Inter-electrode Coherence

Inter-electrode coherence is computed separately across frequencies by relating the cross-spectrum of two recordings to the power spectrum of each recording. This normalization produces a bounded value from zero to one. The result emphasizes whether the signals maintain a consistent relationship in amplitude and phase over time within a particular frequency range.

A high value indicates that the relationship between two electrode signals is relatively stable at the frequency being examined, rather than changing unpredictably across the recording. Lower coherence indicates less consistent amplitude and phase relationships at that frequency. Coherence therefore describes coordination in a selected band or frequency range, not a single undifferentiated property of the whole signal.

Two electrodes may record activity arising from the same underlying source, so their signals can appear coordinated even when the electrode pair does not represent a distinct interacting network. Volume conduction adds to this problem by allowing activity to spread across recording locations. Careful electrode placement, preprocessing, and complementary connectivity measures are therefore important before assigning a neural interpretation.

Coherence can differ across frequencies because the consistency of a relationship is evaluated separately for each frequency range. A pair of electrodes may therefore show stronger coordination in one part of the spectrum than another. Reporting the relevant frequency range helps connect the result to the neural activity pattern, behavior, or condition being examined.

Begin with recordings from selected electrode pairs, then apply appropriate signal preprocessing before estimating the cross-spectrum and the individual power spectra. Compare these spectral quantities to obtain coherence across the frequencies of interest. Finally, interpret the values in light of electrode placement and possible shared sources, rather than treating every high value as independent network coordination.

Electrode placement determines which recording sites are being compared and how vulnerable the estimate is to signals shared across locations. Preprocessing helps ensure that the spectral comparison reflects the recordings being studied rather than avoidable signal problems. These choices directly affect whether an observed coherence pattern can be interpreted as meaningful coordinated neural activity.

It is useful when the goal is to examine coordinated activity across brain recording sites in relation to behavior, cognition, sleep, or neurological disorders. Researchers can use frequency-specific patterns to characterize functional connectivity and to compare how coordination changes across tasks, states, or clinical conditions represented in the recordings.

Patterns across electrode pairs can help distinguish synchronized brain networks and identify changes in coordination associated with a behavioral or cognitive condition, sleep, or a neurological disorder. The measure is most informative when interpreted with the recording context and complementary connectivity analyses, because coherence alone can also reflect shared sources or volume conduction.