def createTFCEfMRIOverlayImages(folder,suffix,title='',vmax=8,display_mode='z',slices=range(-20,50,10),threshold=0.94999,plotToAxis=False,f=[],axes=[],colorbar=True,tight_layout=False,draw_cross=False,annotate=False):
TFCEposImg,posImg,TFCEnegImg,negImg=getFileNamesfromFolder(folder,suffix)
bg_img='./Templates/MNI152_.5mm_masked_edged.nii.gz'
# threshold=0.949
pos=image.math_img("np.multiply(img1,img2)",
img1=image.threshold_img(TFCEposImg,threshold=threshold),img2=posImg)
neg=image.math_img("np.multiply(img1,img2)",
img1=image.threshold_img(TFCEnegImg,threshold=threshold),img2=negImg)
fw=image.math_img("img1-img2",img1=pos,img2=neg)
if plotToAxis:
display=plotting.plot_stat_map(fw,display_mode=display_mode,threshold=0,
cut_coords=slices,vmax=vmax,colorbar=colorbar,
bg_img=bg_img,black_bg=False,title=title,dim=0,
figure=f,axes=axes,draw_cross=draw_cross,
annotate=annotate)
else:
display=plotting.plot_stat_map(fw,display_mode=display_mode,threshold=0,
cut_coords=slices,vmax=vmax,colorbar=colorbar,bg_img=bg_img,
black_bg=False,title=title,dim=0,annotate=annotate)
if tight_layout:
display.tight_layout()
return display
def stat_function_tst(conn, prefix='', OUTPUT_PATH=None, threshold=0.05):
fc = conn.hurst
tst = Parallel(n_jobs=3, verbose=5)(delayed(ttest_group)(group, threshold, fc)
for group in groups)
if OUTPUT_PATH is None:
font = {'family' : 'normal',
'size' : 20}
changefont('font', **font)
gr = ['v', 'av', 'avn']
for i in range(3):
title = prefix + '_'.join(groups[i])
try:
img = conn.masker.inverse_transform(tst[i])
print title
plot_stat_map(img, cut_coords=(3, -63, 36))
plt.show()
except ValueError:
print "problem with tst " + title
changefont.func_defaults
else:
for i in range(3):
title = prefix + '_'.join(groups[i])
output_file = os.path.join(OUTPUT_PATH, title)
try:
img = conn.masker.inverse_transform(tst[i])
plot_stat_map(img, cut_coords=(3, -63, 36), output_file=output_file + '.pdf')
except ValueError:
print "problem with tst " + title
def p_map(task, run, p_values_3d, threshold=0.05):
"""
Generate three thresholded p-value maps.
Parameters
----------
task: int
Task number
run: int
Run number
p_value_3d: 3D array of p_value.
threshold: The cutoff value to determine significant voxels.
Returns
-------
threshold p-value images
"""
fmri_img = image.smooth_img('../../../data/sub001/BOLD/' + 'task00' +
str(task) + '_run00' + str(run) +
'/filtered_func_data_mni.nii.gz',
fwhm=6)
mean_img = image.mean_img(fmri_img)
log_p_values = -np.log10(p_values_3d)
log_p_values[np.isnan(log_p_values)] = 0.
log_p_values[log_p_values > 10.] = 10.
log_p_values[log_p_values < -np.log10(threshold)] = 0
plot_stat_map(nib.Nifti1Image(log_p_values, fmri_img.get_affine()),
mean_img, title="Thresholded p-values",
annotate=False, colorbar=True)
def draw_brain_map(self):
cmap = plt.get_cmap('Accent')
self.fig = plt.figure('brain_map')
plot_stat_map(self.cluster_img, cut_coords=(0, 0, 0), output_file=None,
display_mode='ortho', colorbar=False, figure=self.fig,
axes=None, title=None, threshold=0.1, annotate=True,
draw_cross=False, black_bg='auto', symmetric_cbar="auto",
dim=True, vmax=None, cmap=cmap)
def qc_image_data(dataset, images, plot_dir='qc'):
# Get ready
masker = GreyMatterNiftiMasker(memory=Memory(cachedir='nilearn_cache')).fit()
if op.exists(plot_dir): # Delete old plots.
shutil.rmtree(plot_dir)
# Dataframe to contain summary metadata for neurovault images
if dataset == 'neurovault':
fetch_summary = pd.DataFrame(
columns=('Figure #', 'col_id', 'image_id', 'name',
'modality', 'map_type', 'analysis_level',
'is_thresholded', 'not_mni', 'brain_coverage',
'perc_bad_voxels', 'perc_voxels_outside'))
for ii, image in enumerate(images):
im_path = image['absolute_path']
if im_path is None:
continue
ri = ii % 4 # row i
ci = (ii / 4) % 4 # column i
pi = ii % 16 + 1 # plot i
fi = ii / 16 # figure i
if ri == 0 and ci == 0:
fh = plt.figure(figsize=(16, 10))
print('Plot %03d of %d' % (fi + 1, np.ceil(len(images) / 16.)))
ax = fh.add_subplot(4, 4, pi)
title = "%s%s" % (
'(X) ' if image['rejected'] else '', op.basename(im_path))
if dataset == 'neurovault':
fetch_summary.loc[ii] = [
'fig%03d' % (fi + 1), image.get('collection_id'),
image.get('id'), title, image.get('modality'),
image.get('map_type'), image.get('analysis_level'),
image.get('is_thresholded'), image.get('not_mni'),
image.get('brain_coverage'), image.get('perc_bad_voxels'),
image.get('perc_voxels_outside')]
# Images may fail to be transformed, and are of different shapes,
# so we need to trasnform one-by-one and keep track of failures.
img = cast_img(im_path, dtype=np.float32)
img = clean_img(img)
try:
img = masker.inverse_transform(masker.transform(img))
except Exception as e:
print("Failed to mask/reshape image %s: %s" % (title, e))
plot_stat_map(img, axes=ax, black_bg=True, title=title, colorbar=False)
if (ri == 3 and ci == 3) or ii == len(images) - 1:
out_path = op.join(plot_dir, 'fig%03d.png' % (fi + 1))
save_and_close(out_path)
# Save fetch_summary
if dataset == 'neurovault':
fetch_summary.to_csv(op.join(plot_dir, 'fetch_summary.csv'))
def plot_stat_map2(**kwargs):
cut_coords = kwargs['cut_coords']
row_l = kwargs['row_l']
lines_nb = int(len(cut_coords) / row_l)
for line in xrange(lines_nb):
opt = dict(kwargs)
opt.pop('row_l')
opt['cut_coords'] = cut_coords[line * row_l: (line +1) *row_l]
plotting.plot_stat_map(**opt)
def compute_hurst_and_stat(metric='dfa', regu='off', OUTPUT_PATH = '/volatile/hubert/beamer/test_hurst/', plot=False):
conn = Hurst_Estimator(metric=metric, mask=dataset.mask,smoothing_fwhm=0, regu=regu, n_jobs=5)
os.write(1,'fit\n')
fc = conn.fit(dataset.func1)
#conn.load_map(INPUT_PATH)
os.write(1,'save\n')
#stat_function_tst(conn, metric+' '+regu+' ', OUTPUT_PATH)
conn.save(save_path=OUTPUT_PATH)
if plot:
os.write(1,'plot\n')
a = Parallel(n_jobs=3, verbose=5)(delayed(classify_group)(group, fc)
for group in groups)
tst = Parallel(n_jobs=3, verbose=5)(delayed(ttest_group)(group, .05, fc)
for group in groups)
ost = Parallel(n_jobs=3, verbose=5)(delayed(ttest_onesample)(group, 0.05, fc)
for group in ['v', 'av', 'avn'])
mht = Parallel(n_jobs=3, verbose=5)(delayed(ttest_onesample_Hmean)(group, 0.05, fc)
for group in ['v', 'av', 'avn'])
mpt = Parallel(n_jobs=3, verbose=5)(delayed(mne_permutation_ttest)(group,0.05, fc, 1)
for group in ['v', 'av', 'avn'])
cot = Parallel(n_jobs=3, verbose=5)(delayed(ttest_onesample_coef)(np.reshape(coef['coef'], (coef['coef'].shape[0], coef['coef'].shape[-1])),
0.05, fc)
for coef in a)
gr = ['v', 'av', 'avn']
if regu=='off':
OUTPUT_PATH = os.path.join(OUTPUT_PATH, metric)
else:
OUTPUT_PATH = os.path.join(OUTPUT_PATH, metric, regu)
for i in range(3):
title = '_'.join(groups[i])
output_file = os.path.join(OUTPUT_PATH, title)
img = conn.masker.inverse_transform(tst[i])
plot_stat_map(img, cut_coords=(3, -63, 36), title=title, output_file=output_file + '.pdf')
