.. DO NOT EDIT. .. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY. .. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE: .. "auto_examples/03_connectivity/plot_data_driven_parcellations.py" .. LINE NUMBERS ARE GIVEN BELOW. .. only:: html .. note:: :class: sphx-glr-download-link-note :ref:`Go to the end ` to download the full example code or to run this example in your browser via Binder .. rst-class:: sphx-glr-example-title .. _sphx_glr_auto_examples_03_connectivity_plot_data_driven_parcellations.py: Clustering methods to learn a brain parcellation from fMRI ========================================================== We use spatially-constrained Ward-clustering, KMeans, Hierarchical KMeans and Recursive Neighbor Agglomeration (ReNA) to create a set of parcels. In a high dimensional regime, these methods can be interesting to create a 'compressed' representation of the data, replacing the data in the :term:`fMRI` images by mean signals on the parcellation, which can subsequently be used for statistical analysis or machine learning. Also, these methods can be used to learn functional connectomes and subsequently for classification tasks or to analyze data at a local level. See Also -------- Which clustering method to use, an empirical comparison can be found in :footcite:t:`Thirion2014`. This :term:`parcellation` may be useful in a supervised learning, see for instance :footcite:t:`Michel2011b`. The big picture discussion corresponding to this example can be found in the documentation section :ref:`parcellating_brain`. .. GENERATED FROM PYTHON SOURCE LINES 30-34 Download a brain development fMRI dataset and turn it to a data matrix ---------------------------------------------------------------------- We download one subject of the movie watching dataset from Internet .. GENERATED FROM PYTHON SOURCE LINES 34-50 .. code-block:: Python import time import numpy as np from matplotlib import patches, ticker from nilearn import datasets, plotting from nilearn.image import get_data, index_img, mean_img from nilearn.regions import Parcellations dataset = datasets.fetch_development_fmri(n_subjects=1) # print basic information on the dataset print(f"First subject functional nifti image (4D) is at: {dataset.func[0]}") .. rst-class:: sphx-glr-script-out .. code-block:: none First subject functional nifti image (4D) is at: /home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz .. GENERATED FROM PYTHON SOURCE LINES 51-56 Brain parcellations with Ward Clustering ---------------------------------------- Transforming list of images to data matrix and build brain parcellations, all can be done at once using `Parcellations` object. .. GENERATED FROM PYTHON SOURCE LINES 56-97 .. code-block:: Python # Computing ward for the first time, will be long... This can be seen by # measuring using time start = time.time() # Agglomerative Clustering: ward # We build parameters of our own for this object. Parameters related to # masking, caching and defining number of clusters and specific parcellations # method. ward = Parcellations( method="ward", n_parcels=1000, standardize=False, smoothing_fwhm=2.0, memory="nilearn_cache", memory_level=1, verbose=1, ) # Call fit on functional dataset: single subject (less samples). ward.fit(dataset.func) print(f"Ward agglomeration 1000 clusters: {time.time() - start:.2f}s") # We compute now ward clustering with 2000 clusters and compare # time with 1000 clusters. To see the benefits of caching for second time. # We initialize class again with n_parcels=2000 this time. start = time.time() ward = Parcellations( method="ward", n_parcels=2000, standardize=False, smoothing_fwhm=2.0, memory="nilearn_cache", memory_level=1, verbose=1, ) ward.fit(dataset.func) print(f"Ward agglomeration 2000 clusters: {time.time() - start:.2f}s") .. rst-class:: sphx-glr-script-out .. code-block:: none [MultiNiftiMasker.fit] Loading data from [/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz]. [{self.__class__.