Abstract
Time-frequency coherence has been widely used to quantify statistical dependencies in bivariate data and has proven to be vital for the study of neural interactions in electrophysiological recordings. Conventional methods establish time-frequency coherence by smoothing the cross and power spectra using identical smoothing procedures. Smoothing entails a trade-off between time-frequency resolution and statistical consistency and is critical for detecting instantaneous coherence in single-trial data. Here, we propose a generalized method to estimate time-frequency coherency by using different smoothing procedures for the cross spectra versus power spectra. This novel method has an improved trade-off between time resolution and statistical consistency compared to conventional methods, as verified by two simulated data sets. The methods are then applied to single-trial surface encephalography recorded from human subjects for comparative purposes. Our approach extracted robust alpha- and gamma-band synchronization over the visual cortex that was not detected by conventional methods, demonstrating the efficacy of this method.
Similar content being viewed by others
Article PDF
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mehrkanoon, S., Breakspear, M., Daffertshofer, A. et al. Generalized time-frequency coherency for assessing neural interactions in electrophysiological recordings. Nat Prec (2011). https://doi.org/10.1038/npre.2011.6615.1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/npre.2011.6615.1