Statistical map viewer

Generate a brainsprite viewer for an activation map with an anatomical background, by populating an html template.

First fetch the MNI high-resolution template and a functional statistical map.

from nilearn import datasets
anat = datasets.MNI152_FILE_PATH

# one motor contrast map from NeuroVault
motor_images = datasets.fetch_neurovault_motor_task()
stat_img = motor_images.images[0]

Out:

Dataset created in /home/circleci/nilearn_data/neurovault

We are going to use the same template and instruction as in the plot_anat tutorial. The defaults are set for a functional map, so there is not much to do. We still tweak a couple parameters to get a clean map:

• apply a threshold to get rid of small activation (threshold),

• reduce the opacity of the overlay to see the underlying anatomy (opacity)

• Put a title inside the figure (title)

• manually specify the cut coordinates (cut_coords)

from brainsprite import viewer_substitute

bsprite = viewer_substitute(threshold=3, opacity=0.5, title="plot_stat_map",
                         cut_coords=[36, -27, 66])
bsprite.fit(stat_img, bg_img=anat)

We can now open the template with tempita, and fill it with the required information. The parameters indicate which tempita names we used in the template for the javascript, html and library code, respectively.

import tempita
file_template = '../docs/source/_static/viewer_template.html'
template = tempita.Template.from_filename(file_template, encoding="utf-8")

viewer = bsprite.transform(template, javascript='js', html='html', library='bsprite')

# In a Jupyter notebook, if ``view`` is the output of a cell, it will
# be displayed below the cell
viewer

The following instruction can be used to save the viewer in a stand-alone, html document:

viewer.save_as_html('plot_stat_map.html')

There are a lot more control one can use to modify the appearance of the brain viewer. Check the Python API for more information.