__name__}.fit] Computing mask /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/decomposition/_base.py:429: UserWarning: memory_level is currently set to 0 but a Memory object has been provided. Setting memory_level to 1. self.masker_.fit(imgs) [MultiNiftiMasker.transform] Resampling mask [Parcellations] Loading data [MultiNiftiMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [MultiNiftiMasker.transform_single_imgs] Smoothing images [MultiNiftiMasker.transform_single_imgs] Extracting region signals [MultiNiftiMasker.transform_single_imgs] Cleaning extracted signals [Parcellations] computing ward ________________________________________________________________________________ [Memory] Calling nilearn.regions.parcellations._estimator_fit... _estimator_fit(array([[-0.002728, ..., 0.002821], ..., [-0.001877, ..., 0.003343]]), AgglomerativeClustering(connectivity=<24256x24256 sparse matrix of type '' with 162682 stored elements in COOrdinate format>, memory=Memory(location=nilearn_cache/joblib), n_clusters=1000)) ________________________________________________________________________________ [Memory] Calling sklearn.cluster._agglomerative.ward_tree... ward_tree(array([[-0.002728, ..., -0.001877], ..., [ 0.002821, ..., 0.003343]]), connectivity=<24256x24256 sparse matrix of type '' with 162682 stored elements in COOrdinate format>, n_clusters=1000, return_distance=False) ________________________________________________________ward_tree - 2.5s, 0.0min ____________________________________________________estimator_fit - 2.7s, 0.0min /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/masking.py:980: UserWarning: Data array used to create a new image contains 64-bit ints. This is likely due to creating the array with numpy and passing `int` as the `dtype`. Many tools such as FSL and SPM cannot deal with int64 in Nifti images, so for compatibility the data has been converted to int32. return new_img_like(mask_img, unmasked, affine) Ward agglomeration 1000 clusters: 11.72s [MultiNiftiMasker.fit] Loading data from [/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz]. [{self.__class__.__name__}.fit] Computing mask /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/decomposition/_base.py:429: UserWarning: memory_level is currently set to 0 but a Memory object has been provided. Setting memory_level to 1. self.masker_.fit(imgs) [MultiNiftiMasker.transform] Resampling mask [Parcellations] Loading data [Parcellations] computing ward ________________________________________________________________________________ [Memory] Calling nilearn.regions.parcellations._estimator_fit... _estimator_fit(array([[-0.002728, ..., 0.002821], ..., [-0.001877, ..., 0.003343]]), AgglomerativeClustering(connectivity=<24256x24256 sparse matrix of type '' with 162682 stored elements in COOrdinate format>, memory=Memory(location=nilearn_cache/joblib), n_clusters=2000)) ________________________________________________________________________________ [Memory] Calling sklearn.cluster._agglomerative.ward_tree... ward_tree(array([[-0.002728, ..., -0.001877], ..., [ 0.002821, ..., 0.003343]]), connectivity=<24256x24256 sparse matrix of type '' with 162682 stored elements in COOrdinate format>, n_clusters=2000, return_distance=False) ________________________________________________________ward_tree - 2.1s, 0.0min ____________________________________________________estimator_fit - 2.3s, 0.0min /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/masking.py:980: UserWarning: Data array used to create a new image contains 64-bit ints. This is likely due to creating the array with numpy and passing `int` as the `dtype`. Many tools such as FSL and SPM cannot deal with int64 in Nifti images, so for compatibility the data has been converted to int32. return new_img_like(mask_img, unmasked, affine) Ward agglomeration 2000 clusters: 9.47s .. GENERATED FROM PYTHON SOURCE LINES 98-103 Visualize: Brain parcellations (Ward) ..................................... First, we display the parcellations of the brain image stored in attribute `labels_img_` .. GENERATED FROM PYTHON SOURCE LINES 103-122 .. code-block:: Python ward_labels_img = ward.labels_img_ # Now, ward_labels_img are Nifti1Image object, it can be saved to file # with the following code: from pathlib import Path output_dir = Path.cwd() / "results" / "plot_data_driven_parcellations" output_dir.mkdir(exist_ok=True, parents=True) print(f"Output will be saved to: {output_dir}") ward_labels_img.to_filename(output_dir / "ward_parcellation.nii.gz") first_plot = plotting.plot_roi( ward_labels_img, title="Ward parcellation", display_mode="xz" ) # Grab cut coordinates from this plot to use as a common for all plots cut_coords = first_plot.cut_coords .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_001.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_001.png :class: sphx-glr-single-img .. rst-class:: sphx-glr-script-out .. code-block:: none Output will be saved to: /home/himanshu/Desktop/nilearn_work/nilearn/examples/03_connectivity/results/plot_data_driven_parcellations .. GENERATED FROM PYTHON SOURCE LINES 123-129 Compressed representation of Ward clustering ............................................ Second, we illustrate the effect that the clustering has on the signal. We show the original data, and the approximation provided by the clustering by averaging the signal on each parcel. .. GENERATED FROM PYTHON SOURCE LINES 129-172 .. code-block:: Python # Grab number of voxels from attribute mask image (mask_img_). original_voxels = np.sum(get_data(ward.mask_img_)) # Compute mean over time on the functional image to use the mean # image for compressed representation comparisons mean_func_img = mean_img(dataset.func[0]) # Compute common vmin and vmax vmin = np.min(get_data(mean_func_img)) vmax = np.max(get_data(mean_func_img)) plotting.plot_epi( mean_func_img, cut_coords=cut_coords, title=f"Original ({int(original_voxels)} voxels)", vmax=vmax, vmin=vmin, display_mode="xz", ) # A reduced dataset can be created by taking the parcel-level average: # Note that Parcellation objects with any method have the opportunity to # use a `transform` call that modifies input features. Here it reduces their # dimension. Note that we `fit` before calling a `transform` so that average # signals can be created on the brain parcellations with fit call. fmri_reduced = ward.transform(dataset.func) # Display the corresponding data compressed # using the parcellation using parcels=2000. fmri_compressed = ward.inverse_transform(fmri_reduced) plotting.plot_epi( index_img(fmri_compressed, 0), cut_coords=cut_coords, title="Ward compressed representation (2000 parcels)", vmin=vmin, vmax=vmax, display_mode="xz", ) # As you can see below, this approximation is almost good, although there # are only 2000 parcels, instead of the original 60000 voxels .. rst-class:: sphx-glr-horizontal * .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_002.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_002.png :class: sphx-glr-multi-img * .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_003.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_003.png :class: sphx-glr-multi-img .. rst-class:: sphx-glr-script-out .. code-block:: none [Parcellations.wrapped] loading data from Nifti1Image('/home/himanshu/Desktop/nilearn_work/nilearn/examples/03_connectivity/results/plot_data_driven_parcellations/ward_parcellation.nii.gz') [Parcellations.wrapped] loading data from Nifti1Image( shape=(50, 59, 50), affine=array([[ 4., 0., 0., -96.], [ 0., 4., 0., -132.], [ 0., 0., 4., -78.], [ 0., 0., 0., 1.]]) ) ________________________________________________________________________________ [Memory] Calling nilearn.maskers.base_masker._filter_and_extract... _filter_and_extract(, , { 'background_label': 0, 'clean_kwargs': {}, 'detrend': False, 'dtype': None, 'high_pass': None, 'high_variance_confounds': False, 'keep_masked_labels': True, 'labels': None, 'labels_img': , 'low_pass': None, 'mask_img': , 'reports': True, 'smoothing_fwhm': 2.0, 'standardize': False, 'standardize_confounds': True, 'strategy': 'mean', 't_r': None, 'target_affine': None, 'target_shape': None}, confounds=None, sample_mask=None, dtype=None, memory=Memory(location=nilearn_cache/joblib), memory_level=1, verbose=1) [NiftiLabelsMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [NiftiLabelsMasker.transform_single_imgs] Smoothing images [NiftiLabelsMasker.transform_single_imgs] Extracting region signals [NiftiLabelsMasker.transform_single_imgs] Cleaning extracted signals _______________________________________________filter_and_extract - 1.6s, 0.0min .. GENERATED FROM PYTHON SOURCE LINES 173-180 Brain parcellations with KMeans Clustering ------------------------------------------ We use the same approach as with building parcellations using Ward clustering. But, in the range of a small number of clusters, it is most likely that we want to use standardization. Indeed with standardization and smoothing, the clusters will form as regions. .. GENERATED FROM PYTHON SOURCE LINES 180-199 .. code-block:: Python # class/functions can be used here as they are already imported above. # This object uses method='kmeans' for KMeans clustering with 10mm smoothing # and standardization ON start = time.time() kmeans = Parcellations( method="kmeans", n_parcels=50, standardize="zscore_sample", smoothing_fwhm=10.0, memory="nilearn_cache", memory_level=1, verbose=1, ) # Call fit on functional dataset: single subject (less samples) kmeans.fit(dataset.func) print(f"KMeans clusters: {time.time() - start:.2f}s") .. rst-class:: sphx-glr-script-out .. code-block:: none [MultiNiftiMasker.fit] Loading data from [/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz]. [{self.__class__.