eeg_decompose: Generalized eigenvalue decomposition based methods for EEG...

In neuroconductor/eegUtils: Utilities for Electroencephalographic (EEG) Analysis

Generalized eigenvalue decomposition based methods for EEG data

Description

Implements a selection of Generalized Eigenvalue based decomposition methods for EEG signals. Intended for isolating oscillations at specified frequencies, decomposing channel-based data into components reflecting distinct or combinations of sources of oscillatory signals. Currently, spatio-spectral decomposition (Nikulin, Nolte, & Curio, 2011) and Rhythmic Entrainment Source Separation (Cohen & Gulbinate, 2017) are implemented. The key difference between the two is that the former returns the results of the data-derived spatial filters applied to the bandpass-filtered "signal" data, whereas the latter returns the results of the filters applied to the original, broadband data.

Usage

eeg_decompose(data, ...)

## S3 method for class 'eeg_epochs'
eeg_decompose(
  data,
  sig_range,
  noise_range,
  method = "ssd",
  verbose = TRUE,
  order = 2,
  ...
)

Arguments

data	An eeg_data object
...	Additional parameters
sig_range	Vector with two inputs, the lower and upper bounds of the frequency range of interest
noise_range	Range of frequencies to be considered noise (e.g. bounds of flanker frequencies)
method	Type of decomposition to apply. Currently only "ssd" is supported.
verbose	Informative messages printed to console. Defaults to TRUE.
order	Filter order for filter applied to signal/noise

Value

An eeg_ICA object. Note that

Methods (by class)

• eeg_epochs: method for eeg_epochs objects

Author(s)

Matt Craddock [e-mail]

References

Cohen, M. X., & Gulbinate, R. (2017). Rhythmic entrainment source separation: Optimizing analyses of neural responses to rhythmic sensory stimulation. NeuroImage, 147, 43-56. https://doi.org/10.1016/j.neuroimage.2016.11.036

Haufe, S., Dähne, S., & Nikulin, V. V. (2014). Dimensionality reduction for the analysis of brain oscillations. NeuroImage, 101, 583–597. https://doi.org/10.1016/j.neuroimage.2014.06.073

Nikulin, V. V., Nolte, G., & Curio, G. (2011). A novel method for reliable and fast extraction of neuronal EEG/MEG oscillations on the basis of spatio-spectral decomposition. NeuroImage, 55(4), 1528–1535. https://doi.org/10.1016/j.neuroimage.2011.01.057

See Also

Other decompositions: run_ICA()

Examples

# The default method is Spatio-Spectral Decomposition, which returns
# spatially and temporally filtered source timecourses.
 decomposed <-
   eeg_decompose(demo_epochs,
                 sig_range = c(9, 11),
                 noise_range = c(8, 12),
                 method = "ssd")
 plot_psd(decomposed)
 # We can plot the spatial filters using `topoplot()`
 topoplot(decomposed, 1:2)
 plot_timecourse(decomposed, 1)
# method = "ress" returns spatially but not temporally filtered timecourses.
 with_RESS <-
   eeg_decompose(demo_epochs,
                 sig_range = c(9, 11),
                 noise_range = c(8, 12),
                 method = "ress")
 plot_psd(with_RESS)
 # The topographical plots are identical to those using "ssd", as the
 # spatial filters are the same.
 topoplot(with_RESS, 1:2)
 plot_timecourse(with_RESS, 1)

neuroconductor/eegUtils documentation built on Feb. 3, 2023, 5:33 p.m.