Parcellation Schaefer 1000 - Support request

[Yvonneser]

Hey Team,

• What are you trying to do? I am using Brainspace to run PCA with Alignment on set of subjects in different sessions, I want to use Schaefer 1000 parcellation, As @OualidBenkarim describe previously in closed issue 17, I Download the relevant files Schaefer parcellation in fsa5.

• If you've tried running this with BrainSpace code already, how did you do so and what went wrong? I am trying to visualizing the Brains in separate hemi and views,

The labels that I have (That are not Zero ) = 18741 The fsaverage5 data lh/rh.inflated are 18715(9354 +9361)

This is the code that I used using Brain space:

Here is an example of the plot that I am trying to create and the output:

surf_labels_lh = nib.freesurfer.read_annot('parcellation/lh.Schaefer2018_1000Parcels_7Networks_order.annot')[0]
surf_labels_rh = nib.freesurfer.read_annot('parcellation/rh.Schaefer2018_1000Parcels_7Networks_order.annot')[0]
surf_labels_rh[surf_labels_rh != 0] += 500  # different labels for lh and rh
surf_labels = np.concatenate([surf_labels_lh, surf_labels_rh])

mask = surf_labels != 0  # discard medial wall
cort_lh = nib.freesurfer.read_label('parcellation/lh.cortex.label')
cort_rh = nib.freesurfer.read_label('parcellation/rh.cortex.label')
surf_lh_loc = nib.freesurfer.read_geometry('parcellation/lh.inflated')
surf_rh_loc = nib.freesurfer.read_geometry('parcellation/rh.inflated')
cort = np.concatenate((cort_lh, cort_rh + 10242))
""" 
#     Visualize group-level average components 
#     """

for v,se in enumerate(ses):
    for k, s in enumerate(subjs):
        n_comps = 5
        
        n_views = 4  # number of brains to show per component
        fig_width = 10
        
        eigenVal = []
        from nilearn.plotting import plot_surf_stat_map
        
        fig = plt.figure(figsize=(fig_width, fig_width * n_comps * 0.25), dpi=400)
        for component in range(n_comps):
            ax1 = fig.add_subplot(n_comps, n_views, component * n_views + 1, projection='3d')
            ax2 = fig.add_subplot(n_comps, n_views, component * n_views + 2, projection='3d')
            ax3 = fig.add_subplot(n_comps, n_views, component * n_views + 3, projection='3d')
            ax4 = fig.add_subplot(n_comps, n_views, component * n_views + 4, projection='3d')
            res = np.zeros(10242)
            res[cort_lh] =  np.asarray(split_sets[v][k])[component][cort_lh]
            plax1 = plot_surf_stat_map(surf_lh_loc, res, hemi='left', view='medial', axes=ax1, colorbar=False, cmap=mymap1)
            plax2 = plot_surf_stat_map(surf_lh_loc, res, hemi='left', view='lateral', axes=ax2, colorbar=False, cmap=mymap1)
            res = np.zeros(10242)
            res[cort_rh] = np.asarray(split_sets[v][k])[component][cort_rh]
        
            plax3 = plot_surf_stat_map(surf_rh_loc, res, hemi='right', view='lateral', axes=ax3, colorbar=False, cmap=mymap1)
            plax4 = plot_surf_stat_map(surf_rh_loc, res, hemi='right', view='medial', axes=ax4, colorbar=False, cmap=mymap1)
            plt.figtext(0.5, 0.96 - (1. / n_comps) * component, 'G %i' % int(int(component) + 1),
                        horizontalalignment='center', fontsize='large',fontweight='bold')
        
        plt.tight_layout()
        plt.subplots_adjust(hspace=0.1,wspace=0.05)
        filename = f'PCA-{s}-{se}-procrustes.png'
        plt.savefig(filename, dpi=400)
        plt.show()

  | ReplyForward -- | --

This is the output:

I am trying to understand why I am getting these "NAN" values in the plot, and how can I fix it , Is there something in the files I am using wring or the flow is not correct?

Thanks In advance , and Appreciate any help

Yvonne Serhan

• Other information

[Yvonneser]

Fixed Thanks