img = conn.masker.inverse_transform(cot[i])
plot_stat_map(img, title='coef_map ' + title, output_file=output_file + 'coef_map.pdf')
title = gr[i]
output_file = os.path.join(OUTPUT_PATH, title)
img = conn.masker.inverse_transform(ost[i])
plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title, output_file= output_file + '.pdf')
img = conn.masker.inverse_transform(mht[i])
plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title, output_file= output_file + 'meanH.pdf')
img = conn.masker.inverse_transform(mpt[i])
plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title, output_file= output_file + 'mnepermutH.pdf')
plt.figure()
plt.boxplot(map(lambda x: x['accuracy'], a))
plt.savefig(os.path.join(OUTPUT_PATH, 'boxplot.pdf'))
def montage(img, thr=0, mode='coronal', rows=5, cloumns=6, fsz=(10, 20)):
"""
Make a montage using nilearn for the background
The output figure will be 5 slices wide and 6
slices deep
:param img: nilearn image containing the data
:param thr: threshold for the image
:param mode: view mode. saggital, coronal, axial
:param rows: number of rows in the figure
:param cloumns: number of columns in the figure
:param fsz: size of the figure
:return fig: figure handle for saving or whatnot
"""
# Hardwired view range
sag_rng = [-65, 65]
cor_rng = [-100, 65]
axi_rng = [-71, 85]
# Get the number of slices
n_slices = rows * cloumns
if mode == 'coronal':
# Get the slice indices
view_range = np.floor(np.linspace(cor_rng[0], cor_rng[1], n_slices))
view_mode = 'y'
if mode == 'axial':
# Get the slice indices
view_range = np.floor(np.linspace(axi_rng[0], axi_rng[1], n_slices))
view_mode = 'z'
if mode == 'saggital':
# Get the slice indices
view_range = np.floor(np.linspace(sag_rng[0], sag_rng[1], n_slices))
view_mode = 'x'
# Prepare the figure
fig = plt.figure(figsize=fsz)
gs = gridspec.GridSpec(cloumns, 1, hspace=0, wspace=0)
# Loop through the rows of the image
for row_id in range(cloumns):
# Create the axis to show
ax = fig.add_subplot(gs[row_id, 0])
# Get the slices in the column direction
row_range = view_range[row_id*rows:(row_id+1)*rows]
# Display the thing
nlp.plot_stat_map(img, cut_coords=row_range,
display_mode=view_mode, threshold=thr,
axes=ax, black_bg=True)
return fig
def run(idx, reduction, alpha, mask, raw, n_components, init, func_filenames):
output_dir = join(trace_folder, 'experiment_%i' % idx)
try:
os.makedirs(output_dir)
except OSError:
pass
dict_fact = SpcaFmri(mask=mask,
smoothing_fwhm=3,
batch_size=40,
shelve=not raw,
n_components=n_components,
replacement=False,
dict_init=fetch_atlas_smith_2009().rsn70 if
init else None,
reduction=reduction,
alpha=alpha,
random_state=0,
n_epochs=2,
l1_ratio=0.5,
backend='c',
memory=expanduser("~/nilearn_cache"), memory_level=2,
verbose=5,
n_jobs=1,
trace_folder=output_dir
)
print('[Example] Learning maps')
t0 = time.time()
dict_fact.fit(func_filenames, raw=raw)
t1 = time.time() - t0
print('[Example] Dumping results')
# Decomposition estimator embeds their own masker
masker = dict_fact.masker_
components_img = masker.inverse_transform(dict_fact.components_)
components_img.to_filename(join(output_dir, 'components_final.nii.gz'))
print('[Example] Run in %.2f s' % t1)
# Show components from both methods using 4D plotting tools
import matplotlib.pyplot as plt
from nilearn.plotting import plot_prob_atlas, show
print('[Example] Displaying')
fig, axes = plt.subplots(2, 1)
plot_prob_atlas(components_img, view_type="filled_contours",
axes=axes[0])
plot_stat_map(index_img(components_img, 0),
axes=axes[1],
colorbar=False,
threshold=0)
plt.savefig(join(output_dir, 'components.pdf'))
show()
def diff_computed_hurst(metric='wavelet', regu='off', INPUT_PATH = '/volatile/hubert/beamer/test_hurst/', OUTPUT_PATH=''):
conn = Hurst_Estimator(metric=metric, mask=dataset.mask, regu=regu, n_jobs=5)
os.write(1,'load\n')
conn.load_map(INPUT_PATH)
fc = conn.hurst
os.write(1,'stat\n')
tst = ttest_group(['av', 'v'], .05, fc)
vmean_avmean = np.mean([fc[i] for i in dataset.group_indices['v']], axis=0) - np.mean([fc[i] for i in dataset.group_indices['av']], axis=0)
vmean_avmean[tst == 0] = 0
img = conn.masker.inverse_transform(vmean_avmean)
plot_stat_map(img)
plt.show()
def plot_contrast(first_level_model):
""" Given a first model, specify, enstimate and plot the main contrasts"""
design_matrix = first_level_model.design_matrices_[0]
# Call the contrast specification within the function
contrasts = make_localizer_contrasts(design_matrix)
fig = plt.figure(figsize=(11, 3))
# compute the per-contrast z-map
for index, (contrast_id, contrast_val) in enumerate(contrasts.items()):
ax = plt.subplot(1, len(contrasts), 1 + index)
z_map = first_level_model.compute_contrast(
contrast_val, output_type='z_score')
plotting.plot_stat_map(
z_map, display_mode='z', threshold=3.0, title=contrast_id, axes=ax,
cut_coords=1)
def make_thresholded_slices(regions, colors, display_mode='z', overplot=True, binarize=True, **kwargs):
""" Plots on axial slices numerous images
regions: Nibabel images
colors: List of colors (rgb tuples)
overplot: Overlay images?
binarize: Binarize images or plot full stat maps
"""
from matplotlib.colors import LinearSegmentedColormap
from nilearn import plotting as niplt
if binarize:
for reg in regions:
reg.get_data()[reg.get_data().nonzero()] = 1
for i, reg in enumerate(regions):
reg_color = LinearSegmentedColormap.from_list('reg1', [colors[i], colors[i]])
if i == 0:
plot = niplt.plot_stat_map(reg, draw_cross=False, display_mode=display_mode, cmap = reg_color, alpha=0.9, colorbar=False, **kwargs)
else:
if overplot:
plot.add_overlay(reg, cmap = reg_color, alpha=.72)
else:
plt.plot_stat_map(reg, draw_cross=False, display_mode=display_mode, cmap = reg_color, colorbar=False, **kwargs)
return plot
def make_stat_image(nifti_file,png_img_file=None):
"""Make statmap image"""
nifti_file = str(nifti_file)
brain = plot_stat_map(nifti_file)
if png_img_file:
brain.savefig(png_img_file)
plt.close('all')
return brain
def ica_vis(subj_num): # Use the mean as a background mean_img_1 = image.mean_img(BOLD_file_1) mean_img_2 = image.mean_img(BOLD_file_2) mean_img_3 = image.mean_img(BOLD_file_3) plot_stat_map(image.index_img(component_img_1, 5), mean_img_1, output_file=os.path.join(data_path,'sub'+subj_num+'_BOLD','task001_run001'+'ica_1'+'.jpg')) plot_stat_map(image.index_img(component_img_1, 12), mean_img_1, output_file=os.path.join(data_path,'sub'+subj_num+'_BOLD','task001_run001'+'ica_2'+'.jpg')) plot_stat_map(image.index_img(component_img_2, 5), mean_img_2, output_file=os.path.join(data_path,'sub'+subj_num+'_BOLD','task002_run001'+'ica_1'+'.jpg')) plot_stat_map(image.index_img(component_img_2, 12), mean_img_2, output_file=os.path.join(data_path,'sub'+subj_num+'_BOLD','task002_run001'+'ica_2'+'.jpg'))
def plot_stat_overlay(stat_img, contour_img, bg_img, **kwargs):
"""Plot over bg_img a stat_img and the countour."""
