Overlaying multiple labels on a surface in mne_analyze

✉️ Mailing List Archive (read-only)
1

Hi,

Is it possible to overlay multiple layers at the same time (ideally in
different colors) on a surface in mne_analyze?

Thanks,
Steve

Stephen Politzer-Ahles
New York University, Abu Dhabi
Psychology Department
http://files.nyu.edu/spa268/public
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20130813/74ca0151/attachment.html

2

I have not found a way to do it in mne_analyze, but it is quite easy in
PySurfer.

Best,
Eric

3

Hi Eric,

Thanks for the feedback. Is it also possible to draw labels in PySurfer?
The reason I ask is because I have some funky labels (labels which were
defined on a decimated source space instead of a Freesurfer surface, so
when I plot them they show up as a bunch of disconnected points rather than
as nice contiguous areas), so I was hoping to just draw labels around them
to make nicer-looking plots; but to keep them from overlapping with
neighboring labels, I would probably need to be able to overlay a few at
once. (I tried doing this in Freesurfer by turning my labels into an
annotation and plotting the annotation, but I've been running into issues
with that pipeline, so I thought I'd try it in MNE instead.)

Best,
Steve

Stephen Politzer-Ahles
New York University, Abu Dhabi
Psychology Department
http://files.nyu.edu/spa268/public

4

Hey Steve,

In mne-python, you can load your label using 'label = mne.read_label(...)'
and then just use the 'label.smooth(...)' method -- it will do exactly what
you want, namely fill in a label so it becomes contiguous if it was
produced on a decimated surface. If the label was produced on fsaverage, 2
smoothing steps should be enough (although it will look jagged so 3-5 might
look better), otherwise you might need more. Drawing labels in PySurfer is
then as easy as doing brain.add_label(label, color='r', alpha=0.5), because
it has been coded to understand mne-python Label objects. One of the
options for "add_label" is to use outlines. PySurfer then also has a
brain.save_image() method to save the image. This makes scripting figures
easy, if you're into that. This is the pipeline I've adopted recently with
seemingly nice results.

Eric

5

Thanks, I'll give that a shot!

Stephen Politzer-Ahles
New York University, Abu Dhabi
Psychology Department
http://files.nyu.edu/spa268/public

6

Hi Eric,

What's the best way to install pysurfer to get this label ability. I've been
using the version the gets installed via pip. But that is pysurfer 0.3.1 is
there a newer release that has this change. Should I just grab the master
branch from git hub and install from source? How stable is the master
branch?

Thanks.

Luke

[mailto:mne_analysis-bounces at nmr.mgh.harvard.edu] On Behalf Of Eric Larson

7

Support for mne-python label objects is in master on the PySurfer github
site. It hasn't been released yet, but I /think/ PySurfer might be planning
another release soon, if it matters in your decision. The master branch has
been stable in my experience (it's what I use).

Installing from source should be fairly straightforward ("git clone git://
github.com/nipy/PySurfer.git", then "python setup.py install", maybe with a
"--user" if appropriate).

Eric