nifti_files: a string or a list/tuple of strings. :
The full path(s) to the file(s) from which the time-series is (are)
extracted
coords: ndarray or list/tuple of ndarray, optional. :
x,y,z (inplane,inplane,slice) coordinates of the ROI(s) from which the
time-series is (are) derived. If coordinates are provided, the
resulting time-series object will have 2 dimentsions. The first is the
coordinate dimension, in order of the provided coordinates and the
second is time. If set to None, all the coords in the volume will be
used and the coordinate system will be preserved - the output will be 4
dimensional, with time as the last dimension.
TR: float or TimeArray, optional :
The TR of the fmri measurement. The units are seconds, if provided as a float
argument. Otherwise, in the units of the TimeArray object
provided. Default: 1 second.
normalize: bool, optional :
Whether to normalize the activity in each voxel, defaults to
None, in which case the original fMRI signal is used. Other options
are: ‘percent’: the activity in each voxel is converted to percent
change, relative to this scan. ‘zscore’: the activity is converted to a
zscore relative to the mean and std in this voxel in this scan.
average: bool, optional whether to average the time-series across the :
voxels in the ROI (assumed to be the first dimension). In which
case, TS.data will be 1-d
filter: dict, optional :
If provided with a dict of the form:
{‘lb’:float or 0, ‘ub’:float or None, ‘method’:’fourier’,’boxcar’ ‘fir’
or ‘iir’ }
each voxel’s data will be filtered into the frequency range [lb,ub] with
nitime.analysis.FilterAnalyzer, using the method chosen here (defaults
to ‘fir’)
verbose: Whether to report on ROI and file being read. :
|