__name__}.fit] Computing mask /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/decomposition/_base.py:429: UserWarning: memory_level is currently set to 0 but a Memory object has been provided. Setting memory_level to 1. self.masker_.fit(imgs) [MultiNiftiMasker.transform] Resampling mask [Parcellations] Loading data [MultiNiftiMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [MultiNiftiMasker.transform_single_imgs] Smoothing images [MultiNiftiMasker.transform_single_imgs] Extracting region signals [MultiNiftiMasker.transform_single_imgs] Cleaning extracted signals [Parcellations] computing kmeans ________________________________________________________________________________ [Memory] Calling nilearn.regions.parcellations._estimator_fit... _estimator_fit(array([[-0.002242, ..., -0.008556], ..., [-0.004365, ..., 0.012342]]), MiniBatchKMeans(n_clusters=50, n_init=3, random_state=0)) ____________________________________________________estimator_fit - 0.8s, 0.0min KMeans clusters: 13.27s .. GENERATED FROM PYTHON SOURCE LINES 200-204 Visualize: Brain parcellations (KMeans) ....................................... Grab parcellations of brain image stored in attribute `labels_img_` .. GENERATED FROM PYTHON SOURCE LINES 204-217 .. code-block:: Python kmeans_labels_img = kmeans.labels_img_ display = plotting.plot_roi( kmeans_labels_img, mean_func_img, title="KMeans parcellation", display_mode="xz", ) # kmeans_labels_img is a Nifti1Image object, it can be saved to file with # the following code: kmeans_labels_img.to_filename(output_dir / "kmeans_parcellation.nii.gz") .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_004.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_004.png :class: sphx-glr-single-img .. GENERATED FROM PYTHON SOURCE LINES 218-234 Brain parcellations with Hierarchical KMeans Clustering ------------------------------------------------------- As the number of images from which we try to cluster grows, voxels display more and more specific activity patterns causing KMeans clusters to be very unbalanced with a few big clusters and many voxels left as singletons. Hierarchical Kmeans algorithm is tailored to enforce more balanced clusterings. To do this, Hierarchical Kmeans does a first Kmeans clustering in square root of n_parcels. In a second step, it clusters voxels inside each of these parcels in m pieces with m adapted to the size of the cluster in order to have n balanced clusters in the end. This object uses method='hierarchical_kmeans' for Hierarchical KMeans clustering and 10mm smoothing and standardization to compare with the previous method. .. GENERATED FROM PYTHON SOURCE LINES 234-247 .. code-block:: Python start = time.time() hkmeans = Parcellations( method="hierarchical_kmeans", n_parcels=50, standardize="zscore_sample", smoothing_fwhm=10, memory="nilearn_cache", memory_level=1, verbose=1, ) # Call fit on functional dataset: single subject (less samples) hkmeans.fit(dataset.func) .. rst-class:: sphx-glr-script-out .. code-block:: none [MultiNiftiMasker.fit] Loading data from [/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz]. [{self.__class__.__name__}.fit] Computing mask /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/decomposition/_base.py:429: UserWarning: memory_level is currently set to 0 but a Memory object has been provided. Setting memory_level to 1. self.masker_.fit(imgs) [MultiNiftiMasker.transform] Resampling mask [Parcellations] Loading data [MultiNiftiMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [MultiNiftiMasker.transform_single_imgs] Smoothing images [MultiNiftiMasker.transform_single_imgs] Extracting region signals [MultiNiftiMasker.transform_single_imgs] Cleaning extracted signals [Parcellations] computing hierarchical_kmeans ________________________________________________________________________________ [Memory] Calling nilearn.regions.parcellations._estimator_fit... _estimator_fit(array([[-0.002242, ..., -0.008556], ..., [-0.004365, ..., 0.012342]]), HierarchicalKMeans(n_clusters=50)) ____________________________________________________estimator_fit - 5.3s, 0.1min /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/masking.py:980: UserWarning: Data array used to create a new image contains 64-bit ints. This is likely due to creating the array with numpy and passing `int` as the `dtype`. Many tools such as FSL and SPM cannot deal with int64 in Nifti images, so for compatibility the data has been converted to int32. return new_img_like(mask_img, unmasked, affine) .. raw:: html 
Parcellations(memory=Memory(location=nilearn_cache/joblib), memory_level=1,
                  method='hierarchical_kmeans', smoothing_fwhm=10,
                  standardize='zscore_sample')
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.