import nilearn.plotting as niplot
if bg_img is not None:
kwargs['bg_img'] = bg_img
display = niplot.plot_stat_map(stat_img, **kwargs)
display.add_contours(contour_img, filled=True, alpha=0.6, levels=[0.5], colors='g')
return display
def dump_comps(masker, compressor, components, threshold=2, fwhm=None,
perc=None):
from scipy.stats import zscore
from nilearn.plotting import plot_stat_map
from nilearn.image import smooth_img
from scipy.stats import scoreatpercentile
if isinstance(compressor, basestring):
comp_name = compressor
else:
comp_name = compressor.__str__().split('(')[0]
for i_c, comp in enumerate(components):
path_mask = op.join(WRITE_DIR, '%s_%i-%i' % (comp_name,
n_comp, i_c + 1))
nii_raw = masker.inverse_transform(comp)
nii_raw.to_filename(path_mask + '.nii.gz')
comp_z = zscore(comp)
if perc is not None:
cur_thresh = scoreatpercentile(np.abs(comp_z), per=perc)
path_mask += '_perc%i' % perc
print('Applying percentile %.2f (threshold: %.2f)' % (perc, cur_thresh))
else:
cur_thresh = threshold
path_mask += '_thr%.2f' % cur_thresh
print('Applying threshold: %.2f' % cur_thresh)
nii_z = masker.inverse_transform(comp_z)
gz_path = path_mask + '_zmap.nii.gz'
nii_z.to_filename(gz_path)
plot_stat_map(gz_path, bg_img='colin.nii', threshold=cur_thresh,
cut_coords=(0, -2, 0), draw_cross=False,
output_file=path_mask + 'zmap.png')
# optional: do smoothing
if fwhm is not None:
nii_z_fwhm = smooth_img(nii_z, fwhm=fwhm)
plot_stat_map(nii_z_fwhm, bg_img='colin.nii', threshold=cur_thresh,
cut_coords=(0, -2, 0), draw_cross=False,
output_file=path_mask +
('zmap_%imm.png' % fwhm))
def plot_components(ica_image, hemi='', out_dir=None,
bg_img=datasets.load_mni152_template()):
print("Plotting %s components..." % hemi)
# Determine threshoold and vmax for all the plots
# get nonzero part of the image for proper thresholding of
# r- or l- only component
nonzero_img = ica_image.get_data()[np.nonzero(ica_image.get_data())]
thr = stats.scoreatpercentile(np.abs(nonzero_img), 90)
vmax = stats.scoreatpercentile(np.abs(nonzero_img), 99.99)
for ci, ic_img in enumerate(iter_img(ica_image)):
title = _title_from_terms(terms=ica_image.terms, ic_idx=ci, label=hemi)
fh = plt.figure(figsize=(14, 6))
plot_stat_map(ic_img, axes=fh.gca(), threshold=thr, vmax=vmax,
colorbar=True, title=title, black_bg=True, bg_img=bg_img)
# Save images instead of displaying
if out_dir is not None:
save_and_close(out_path=op.join(
out_dir, '%s_component_%i.png' % (hemi, ci)))
def dump_comps(masker, compressor, components, threshold=2):
from scipy.stats import zscore
from nilearn.plotting import plot_stat_map
if isinstance(compressor, basestring):
comp_name = compressor
else:
comp_name = compressor.__str__().split('(')[0]
for i_c, comp in enumerate(components):
path_mask = op.join(WRITE_DIR, '%s_%i-%i' % (comp_name,
n_comp, i_c + 1))
nii_raw = masker.inverse_transform(comp)
nii_raw.to_filename(path_mask + '.nii.gz')
nii_z = masker.inverse_transform(zscore(comp))
gz_path = path_mask + '_zmap.nii.gz'
nii_z.to_filename(gz_path)
plot_stat_map(gz_path, bg_img='colin.nii', threshold=threshold,
cut_coords=(0, -2, 0), draw_cross=False,
output_file=path_mask + 'zmap.png')
def main(stat_map="nii/tstat.nii.gz", template="~/NIdata/templates/medres_QBI_chr.nii.gz", black_bg=False, cut_coords=(-50,8,45)):
template = path.expanduser(template)
colors_plus = plt.cm.autumn(np.linspace(0., 1, 128))
colors_minus = plt.cm.winter(np.linspace(0, 1, 128))
colors = np.vstack((colors_minus, colors_plus[::-1]))
mymap = mcolors.LinearSegmentedColormap.from_list('my_colormap', colors)
for i in ['none', 'nearest', 'bilinear', 'bicubic', 'spline16','spline36', 'hanning', 'hamming', 'hermite', 'kaiser', 'quadric','catrom', 'gaussian', 'bessel', 'mitchell', 'sinc', 'lanczos']:
display = plotting.plot_stat_map(stat_map, bg_img=template,threshold=2.5, vmax=40, cmap=mymap, black_bg=black_bg, cut_coords=cut_coords, annotate=True, title=i+" interpolation", draw_cross=False, interpolation=i)
plt.savefig(i, dpi=700)
def generate_images(components_img, n_components, images_dir, glass=False):
# Remove existing images
if os.path.exists(images_dir):
shutil.rmtree(images_dir)
os.makedirs(images_dir)
output_filenames = [osp.join(images_dir, 'IC_{}.png'.format(i))
for i in range(n_components)]
for i, output_file in enumerate(output_filenames):
plot_stat_map(nibabel.Nifti1Image(components_img.get_data()[..., i],
components_img.get_affine()),
display_mode="z", title="IC %d" % i, cut_coords=7,
colorbar=False, output_file=output_file)
if glass:
output_filenames = [osp.join(images_dir, 'glass_IC_{}.png'.format(i))
for i in range(n_components)]
for i, output_file in enumerate(output_filenames):
plot_glass_brain(nibabel.Nifti1Image(
components_img.get_data()[..., i],
components_img.get_affine()),
display_mode="ortho", title="IC %d" % i,
output_file=output_file)
def make_plot(mr, color, png_image, black_bg=False):
"""make_plot
:param mr: a nibabel.Nifti1Image
:param color: a matplotlib color map
:param png_image: file name for output png
:black_bg: True saves image with black background (default False)
"""
plot_stat_map(
mr,
display_mode="z",
colorbar=False,
annotate=False,
draw_cross=False,
cmap=color,
cut_coords=1,
black_bg=black_bg,
)
if black_bg == False:
plt.savefig(png_image)
else:
plt.savefig(png_image, facecolor="k", edgecolor="k")
plt.close()
def stat_ttest_function(conn, prefix='', OUTPUT_PATH=None):
fc = conn.hurst
ost = Parallel(n_jobs=3, verbose=5)(delayed(ttest_onesample)(group, 0.05, fc)
for group in ['v', 'av', 'avn'])
mht = Parallel(n_jobs=3, verbose=5)(delayed(ttest_onesample_Hmean)(group, 0.05, fc)
for group in ['v', 'av', 'avn'])
gr = ['v', 'av', 'avn']
if OUTPUT_PATH is None:
for i in range(3):
title = prefix + gr[i]
try:
img = conn.masker.inverse_transform(ost[i])
plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title)
except ValueError:
print "problem with ost " + title
try:
img = conn.masker.inverse_transform(mht[i])
plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title)
except ValueError:
print "problem with mht " + title
else:
for i in range(3):
title = prefix + gr[i]
output_file = os.path.join(OUTPUT_PATH, title)
try:
img = conn.masker.inverse_transform(ost[i])
#plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title, output_file= output_file + '.pdf')
plot_stat_map(img, output_file= output_file + '.pdf')
except ValueError:
print "problem with ost " + title
try:
img = conn.masker.inverse_transform(mht[i])
#plot_stat_map(img, title='t-test H0 : H = 0.5 pvalue in -log10 scale groupe : ' + title, output_file= output_file + 'meanH.pdf')
plot_stat_map(img, output_file= output_file + 'meanH.pdf')
except ValueError:
print "problem with mht " + title
def _generate_thumbnail(build_dir, img, cut_coords):
threshold = np.percentile(nb.load(img).get_data(), 97)
display = plot_stat_map(img, threshold=threshold,
cut_coords=cut_coords, display_mode='z',
annotate=False, colorbar=False,
draw_cross=False, black_bg=False)
study_id = img.split(os.path.sep)[-2]
task_id, map_id = os.path.split(img)[-1].split('.nii.gz')[0].split('_', 1)
fname = '%s_%s_%s.png' % (
img.split(os.path.sep)[-2],
os.path.split(img)[-1].split('.nii.gz')[0], cut_coords)
fname = os.path.join(build_dir, 'thumbnails', fname)
display.savefig(fname, dpi=200)
return study_id, task_id, map_id, fname
def plot_components_summary(ica_image, hemi='', out_dir=None,
bg_img=datasets.load_mni152_template()):
print("Plotting %s components summary..." % hemi)
n_components = ica_image.get_data().shape[3]
# Determine threshoold and vmax for all the plots
# get nonzero part of the image for proper thresholding of
# r- or l- only component
nonzero_img = ica_image.get_data()[np.nonzero(ica_image.get_data())]
thr = stats.scoreatpercentile(np.abs(nonzero_img), 90)
vmax = stats.scoreatpercentile(np.abs(nonzero_img), 99.99)
for ii, ic_img in enumerate(iter_img(ica_image)):
ri = ii % 5 # row i
ci = (ii / 5) % 5 # column i
pi = ii % 25 + 1 # plot i
fi = ii / 25 # figure i
if ri == 0 and ci == 0:
fh = plt.figure(figsize=(30, 20))
print('Plot %03d of %d' % (fi + 1, np.ceil(n_components / 25.)))
