Current Source Density and Non-Current Source Density Transformed Spectral Measure Comparability: A Methodological Study
Disciplines
Biological Psychology | Cognition and Perception
Abstract (300 words maximum)
Current source density (CSD) transformation is a process by which researchers can increase the spatial frequency of electroencephalographic (EEG) measures by imposing a high-pass spatial filter on the data to attenuate the lateral spread of electrical signals across the skull and scalp. The present study examines the relationship between spectral power measures taken from CSD transformed and non-CSD transformed data to assess whether the CSD transformation limits researchers’ ability to compare spectral measures from studies that do and do not use CSD transformation. Data from multiple electrodes across multiple cognitive states (eyes open/closed and before/after a cognitive task) were compared to the CSD transformed versions of the same data using bivariate analysis. Further, correlations between CSD/non-CSD transformed measures for frontal alpha asymmetry (FAA), theta/beta ratio (TBR), and frontal beta asymmetry (FBA), were assessed using bivariate correlation. The relationship between CSD and non-CSD transformed data in the present study varied such that few correlations between CSD/non-CSD transformed data for the same spectral measure were relatively high, while the majority of bivariate analyses yielded correlations too low to support comparing CSD and non-CSD transformed spectral data. Overall, our study suggests that the CSD transformation changes the original measures to such a degree that data which have been filtered using CSD should not be compared to data that have not been transformed.
Academic department under which the project should be listed
RCHSS - Psychological Science
Primary Investigator (PI) Name
Tim Martin
Current Source Density and Non-Current Source Density Transformed Spectral Measure Comparability: A Methodological Study
Current source density (CSD) transformation is a process by which researchers can increase the spatial frequency of electroencephalographic (EEG) measures by imposing a high-pass spatial filter on the data to attenuate the lateral spread of electrical signals across the skull and scalp. The present study examines the relationship between spectral power measures taken from CSD transformed and non-CSD transformed data to assess whether the CSD transformation limits researchers’ ability to compare spectral measures from studies that do and do not use CSD transformation. Data from multiple electrodes across multiple cognitive states (eyes open/closed and before/after a cognitive task) were compared to the CSD transformed versions of the same data using bivariate analysis. Further, correlations between CSD/non-CSD transformed measures for frontal alpha asymmetry (FAA), theta/beta ratio (TBR), and frontal beta asymmetry (FBA), were assessed using bivariate correlation. The relationship between CSD and non-CSD transformed data in the present study varied such that few correlations between CSD/non-CSD transformed data for the same spectral measure were relatively high, while the majority of bivariate analyses yielded correlations too low to support comparing CSD and non-CSD transformed spectral data. Overall, our study suggests that the CSD transformation changes the original measures to such a degree that data which have been filtered using CSD should not be compared to data that have not been transformed.