Fig. 6: Event-related potentials for successful response inhibition.
From: Functional architecture of executive control and associated event-related potentials in macaques
Conventions as in Fig. 3. a Grand average z-transformed EEG on canceled (thick) and latency-matched no stop-signal (thin) trials. b Difference functions (top) remove stimulus-evoked ERP to highlight N2 and P3 components in 3 SSD bins. Shaded intervals show ±50 ms sampling interval around N2 (orange) and P3 (gray) peaks. Concomitant recruitment of the three neuron classes (bottom). c Model comparison shows N2 amplitude variation was best described by the error-likelihood, and P3 amplitude was best described by log(Ttone). Full statistics in Supplementary Table 6. d Significant variation of N2 amplitude as a function of p(NCerror|SSD)/p(SSseen|SSD) in 29 sessions contributing 87 points across early, intermediate, or late SSD. e Significant variation of P3 amplitude as a function of Ttone in 29 sessions contributing 87 points across Ttone. f Partial regression between N2 amplitude and spike rate for Conflict (left) and Event Timing (right) neurons in L2/3 (top) and L5/6 (bottom) for sessions with both L2/3 and L5/6 neurons sampled. Ordinate scale plots, with EEG convention, residual from fixed-effects-adjusted ERP amplitude controlling for activity in the opposite layer. Abscissa scale plots residual fixed-effects-adjusted neuronal discharge rate in the identified layer controlling for the activity in the opposite layer and stop-signal delay. Each point plots the average EEG voltage and associated spiking rate in one of 20 bins with equal numbers of trials per session. Plotted are 120 points from 6 sessions for Conflict (left) and 100 points from 5 sessions for Event Timing (right) neurons. N2 amplitude variation was predicted by spiking rate variation of Conflict and Event Timing neurons in L2/3 but not in L5/6. g P3 amplitude variation was predicted by spiking rate variation of Goal Maintenance neurons in L2/3 but not in L5/6.