.. GENERATED FROM PYTHON SOURCE LINES 248-252 Visualize: Brain parcellations (Hierarchical KMeans) .................................................... Grab parcellations of brain image stored in attribute `labels_img_` .. GENERATED FROM PYTHON SOURCE LINES 252-268 .. code-block:: Python hkmeans_labels_img = hkmeans.labels_img_ plotting.plot_roi( hkmeans_labels_img, mean_func_img, title="Hierarchical KMeans parcellation", display_mode="xz", cut_coords=display.cut_coords, ) # kmeans_labels_img is a :class:`nibabel.nifti1.Nifti1Image` object, it can be # saved to file with the following code: hkmeans_labels_img.to_filename( output_dir / "hierarchical_kmeans_parcellation.nii.gz" ) .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_005.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_005.png :class: sphx-glr-single-img .. GENERATED FROM PYTHON SOURCE LINES 269-278 Compare Hierarchical Kmeans clusters with those from Kmeans ........................................................... To compare those, we'll first count how many voxels are contained in each of the 50 clusters for both algorithms and compare those sizes distribution. Hierarchical KMeans should give clusters closer to average (600 here) than KMeans. First count how many voxels have each label (except 0 which is the background). .. GENERATED FROM PYTHON SOURCE LINES 278-289 .. code-block:: Python _, kmeans_counts = np.unique(get_data(kmeans_labels_img), return_counts=True) _, hkmeans_counts = np.unique(get_data(hkmeans_labels_img), return_counts=True) voxel_ratio = np.round(np.sum(kmeans_counts[1:]) / 50) # If all voxels not in background were balanced between clusters ... print(f"... each cluster should contain {voxel_ratio} voxels") .. rst-class:: sphx-glr-script-out .. code-block:: none ... each cluster should contain 485.0 voxels .. GENERATED FROM PYTHON SOURCE LINES 290-293 Let's plot clusters sizes distributions for both algorithms You can just skip the plotting code, the important part is the figure .. GENERATED FROM PYTHON SOURCE LINES 293-320 .. code-block:: Python import matplotlib.pyplot as plt bins = np.concatenate( [ np.linspace(0, 500, 11), np.linspace(600, 2000, 15), np.linspace(3000, 10000, 8), ] ) fig, axes = plt.subplots( nrows=2, sharex=True, gridspec_kw={"height_ratios": [4, 1]} ) plt.semilogx() axes[0].hist(kmeans_counts[1:], bins, color="blue") axes[1].hist(hkmeans_counts[1:], bins, color="green") axes[0].set_ylim(0, 16) axes[1].set_ylim(4, 0) axes[1].xaxis.set_major_formatter(ticker.ScalarFormatter()) axes[1].yaxis.set_label_coords(-0.08, 2) fig.subplots_adjust(hspace=0) plt.xlabel("Number of voxels (log)", fontsize=12) plt.ylabel("Number of clusters", fontsize=12) handles = [ patches.Rectangle((0, 0), 1, 1, color=c, ec="k") for c in ["blue", "green"] ] labels = ["Kmeans", "Hierarchical Kmeans"] fig.legend(handles, labels, loc=(0.5, 0.8)) .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_006.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_006.png :class: sphx-glr-single-img .. rst-class:: sphx-glr-script-out .. code-block:: none .. GENERATED FROM PYTHON SOURCE LINES 321-326 As we can see, half of the 50 KMeans clusters contain less than 100 voxels whereas three contain several thousands voxels Hierarchical KMeans yield better balanced clusters, with a significant proportion of them containing hundreds to thousands of voxels. .. GENERATED FROM PYTHON SOURCE LINES 328-340 Brain parcellations with :term:`ReNA` Clustering ------------------------------------------------ One interesting algorithmic property of :term:`ReNA` (see References) is that it is very fast for a large number of parcels (notably faster than Ward). As before, the :term:`parcellation` is done with a ``Parcellations`` object. The spatial constraints are implemented inside the ``Parcellations`` object. More about :term:`ReNA` clustering algorithm in the original paper (:footcite:t:`Hoyos2019`). .. GENERATED FROM PYTHON SOURCE LINES 340-355 .. code-block:: Python start = time.time() rena = Parcellations( method="rena", n_parcels=5000, standardize=False, smoothing_fwhm=2.0, scaling=True, memory="nilearn_cache", memory_level=1, verbose=1, ) rena.fit_transform(dataset.func) print(f"ReNA 5000 clusters: {time.time() - start:.2f}s") .. rst-class:: sphx-glr-script-out .. code-block:: none [MultiNiftiMasker.fit] Loading data from [/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz]. [{self.