Use NeuroVault statistical maps in NiMARE

Download statistical maps from NeuroVault, then use them in a meta-analysis, with NiMARE.

import matplotlib.pyplot as plt
from nilearn.plotting import plot_stat_map

Neurovault + NiMARE: Load freely shared statistical maps for Meta-Analysis

Neurovault is an online platform that hosts unthresholded statistical maps, including group statistical maps. NiMARE can read these statistical maps when given a list of collection_ids. I search “working memory” on neurovault, and find these relevant collections:

• 2884

• 2621

• 3085

• 5623

• 3264

• 3192

• 457

I can load specific statistical maps from these collections into a NiMARE dataset:

from nimare.io import convert_neurovault_to_dataset

# The specific collections I would like to download group level
# statistical maps from
collection_ids = (2884, 2621, 3085, 5623, 3264, 3192, 457)

# A mapping between what I want the contrast(s) to be
# named in the dataset and what their respective group
# statistical maps are named on neurovault
contrasts = {
    "working_memory": (
        "Working memory load of 2 faces versus 1 face - NT2_Tstat|"
        "t-value contrast 2-back minus 0-back|"
        "Searchlight multivariate Decoding 2: visual working memory|"
        "Context-dependent group-specific WM information|"
        "WM working memory zstat1|"
        "WM task over CRT task map|"
        "tfMRI WM 2BK PLACE zstat1"
    )
}

# Convert how the statistical maps on neurovault are represented
# in a NiMARE dataset.
map_type_conversion = {"Z map": "z", "T map": "t"}

dset = convert_neurovault_to_dataset(
    collection_ids,
    contrasts,
    img_dir=None,
    map_type_conversion=map_type_conversion,
)

Conversion of Statistical Maps

Some of the statistical maps are T statistics and others are Z statistics. To perform a Fisher’s meta analysis, we need all Z maps. Thoughtfully, NiMARE has a class named ImageTransformer that will help us.

from nimare.transforms import ImageTransformer

# Not all studies have Z maps!
dset.images[["z"]]

	z
1	None
0	None
2	None
5	/home/docs/.nimare/working_memory/collection-3...
4	/home/docs/.nimare/working_memory/collection-3...
6	/home/docs/.nimare/working_memory/collection-4...
3	None

z_transformer = ImageTransformer(target="z")
dset = z_transformer.transform(dset)

Out:

/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/maskers/nifti_masker.py:108: UserWarning: imgs are being resampled to the mask_img resolution. This process is memory intensive. You might want to provide a target_affine that is equal to the affine of the imgs or resample the mask beforehand to save memory and computation time.
  warnings.warn(
/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/image/resampling.py:663: RuntimeWarning: NaNs or infinite values are present in the data passed to resample. This is a bad thing as they make resampling ill-defined and much slower.
  _resample_one_img(
/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/maskers/nifti_masker.py:108: UserWarning: imgs are being resampled to the mask_img resolution. This process is memory intensive. You might want to provide a target_affine that is equal to the affine of the imgs or resample the mask beforehand to save memory and computation time.
  warnings.warn(
/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/maskers/nifti_masker.py:108: UserWarning: imgs are being resampled to the mask_img resolution. This process is memory intensive. You might want to provide a target_affine that is equal to the affine of the imgs or resample the mask beforehand to save memory and computation time.
  warnings.warn(
/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/maskers/nifti_masker.py:108: UserWarning: imgs are being resampled to the mask_img resolution. This process is memory intensive. You might want to provide a target_affine that is equal to the affine of the imgs or resample the mask beforehand to save memory and computation time.
  warnings.warn(

All studies now have Z maps!

dset.images[["z"]]

	z
1	/home/docs/.nimare/working_memory/study-2621-w...
0	/home/docs/.nimare/working_memory/study-2884-w...
2	/home/docs/.nimare/working_memory/study-3085-w...
5	/home/docs/.nimare/working_memory/collection-3...
4	/home/docs/.nimare/working_memory/collection-3...
6	/home/docs/.nimare/working_memory/collection-4...
3	/home/docs/.nimare/working_memory/study-5623-w...

Run a Meta-Analysis

With the missing Z maps filled in, we can run a Meta-Analysis and plot our results

from nimare.meta.ibma import Fishers

# The default template has a slightly different, but completely compatible,
# affine than the NeuroVault images, so we allow the Estimator to resample
# images during the fitting process.
meta = Fishers(resample=True)

meta_res = meta.fit(dset)

fig, ax = plt.subplots()
display = plot_stat_map(meta_res.get_map("z"), threshold=3.3, axes=ax, figure=fig)
fig.show()
# The result may look questionable, but this code provides
# a template on how to use neurovault in your meta analysis.

Out:

/home/docs/checkouts/readthedocs.org/user_builds/nimare/envs/0.2.0/lib/python3.8/site-packages/nilearn/_utils/niimg.py:61: UserWarning: Non-finite values detected. These values will be replaced with zeros.
  warn(