Cross-correlation study of the temporal interactions between areas V1 and V2 of the macaque monkey

Cross-correlation studies performed in cat visual cortex have shown that ne urons in different cortical areas of the same hemisphere or in correspondin g areas of opposite hemispheres tend to synchronize their activities. The p resence of synchronization may be related to the parallel organization of t he cat visual system, in which different cortical areas can be activated in parallel from the lateral geniculate nucleus. We wanted to determine wheth er interareal synchronization of firing can also be observed in the monkey, in which cortical areas are thought to be organized in a hierarchy spannin g different levels. Cross-correlation histograms (CCHs) were calculated fro m pairs of single or pairs of multiunit activities simultaneously recorded in areas V1 and V2 of paralyzed and anesthetized macaque monkeys. Moving ba rs and flashed bars were used as stimuli. The shift predictor was calculate d and subtracted from the raw CCH to reveal interactions of neuronal origin in isolation. Significant CCH peaks, indicating interactions of neuronal o rigin, were obtained in 11% of the dual single-unit recordings and 46% of t he dual multiunit recordings with moving bars. The incidence of nonflat CCH s with flashed bars was 29 and 78%, respectively. For the pairs of recordin g sites where both flashed and moving stimuli were used, the incidences of significant CCHs were very similar. Three types of peaks were distinguished on the basis of their width at half-height: T (<16 ms), C (between 16 and 180 ms), and H peaks (>180 ms). T peaks were very rarely observed (<1% in s ingle-unit recordings). H peaks were observed in 7-16% of the single-unit C CHs, and C peaks in 6-16%, depending on the stimulus used. C and H peaks we re observed more often when the receptive fields were overlapping or distan t by <2 degrees. To test for the presence of synchronization between neuron s in areas V1 and V2, we measured the position of the CCH peak with respect to the origin of the time axis of the CCH. Only in the case of a few T pea ks did we find displaced peaks, indicating a possible drive of the V2 neuro n by the simultaneously recorded V1 cell. All the other peaks were either c entered on the origin or overlapped the origin of time with their upper hal ves. Thus similarly to what has been reported for the cat, neurons belongin g to different cortical areas in the monkey tend to synchronize the time of emission of their action potentials with three different levels of tempora l precision. For peaks calculated from hashed stimuli, we compared the peak position with the difference between latencies of V1 and V2 neurons. There was a clear correlation for single-unit pairs in the case of C peaks. Thus the position of a C peak on the time axis appears to reflect the order of visual activation of the correlated neurons. The coupling strength for H pe aks was smaller during Visual drive compared with spontaneous activity. On the contrary, C peaks were seen more often and were stronger during visual stimulation than during spontaneous activity. This suggests that C-type syn chronization is associated with the processing of visual information. The o rigin of synchronized activity in a serially organized system is discussed.