Description¶
This pipeline generates an individual session quality report for each file passed to it that reveals topoplots and tables showing the signal quality of all channels across the scalp.
In order to assess the signal quality of the session, the pipeline measures various metrics like line noise, high frequency noise, channel artifact, and channel correlation to determine the ratio of good to bad channels across the session. The metric values are also included in tables in the html report and can optionally be written out to CSV files.
PARAMETERS¶
Category: Input¶
This pipeline expects one file (or a Path Iterator formatted path name) and its path is relative to your Neuroscale account storage bucket. (The easiest way to run it is to go to your Storage page, select the file, click the Task button which will take you to the Create Task page, and then select this pipeline. The filename will be filed in automatically.)
input path¶
The input data files that the pipeline will process. The filenames can be entered by name, as a wild card expression (studies/Subject*/myfile.set), or by passing an ESS (EEG Study Schema) manifest XML file.
eeg stream¶
If the data contains multiple streams of signal data, enter the name of the stream containing the EEG data to perform the quality assessment on. Leave empty if there is only one stream of signal data in the file.
trim markers¶
Trim the data to these two markers (begin/end) before computing the quality metrics. Must be a list of two markers. If both markers are not found, no trimming will be performed.
trim time¶
Trim these time periods from the beginning and end of the data before computing the quality metrics. Must be a list of two numeric values. The first represents the number of seconds to trim from the start of the file, the second the number of seconds to trim from the end of the file.
Category: Output¶
group report¶
If multiple files are passed to this pipeline, generate a single group averages report for all files (=True). If False, an individual report will be generated for each file.
Category: Settings¶
line noise frequency¶
Frequency in Hz that line noise is expected due to AC power (e.g. USA is 60 Hz, Europe is 50 Hz). Use 'auto' to detect automatically.
correlation threshold¶
Correlation threshold for determining bad channels. Higher values (above 0.7) are more aggressive and mark more channels as being bad (red) or only acceptable (yellow). Values below 0.5 would be considered rather lax. This threshold is based on the correlation between a channel and what one would expect the channel to be based on the other channels.
bad channel subset size¶
Size of random channel subsets to use when computing the Bad Channel metric. This is for use in a RANSAC estimator, and is given as a fraction of the total number of channels. This may need to be higher for a montage with a small number of channels and/or where the montage is sparse (i.e., a circular montage with no electrode(s) in the middle), so that each subset contains at least 2 channels.
bad channel num samples¶
Number of random channel subsets to use when computing the Bad Channel metric (see bad_channel_subset_size parameter). Higher numbers together with smaller subset sizes give a more robust estimate, at higher computational cost during calibration. As with the bad_channel_subset_size, this may need to be higher for a montage with a small number of channels.
bad channel noise threshold¶
High-frequency noise threshold. Lower values (below 3.5) are more stringent and will remove more channels (i.e., moderately bad channels will get removed). Values above 5 would be considered very lax (i.e., only the worst channels get removed). This threshold is based on the amount of high frequency noise compared to other channels, and is measured in standard deviations.
Use clean time windows for Bad Channel metric¶
Use clean time windows for calibration for the bad channel metric computation. Setting this to False will flag channels as bad if they are across the entire file including time windows when the data is considered 'bad'. Set to True (default) is a more stringent test that only flags channels as bad if they are deemed so (uncorrelated with neighboring channels, HF noise) during time windows when the data is considered 'good' overall (other channels are ok). When set to True, the flagged channels represent those that are removed and re-interpolated by the standard Neuroscale Insight preprocessing chain.
remove unlocalized channels¶
Only plot channels which have a channel position or which can be matched to an available montage. (Set to False if you want to process channels anyway without a location. Note that in such a case, metrics which depend on channel location, such as inter-channel correlation, will be invalid.)