ax = fh.add_subplot(5, 5, pi)
title = _title_from_terms(terms=ica_image.terms, ic_idx=ii, label=hemi)
colorbar = ci == 4
plot_stat_map(
ic_img, axes=ax, threshold=thr, vmax=vmax, colorbar=colorbar,
title=title, black_bg=True, bg_img=bg_img)
if (ri == 4 and ci == 4) or ii == n_components - 1:
out_path = op.join(
out_dir, '%s_components_summary%02d.png' % (hemi, fi + 1))
save_and_close(out_path)
def plot_diff_avg(scores, threshold=0.01, coords=None, cross=False, **kwargs):
'''plots subject scoremap using nilearn and returns display object'''
mask = spenc_dir+'temporal_lobe_mask_grp_7T_test.nii.gz'
background_img = '/home/data/psyinf/forrest_gump/anondata/templates/' + \
'grpbold7Tp1/brain.nii.gz'
scores = scores.copy()
scores[np.abs(scores)<threshold] = 0.0
unmasked = unmask(scores, mask)
unmasked = threshold_img(unmasked, 0.001)
display = plot_stat_map(
unmasked, cut_coords=coords, bg_img=background_img,
title='metric per voxel', dim=-1, aspect=1.25,
threshold=0.001, draw_cross=cross, **kwargs)
fig = plt.gcf()
fig.set_size_inches(12, 4)
return display
def plot_subj_ko(subj, scores, threshold=0.01, coords=None):
'''plots subject scoremap using nilearn and returns display object'''
subj_mask = './masks/template_mask_thick.nii.gz'
background_img = os.path.join(DATA_DIR, 'templates', 'grpbold7Tp1', 'brain.nii.gz')
scores = scores.copy()
scores[scores<threshold] = 0.0
unmasked = unmask(scores, subj_mask)
unmasked = threshold_img(unmasked, 0.001)
display = plot_stat_map(
unmasked, cut_coords=coords, bg_img=background_img,
symmetric_cbar=False,
title='metric per voxel', dim=-1, aspect=1.25,
threshold=0.001, draw_cross=False)
fig = plt.gcf()
fig.set_size_inches(12, 4)
return display
def plot_subj(subj, scores, threshold=0.01, coords=None):
'''plots subject scoremap using nilearn and returns display object'''
subj_mask = spenc_dir+'temporal_lobe_mask_brain_subj{0:02}bold.nii.gz'.format(subj)
background_img = '/home/data/psyinf/forrest_gump/anondata/sub{0:03}/'.format(subj)+\
'templates/bold7Tp1/brain.nii.gz'
scores = scores.copy()
scores[scores<threshold] = 0.0
unmasked = unmask(scores, subj_mask)
unmasked = threshold_img(unmasked, 0.001)
display = plot_stat_map(
unmasked, cut_coords=coords, bg_img=background_img,
symmetric_cbar=False,
title='metric per voxel', dim=-1, aspect=1.25,
threshold=0.001, draw_cross=False)
fig = plt.gcf()
fig.set_size_inches(12, 4)
return display
def _save_plot(self, predictor):
""" Save Plots.
Args:
predictor: predictor instance
Returns:
predicter_weightmap_montage.png: Will output a montage of axial slices of weightmap
predicter_prediction.png: Will output a plot of prediction
"""
if not os.path.isdir(self.output_dir):
os.makedirs(self.output_dir)
if self.algorithm == 'lassopcr':
coef = np.dot(self._pca.components_.T,self._lasso.coef_)
coef_img = self.nifti_masker.inverse_transform(np.transpose(coef))
elif self.algorithm == 'pcr':
coef = np.dot(self._pca.components_.T,self._regress.coef_)
coef_img = self.nifti_masker.inverse_transform(np.transpose(coef))
else:
coef_img = self.nifti_masker.inverse_transform(predictor.coef_)
overlay_img = nib.load(os.path.join(get_resource_path(),'MNI152_T1_2mm_brain.nii.gz'))
fig1 = plot_stat_map(coef_img, overlay_img, title=self.algorithm + " weights",
cut_coords=range(-40, 40, 10), display_mode='z')
fig1.savefig(os.path.join(self.output_dir, self.algorithm + '_weightmap_axial.png'))
if self.prediction_type == 'classification':
if self.algorithm not in ['svm','ridgeClassifier','ridgeClassifierCV']:
fig2 = probability_plot(self.stats_output)
fig2.savefig(os.path.join(self.output_dir, self.algorithm + '_prob_plot.png'))
else:
fig2 = dist_from_hyperplane_plot(self.stats_output)
fig2.savefig(os.path.join(self.output_dir, self.algorithm +
'_Distance_from_Hyperplane_xval.png'))
if self.algorithm == 'svm' and self.predictor.probability:
fig3 = probability_plot(self.stats_output)
fig3.savefig(os.path.join(self.output_dir, self.algorithm + '_prob_plot.png'))
elif self.prediction_type == 'prediction':
fig2 = scatterplot(self.stats_output)
fig2.savefig(os.path.join(self.output_dir, self.algorithm + '_scatterplot.png'))
def plot_ica_components(components_img, **kwargs):
""" Plot the components IC spatial maps in a grid."""
import math
from nilearn.image import iter_img
from nilearn.plotting import plot_stat_map
from matplotlib import pyplot as plt
from matplotlib import gridspec
n_ics = len(list(iter_img(components_img)))
n_rows = math.ceil(n_ics/2)
fig = plt.figure(figsize=(6, 3*n_rows), facecolor='black')
gs = gridspec.GridSpec(n_rows, 2)
plots = []
for i, ic_img in enumerate(iter_img(components_img)):
ax = plt.subplot(gs[i])
p = plot_stat_map(ic_img, display_mode="z", title="IC {}".format(i+1),
cut_coords=1, colorbar=False, figure=fig, axes=ax, **kwargs)
plots.append(p)
for p in plots:
p.close()
return fig
# In[14]:
fig, axes = plt.subplots(nrows=5, ncols=2, figsize=(12, 10))
topic_img_4d = neurosynth_dset_first_500.masker.inverse_transform(
gclda_model.p_voxel_g_topic_.T)
# Plot first ten topics
topic_counter = 0
for i_row in range(5):
for j_col in range(2):
topic_img = image.index_img(topic_img_4d, index=topic_counter)
display = plotting.plot_stat_map(
topic_img,
annotate=False,
cmap="Reds",
draw_cross=False,
figure=fig,
axes=axes[i_row, j_col],
)
axes[i_row, j_col].set_title(f"Topic {str(topic_counter).zfill(3)}")
topic_counter += 1
colorbar = display._cbar
colorbar_ticks = colorbar.get_ticks()
if colorbar_ticks[0] < 0:
new_ticks = [colorbar_ticks[0], 0, colorbar_ticks[-1]]
else:
new_ticks = [colorbar_ticks[0], colorbar_ticks[-1]]
colorbar.set_ticks(new_ticks, update_ticks=True)
glue("figure_gclda_topics", fig, display=False)
def plot_t_brain(objIn,
how="full",
thr="unc",
alpha=None,
nperm=None,
cut_coords=[],
**kwargs):
"""
Takes a brain data object and computes a 1 sample t-test across it's first axis. If a list is provided will compute difference between brain data objects in list (i.e. paired samples t-test).
Args:
objIn:(list/Brain_Data) if list will compute difference map first
how: (list) whether to plot a glass brain 'glass', 3 view-multi-slice mni 'mni', or both 'full'
thr: (str) what method to use for multiple comparisons correction unc, fdr, or tfce
alpha: (float) p-value threshold
nperm: (int) number of permutations for tcfe; default 1000
cut_coords: (list) x,y,z coords to plot brain slice
kwargs: optionals args to nilearn plot functions (e.g. vmax)
"""
if thr not in ["unc", "fdr", "tfce"]:
raise ValueError("Acceptable threshold methods are 'unc','fdr','tfce'")
views = ["x", "y", "z"]
if len(cut_coords) == 0:
cut_coords = [
range(-40, 50, 10),
[-88, -72, -58, -38, -26, 8, 20, 34, 46],
[-34, -22, -10, 0, 16, 34, 46, 56, 66],
]
else:
if len(cut_coords) != 3:
raise ValueError(
"cut_coords must be a list of coordinates like [[xs],[ys],[zs]]"
)
cmap = "RdBu_r"
if isinstance(objIn, list):
if len(objIn) == 2:
obj = objIn[0] - objIn[1]
else:
raise ValueError("Contrasts should contain only 2 list items!")
thrDict = {}
if thr == "tfce":
thrDict["permutation"] = thr
if nperm is None:
nperm = 1000
thrDict["n_permutations"] = nperm
print("1-sample t-test corrected using: TFCE w/ %s permutations" %
nperm)
else:
if thr == "unc":
if alpha is None:
alpha = 0.001
thrDict[thr] = alpha
print("1-sample t-test uncorrected at p < %.3f " % alpha)
elif thr == "fdr":
if alpha is None:
alpha = 0.05
thrDict[thr] = alpha
print("1-sample t-test corrected at q < %.3f " % alpha)
else:
thrDict = None
print("1-sample test unthresholded")
out = objIn.ttest(threshold_dict=thrDict)
if thrDict is not None:
obj = out["thr_t"]
else:
obj = out["t"]
if how == "full":
plot_glass_brain(obj.to_nifti(),
display_mode="lzry",
colorbar=True,
cmap=cmap,
plot_abs=False,
**kwargs)
for v, c in zip(views, cut_coords):
plot_stat_map(obj.to_nifti(),
cut_coords=c,
display_mode=v,
cmap=cmap,
bg_img=resolve_mni_path(MNI_Template)["brain"],
**kwargs)
elif how == "glass":
plot_glass_brain(obj.to_nifti(),
display_mode="lzry",
colorbar=True,
cmap=cmap,
plot_abs=False,
**kwargs)
elif how == "mni":
for v, c in zip(views, cut_coords):
plot_stat_map(obj.to_nifti(),
cut_coords=c,
display_mode=v,
cmap=cmap,
bg_img=resolve_mni_path(MNI_Template)["brain"],
**kwargs)
del obj
del out
return
haxby_func_filename = haxby_dataset.func[0]
# one motor contrast map from NeuroVault
motor_images = datasets.fetch_neurovault_motor_task()
stat_img = motor_images.images[0]
###############################################################################
# Plotting statistical maps with function `plot_stat_map`
# --------------------------------------------------------
from nilearn import plotting
# Visualizing t-map image on EPI template with manual
# positioning of coordinates using cut_coords given as a list
plotting.plot_stat_map(stat_img,
threshold=3,
title="plot_stat_map",
cut_coords=[36, -27, 66])
