Deriving spatial maps from group fMRI data using ICA and Dictionary Learning

Various approaches exist to derive spatial maps or networks from group fmr data. The methods extract distributed brain regions that exhibit similar BOLD fluctuations over time. Decomposition methods allow for generation of many independent maps simultaneously without the need to provide a priori information (e.g. seeds or priors.)

This example will apply two popular decomposition methods, ICA and Dictionary learning, to fMRI data measured while children and young adults watch movies. The resulting maps will be visualized using atlas plotting tools.

CanICA is an ICA method for group-level analysis of fMRI data. Compared to other strategies, it brings a well-controlled group model, as well as a thresholding algorithm controlling for specificity and sensitivity with an explicit model of the signal.

The reference paper is Varoquaux et al.[1].

Load brain development fMRI dataset

from nilearn import datasets

rest_dataset = datasets.fetch_development_fmri(n_subjects=30)
func_filenames = rest_dataset.func  # list of 4D nifti files for each subject

# print basic information on the dataset
print(f"First functional nifti image (4D) is at: {rest_dataset.func[0]}")
[get_dataset_dir] Dataset found in /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri
[get_dataset_dir] Dataset found in /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri
[get_dataset_dir] Dataset found in /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri
First functional nifti image (4D) is at: /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz

Apply CanICA on the data

We use “whole-brain-template” as a strategy to compute the mask, as this leads to slightly faster and more reproducible results. However, the images need to be in MNI template space.

from nilearn.decomposition import CanICA

canica = CanICA(
    n_components=20,
    memory="nilearn_cache",
    memory_level=2,
    verbose=10,
    mask_strategy="whole-brain-template",
    random_state=0,
    standardize="zscore_sample",
    n_jobs=2,
)
canica.fit(func_filenames)

# Retrieve the independent components in brain space. Directly
# accessible through attribute `components_img_`.
canica_components_img = canica.components_img_
# components_img is a Nifti Image object, and can be saved to a file with
# the following lines:
from pathlib import Path

output_dir = Path.cwd() / "results" / "plot_compare_decomposition"
output_dir.mkdir(exist_ok=True, parents=True)
print(f"Output will be saved to: {output_dir}")
canica_components_img.to_filename(output_dir / "canica_resting_state.nii.gz")
[CanICA.fit] Loading data from [/home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar124_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar125_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar126_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar127_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar128_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar001_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar002_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar003_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar004_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar005_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar006_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar007_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar008_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar009_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar010_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar011_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar012_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar013_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar014_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar015_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar016_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar017_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar018_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar019_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar020_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar021_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar022_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar023_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar024_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz].
[CanICA.fit] Computing mask
/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/joblib/memory.py:616: UserWarning:

Cannot inspect object functools.partial(<function compute_multi_brain_mask at 0x7f42af50e480>, mask_type='whole-brain'), ignore list will not work.

[CanICA.fit] Template whole-brain mask computation
/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/joblib/memory.py:813: UserWarning:

Cannot inspect object functools.partial(<function compute_multi_brain_mask at 0x7f42af50e480>, mask_type='whole-brain'), ignore list will not work.

[CanICA.fit] Resampling mask
[CanICA.fit] Loading data
________________________________________________________________________________
[Memory] Calling sklearn.utils.extmath.randomized_svd...
randomized_svd(array([[0.003659, ..., 0.013254],
       ...,
       [0.012477, ..., 0.002881]]), n_components=20, transpose=True, random_state=0, n_iter=3)
___________________________________________________randomized_svd - 0.8s, 0.0min
[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.
[Parallel(n_jobs=2)]: Done   1 tasks      | elapsed:    0.6s
[Parallel(n_jobs=2)]: Done   4 tasks      | elapsed:    1.3s
[Parallel(n_jobs=2)]: Done  10 out of  10 | elapsed:    2.9s finished
Output will be saved to: /home/runner/work/nilearn/nilearn/examples/03_connectivity/results/plot_compare_decomposition

To visualize we plot the outline of all components on one figure

from nilearn.plotting import plot_prob_atlas

# Plot all ICA components together
plot_prob_atlas(canica_components_img, title="All ICA components")
plot compare decomposition
/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/nilearn/plotting/displays/_axes.py:74: UserWarning:

linewidths is ignored by contourf

/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/numpy/ma/core.py:2881: UserWarning:

Warning: converting a masked element to nan.


<nilearn.plotting.displays._slicers.OrthoSlicer object at 0x7f42adce3170>

Finally, we plot the map for each ICA component separately

from nilearn.image import iter_img
from nilearn.plotting import plot_stat_map, show

for i, cur_img in enumerate(iter_img(canica_components_img)):
    plot_stat_map(
        cur_img,
        display_mode="z",
        title=f"IC {int(i)}",
        cut_coords=1,
        colorbar=False,
    )
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition

Compare CanICA to dictionary learning

Dictionary learning is a sparsity based decomposition method for extracting spatial maps. It extracts maps that are naturally sparse and usually cleaner than ICA. Here, we will compare networks built with CanICA to networks built with Dictionary learning.

For more detailse see Mensch et al.[2].

