xcp_d.utils.confounds module

Confound matrix selection based on Ciric et al. 2007.

xcp_d.utils.confounds.get_custom_confounds(custom_confounds_folder, fmriprep_confounds_file)

Identify a custom confounds file.

Parameters:

• custom_confounds_folder (str or None) – Path to folder containing custom nuisance regressors. Must be a folder containing confounds files, in which case the file with the name matching the preprocessing confounds file will be selected.

• fmriprep_confounds_file (str) – Confounds TSV file from preprocessing derivatives. We expect the custom confounds file to have the same name.

Returns:

custom_confounds_file – Path to custom nuisance regressors.

Return type:

str or None

xcp_d.utils.confounds.load_confound_matrix(params, img_file, confounds_file, confounds_json_file, custom_confounds=None)

Load a subset of the confounds associated with a given file.

Parameters:

• params ({“36P”, “24P”, “27P”, “acompcor”, “acompcor_gsr”, “aroma”, “aroma_gsr”, “custom”}, optional) – Shorthand for the parameter set to extract from the confounds TSV. Default is “36P”, most expansive option.

• img_file (str) – The path to the bold file. Used to load the AROMA mixing matrix, if necessary.

• confounds_file (str) – The fMRIPrep confounds file. Used to load most confounds.

• confounds_json_file (str) – The JSON file associated with the fMRIPrep confounds file.

• custom_confounds (str or None, optional) – Custom confounds TSV if there is one. Default is None.

Returns:

• confounds_df (pandas.DataFrame or None) – The loaded and selected confounds. If “AROMA” is requested, then this DataFrame will include signal components as well. These will be named something like “signal_[XX]”. If params is “none”, confounds_df will be None.

• confounds_metadata (dict) – Metadata for the columns in the confounds file.

xcp_d.utils.confounds.load_motion(confounds_df, TR, motion_filter_type=None, band_stop_min=None, band_stop_max=None, motion_filter_order=4)

Load the six basic motion regressors (three rotations, three translations).

Parameters:

• confounds_df (pandas.DataFrame) – The confounds DataFrame from which to extract the six basic motion regressors.

• TR (float) – Repetition time of the BOLD run, in seconds.

• motion_filter_type ({None, “lp”, “notch”}) – Type of filter to use for removing respiratory artifact from motion regressors.

  If None, no filter will be applied.

  If the filter type is set to “notch”, frequencies between band_stop_min and band_stop_max will be removed with a notch filter. In this case, both band_stop_min and band_stop_max must be defined.

  If “lp”, frequencies above band_stop_min will be removed with a Butterworth filter. In this case, only band_stop_min must be defined. If “lp” or “notch”, that filtering will be done in this function. Otherwise, no filtering will be applied.

• band_stop_min (float or None) – Lower frequency for the motion parameter filter, in breaths-per-minute (bpm). Motion filtering is only performed if motion_filter_type is not None. If used with the “lp” motion_filter_type, this parameter essentially corresponds to a low-pass filter (the maximum allowed frequency in the filtered data). This parameter is used in conjunction with motion_filter_order and band_stop_max.

  Here is a list of recommended values, based on participant age:

  Age Range (years)	Recommended Value
  < 1	30
  1 - 2	25
  2 - 6	20
  6 - 12	15
  12 - 18	12
  19 - 65	12
  65 - 80	12
  > 80	10

  When motion_filter_type is set to “lp” (low-pass filter), another commonly-used value for this parameter is 6 BPM (equivalent to 0.1 Hertz), based on Gratton et al.[1].

• band_stop_max (float or None) – Upper frequency for the motion parameter filter, in breaths-per-minute (bpm). Motion filtering is only performed if motion_filter_type is not None. This parameter is only used if motion-filter-type is set to “notch”. This parameter is used in conjunction with motion_filter_order and band_stop_min.

  Here is a list of recommended values, based on participant age:

  Age Range (years)	Recommended Value
  < 1	60
  1 - 2	50
  2 - 6	35
  6 - 12	25
  12 - 18	20
  19 - 65	18
  65 - 80	28
  > 80	30

• motion_filter_order (int) – Number of filter coefficients for the motion parameter filter. Motion filtering is only performed if motion_filter_type is not None. This parameter is used in conjunction with band_stop_max and band_stop_min.

Returns:

motion_confounds_df – The six motion regressors. The three rotations are listed first, then the three translations.

