Correl Mat¶

Functions used for the generation of weighted correlation matrices

ExtractTS¶

class neuropype_graph.nodes.correl_mat.ExtractTS(from_file=None, **inputs)¶

Description: Extract time series from a labelled mask in Nifti Format where all ROIs have the same index

Inputs:

indexed_rois_file:

type = File, exists=True, desc=’indexed mask where all voxels belonging to the same ROI have the same value (! starting from 1)’, mandatory=True

file_4D:

type = File, exists=True, desc=‘4D volume to be extracted’, mandatory=True

coord_rois_file:

type = File, desc=’ROI coordinates’

min_BOLD_intensity:

type = Float, default = 50.0, desc=’BOLD signal below this value will be set to zero’,usedefault = True

percent_signal:

type = Float, default = 0.5, desc = “Percent of voxels in a ROI with signal higher that min_BOLD_intensity to keep this ROI”,usedefault = True

plot_fig:

type = Bool, defaults = False, desc = “Plotting mean signal or not”, usedefault = True)

Outputs:

mean_masked_ts_file:

type = File, exists=True, desc=”mean ts in .npy (pickle format)”

subj_coord_rois_file:

type = File, exists=True, desc=”ROI coordinates retained for the subject”

IntersectMask¶

class neuropype_graph.nodes.correl_mat.IntersectMask(from_file=None, **inputs)¶

Description:

Keep only values of indexed mask where filter_mask is present. Optionnally, keep only ijk_coords, MNI_coords and labels that are kept in filtered mask

Inputs:

indexed_rois_file:

type = File, exists=True, desc=’nii file with indexed mask where all voxels belonging to the same ROI have the same value (! starting from 0)’, mandatory=True

filter_mask_file:

type = File, exists=True, desc=’nii file with (binary) mask - e.g. grey matter mask’, mandatory=True

coords_rois_file:

type = File, desc=’ijk coords txt file’

labels_rois_file:

type = File, desc=’labels txt file’

MNI_coords_rois_file:

type = File, desc=’MNI coords txt file’

filter_thr:

type = Float, default = 0.99, usedefault = True, desc=’Value to threshold filter_mask’

Outputs:

filtered_indexed_rois_file:

type = File, exists=True, desc=’nii file with indexed mask where all voxels belonging to the same ROI have the same value (! starting from 0)’

filtered_coords_rois_file:

type = File, exists=False, desc=’filtered ijk coords txt file’

filtered_labels_rois_file:

type = File, exists=False, desc=’filtered labels txt file’

filtered_MNI_coords_rois_file:

type = File, exists=False, desc=’filtered MNI coords txt file’

ExtractMeanTS¶

class neuropype_graph.nodes.correl_mat.ExtractMeanTS(from_file=None, **inputs)¶

Description:

Extract mean time series from a labelled mask in Nifti Format where the voxels of interest have values 1 (mask_file), or from a percent mask (filter_mask_file) with values higher than threshold (filter_thr)

Inputs:

file_4D:

type = File, exists=True, desc=‘4D volume to be extracted’, mandatory=True

mask_file:

type = File, xor = [‘filter_mask_file’], exists=True, desc=’mask file where all voxels belonging to the selected region have index 1’, mandatory=True

filter_mask_file:

type = File, xor = [‘mask_file’],requires = [‘filter_thr’], exists=True, desc=’mask file where all voxels belonging to the selected region have values higher than threshold’, mandatory=True

filter_thr:

type = Float, default = 0.99, usedefault = True, desc=’Value to threshold filter_mask’

suffix:

type = String, default = “suf”,desc=’Suffix added to describe the extracted time series’,mandatory=False,usedefault = True

plot_fig:

type = Bool, default = False, desc = “Plotting mean signal or not”, usedefault = True

Outputs:

mean_masked_ts_file:

type = File, exists=True, desc=”mean ts in .npy format”

ConcatTS¶

class neuropype_graph.nodes.correl_mat.ConcatTS(from_file=None, **inputs)¶

Description:

Concatenate time series

Inputs:

all_ts_file:

type = File, exists=True, desc=’npy file containing all ts to be concatenated’, mandatory=True

Outputs:

concatenated_ts_file:

type = File, exists=True, desc=”ts after concatenation”

Comments:

Not sure where it is used

MergeTS¶

class neuropype_graph.nodes.correl_mat.MergeTS(from_file=None, **inputs)¶

Description:

Merges time series from several files

Inputs:

all_ts_files:

type = List of Files, exists=True, desc=’list of npy files containing all ts to be merged’, mandatory=True

Outputs:

merged_ts_file:

type = File, exists=True, desc=”ts after merge”

Comments:

Used for multiple-session merges

SeparateTS¶

class neuropype_graph.nodes.correl_mat.SeparateTS(from_file=None, **inputs)¶

Description:

Save all time series from a npy file to several single time series npy files

Inputs:

all_ts_file:

type = File, exists=True, desc=’npy file containing all ts to be concatenated’, mandatory=True

Outputs:

separated_ts_files

type = List of Files, exists=True, desc=”ts files after separation”

Comments:

Not sure where it is used...

FindSPMRegressor¶

class neuropype_graph.nodes.correl_mat.FindSPMRegressor(from_file=None, **inputs)¶

Description:

Find regressor in SPM.mat and save it as timeseries txt file

Inputs:

spm_mat_file:

type = File, exists=True, desc=’SPM design matrix after generate model’, mandatory=True

regressor_name:

type = String, exists=True, desc=’Name of the regressor in SPM design matrix to be looked after’, mandatory=True

run_index:

type = Int, default = 1 , usedefault = True , desc = “Run (session) index, default is one in SPM”

only_positive_values:

type = Bool, default = True, usedefault = True , desc = “Return only positive values of the regressor (negative values are set to 0); Otherwise return all values”

concatenated_runs:

type = Bool, default = False , usedefault = True , desc = “If concatenate runs, need to search for the length of the session”

Deprecation: #concatenate_runs = traits.Int(1, usedefault = True , desc = “If concatenate runs, how many runs there is (needed to return the part of the regressors that is active for the session only)”)

Outputs:

regressor_file:

type = File,exists=True, desc=”txt file containing the regressor”

ComputeConfCorMat¶

class neuropype_graph.nodes.correl_mat.ComputeConfCorMat(from_file=None, **inputs)¶

Description:

Compute correlation between time series, with a given confidence interval. If weight_file is specified, used for weighted correlation

Inputs:

ts_file:

type = File, exists=True, desc=’Numpy files with time series to be correlated’,mandatory=True

transpose_ts:

type = Bool, default=True,usedefault = True,desc = ‘whether to transpose timeseries’, mandatory = True

weight_file:

type = File, exists=True, desc=’Weight of the correlation (normally, condition regressor file)’, mandatory=False

conf_interval_prob:

type = Float, default = 0.05, usedefault = True, desc=’Confidence interval’, mandatory=True

plot_mat:

type = Bool, default = True, usedefault = True, desc=’Confidence interval’, mandatory=False

labels_file:

type = File, exists=True, desc=’Name of the nodes (used only if plot = true)’, mandatory=False

Outputs:

cor_mat_file:

type = File, exists=True, desc=”npy file containing the R values of correlation”

Z_cor_mat_file:

type = File, exists=True, desc=”npy file containing the Z-values (after Fisher’s R-to-Z trasformation) of correlation”

conf_cor_mat_file:

type = File, exists=True, desc=”npy file containing the confidence interval around R values”