Table 1.
Overview of the various steps comprising the EEG-Pype preprocessing and quantitative analysis workflow. AECc: corrected Amplitude Envelope Correlation, JPE: Joint Permutation Entropy, PLI: Phase Lag Index.
Fig 1.
Raw and Filtered Power Spectrum for the test EEG measurement.
(A) The power spectral density of the raw sample data, showing a prominent 50 Hz line noise artifact. (B) The power spectral density of the same data after the application of a 0.5-47 Hz band-pass filter, demonstrating the effective removal of line noise. The dotted lines correspond to the applied broadband filter cutoff values. dB: decibels, Hz: hertz, μV: microvolt.
Fig 2.
Independent Component Analysis for the test EEG measurement.
(A) The time series plot shows the EEG signal over the entire measurement duration after ICA decomposition. The window shown here contains (eye) movement artifacts, with eye blinks clearly distinguishable in component ICA000 and muscle activity in multiple other components. The automatically generated IC labels and the algorithm confidence level in percentages, seen in (B), can help to support ICA decision making. The topographical plot (C) confirms that component ICA000 corresponds to ocular artifacts, as can be seen from the field distribution with a maximum over frontal regions. ICA: Independent Component Analysis, s: seconds.
Fig 3.
MNE-Python interactive epoch selection windows.
(A) An example of an eight-second epoch with adequate quality. Magenta vertical markers denote whole seconds. (B) An example of an epoch with more artifacts present. The red color means the epoch will be discarded.
Fig 4.
The graphical user interface used for the quantitative analysis module of EEG-Pype.
Here, it can be seen that all possible quantitative metrics were selected. In addition, the optional output of connectivity matrices, MST matrices and channel-level averages (metric values per channel or brain area) was selected.
Table 2.
For the metrics calculated on the test EEG file (except minimum spanning tree metrics), the whole brain, epoch-averaged values are shown. Where applicable, these results are limited to theta and alpha bands, while the actual software output contains many more bands and non-epoch-averaged values. EEG: electroencephalography, Hz: Hertz, PSD: power spectral density.