mne.preprocessing.annotate_muscle_zscore#
- mne.preprocessing.annotate_muscle_zscore(raw, threshold=4, ch_type=None, min_length_good=0.1, filter_freq=(110, 140), n_jobs=None, verbose=None)[source]#
Create annotations for segments that likely contain muscle artifacts.
Detects data segments containing activity in the frequency range given by
filter_freqwhose envelope magnitude exceeds the specified z-score threshold, when summed across channels and divided bysqrt(n_channels). False-positive transient peaks are prevented by low-pass filtering the resulting z-score time series at 4 Hz. Only operates on a single channel type, ifch_typeisNoneit will select the first type in the listmag,grad,eeg. See [1] for background on choosingfilter_freqandthreshold.- Parameters:
- rawinstance of
Raw Data to estimate segments with muscle artifacts.
- threshold
float The threshold in z-scores for marking segments as containing muscle activity artifacts.
- ch_type‘mag’ | ‘grad’ | ‘eeg’ |
None The type of sensors to use. If
Noneit will take the first type inmag,grad,eeg.- min_length_good
float|None The shortest allowed duration of “good data” (in seconds) between adjacent annotations; shorter segments will be incorporated into the surrounding annotations.``None`` is equivalent to
0. Default is0.1.- filter_freqarray_like, shape (2,)
The lower and upper frequencies of the band-pass filter. Default is
(110, 140).- n_jobs
int|str Number of jobs to run in parallel. Can be
'cuda'ifcupyis installed properly.Changed in version 1.13: Added support for CUDA.
- verbosebool |
str|int|None Control verbosity of the logging output. If
None, use the default verbosity level. See the logging documentation andmne.verbose()for details. Should only be passed as a keyword argument.
- rawinstance of
- Returns:
- annot
mne.Annotations Periods with muscle artifacts annotated as BAD_muscle.
- scores_muscle
array Z-score values averaged across channels for each sample.
- annot
Notes
Because the z-scores are summed across channels and divided by
sqrt(n_channels), an artifact present on only a few channels is diluted: one channel with a z-score of 10 among 270 channels contributes only10 / sqrt(270), about 0.6, to the score. To detect focal muscle activity, such as temporal muscle seen on a few temporal sensors, pass a copy restricted to those channels, e.g.annotate_muscle_zscore(raw.copy().pick(ch_names)).Each channel is z-scored against its own mean and standard deviation over the whole recording (excluding segments annotated as bad). If a channel shows muscle activity during a large part of the recording, that activity inflates its mean and standard deviation, so the artifact no longer stands out and may not be annotated. In that case, consider marking the channel as bad rather than annotating time spans.
References
