Effect of amplitude correlations on coherence in the local field potential

Abstract

Neural activity across the brain shows both spatial and temporal correlations at multiple scales, and understanding these correlations is a key step toward understanding cortical processing. Correlation in the local field potential (LFP) recorded from two brain areas is often characterized by computing the coherence, which is generally taken to reflect the degree of phase consistency across trials between two sites. Coherence, however, depends on two factors—phase consistency as well as amplitude covariation across trials—but the spatial structure of amplitude correlations across sites and its contribution to coherence are not well characterized. We recorded LFP from an array of microelectrodes chronically implanted in the primary visual cortex of monkeys and studied correlations in amplitude across electrodes as a function of interelectrode distance. We found that amplitude correlations showed a similar trend as coherence as a function of frequency and interelectrode distance. Importantly, even when phases were completely randomized between two electrodes, amplitude correlations introduced significant coherence. To quantify the contributions of phase consistency and amplitude correlations to coherence, we simulated pairs of sinusoids with varying phase consistency and amplitude correlations. These simulations confirmed that amplitude correlations can significantly bias coherence measurements, resulting in either over- or underestimation of true phase coherence. Our results highlight the importance of accounting for the correlations in amplitude while using coherence to study phase relationships across sites and frequencies.

Spatial spread of local field potential is band-pass in the primary visual cortex

“…1). Other studies from the same dataset have been reported previously (Ray and Maunsell 2010, 2011a, b; Srinath and Ray 2014). …”

“…We recorded spike and LFP signals from two male rhesus monkeys using a 10 × 10 array (Blackrock Microsystems) implanted in the V1, while they attended to a Gabor stimulus presented in the opposite hemifield of the RFs of the recorded sites and performed an orientation-change detection task [same task as used in previous studies: Ray and Maunsell (2010, 2011a); Srinath and Ray (2014); see materials and methods for details]. To estimate the spatial spread, we first estimated the RF sizes by flashing a small Gabor stimulus in a random sequence at 81 ( Monkey 1 , 9 × 9 rectangular grid spanning 2 × 2° of visual space) or 121 ( Monkey 2 , 11 × 11 grid spanning 3 × 3° of visual space) locations spanning all of the RFs.…”

“…In the primary visual cortex (V1), the correlation in the network (as measured indirectly using field-field or spike-field coherence) during spontaneous activity, as well as the change in correlation upon stimulus presentation, is frequency dependent, with a greater change in correlation in alpha (8–12 Hz), gamma (30–60 Hz), and high-gamma (60–150 Hz) bands (Jia et al 2013a; Srinath and Ray 2014). However, whether the spatial spread is frequency dependent in V1 in a manner that depends on the correlation profile has not been studied in detail.…”

Theta Phase Segregation of Input-Specific Gamma Patterns in Entorhinal-Hippocampal Networks

“…While principal cells in the superficial layers of neocortex are more electrotonically compact (Zador et al, 1995), which could perhaps enable cortical circuits to more effectively employ gamma synchronization as a mechanism for selective interregional communication (Singer, 1993; Fries, 2005), this hypothesis remains hotly debated (Shadlen and Movshon, 1999; Kopell et al, 2000; Fries, 2009; Ray and Maunsell, 2010; Fell and Axmacher, 2011; Bosman et al, 2012; Akam and Kullmann, 2012; Jia et al, 2013; Srinath and Ray, 2014). In our analysis, LFP-spike coupling between CA1 units and CA3 or EC3 LFP was mostly attributable to phase-locking of interneurons (Figure 6).…”

What does gamma coherence tell us about inter-regional neural communication?

“…Another spurious source of gamma coherence is amplitude covariations between signals in the same frequency band 65 . Given that power increases at gamma frequencies can also be brought about by both spiking activity and nonoscillatory transient events 66 , this is another artificial source of increased gamma coherence.…”