The Spatiotemporal Dynamics of Oscillatory Activity in Humans Across Micro, Meso, and Macro Scales

Brain activity is characterized by oscillatory activity that spans at least two orders of magnitude. Previous investigations of the spatiotemporal dynamics of this wide range of oscillatory behavior has led to the concept that long-range intercortical interactions are expressed in low-frequency patterns while higher frequencies reflect more local intracortical connectivity. This inverse relationship between frequency and spatial coherence is seen as one of a small number of nearly universal rules governing brain activity. While there is substantial evidence for this proposition, there are surprisingly few direct, quantitative investigations of this phenomenon, especially in human cortex. To more completely characterize the spatial characteristics of ongoing brain activity, we investigated the coherence in different brain statesawake and sleep, at different frequencies, and with respect to a wide range of distances using standard pial surface macroelectrode arrays (1 cm spacing in an 8x8 cm grid), mesogrids (5 mm spacing), microgrids (1 mm spacing), and microelectrode arrays (400 micron spacing in 4x4mm arrays). As expected, we found that correlations and coherence decreases as a function of increasing interelectrode distance and as a function of frequency. We observed a robust linear relationship up until 1 cm; for distances 1 cm and greater, the relationship was largely non-linear. This relationship was not strongly affected by specific cortical lobe, nor was the overall coherence significantly different between awake and asleep states. These data are congruent with the overall notion that frequency and spatial relationships are inversely related with faster frequencies being more focal and provides an important quantitative assessment of that relationship with implications for the spatial scale of neural processing and recordings.