###############################################################################
# Making interactive visualizations with function `view_img`
# ----------------------------------------------------------
# An alternative to :func:`nilearn.plotting.plot_stat_map` is to use
# :func:`nilearn.plotting.view_img` that gives more interactive
# visualizations in a web browser. See :ref:`interactive-stat-map-plotting`
# for more details.
view = plotting.view_img(stat_img, threshold=3)
# In a Jupyter notebook, if ``view`` is the output of a cell, it will
# be displayed below the cell
view
def plot_n_save_3plane(in_dict):
'''
in_dict fields:
in_dir
in_fn
cut_coords
do_save (boolean)
formats_used (a list-like object)
area (brain area, for naming the output file)
cmap (optional): a matplotlib colorbar object
symmetric_cbar (optional): whether to treat cmap as symetrical
'''
if 'symmetric_cbar' not in in_dict.keys():
in_dict['symmetric_cbar'] = 'auto'
if 'cmap' not in in_dict.keys():
in_dict['cmap'] = mymap
if in_dict['symmetric_cbar'] == 'auto':
in_dict['cmap'] = cm.autumn
if 'vmax' not in in_dict.keys():
in_dict['vmax'] = None
if 'draw_cross' not in in_dict.keys():
in_dict['draw_cross'] = True
if 'cluster_dict' not in in_dict.keys():
cluster_dict = {}
else:
cluster_dict = in_dict['cluster_dict']
cluster_dict['in_nii_fn'] = opj(in_dict['in_dir'], in_dict['in_fn'])
if cluster_dict:
thresholder = NiiThresholder(**cluster_dict)
used_fn = thresholder.make_thresholded_image()
else:
used_fn = opj(in_dict['in_dir'], in_dict['in_fn'])
pv = plotting.plot_stat_map(
used_fn,
black_bg=True,
bg_img=(os.getenv('FSLDIR') +
'/data/standard/MNI152_T1_1mm_brain.nii.gz'),
threshold=in_dict['threshold'],
cut_coords=in_dict['cut_coords'],
cmap=in_dict['cmap'],
symmetric_cbar=in_dict['symmetric_cbar'],
draw_cross=in_dict['draw_cross'],
vmax=in_dict['vmax'])
set_fig_bg_contrast(pv)
plt.show()
if in_dict['do_save']:
for fmt in in_dict['formats_used']:
pv.savefig(
opj(
in_dict['out_dir'], in_dict['in_fn'].split('.')[0] + '_' +
in_dict['area'] + '.' + fmt))
try:
thresholder.clean_out_fn()
except NameError:
pass
return pv
def plot_spm(
zmaps,
roi_dict,
bg_img=None,
z_threshold=0,
f=None,
axes=None,
# brain_mask='../Templates/mni_icbm152_nlin_asym_09c_nifti/mni_icbm152_nlin_asym_09c.nii.gz',
roi_to_plot=('PreSMA', 'M1', 'ACC', 'rIFG', 'STR', 'GPe', 'GPi',
'STN'),
cut_coords=[None, None, None, None, None, None, None, None],
contrasts=('failed_stop - go_trial', 'successful_stop - go_trial',
'failed_stop - successful_stop'),
plot_columns=(0, 1, 3, 4, 6, 7),
empty_plots=False,
skip_all_but_last=False,
**kwargs):
if f is None:
gridspec = dict(hspace=0.0,
wspace=0.0,
width_ratios=[1, 1, 0.05, 1, 1, .05, 1, 1, .1])
f, axes = plt.subplots(
len(roi_to_plot), len(zmaps) + 3, gridspec_kw=gridspec
) # add 3 columns: 2 interspace, 1 on the right for the colorbar
if empty_plots:
f.set_size_inches(len(zmaps) * 4, len(roi_to_plot) * 4)
return f, axes
all_cut_coords = []
all_disps = []
for row_n, roi in enumerate(roi_to_plot):
# for debugging
if skip_all_but_last:
if row_n < (len(roi_to_plot) - 1):
continue
# get cut coordinates based on 1 hemisphere (if applicable)
if roi in ['STR', 'STN', 'PreSMA', 'GPe', 'GPi']:
roi_map = roi_dict['l' + roi]
else:
roi_map = roi_dict[roi]
# roi_map = make_conjunction_mask(roi_map['fn'], brain_mask)
if roi == 'rIFG':
## saggital
if cut_coords[row_n] is None:
this_cut_coords = plotting.find_xyz_cut_coords(
roi_map['fn'])[0:1]
else:
this_cut_coords = cut_coords[row_n]
display_mode = 'x'
plot_rois = ['rIFG'] #, 'M1', 'rPreSMA']
elif roi == 'STR':
## axial view
if cut_coords[row_n] is None:
this_cut_coords = plotting.find_xyz_cut_coords(
roi_map['fn'])[2:3]
else:
this_cut_coords = cut_coords[row_n]
display_mode = 'z'
plot_rois = [
'rIFG', 'M1', 'lSTR', 'lGPe', 'lGPi', 'lSTN', 'rSTR', 'rGPe',
'rGPi', 'rSTN'
]
elif roi == 'STN':
## plot coronal view
if cut_coords[row_n] is None:
this_cut_coords = plotting.find_xyz_cut_coords(
roi_map['fn'])[1:2]
else:
this_cut_coords = cut_coords[row_n]
display_mode = 'y'
plot_rois = [
'rIFG', 'M1', 'lSTR', 'lGPe', 'lGPi', 'lSTN', 'rSTR', 'rGPe',
'rGPi', 'rSTN'
]
all_cut_coords.append({display_mode: this_cut_coords[0]})
# loop over contrasts for columns
for col_n, map_n in zip(plot_columns, np.arange(len(zmaps))):
zmap = zmaps[map_n]
if skip_all_but_last:
if col_n < (len(zmaps) - 1):
continue
if row_n == (len(roi_to_plot) - 1) and col_n == (len(zmaps) - 1):
# plot colobar in the last plot
cbar = False
else:
cbar = False
# # do not plot in column 2 or 5
# plot_col = col_n
# if col_n > 1:
# plot_col = col_n + 1
# if col_n > 3:
# plot_col = col_n + 2
if isinstance(z_threshold, list):
this_threshold = z_threshold[map_n]
else:
this_threshold = z_threshold
ax = axes[row_n, col_n]
# print(cbar)
disp = plotting.plot_stat_map(zmap,
bg_img=bg_img,
threshold=this_threshold,
cut_coords=this_cut_coords,
display_mode=display_mode,
axes=ax,
colorbar=cbar,
**kwargs)
# just plot *all* contours, always
for roi_ in plot_rois:
roi_map = roi_dict[roi_]
# for roi_, roi_map in roi_dict.items():
# print(roi_map)
add_contours(disp,
roi=roi_map['fn'],
thr=roi_map['threshold'],
color=roi_map['color'])
# determine limits (xlim/ylim) based on first column, and apply to all others
this_key = list([x for x in disp.axes.keys()])[0]
# Determine new xlim/ylim based on first column
if col_n == plot_columns[0]:
# extract old/current limits
cur_xlim = disp.axes[this_key].ax.get_xlim()
cur_ylim = disp.axes[this_key].ax.get_ylim()
if display_mode == 'x':
new_xlim = get_prop_limits([0, 1], cur_xlim)
new_ylim = get_prop_limits([0, 1], cur_ylim)
elif display_mode == 'z' and 'STN' in roi:
new_xlim = get_prop_limits([.25, .75], cur_xlim)
new_ylim = get_prop_limits([.40, .90], cur_ylim)
elif display_mode == 'z' and 'STR' in roi:
new_xlim = get_prop_limits([0, 1], cur_xlim)
new_ylim = get_prop_limits([0.3, 1], cur_ylim)
elif display_mode == 'y':
new_xlim = get_prop_limits([.26, .74], cur_xlim)
new_ylim = get_prop_limits([.25, .75], cur_ylim)
# Change axes limits
disp.axes[this_key].ax.set_xlim(new_xlim[0], new_xlim[1])
disp.axes[this_key].ax.set_ylim(new_ylim[0], new_ylim[1])
all_disps.append(disp)
# # set new xlimits if necessary (ie zoom for STN view)
# if 'STN' in roi and display_mode == 'z':
# this_key = [x for x in disp.axes.keys()]
# this_key = this_key[0]
# cur_xlim = disp.axes[this_key].ax.get_xlim()
# cur_ylim = disp.axes[this_key].ax.get_ylim()
# new_xlim = get_prop_limits([.25, .75], cur_xlim)
# new_ylim = get_prop_limits([.40, .90], cur_ylim)
# disp.axes[this_key].ax.set_xlim(new_xlim[0], new_xlim[1])
# disp.axes[this_key].ax.set_ylim(new_ylim[0], new_ylim[1])
# elif 'STN' in roi and display_mode == 'y':
# this_key = [x for x in disp.axes.keys()]
# this_key = this_key[0]
# cur_xlim = disp.axes[this_key].ax.get_xlim()
# cur_ylim = disp.axes[this_key].ax.get_ylim()
# new_xlim = get_prop_limits([.25, .75], cur_xlim)
# new_ylim = get_prop_limits([.25, .75], cur_ylim)
# disp.axes[this_key].ax.set_xlim(new_xlim[0], new_xlim[1])
# disp.axes[this_key].ax.set_ylim(new_ylim[0], new_ylim[1])
# elif 'STR' in roi and display_mode == 'z':
# this_key = [x for x in disp.axes.keys()]
# this_key = this_key[0]
# cur_xlim = disp.axes[this_key].ax.get_xlim()
# cur_ylim = disp.axes[this_key].ax.get_ylim()
# new_xlim = get_prop_limits([0, 1], cur_xlim)
# new_ylim = get_prop_limits([.3, 1], cur_ylim)
# disp.axes[this_key].ax.set_xlim(new_xlim[0], new_xlim[1])
# disp.axes[this_key].ax.set_ylim(new_ylim[0], new_ylim[1])
# all_disps.append(disp)
# add labels
if not skip_all_but_last:
for row_n, ax in enumerate(axes[:, 0]):
cc = all_cut_coords[row_n]
disp_mode = [x for x in cc.keys()][0]
coord = cc[disp_mode]
ax.annotate('%s = %d' % (disp_mode, int(coord)),
xy=(0, 0.5),
xytext=(-ax.yaxis.labelpad - 0.5, 0),
xycoords=ax.yaxis.label,
textcoords='offset points',
rotation=90,
ha='right',
va='center')
f.set_size_inches(len(zmaps) * 4, len(roi_to_plot) * 4)
return f, axes, all_disps
def denoise(img_file,
tsv_file,
out_path,
col_names=False,
hp_filter=False,
lp_filter=False,
out_figure_path=False):
nii_ext = '.nii.gz'
FD_thr = [.5]
sc_range = np.arange(-1, 3)
constant = 'constant'
# read in files
img = load_niimg(img_file)
# get file info
img_name = os.path.basename(img.get_filename())
file_base = img_name[0:img_name.find('.')]
