Figure 4.

Spatial and temporal cross-contamination between LFP-generators with a strongly unbalanced contribution. (A) The example corresponds to LFPs modeled by combining spatially overlapping of GABA_B dendritic inhibition (G1, blue) and perisomatic GABA_A inhibition (G2, green). Irregular series of synaptic inputs were injected in the two generators with a variable degree of temporal coincidence between them (case I: 0% and case II: 25%). The spatial distribution of the strong generator (G1) was stable, but that of the weak generator (G2) underwent a slight spatial shift when the two inputs had a partial temporal coincidence (25% of the time), which did not however modify the estimation of the synaptic loci (CSD). Note the small bump in the CSD curve of the weak generator at the site where the stronger generator peaked. (B) Cross-contamination can also be observed in the respective time courses of the generators (arrows mark the timings of synaptic inputs and asterisks show coincident inputs). The cross-contamination grew larger as the inputs coincided for a longer period (case I vs. case II). G1 got extra power (variance) from G2. (C) The strong generator (G1) practically does not suffer from coincident inputs. However, the time course of the weak generator (G2) may be severely distorted by coincident inputs and a significant part of its contribution lost. The spatial distribution of the weak generator also becomes worse for higher coincidence level, but at 100% coincidence G2 recovers its original spatial distribution (see main text).