[Mne_analysis] how to save source estamates from sphere label generated by mne.label.grow_labels

[David Leitman]

Hi Alex,

 sorry to keep on pestering but, in trying to grow my label i do the following and get the following result:

# read inverse solution
stc = mne.read_source_estimate(fname_meeg_stc)

# load the anatomical ROI
anat_label1 = mne.labels_from_parc(subject, parc='aparc',
                                  subjects_dir=data_path,
                                  regexp='superiortemporal-rh')[0][0]

# extract the anatomical time course for each label                                                                                  
stc_anat_label1 = stc.in_label(anat_label1) 

# calculate center of mass and transform to mni coordinates
vtx, _, t_anat_label1 = stc_anat_label1.center_of_mass(subject)
mni_anat_label1 = mne.vertex_to_mni(vtx, 1, subject)[0]

#make 3mm radius sphere around vertex
anat_label1_sphere=mne.label.grow_labels(subject=subject, seeds=vtx, extents=3.0, hemis=1, subjects_dir=data_path, n_jobs=1)

# extract the vertex-sphere time course for label  
stc_anat_label1_sphere=stc.in_label(anat_label1_sphere)

i get the following error:

/home/leitman/Enthought/Canopy_64bit/User/lib/python2.7/site-packages/mne-0.7.1-py2.7.egg/mne/source_estimate.pyc in in_label(self, label)
   1089         """
   1090         # make sure label and stc are compatible
-> 1091         if label.subject is not None and self.subject is not None \
   1092                 and label.subject != self.subject:
   1093             raise RuntimeError('label and stc must have same subject names, '

AttributeError: 'list' object has no attribute 'subject' 


It does not seem that I can add the subject attribute mannually as there is no __dict__ object

so what should I do?

dave




----- Original Message -----
From: "Alexandre Gramfort" <alexandre.gramfort at telecom-paristech.fr>
To: "David I Leitman" <leitman at mail.med.upenn.edu>
Sent: Tuesday, January 14, 2014 5:32:40 AM
Subject: Re: [Mne_analysis] how to save source estamates from epoched data set with multiple events

> thanks I have been using this example as a base but I was curious about the
> following:
> 1. can stc.extract_label_time_course  work on a epoched data . fif file
> containing all my event condtions

no. You will need individual STCs and lose track of event conditions.
Maybe we could do better ...

> 2.  lets say I want to extract  as a time course just the vertices with the
> center of mass or perhaps an roi reflecting this vertices and a small number
> of neighboring vertices instead of the mean of a whole anatomic roi label.
> how could I do that?

you can grow labels given a seed but you'll need to define a label.

stc.data contains the time course of all vertices in the label. It's a plain
numpy array so you can manipulate it as you want.

> 3 ) with np.save can I also ad a vector that corresponds to the event type
> that each epoch represents?

to save multiple arrays in one file use scipy.io.savemat (to write a
matlab file)

HTH
Alex