__class__.__name__}.fit] Computing mask /home/himanshu/Desktop/nilearn_work/nilearn/nilearn/decomposition/_base.py:429: UserWarning: memory_level is currently set to 0 but a Memory object has been provided. Setting memory_level to 1. self.masker_.fit(imgs) [MultiNiftiMasker.transform] Resampling mask [Parcellations] Loading data [Parcellations] computing rena ________________________________________________________________________________ [Memory] Calling nilearn.regions.parcellations._estimator_fit... _estimator_fit(array([[-0.002728, ..., 0.002821], ..., [-0.001877, ..., 0.003343]]), ReNA(mask_img=, memory=Memory(location=nilearn_cache/joblib), n_clusters=5000, scaling=True), 'rena') ________________________________________________________________________________ [Memory] Calling nilearn.regions.rena_clustering.recursive_neighbor_agglomeration... recursive_neighbor_agglomeration(array([[-0.002728, ..., 0.002821], ..., [-0.001877, ..., 0.003343]]), , 5000, n_iter=10, threshold=1e-07, verbose=0) _________________________________recursive_neighbor_agglomeration - 1.4s, 0.0min ____________________________________________________estimator_fit - 1.5s, 0.0min [Parcellations.wrapped] loading data from Nifti1Image( shape=(50, 59, 50), affine=array([[ 4., 0., 0., -96.], [ 0., 4., 0., -132.], [ 0., 0., 4., -78.], [ 0., 0., 0., 1.]]) ) [Parcellations.wrapped] loading data from Nifti1Image( shape=(50, 59, 50), affine=array([[ 4., 0., 0., -96.], [ 0., 4., 0., -132.], [ 0., 0., 4., -78.], [ 0., 0., 0., 1.]]) ) ________________________________________________________________________________ [Memory] Calling nilearn.maskers.base_masker._filter_and_extract... _filter_and_extract(, , { 'background_label': 0, 'clean_kwargs': {}, 'detrend': False, 'dtype': None, 'high_pass': None, 'high_variance_confounds': False, 'keep_masked_labels': True, 'labels': None, 'labels_img': , 'low_pass': None, 'mask_img': , 'reports': True, 'smoothing_fwhm': 2.0, 'standardize': False, 'standardize_confounds': True, 'strategy': 'mean', 't_r': None, 'target_affine': None, 'target_shape': None}, confounds=None, sample_mask=None, dtype=None, memory=Memory(location=nilearn_cache/joblib), memory_level=1, verbose=1) [NiftiLabelsMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [NiftiLabelsMasker.transform_single_imgs] Smoothing images [NiftiLabelsMasker.transform_single_imgs] Extracting region signals [NiftiLabelsMasker.transform_single_imgs] Cleaning extracted signals _______________________________________________filter_and_extract - 2.1s, 0.0min ReNA 5000 clusters: 13.89s .. GENERATED FROM PYTHON SOURCE LINES 356-361 Visualize: Brain parcellations (ReNA) ..................................... First, we display the parcellations of the brain image stored in attribute `labels_img_` .. GENERATED FROM PYTHON SOURCE LINES 361-374 .. code-block:: Python rena_labels_img = rena.labels_img_ # Now, rena_labels_img are Nifti1Image object, it can be saved to file # with the following code: rena_labels_img.to_filename(output_dir / "rena_parcellation.nii.gz") plotting.plot_roi( ward_labels_img, title="ReNA parcellation", display_mode="xz", cut_coords=cut_coords, ) .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_007.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_007.png :class: sphx-glr-single-img .. rst-class:: sphx-glr-script-out .. code-block:: none .. GENERATED FROM PYTHON SOURCE LINES 375-384 Compressed representation of :term:`ReNA` clustering .................................................... We illustrate the effect that the clustering has on the signal. We show the original data, and the approximation provided by the clustering by averaging the signal on each parcel. We can then compare the results with the compressed representation obtained with Ward. .. GENERATED FROM PYTHON SOURCE LINES 384-415 .. code-block:: Python # Display the original data