It is theoretically possible for the phase alone (Figure 2B, right) or amplitude alone (Figure 2B, left) Bortezomib price to carry all the information, or they can both contribute in part. A central goal of our study is to determine the prevalence of these different response types in the medial temporal and frontal areas of the human brain. We also aim to better understand these electrophysiological signals by asking which component carries the most information about behavioral events. We used the LFP measurements, triggered on the first and second card presentations,

to calculate the discriminability index d  ′ between correct and incorrect trials. This was done using the full LFP signal (dLFP’) and using the amplitude (damp’) and phase (dphase’) of the signal at a given frequency after decomposing the LFP using a wavelet transform ( Figure 2A; see Experimental Procedures). There was a clear dependence www.selleckchem.com/products/E7080.html of d  ′ on frequency ( Figure 3A). Discriminability

was low for phase and amplitude after the first click ( Figure 3A, black lines), but it was substantially higher for phase than amplitude after the second click ( Figure 3A, red lines). The differences between damp’ and dphase’ were greatest for frequencies below 4 Hz (Wilcoxon sign-rank test; p = 1 × 10−36 at 2.14 Hz; see also Figure S2A), and the largest average value for dphase’ occurred at 2.14 Hz. Interestingly, in addition to differences between phase and amplitude classifiers, there were differences between brain regions. The values of dphase’ in the temporal lobe (n = 1,008) Mannose-binding protein-associated serine protease were significantly larger than those in the frontal lobe (n = 644) when measured after the second

click (Figure 3B). Again, the largest average dphase’ value occurred at a frequency of 2.14 Hz, where the difference between temporal and frontal values was greatest (two-sample t test; p = 1 × 10−39; see also Figure S2B). Looking specifically at 2.14 Hz, a scatter plot of all d′ values in the temporal lobe confirms that classification using the phase of the LFP is better than classification using the amplitude, and it demonstrates that the d′ values based on phase rival those obtained using the full LFP signal ( Figure 3C, top left). No such relationships were found in the frontal lobe regions, where the d′ values were lower ( Figure 3C, bottom). To assess the significance of individual d′ values, we employed the technique of permutation resampling. For each electrode, all correct and incorrect trials were pooled together. Then, two new groups (of equal size to the original correct and incorrect groups) were chosen randomly without replacement by random assignment of the correct/incorrect labels to each waveform. These two new groups were used to calculate a classifier and an associated d′ value.