First level analysis of a complete BIDS dataset from openneuro

Full step-by-step example of fitting a GLM to perform a first level analysis in an openneuro BIDS dataset. We demonstrate how BIDS derivatives can be exploited to perform a simple one subject analysis with minimal code. Details about the BIDS standard are available at https://bids.neuroimaging.io/. We also demonstrate how to download individual groups of files from the Openneuro s3 bucket.

More specifically:

1. Download an fMRI BIDS dataset with derivatives from openneuro.

2. Extract first level model objects automatically from the BIDS dataset.

3. Demonstrate Quality assurance of Nilearn estimation against available FSL. estimation in the openneuro dataset.

4. Display contrast plot and uncorrected first level statistics table report.

from nilearn import plotting

Fetch openneuro BIDS dataset

We download one subject from the stopsignal task in the ds000030 V4 BIDS dataset available in openneuro. This dataset contains the necessary information to run a statistical analysis using Nilearn. The dataset also contains statistical results from a previous FSL analysis that we can employ for comparison with the Nilearn estimation.

from nilearn.datasets import (
    fetch_ds000030_urls,
    fetch_openneuro_dataset,
    select_from_index,
)

_, urls = fetch_ds000030_urls()

exclusion_patterns = [
    "*group*",
    "*phenotype*",
    "*mriqc*",
    "*parameter_plots*",
    "*physio_plots*",
    "*space-fsaverage*",
    "*space-T1w*",
    "*dwi*",
    "*beh*",
    "*task-bart*",
    "*task-rest*",
    "*task-scap*",
    "*task-task*",
]
urls = select_from_index(
    urls, exclusion_filters=exclusion_patterns, n_subjects=1
)

data_dir, _ = fetch_openneuro_dataset(urls=urls)

[get_dataset_dir] Dataset found in /home/remi/nilearn_data/ds000030/ds000030_R1.0.4/uncompressed
[get_dataset_dir] Dataset found in /home/remi/nilearn_data/ds000030/ds000030_R1.0.4/uncompressed

Obtain FirstLevelModel objects automatically and fit arguments

From the dataset directory we automatically obtain FirstLevelModel objects with their subject_id filled from the BIDS dataset. Moreover we obtain, for each model, the list of run images and their respective events and confound regressors. Those are inferred from the confounds.tsv files available in the BIDS dataset. To get the first level models we have to specify the dataset directory, the task_label and the space_label as specified in the file names. We also have to provide the folder with the desired derivatives, that in this case were produced by the fMRIPrep BIDS app.

from nilearn.glm.first_level import first_level_from_bids

task_label = "stopsignal"
space_label = "MNI152NLin2009cAsym"
derivatives_folder = "derivatives/fmriprep"
(
    models,
    models_run_imgs,
    models_events,
    models_confounds,
) = first_level_from_bids(
    data_dir,
    task_label,
    space_label,
    smoothing_fwhm=5.0,
    derivatives_folder=derivatives_folder,
    n_jobs=2,
)

Access the model and model arguments of the subject and process events.

model, imgs, events, confounds = (
    models[0],
    models_run_imgs[0],
    models_events[0],
    models_confounds[0],
)
subject = f"sub-{model.subject_label}"
model.minimize_memory = False  # override default

from pathlib import Path

from nilearn.interfaces.fsl import get_design_from_fslmat

fsl_design_matrix_path = (
    Path(data_dir)
    / "derivatives"
    / "task"
    / subject
    / "stopsignal.feat"
    / "design.mat"
)
design_matrix = get_design_from_fslmat(
    fsl_design_matrix_path, column_names=None
)

We identify the columns of the Go and StopSuccess conditions of the design matrix inferred from the FSL file, to use them later for contrast definition.

design_columns = [
    f"cond_{int(i):02}" for i in range(len(design_matrix.columns))
]
design_columns[0] = "Go"
design_columns[4] = "StopSuccess"
design_matrix.columns = design_columns

First level model estimation (one subject)

We fit the first level model for one subject.

model.fit(imgs, design_matrices=[design_matrix])

FirstLevelModel(fir_delays=[0], memory=Memory(location=None),
                minimize_memory=False, n_jobs=2, smoothing_fwhm=5.0,
                subject_label='10159', t_r=2)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Then we compute the StopSuccess - Go contrast. We can use the column names of the design matrix.

z_map = model.compute_contrast("StopSuccess - Go")

We show the agreement between the Nilearn estimation and the FSL estimation available in the dataset.

import nibabel as nib

fsl_z_map = nib.load(
    Path(data_dir)
    / "derivatives"
    / "task"
    / subject
    / "stopsignal.feat"
    / "stats"
    / "zstat12.nii.gz"
)

import matplotlib.pyplot as plt
from scipy.stats import norm

plotting.plot_glass_brain(
    z_map,
    colorbar=True,
    threshold=norm.isf(0.001),
    title='Nilearn Z map of "StopSuccess - Go" (unc p<0.001)',
    plot_abs=False,
    display_mode="ortho",
)
plotting.plot_glass_brain(
    fsl_z_map,
    colorbar=True,
    threshold=norm.isf(0.001),
    title='FSL Z map of "StopSuccess - Go" (unc p<0.001)',
    plot_abs=False,
    display_mode="ortho",
)
plt.show()

from nilearn.plotting import plot_img_comparison

plot_img_comparison(
    [z_map], [fsl_z_map], model.masker_, ref_label="Nilearn", src_label="FSL"
)
plt.show()

• 
• 
• 

Simple statistical report of thresholded contrast

We display the contrast plot and table with cluster information.