plotting.plot_epi( mean_func_img, cut_coords=cut_coords, title=f"Original ({int(original_voxels)} voxels)", vmax=vmax, vmin=vmin, display_mode="xz", ) # A reduced data can be created by taking the parcel-level average: # Note that, as many scikit-learn objects, the ``rena`` object exposes # a transform method that modifies input features. Here it reduces their # dimension. # However, the data are in one single large 4D image, we need to use # index_img to do the split easily: fmri_reduced_rena = rena.transform(dataset.func) # Display the corresponding data compression using the parcellation compressed_img_rena = rena.inverse_transform(fmri_reduced_rena) plotting.plot_epi( index_img(compressed_img_rena, 0), cut_coords=cut_coords, title="ReNA compressed representation (5000 parcels)", vmin=vmin, vmax=vmax, display_mode="xz", ) .. rst-class:: sphx-glr-horizontal * .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_008.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_008.png :class: sphx-glr-multi-img * .. image-sg:: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_009.png :alt: plot data driven parcellations :srcset: /auto_examples/03_connectivity/images/sphx_glr_plot_data_driven_parcellations_009.png :class: sphx-glr-multi-img .. rst-class:: sphx-glr-script-out .. code-block:: none [Parcellations.wrapped] loading data from Nifti1Image('/home/himanshu/Desktop/nilearn_work/nilearn/examples/03_connectivity/results/plot_data_driven_parcellations/rena_parcellation.nii.gz') [Parcellations.wrapped] loading data from Nifti1Image( shape=(50, 59, 50), affine=array([[ 4., 0., 0., -96.], [ 0., 4., 0., -132.], [ 0., 0., 4., -78.], [ 0., 0., 0., 1.]]) ) ________________________________________________________________________________ [Memory] Calling nilearn.maskers.base_masker._filter_and_extract... _filter_and_extract(, , { 'background_label': 0, 'clean_kwargs': {}, 'detrend': False, 'dtype': None, 'high_pass': None, 'high_variance_confounds': False, 'keep_masked_labels': True, 'labels': None, 'labels_img': , 'low_pass': None, 'mask_img': , 'reports': True, 'smoothing_fwhm': 2.0, 'standardize': False, 'standardize_confounds': True, 'strategy': 'mean', 't_r': None, 'target_affine': None, 'target_shape': None}, confounds=None, sample_mask=None, dtype=None, memory=Memory(location=nilearn_cache/joblib), memory_level=1, verbose=1) [NiftiLabelsMasker.transform_single_imgs] Loading data from Nifti1Image('/home/himanshu/nilearn_data/development_fmri/development_fmri/sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz') [NiftiLabelsMasker.transform_single_imgs] Smoothing images [NiftiLabelsMasker.transform_single_imgs] Extracting region signals [NiftiLabelsMasker.transform_single_imgs] Cleaning extracted signals _______________________________________________filter_and_extract - 2.2s, 0.0min .. GENERATED FROM PYTHON SOURCE LINES 416-422 Even if the compressed signal is relatively close to the original signal, we can notice that Ward Clustering gives a slightly more accurate compressed representation. However, as said in the previous section, the computation time is reduced which could still make :term:`ReNA` more relevant than Ward in some cases. .. GENERATED FROM PYTHON SOURCE LINES 424-428 References ---------- .. footbibliography:: .. rst-class:: sphx-glr-timing **Total running time of the script:** (1 minutes 24.008 seconds) **Estimated memory usage:** 2852 MB .. _sphx_glr_download_auto_examples_03_connectivity_plot_data_driven_parcellations.py: .. only:: html .. container:: sphx-glr-footer sphx-glr-footer-example .. container:: binder-badge .. image:: images/binder_badge_logo.svg :target: https://mybinder.org/v2/gh/nilearn/nilearn/0.10.4?urlpath=lab/tree/notebooks/auto_examples/03_connectivity/plot_data_driven_parcellations.ipynb :alt: Launch binder :width: 150 px .. container:: sphx-glr-download sphx-glr-download-jupyter :download:`Download Jupyter notebook: plot_data_driven_parcellations.ipynb ` .. container:: sphx-glr-download sphx-glr-download-python :download:`Download Python source code: plot_data_driven_parcellations.py ` .. only:: html .. rst-class:: sphx-glr-signature `Gallery generated by Sphinx-Gallery `_