Δr=2πrρΔr of neurons on a narrow ring of radius r   and width ΔrΔr will then grow linearly with the population radius ( Figure 1C). If the single-cell contributions to the LFP are uncorrelated, the variances of the signals generated by the individual cells positioned on a narrow ring of radius r   will sum up, so that the amplitude σ of the compound signal will be proportional to N(r)f(r).