from nilearn.plotting import plot_contrast_matrix

plot_contrast_matrix("StopSuccess - Go", design_matrix)
plotting.plot_glass_brain(
    z_map,
    colorbar=True,
    threshold=norm.isf(0.001),
    plot_abs=False,
    display_mode="z",
    figure=plt.figure(figsize=(4, 4)),
)
plt.show()

• 
• 

We can get a latex table from a Pandas Dataframe for display and publication purposes

from nilearn.reporting import get_clusters_table

table = get_clusters_table(z_map, norm.isf(0.001), 10)
print(table.to_latex())

\begin{tabular}{llrrrrl}
\toprule
 & Cluster ID & X & Y & Z & Peak Stat & Cluster Size (mm3) \\
\midrule
0 & 1 & -66.000000 & -45.000000 & 22.000000 & 5.307532 & 6300 \\
1 & 1a & -66.000000 & -33.000000 & 18.000000 & 4.668929 &  \\
2 & 1b & -48.000000 & -36.000000 & 14.000000 & 4.534376 &  \\
3 & 1c & -57.000000 & -48.000000 & 10.000000 & 4.254210 &  \\
4 & 2 & -42.000000 & 15.000000 & 26.000000 & 4.918703 & 2520 \\
5 & 2a & -51.000000 & 9.000000 & 34.000000 & 4.715845 &  \\
6 & 2b & -42.000000 & 9.000000 & 30.000000 & 4.683343 &  \\
7 & 2c & -57.000000 & 12.000000 & 38.000000 & 4.587956 &  \\
8 & 3 & 57.000000 & -27.000000 & 2.000000 & 4.692869 & 504 \\
9 & 3a & 66.000000 & -27.000000 & 2.000000 & 3.664250 &  \\
10 & 4 & 42.000000 & 9.000000 & 34.000000 & 4.461193 & 540 \\
11 & 5 & 6.000000 & 18.000000 & 34.000000 & 4.257986 & 2520 \\
12 & 5a & -3.000000 & 15.000000 & 46.000000 & 4.078390 &  \\
13 & 5b & 0.000000 & 0.000000 & 38.000000 & 3.815609 &  \\
14 & 5c & 3.000000 & 9.000000 & 50.000000 & 3.798387 &  \\
15 & 6 & 6.000000 & 6.000000 & 54.000000 & 4.208105 & 468 \\
16 & 6a & 6.000000 & 3.000000 & 62.000000 & 3.348351 &  \\
17 & 7 & -45.000000 & 21.000000 & 2.000000 & 4.190472 & 504 \\
18 & 7a & -54.000000 & 21.000000 & 6.000000 & 3.385929 &  \\
19 & 8 & 45.000000 & -21.000000 & 42.000000 & 4.163956 & 432 \\
20 & 9 & 63.000000 & -24.000000 & 30.000000 & 4.079389 & 360 \\
21 & 10 & -12.000000 & 6.000000 & 6.000000 & 4.056165 & 792 \\
22 & 10a & -9.000000 & -3.000000 & 10.000000 & 3.726486 &  \\
23 & 10b & -9.000000 & 6.000000 & 14.000000 & 3.710553 &  \\
24 & 11 & -27.000000 & 45.000000 & 18.000000 & 4.043724 & 432 \\
25 & 12 & 3.000000 & -24.000000 & 30.000000 & 3.950054 & 360 \\
26 & 13 & 12.000000 & -72.000000 & 22.000000 & 3.937283 & 360 \\
27 & 14 & 33.000000 & 42.000000 & 34.000000 & 3.906274 & 756 \\
28 & 14a & 30.000000 & 45.000000 & 26.000000 & 3.882906 &  \\
29 & 15 & 51.000000 & -30.000000 & 14.000000 & 3.776293 & 648 \\
\bottomrule
\end{tabular}

Saving model outputs to disk

We can now easily save the main results, the model metadata and an HTML report to the disk.

output_dir = Path.cwd() / "results" / "plot_bids_features"
output_dir.mkdir(exist_ok=True, parents=True)

from nilearn.interfaces.bids import save_glm_to_bids

save_glm_to_bids(
    model,
    contrasts="StopSuccess - Go",
    contrast_types={"StopSuccess - Go": "t"},
    out_dir=output_dir / "derivatives" / "nilearn_glm",
    prefix=f"{subject}_task-stopsignal",
)

• 
• 

/home/remi/github/nilearn/nilearn_doc_build/examples/04_glm_first_level/plot_bids_features.py:227: UserWarning:

Contrast name "StopSuccess - Go" changed to "stopsuccessMinusGo"

View the generated files

files = sorted((output_dir / "derivatives" / "nilearn_glm").glob("**/*"))
print("\n".join([str(x.relative_to(output_dir)) for x in files]))

derivatives/nilearn_glm/dataset_description.json
derivatives/nilearn_glm/sub-10159
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_design.svg
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_stat-effect_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_stat-p_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_stat-t_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_stat-variance_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_contrast-stopsuccessMinusGo_stat-z_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_design.json
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_design.svg
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_design.tsv
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_report.html
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_stat-errorts_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_stat-rsquared_statmap.nii.gz
derivatives/nilearn_glm/sub-10159/sub-10159_task-stopsignal_statmap.json

Only generating the HTML report

Using the computed FirstLevelModel and contrast information, we can quickly also also only create a summary report.

from nilearn.reporting import make_glm_report

report = make_glm_report(
    model=model,
    contrasts="StopSuccess - Go",
)

We have several ways to access the report:

# report  # This report can be viewed in a notebook
# report.open_in_browser()

# or we can save as an html file
# report.save_as_html(output_dir / 'report.html')

Estimated memory usage: 900 MB