Create a dictionary learning estimator

from nilearn.decomposition import DictLearning

dict_learning = DictLearning(
    n_components=20,
    memory="nilearn_cache",
    memory_level=2,
    verbose=1,
    random_state=0,
    n_epochs=1,
    mask_strategy="whole-brain-template",
    standardize="zscore_sample",
    n_jobs=2,
)

print("[Example] Fitting dictionary learning model")
dict_learning.fit(func_filenames)
print("[Example] Saving results")
# Grab extracted components umasked back to Nifti image.
# Note: For older versions, less than 0.4.1. components_img_
# is not implemented. See Note section above for details.
dictlearning_components_img = dict_learning.components_img_
dictlearning_components_img.to_filename(
    output_dir / "dictionary_learning_resting_state.nii.gz"
)
[Example] Fitting dictionary learning model
[DictLearning.fit] Loading data from [/home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar124_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar125_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar126_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar127_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar128_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar001_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar002_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar003_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar004_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar005_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar006_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar007_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar008_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar009_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar010_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar011_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar012_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar013_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar014_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar015_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar016_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar017_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar018_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar019_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar020_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar021_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar022_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar023_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz,
 /home/runner/work/nilearn/nilearn/nilearn_data/development_fmri/development_fmri/sub-pixar024_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz].
[DictLearning.fit] Computing mask
/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/joblib/memory.py:616: UserWarning:

Cannot inspect object functools.partial(<function compute_multi_brain_mask at 0x7f42af50e480>, mask_type='whole-brain'), ignore list will not work.

/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/joblib/memory.py:813: UserWarning:

Cannot inspect object functools.partial(<function compute_multi_brain_mask at 0x7f42af50e480>, mask_type='whole-brain'), ignore list will not work.

[DictLearning.fit] Resampling mask
[DictLearning.fit] Loading data
[DictLearning.fit] Learning initial components
________________________________________________________________________________
[Memory] Calling sklearn.utils.extmath.randomized_svd...
randomized_svd(array([[-0.001315, ...,  0.004387],
       ...,
       [ 0.011243, ...,  0.004194]]), n_components=20, transpose=True, random_state=0, n_iter=3)
___________________________________________________randomized_svd - 0.8s, 0.0min
[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.
[DictLearning.fit] Computing initial loadings
________________________________________________________________________________
[Memory] Calling nilearn.decomposition.dict_learning._compute_loadings...
_compute_loadings(array([[ 0.00281 , ..., -0.000886],
       ...,
       [ 0.000697, ..., -0.005029]]),
array([[-0.622651, ...,  5.322742],
       ...,
       [ 0.777205, ...,  0.743122]]))
_________________________________________________compute_loadings - 0.1s, 0.0min
[DictLearning.fit]  Learning dictionary
________________________________________________________________________________
[Memory] Calling sklearn.decomposition._dict_learning.dict_learning_online...
dict_learning_online(array([[-0.622651, ...,  0.777205],
       ...,
       [ 5.322742, ...,  0.743122]]),
20, alpha=10, batch_size=20, method='cd', dict_init=array([[-0.146031, ..., -0.158448],
       ...,
       [ 0.423266, ..., -0.263781]]), verbose=0, random_state=0, return_code=True, shuffle=True, n_jobs=1, max_iter=1090)
_____________________________________________dict_learning_online - 2.4s, 0.0min
[Example] Saving results

Visualize the results

First plot all DictLearning components together

plot_prob_atlas(
    dictlearning_components_img, title="All DictLearning components"
)
plot compare decomposition
/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/nilearn/plotting/displays/_axes.py:74: UserWarning:

linewidths is ignored by contourf

/opt/hostedtoolcache/Python/3.12.6/x64/lib/python3.12/site-packages/numpy/ma/core.py:2881: UserWarning:

Warning: converting a masked element to nan.


<nilearn.plotting.displays._slicers.OrthoSlicer object at 0x7f42af4c9d30>

One plot of each component

for i, cur_img in enumerate(iter_img(dictlearning_components_img)):
    plot_stat_map(
        cur_img,
        display_mode="z",
        title=f"Comp {int(i)}",
        cut_coords=1,
        colorbar=False,
    )
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition
  • plot compare decomposition

Estimate explained variance per component and plot using matplotlib

The fitted object dict_learning can be used to calculate the score per component

scores = dict_learning.score(func_filenames, per_component=True)

# Plot the scores
import numpy as np
from matplotlib import pyplot as plt
from matplotlib.ticker import FormatStrFormatter

plt.figure(figsize=(4, 4), layout="constrained")
positions = np.arange(len(scores))
plt.barh(positions, scores)
plt.ylabel("Component #", size=12)
plt.xlabel("Explained variance", size=12)
plt.yticks(np.arange(20))
plt.gca().xaxis.set_major_formatter(FormatStrFormatter("%.3f"))

show()
plot compare decomposition
________________________________________________________________________________
[Memory] Calling nilearn.decomposition._base._explained_variance...
_explained_variance(array([[-6.227098e-01, ...,  5.322679e+00],
       ...,
       [ 1.345007e-16, ..., -5.703498e-16]]),
array([[0., ..., 0.],
       ...,
       [0., ..., 0.]]), per_component=True)
______________________________________________explained_variance - 13.7s, 0.2min

Note

To see how to extract subject-level timeseries’ from regions created using Dictionary learning, see example Regions extraction using dictionary learning and functional connectomes. # noqa: E501

References

Total running time of the script: (2 minutes 34.707 seconds)

Estimated memory usage: 3116 MB

Gallery generated by Sphinx-Gallery