save_img_file = pjoin(out_path, file_base + \
'_NR' + nii_ext)
data = img.get_data()
df_orig = pandas.read_csv(tsv_file, '\t', na_values='n/a')
df = copy.deepcopy(df_orig)
Ntrs = df.as_matrix().shape[0]
print('# of TRs: ' + str(Ntrs))
assert (Ntrs == data.shape[len(data.shape) - 1])
# select columns to use as nuisance regressors
if col_names:
df = df[col_names]
str_append = ' [SELECTED regressors in CSV]'
else:
col_names = df.columns.tolist()
str_append = ' [ALL regressors in CSV]'
# fill in missing nuisance values with mean for that variable
for col in df.columns:
if sum(df[col].isnull()) > 0:
print('Filling in ' + str(sum(df[col].isnull())) +
' NaN value for ' + col)
df[col] = df[col].fillna(np.mean(df[col]))
print('# of Confound Regressors: ' + str(len(df.columns)) + str_append)
# implement HP filter in regression
TR = img.header.get_zooms()[-1]
frame_times = np.arange(Ntrs) * TR
if hp_filter:
hp_filter = float(hp_filter)
assert (hp_filter > 0)
period_cutoff = 1. / hp_filter
df = make_first_level_design_matrix(frame_times,
period_cut=period_cutoff,
add_regs=df.as_matrix(),
add_reg_names=df.columns.tolist())
# fn adds intercept into dm
hp_cols = [col for col in df.columns if 'drift' in col]
print('# of High-pass Filter Regressors: ' + str(len(hp_cols)))
else:
# add in intercept column into data frame
df[constant] = 1
print('No High-pass Filter Applied')
dm = df.as_matrix()
# prep data
data = np.reshape(data, (-1, Ntrs))
data_mean = np.mean(data, axis=1)
Nvox = len(data_mean)
# setup and run regression
model = regression.OLSModel(dm)
results = model.fit(data.T)
if not hp_filter:
results_orig_resid = copy.deepcopy(
results.resid) # save for rsquared computation
# apply low-pass filter
if lp_filter:
# input to butterworth fn is time x voxels
low_pass = float(lp_filter)
Fs = 1. / TR
if low_pass >= Fs / 2:
raise ValueError(
'Low pass filter cutoff if too close to the Nyquist frequency (%s)'
% (Fs / 2))
temp_img_file = pjoin(out_path, file_base + \
'_temp' + nii_ext)
temp_img = nb.Nifti1Image(np.reshape(
results.resid.T + np.reshape(data_mean, (Nvox, 1)),
img.shape).astype('float32'),
img.affine,
header=img.header)
temp_img.to_filename(temp_img_file)
results.resid = butterworth(results.resid,
sampling_rate=Fs,
low_pass=low_pass,
high_pass=None)
print('Low-pass Filter Applied: < ' + str(low_pass) + ' Hz')
# add mean back into data
clean_data = results.resid.T + np.reshape(
data_mean, (Nvox, 1)) # add mean back into residuals
# save out new data file
print('Saving output file...')
clean_data = np.reshape(clean_data, img.shape).astype('float32')
new_img = nb.Nifti1Image(clean_data, img.affine, header=img.header)
new_img.to_filename(save_img_file)
######### generate Rsquared map for confounds only
if hp_filter:
# first remove low-frequency information from data
hp_cols.append(constant)
model_first = regression.OLSModel(df[hp_cols].as_matrix())
results_first = model_first.fit(data.T)
results_first_resid = copy.deepcopy(results_first.resid)
del results_first, model_first
# compute sst - borrowed from matlab
sst = np.square(
np.linalg.norm(results_first_resid -
np.mean(results_first_resid, axis=0),
axis=0))
# now regress out 'true' confounds to estimate their Rsquared
nr_cols = [col for col in df.columns if 'drift' not in col]
model_second = regression.OLSModel(df[nr_cols].as_matrix())
results_second = model_second.fit(results_first_resid)
# compute sse - borrowed from matlab
sse = np.square(np.linalg.norm(results_second.resid, axis=0))
del results_second, model_second, results_first_resid
elif not hp_filter:
# compute sst - borrowed from matlab
sst = np.square(
np.linalg.norm(data.T - np.mean(data.T, axis=0), axis=0))
# compute sse - borrowed from matlab
sse = np.square(np.linalg.norm(results_orig_resid, axis=0))
del results_orig_resid
# compute rsquared of nuisance regressors
zero_idx = scipy.logical_and(sst == 0, sse == 0)
sse[zero_idx] = 1
sst[zero_idx] = 1 # would be NaNs - become rsquared = 0
rsquare = 1 - np.true_divide(sse, sst)
rsquare[np.isnan(rsquare)] = 0
######### Visualizing DM & outputs
fontsize = 12
fontsize_title = 14
def_img_size = 8
if not out_figure_path:
out_figure_path = save_img_file[0:save_img_file.find('.')] + '_figures'
if not os.path.isdir(out_figure_path):
os.mkdir(out_figure_path)
png_append = '_' + img_name[0:img_name.find('.')] + '.png'
print('Output directory: ' + out_figure_path)
# DM corr matrix
cm = df[df.columns[0:-1]].corr()
curr_sz = copy.deepcopy(def_img_size)
if cm.shape[0] > def_img_size:
curr_sz = curr_sz + ((cm.shape[0] - curr_sz) * .3)
mtx_scale = curr_sz * 100
mask = np.zeros_like(cm, dtype=np.bool)
mask[np.triu_indices_from(mask)] = True
fig, ax = plt.subplots(figsize=(curr_sz, curr_sz))
cmap = sns.diverging_palette(220, 10, as_cmap=True)
sns.heatmap(cm,
mask=mask,
cmap=cmap,
center=0,
vmax=cm[cm < 1].max().max(),
vmin=cm[cm < 1].min().min(),
square=True,
linewidths=.5,
cbar_kws={"shrink": .6})
ax.set_xticklabels(ax.get_xticklabels(),
rotation=60,
ha='right',
fontsize=fontsize)
ax.set_yticklabels(cm.columns.tolist(),
rotation=-30,
va='bottom',
fontsize=fontsize)
ax.set_title('Nuisance Corr. Matrix', fontsize=fontsize_title)
plt.tight_layout()
file_corr_matrix = 'Corr_matrix_regressors' + png_append
fig.savefig(pjoin(out_figure_path, file_corr_matrix))
plt.close(fig)
del fig, ax
# DM of Nuisance Regressors (all)
tr_label = 'TR (Volume #)'
fig, ax = plt.subplots(figsize=(curr_sz - 4.1, def_img_size))
x_scale_html = ((curr_sz - 4.1) / def_img_size) * 890
reporting.plot_design_matrix(df, ax=ax)
ax.set_title('Nuisance Design Matrix', fontsize=fontsize_title)
ax.set_xticklabels(ax.get_xticklabels(),
rotation=60,
ha='right',
fontsize=fontsize)
ax.set_yticklabels(ax.get_yticklabels(), fontsize=fontsize)
ax.set_ylabel(tr_label, fontsize=fontsize)
plt.tight_layout()
file_design_matrix = 'Design_matrix' + png_append
fig.savefig(pjoin(out_figure_path, file_design_matrix))
plt.close(fig)
del fig, ax
# FD timeseries plot
FD = 'FD'
poss_names = ['FramewiseDisplacement', FD, 'framewisedisplacement', 'fd']
fd_idx = [df_orig.columns.__contains__(i) for i in poss_names]
if np.sum(fd_idx) > 0:
FD_name = poss_names[fd_idx == True]
if sum(df_orig[FD_name].isnull()) > 0:
df_orig[FD_name] = df_orig[FD_name].fillna(
np.mean(df_orig[FD_name]))
y = df_orig[FD_name].as_matrix()
Nremove = []
sc_idx = []
for thr_idx, thr in enumerate(FD_thr):
idx = y >= thr
sc_idx.append(copy.deepcopy(idx))
for iidx in np.where(idx)[0]:
for buffer in sc_range:
curr_idx = iidx + buffer
if curr_idx >= 0 and curr_idx <= len(idx):
sc_idx[thr_idx][curr_idx] = True
Nremove.append(np.sum(sc_idx[thr_idx]))
Nplots = len(FD_thr)
sns.set(font_scale=1.5)
sns.set_style('ticks')
fig, axes = plt.subplots(Nplots,
1,
figsize=(def_img_size * 1.5,
def_img_size / 2),
squeeze=False)
