Surface timeseries

Display a surface timeseries using data from nilearn

Tags: experimental

Python (.py)

Notebook (.ipynb)

Extra packages required

This example requires additional packages that are not available in typical napari installations. For this example to work in recommended napari installations, you will need to install additional packages: nilearn. See the plugin manager guide for ways to install additional packages from within napari.

[fetch_surf_nki_enhanced] Added README.md to /home/runner/nilearn_data
[fetch_surf_nki_enhanced] Dataset created in
/home/runner/nilearn_data/nki_enhanced_surface
[fetch_surf_nki_enhanced] Downloading data from
https://www.nitrc.org/frs/download.php/8470/pheno_nki_nilearn.csv ...
[fetch_surf_nki_enhanced]  ...done. (0 seconds, 0 min)

[fetch_surf_nki_enhanced] Downloading data from
https://www.nitrc.org/frs/download.php/8261/A00028185_rh_preprocessed_fsaverage5
_fwhm6.gii ...
[fetch_surf_nki_enhanced]  ...done. (0 seconds, 0 min)

[fetch_surf_nki_enhanced] Downloading data from
https://www.nitrc.org/frs/download.php/8260/A00028185_lh_preprocessed_fsaverage5
_fwhm6.gii ...
[fetch_surf_nki_enhanced]  ...done. (1 seconds, 0 min)

from importlib.metadata import version

from nilearn import datasets, surface
from packaging.version import parse

import napari

if parse(version("numpy")) >= parse('1.24') and parse(version("nilearn")) < parse('0.10.1'):
    raise RuntimeError(
        'Incompatible numpy version. '
        'You must have numpy less than 1.24 for nilearn 0.10.1 and below to '
        'work and download the example data'
    )


# Fetch datasets - this will download dataset if datasets are not found
nki_dataset = datasets.fetch_surf_nki_enhanced(n_subjects=1)
fsaverage = datasets.fetch_surf_fsaverage()

# Load surface data and resting state time series from nilearn
brain_vertices, brain_faces = surface.load_surf_data(fsaverage['pial_left'])
brain_vertex_depth = surface.load_surf_data(fsaverage['sulc_left'])
timeseries = surface.load_surf_data(nki_dataset['func_left'][0])
# nilearn provides data as n_vertices x n_timepoints, but napari requires the
# vertices axis to be placed last to match NumPy broadcasting rules
timeseries = timeseries.transpose((1, 0))

# create an empty viewer
viewer = napari.Viewer(ndisplay=3)

# add the mri
viewer.add_surface((brain_vertices, brain_faces, brain_vertex_depth), name='base')
viewer.add_surface((brain_vertices, brain_faces, timeseries),
                    colormap='turbo', opacity=0.9,
                    contrast_limits=[-1.5, 3.5], name='timeseries')

if __name__ == '__main__':
    napari.run()