Pipeline Labels

Annotation Data Structure

In EEGLAB, the pipeline annotations are stored within the EEG data structure. The variable EEG.marks contains time_info, chan_info, and comp_info fields that store the time, channel, and component annotations, respectively. These fields contain information about the label, colour, and flags for each annotation. The label is the name of the annotation and the colour is the RGB values for visually plotting the annotation. The flags array for channel, component and binary time annotations has either a 1 or 0 for each data point that indicates if the annotation is present (value of 1) for that data point or not (value of 0). For continuous time annotations, the flags array contains a unique value for each data point. The size of EEG.marks.time_info(i).flags is the number of time points (EEG.pnts), the size of the EEG.marks.chan_info(i).flags is the number of channels, and the size of EEG.marks.comp_info(i).flags is the number of components.

Channel Pipeline Labels

Label	Description
manual	The interactive label (modified by analysts interactively or by a pipeline decision along with other labels) typically used to indicate which channels are considered artefactual for any reason.
ch_sd	Pipeline decision flag indicating that channels were too often outliers compared to other channels for the measure of standard deviation of voltage within one second epochs.
low_r	Pipeline decision flag indicating that channels were too often outliers compared to other channels for the measure of correlation coefficient to spatially neighbouring channels within one second epochs.
bridge	Pipeline decision flag indicating that channels were outliers in terms of having high and invariant correlation coefficients to spatially neighbouring channels.
rank	Pipeline decision identifying the channel that has the least amount of unique information (highest average correlation coefficient to spatially neighbouring channels) to be ignored by the independent component analysis to account for the rank deficiency of the average referenced data.

Component Pipeline Labels

Label	Description
manual	The interactive label (modified by analysts interactively or by a pipeline decision along with other labels) typically used to indicate which components are considered artefactual for any reason.
ic_rt	Pipeline decision flag indicating that components were not replicated across the three parallel AMICA runs.
brain	ICLabel 1.0 Mark indicating that the component has cortical characteristics.
muscle	ICLabel 1.0 Mark indicating that the component has EMG characteristics.
eye	ICLabel 1.0 Mark indicating that the component has EOG characteristics.
heart	ICLabel 1.0 Mark indicating that the component has ECG characteristics.
line_noise	ICLabel 1.0 Mark indicating that the component has electrical mains noise characteristics.
chan_noise	ICLabel 1.0 Mark indicating that the component has channel independence characteristics.
other	ICLabel 1.0 Mark indicating that the component could not be confidently classified within any other ICLabel classification.
ambiguous	QC procedure rater markup indicating that the component is difficult to classify as either artefact or not.

Time Pipeline Labels

Label	Description
manual	The interactive label (modified by analysts interactively or by a pipeline decision along with other labels) typically used to indicate which time points are considered artefactual for any reason.
init_ind	Continuous variable indicating the initial time point index within the session.
mark_gap	Pipeline decision flag indicating that time points are within a short gap between other annotations.
ch_sd	Pipeline decision flag indicating that time points were too often outliers across channels compared to other time points for the measure of standard deviation of voltage within one second epochs.
low_r	Pipeline decision flag indicating that time points were too often outliers across channels compared to other time points for the measure of correlation coefficient to spatially neighbouring channels within one second epcohs.
logl_init	Log likelihood of initial AMICA decomposition.
ic_sd1	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component standard deviation of voltage within one second epochs.
logl_A	Log likelihood of final AMICA decomposition (replication A).
logl_B	Log likelihood of final AMICA decomposition (replication B).
logl_C	Log likelihood of final AMICA decomposition (replication C).
ic_sd2	Pipeline decision flag indicating that time points were too often outliers across final components compared to other time points for the measure of component standard deviation of voltage within one second epochs.
ic_dt	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component spectral Theta within one second epochs.
ic_a	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component spectral Alpha within one second epochs.
ic_b	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component spectral Beta within one second epochs.
ic_lg	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component spectral Low Gamma within one second epochs.
ic_hg	Pipeline decision flag indicating that time points were too often outliers across initial components compared to other time points for the measure of component spectral High Gamma within one second epochs.

---

🏠 Return to Main Page