sns.despine()
bound = .4
fd_mean = np.mean(y)
for curr in np.arange(0, Nplots):
axes[curr, 0].plot(y)
axes[curr, 0].plot((-bound, Ntrs + bound),
FD_thr[curr] * np.ones((1, 2))[0],
'--',
color='black')
axes[curr, 0].scatter(np.arange(0, Ntrs), y, s=20)
if Nremove[curr] > 0:
info = scipy.ndimage.measurements.label(sc_idx[curr])
for cluster in np.arange(1, info[1] + 1):
temp = np.where(info[0] == cluster)[0]
axes[curr, 0].axvspan(temp.min() - bound,
temp.max() + bound,
alpha=.5,
color='red')
axes[curr, 0].set_ylabel('Framewise Disp. (' + FD + ')')
axes[curr, 0].set_title(FD + ': ' +
str(100 * Nremove[curr] / Ntrs)[0:4] +
'% of scan (' + str(Nremove[curr]) +
' volumes) would be scrubbed (FD thr.= ' +
str(FD_thr[curr]) + ')')
plt.text(Ntrs + 1,
FD_thr[curr] - .01,
FD + ' = ' + str(FD_thr[curr]),
fontsize=fontsize)
plt.text(Ntrs,
fd_mean - .01,
'avg = ' + str(fd_mean),
fontsize=fontsize)
axes[curr, 0].set_xlim((-bound, Ntrs + 8))
plt.tight_layout()
axes[curr, 0].set_xlabel(tr_label)
file_fd_plot = FD + '_timeseries' + png_append
fig.savefig(pjoin(out_figure_path, file_fd_plot))
plt.close(fig)
del fig, axes
print(FD + ' timeseries plot saved')
else:
print(FD + ' not found: ' + FD + ' timeseries not plotted')
file_fd_plot = None
# Carpet and DVARS plots - before & after nuisance regression
# need to create mask file to input to DVARS function
mask_file = pjoin(out_figure_path, 'mask_temp.nii.gz')
nifti_masker = NiftiMasker(mask_strategy='epi', standardize=False)
nifti_masker.fit(img)
nifti_masker.mask_img_.to_filename(mask_file)
# create 2 or 3 carpet plots, depending on if LP filter is also applied
Ncarpet = 2
total_sz = int(16)
carpet_scale = 840
y_labels = ['Input (voxels)', 'Output \'cleaned\'']
imgs = [img, new_img]
img_files = [img_file, save_img_file]
color = ['red', 'salmon']
labels = ['input', 'cleaned']
if lp_filter:
Ncarpet = 3
total_sz = int(20)
carpet_scale = carpet_scale * (9 / 8)
y_labels = ['Input', 'Clean Pre-LP', 'Clean LP']
imgs.insert(1, temp_img)
img_files.insert(1, temp_img_file)
color.insert(1, 'firebrick')
labels.insert(1, 'clean pre-LP')
labels[-1] = 'clean LP'
dvars = []
print('Computing dvars...')
for in_file in img_files:
temp = nac.compute_dvars(in_file=in_file, in_mask=mask_file)[1]
dvars.append(np.hstack((temp.mean(), temp)))
del temp
small_sz = 2
fig = plt.figure(figsize=(def_img_size * 1.5,
def_img_size + ((Ncarpet - 2) * 1)))
row_used = 0
if np.sum(fd_idx) > 0: # if FD data is available
row_used = row_used + small_sz
ax0 = plt.subplot2grid((total_sz, 1), (0, 0), rowspan=small_sz)
ax0.plot(y)
ax0.scatter(np.arange(0, Ntrs), y, s=10)
curr = 0
if Nremove[curr] > 0:
info = scipy.ndimage.measurements.label(sc_idx[curr])
for cluster in np.arange(1, info[1] + 1):
temp = np.where(info[0] == cluster)[0]
ax0.axvspan(temp.min() - bound,
temp.max() + bound,
alpha=.5,
color='red')
ax0.set_ylabel(FD)
for side in ["top", "right", "bottom"]:
ax0.spines[side].set_color('none')
ax0.spines[side].set_visible(False)
ax0.set_xticks([])
ax0.set_xlim((-.5, Ntrs - .5))
ax0.spines["left"].set_position(('outward', 10))
ax_d = plt.subplot2grid((total_sz, 1), (row_used, 0), rowspan=small_sz)
for iplot in np.arange(len(dvars)):
ax_d.plot(dvars[iplot], color=color[iplot], label=labels[iplot])
ax_d.set_ylabel('DVARS')
for side in ["top", "right", "bottom"]:
ax_d.spines[side].set_color('none')
ax_d.spines[side].set_visible(False)
ax_d.set_xticks([])
ax_d.set_xlim((-.5, Ntrs - .5))
ax_d.spines["left"].set_position(('outward', 10))
ax_d.legend(fontsize=fontsize - 2)
row_used = row_used + small_sz
st = 0
carpet_each = int((total_sz - row_used) / Ncarpet)
for idx, img_curr in enumerate(imgs):
ax_curr = plt.subplot2grid((total_sz, 1), (row_used + st, 0),
rowspan=carpet_each)
fig = plotting.plot_carpet(img_curr, figure=fig, axes=ax_curr)
ax_curr.set_ylabel(y_labels[idx])
for side in ["bottom", "left"]:
ax_curr.spines[side].set_position(('outward', 10))
if idx < len(imgs) - 1:
ax_curr.spines["bottom"].set_visible(False)
ax_curr.set_xticklabels('')
ax_curr.set_xlabel('')
st = st + carpet_each
file_carpet_plot = 'Carpet_plots' + png_append
fig.savefig(pjoin(out_figure_path, file_carpet_plot))
plt.close()
del fig, ax0, ax_curr, ax_d, dvars
os.remove(mask_file)
print('Carpet/DVARS plots saved')
if lp_filter:
os.remove(temp_img_file)
del temp_img
# Display T-stat maps for nuisance regressors
# create mean img
img_size = (img.shape[0], img.shape[1], img.shape[2])
mean_img = nb.Nifti1Image(np.reshape(data_mean, img_size), img.affine)
mx = []
for idx, col in enumerate(df.columns):
if not 'drift' in col and not constant in col:
con_vector = np.zeros((1, df.shape[1]))
con_vector[0, idx] = 1
con = results.Tcontrast(con_vector)
mx.append(np.max(np.absolute([con.t.min(), con.t.max()])))
mx = .8 * np.max(mx)
t_png = 'Tstat_'
file_tstat = []
for idx, col in enumerate(df.columns):
if not 'drift' in col and not constant in col:
con_vector = np.zeros((1, df.shape[1]))
con_vector[0, idx] = 1
con = results.Tcontrast(con_vector)
m_img = nb.Nifti1Image(np.reshape(con, img_size), img.affine)
title_str = col + ' Tstat'
fig = plotting.plot_stat_map(m_img,
mean_img,
threshold=3,
colorbar=True,
display_mode='z',
vmax=mx,
title=title_str,
cut_coords=7)
file_temp = t_png + col + png_append
fig.savefig(pjoin(out_figure_path, file_temp))
file_tstat.append({'name': col, 'file': file_temp})
plt.close()
del fig, file_temp
print(title_str + ' map saved')
# Display R-sq map for nuisance regressors
m_img = nb.Nifti1Image(np.reshape(rsquare, img_size), img.affine)
title_str = 'Nuisance Rsq'
mx = .95 * rsquare.max()
fig = plotting.plot_stat_map(m_img,
mean_img,
threshold=.2,
colorbar=True,
display_mode='z',
vmax=mx,
title=title_str,
cut_coords=7)
file_rsq_map = 'Rsquared' + png_append
fig.savefig(pjoin(out_figure_path, file_rsq_map))
plt.close()
del fig
print(title_str + ' map saved')
######### html report
templateLoader = jinja2.FileSystemLoader(searchpath="/")
templateEnv = jinja2.Environment(loader=templateLoader)
templateVars = {
"img_file": img_file,
"save_img_file": save_img_file,
"Ntrs": Ntrs,
"tsv_file": tsv_file,
"col_names": col_names,
"hp_filter": hp_filter,
"lp_filter": lp_filter,
"file_design_matrix": file_design_matrix,
"file_corr_matrix": file_corr_matrix,
"file_fd_plot": file_fd_plot,
"file_rsq_map": file_rsq_map,
"file_tstat": file_tstat,
"x_scale": x_scale_html,
"mtx_scale": mtx_scale,
"file_carpet_plot": file_carpet_plot,
"carpet_scale": carpet_scale
}
TEMPLATE_FILE = pjoin(os.getcwd(), "report_template.html")
template = templateEnv.get_template(TEMPLATE_FILE)
outputText = template.render(templateVars)
html_file = pjoin(out_figure_path,
img_name[0:img_name.find('.')] + '.html')
with open(html_file, "w") as f:
f.write(outputText)
print('')
print('HTML report: ' + html_file)
