Neural Oscillations

An oscillation is a field signal. It arises when the synaptic currents of many neurons vary together, so that the summed potential varies with them. The effect is typically strongest where pyramidal cells are aligned in parallel. The same rhythm is recorded as the local field potential in the tissue, the ECoG on the cortical surface, and the EEG at the scalp (Buzsáki et al. (2012)). Field power decreases with frequency roughly as 1/f to 1/f², so a frequency band is generally identified as a peak above that background by spectral analysis. The band edges are conventions. The same rhythm can arise from different cellular events, so its source is inferred with supporting evidence.

Oscillations organize the timing of neuronal activity. Neurons that fire together within a gamma cycle form an assembly, and slower rhythms order assemblies into sequences. The phase of a slow rhythm can modulate the power of a faster one (phase–amplitude coupling) (Buzsáki et al., 2012). The phase at which a neuron fires can carry information that its firing rate does not, as shown, for example, for delta phase in visual cortex and theta phase in hippocampus (Quian Quiroga & Panzeri (2009)).

The electrode layout determines, in part, whether the source of a rhythm can be located. A distant reference adds its own rhythm to every channel, and a single site cannot identify the layer that generates a rhythm. Sites at regular pitch can do so by current source density analysis, along one shank or across several (Angotzi et al. (2025)). Cortical oscillations are reviewed by Buzsáki & Draguhn (2004). The book-length treatment is Buzsáki (2006), Rhythms of the Brain, and the literature view of the field is the Nature Index topic Cognitive Neuroscience of Neural Oscillations.