Return type:

pandas.DataFrame

References

[1] (1,2,3)

Caterina Gratton, Ally Dworetsky, Rebecca S Coalson, Babatunde Adeyemo, Timothy O Laumann, Gagan S Wig, Tania S Kong, Gabriele Gratton, Monica Fabiani, Deanna M Barch, and others. Removal of high frequency contamination from motion estimates in single-band fmri saves data without biasing functional connectivity. Neuroimage, 217:116866, 2020. URL: https://doi.org/10.1016/j.neuroimage.2020.116866, doi:10.1016/j.neuroimage.2020.116866.

xcp_d.utils.confounds.motion_regression_filter(data, TR, motion_filter_type, band_stop_min, band_stop_max, motion_filter_order=4)

Filter translation and rotation motion parameters.

Parameters:

• data ((T, R) numpy.ndarray) – Data to filter. T = time, R = motion regressors The filter will be applied independently to each variable, across time.

• TR (float) – Repetition time of the BOLD run, in seconds.

• motion_filter_type ({None, “lp”, “notch”}) – Type of filter to use for removing respiratory artifact from motion regressors.

  If None, no filter will be applied.

  If the filter type is set to “notch”, frequencies between band_stop_min and band_stop_max will be removed with a notch filter. In this case, both band_stop_min and band_stop_max must be defined.

  If “lp”, frequencies above band_stop_min will be removed with a Butterworth filter. In this case, only band_stop_min must be defined. If not “notch” or “lp”, an exception will be raised.

• band_stop_min (float or None) – Lower frequency for the motion parameter filter, in breaths-per-minute (bpm). Motion filtering is only performed if motion_filter_type is not None. If used with the “lp” motion_filter_type, this parameter essentially corresponds to a low-pass filter (the maximum allowed frequency in the filtered data). This parameter is used in conjunction with motion_filter_order and band_stop_max.

  Here is a list of recommended values, based on participant age:

  Age Range (years)	Recommended Value
  < 1	30
  1 - 2	25
  2 - 6	20
  6 - 12	15
  12 - 18	12
  19 - 65	12
  65 - 80	12
  > 80	10

  When motion_filter_type is set to “lp” (low-pass filter), another commonly-used value for this parameter is 6 BPM (equivalent to 0.1 Hertz), based on Gratton et al.[1].

• band_stop_max (float or None) – Upper frequency for the motion parameter filter, in breaths-per-minute (bpm). Motion filtering is only performed if motion_filter_type is not None. This parameter is only used if motion-filter-type is set to “notch”. This parameter is used in conjunction with motion_filter_order and band_stop_min.

  Here is a list of recommended values, based on participant age:

  Age Range (years)	Recommended Value
  < 1	60
  1 - 2	50
  2 - 6	35
  6 - 12	25
  12 - 18	20
  19 - 65	18
  65 - 80	28
  > 80	30

• motion_filter_order (int) – Number of filter coefficients for the motion parameter filter. Motion filtering is only performed if motion_filter_type is not None. This parameter is used in conjunction with band_stop_max and band_stop_min.

Returns:

data – Filtered data. Same shape as the original data.

Return type:

(T, R) numpy.ndarray

Notes

Low-pass filtering (motion_filter_type = "lp") is performed with a Butterworth filter, as in Gratton et al.[1]. The order of the Butterworth filter is determined by motion_filter_order, although the original paper used a first-order filter. The original paper also used zero-padding with a padding size of 100. We use constant-padding, with the default padding size determined by scipy.signal.filtfilt().

Band-stop filtering (motion_filter_type = "notch") is performed with a notch filter, as in Fair et al.[2]. This filter uses the mean of the stopband frequencies as the target frequency, and the range between the two frequencies as the bandwidth. The filter is applied with constant-padding, using the default padding size determined by scipy.signal.filtfilt().

References

[2]

Damien A Fair, Oscar Miranda-Dominguez, Abraham Z Snyder, Anders Perrone, Eric A Earl, Andrew N Van, Jonathan M Koller, Eric Feczko, M Dylan Tisdall, Andre van der Kouwe, and others. Correction of respiratory artifacts in mri head motion estimates. Neuroimage, 208:116400, 2020. URL: https://doi.org/10.1016/j.neuroimage.2019.116400, doi:10.1016/j.neuroimage.2019.116400.