EEG coherence or EEG correlation?

Miguel Angel Guevara,María Corsi-Cabrera
DOI: https://doi.org/10.1016/s0167-8760(96)00038-4
IF: 2.903
1996-10-01
International Journal of Psychophysiology
Abstract:In view of the widespread use of electroencephalographic correlation (r) and coherence (Coh) analyses in studying brain functional relationships, it seems important to illustrate results yielded by both methods. Although they are considered as equivalent, they show important differences. Results obtained from r and Coh were compared using: (a) 2 Hz sinusoidal signals, where phase and amplitude were artificially manipulated and, (b) Coh and r spectra obtained between each pair of derivations (C3, C4, F3, F4, T3 and T4) from wakefulness, stage 2, stage 4 and paradoxical sleep (n = 8). The following results were obtained: significant Pearson product-moment coefficients were obtained between r and Coh spectra in the range of 0.86 to 0.96 for interhemispheric and 0.60 to 0.90, for intrahemispheric paris; as a result of principal component analyses, the same three frequency bands were formed for r and Coh spectra with the exception of one single bin; similar results for r and Coh were obtained by two-way ANOVAs (physiological stages by derivations). In conclusion, as expected, a high degree of comparability between r and Coh was observed under normal physiological conditions and with the use of good quality recordings.
psychology, experimental,physiology,neurosciences, biological
What problem does this paper attempt to address?