Skip to main content
. 2020 Jul 1;40(27):5228–5246. doi: 10.1523/JNEUROSCI.2731-19.2020

Figure 3.

Figure 3.

Subcortical neurons discriminate better the original vocalizations than cortical neurons. A, From bottom to top, neuronal responses were recorded in CN, CNIC, MGv, A1, and VRB simultaneously under 16 electrodes but only two are represented here, with alternated black and red colors. Each dot represents the emission of an action potential, and each line corresponds to each presentation of one of four original whistles. The gray part of rasters corresponds to evoked activity. For each example, the values of the BF (in kHz) and of the BW (in octaves) obtained when testing the responses to pure tones are indicated on the left side. The waveforms of the four original whistles are displayed under the rasters. B, PSTHs of each neuronal response presented in A. For each neuronal recording, the four PSTHs of the four original whistles have been overlayed. C–F, The mean values of the evoked firing rate (in spikes/s; C), the trial-to-trial temporal reliability quantified by the CorrCoef value (D), the neuronal discrimination assessed by MIIndividual values (bits; E), and MIPopulation (bits; F), with populations of nine simultaneous multiunit recordings obtained with the four original vocalizations in CN (in black), CNIC (in green), MGv (in orange), A1 (in blue), and VRB (in purple). The evoked firing rate corresponds to the total number of action potentials occurring during the presentation of the stimulus minus spontaneous activity (200 ms before each acoustic stimulus). In each structure, error bars represent the SEM of the mean values, and black lines represent significant differences between the mean values (unpaired t test, p < 0.05). The evoked firing rate decreases from the CN to VRB, but both the trial-to-trial temporal reliability (CorrCoef) and the discrimination performance (MI) reach a maximal value in MGv. Note also that at the population level, all the subcortical structures discriminate better the original vocalizations than cortical areas. G, Scatter plots showing in each structure, the strong correlations (0.77 < r < 0.88) between the CorrCoef and the MIIndividual (bits) values obtained in original conditions.