return new_img
from nilearn.plotting import plot_stat_map
# Before visualizing, we transform the computed p-values to Nifti-like image
# using function `new_img_like` from nilearn.
from nilearn.image import new_img_like
# First argument being a reference image and second argument should be p-values
# data to convert to a new image as output. This new image will have same header
# information as reference image.
log_p_values_img = new_img_like(fmri_img, log_p_values)
# Now, we visualize log p-values image on functional mean image as background
# with coordinates given manually and colorbar on the right side of plot (by
# default colorbar=True)
plot_stat_map(log_p_values_img,
mean_img,
title="p-values",
cut_coords=cut_coords)
#############################################################################
# **Selecting features using f_classif**: Feature selection method is also
# available in the scikit-learn Python package, where it has been extended to
# several classes, using the `sklearn.feature_selection.f_classif` function.
##############################################################################
# Build a mask from this statistical map (Improving the quality of the mask)
# --------------------------------------------------------------------------
# **Thresholding** - We build the t-map to have better representation of voxels
# of interest and voxels with lower p-values correspond to the most intense
# voxels. This can be done easily by applying a threshold to a t-map data in
# array.
# Now, we show from here how to visualize the retrieved datasets using plotting
# tools from nilearn.
from nilearn import plotting
########################################
# Visualizing in - 'sagittal', 'coronal' and 'axial' with given coordinates
# -------------------------------------------------------------------------
# The first argument is a path to the filename of a constrast map,
# optional argument `display_mode` is given as string 'ortho' to visualize
# the map in three specific directions xyz and the optional `cut_coords`
# argument, is here a list of integers denotes coordinates of each slice
# in the order [x, y, z]. By default the `colorbar` argument is set to True
# in plot_stat_map.
plotting.plot_stat_map(stat_img,
display_mode='ortho',
cut_coords=[36, -27, 60],
title="display_mode='ortho', cut_coords=[36, -27, 60]")
########################################
# Visualizing in - single view 'axial' with number of cuts=5
# -----------------------------------------------------------
# In this type of visualization, the `display_mode` argument is given as
# string 'z' for axial direction and `cut_coords` as integer 5 without a
# list implies that number of cuts in the slices should be maximum of 5.
# The coordinates to cut the slices are selected automatically
plotting.plot_stat_map(stat_img,
display_mode='z',
cut_coords=5,
title="display_mode='z', cut_coords=5")
########################################
logp_thresh = -np.log(.05)
###############################################################################
# Fisher's (using functions)
# --------------------------------------------------
# Get images for analysis
files = dset.get_images(imtype='z')
files = [f for f in files if f]
z_imgs = [nib.load(f) for f in files]
z_data = apply_mask(z_imgs, mask_img)
print('{0} studies found.'.format(z_data.shape[0]))
result = fishers(z_data, mask_img)
fishers_result = unmask(result['z'], mask_img)
plot_stat_map(fishers_result,
cut_coords=[0, 0, -8],
draw_cross=False,
cmap='RdBu_r')
###############################################################################
# Fisher's (using Estimators)
# --------------------------------------------------
# Here is the object-oriented approach
meta = Fishers()
meta.fit(dset)
plot_stat_map(meta.results.get_map('z'),
cut_coords=[0, 0, -8],
draw_cross=False,
cmap='RdBu_r')
###############################################################################
# Stouffer's with fixed-effects inference
for train, test in cv:
svc.fit(fmri_masked[train], target[train])
prediction = svc.predict(fmri_masked[test])
cv_scores.append(
np.sum(prediction == target[test]) / float(np.size(target[test])))
print cv_scores
### Unmasking #################################################################
# Retrieve the SVC discriminating weights
coef_ = svc.coef_
# Reverse masking thanks to the Nifti Masker
coef_img = nifti_masker.inverse_transform(coef_)
# Save the coefficients as a Nifti image
coef_img.to_filename('haxby_svc_weights.nii')
### Visualization #############################################################
import pylab as plt
from nilearn.image.image import mean_img
from nilearn.plotting import plot_roi, plot_stat_map
mean_epi = mean_img(func_filename)
plot_stat_map(coef_img, mean_epi, title="SVM weights", display_mode="yx")
plot_roi(nifti_masker.mask_img_, mean_epi, title="Mask", display_mode="yx")
plt.show()
plotting.plot_connectome(mean_correlations,
coords_connectome,
edge_threshold='90%',
title=title)
################################################################################
# Plot regions extracted for only one specific network
# ----------------------------------------------------
components_img = atlas_harvard_oxford.maps
# First, we plot a network of index=4 without region extraction (left plot)
from nilearn import image
img = image.index_img(components_img, 1)
coords = plotting.find_xyz_cut_coords(img)
display = plotting.plot_stat_map(img,
cut_coords=coords,
colorbar=False,
title='Showing one specific network')
################################################################################
# Now, we plot (right side) same network after region extraction to show that
# connected regions are nicely seperated.
# Each brain extracted region is identified as separate color.
# For this, we take the indices of the all regions extracted related to original
# network given as 4.
regions_indices_of_map3 = np.where(np.array(regions_index) == 1)
display = plotting.plot_anat(cut_coords=coords,
title='Regions from this network')
# Add as an overlay all the regions of index 4
print 'dot product is done'
#maybe save the z-image and run randomise on that?
sbc_hb_z = np.arctanh(sbc_hb)
print 'we have z values'
sbc_hb_img = brain_masker.inverse_transform(sbc_hb.T)
sbc_hb_z_img = brain_masker.inverse_transform(sbc_hb_z.T)
print 'and a z-value image'
output_png = join(sink_dir, level,
'{0}-{1}_sbc_hb.png'.format(s, run))
output_nii = join(sink_dir, level,
'{0}-{1}_sbc_hb.nii.gz'.format(s, run))
output_z_nii = join(sink_dir, level,
'{0}-{1}_sbc_z_hb.nii.gz'.format(s, run))
rmaps.append(output_nii)
zmaps.append(output_z_nii)
sbc_hb_img.to_filename(output_nii)
sbc_hb_z_img.to_filename(output_z_nii)
plotting.plot_stat_map(sbc_hb_img,
bg_img=mean_func,
output_file=output_png)
except Exception as e:
print(e)
try:
avg_z_map = mean_img(zmaps)
avg_z_map.to_filename(
join(sink_dir, level, '{0}_mean_zmap.nii.gz'.format(s)))
except Exception as e:
print(e)
# Print the results
print("Classification accuracy: %.4f / Chance level: %f" %
(classification_accuracy, 1. / len(np.unique(conditions))))
# Classification accuracy: 0.9861 / Chance level: 0.5000
#############################################################################
# Visualize the results
# Look at the SVC's discriminating weights
coef = svc.coef_
# reverse feature selection
coef = feature_selection.inverse_transform(coef)
# reverse masking
weight_img = masker.inverse_transform(coef)
# Use the mean image as a background to avoid relying on anatomical data
from nilearn import image
mean_img = image.mean_img(func_filename)
# Create the figure
from nilearn.plotting import plot_stat_map, show
plot_stat_map(weight_img, mean_img, title='SVM weights')
# Saving the results as a Nifti file may also be important
weight_img.to_filename('haxby_face_vs_house.nii')
show()
nested_cv_scores = cross_val_score(grid, X, y, cv=5)
#NEST_SCORE = np.mean(nested_cv_scores)
print("Nested CV score: %.4f" % np.mean(nested_cv_scores))
# Here is the image
coef = svc.coef_
# reverse feature selection
coef = feature_selection.inverse_transform(coef)
# reverse masking
weight_img = masker.inverse_transform(coef)
# Use the mean image as a background to avoid relying on anatomical data
from nilearn import image
mean_img = image.mean_img(dataset)
mean_img.to_filename(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/milkshake_vs_h2O/images/all_p1_mean_nimask.nii'
)
# Create the figure
from nilearn.plotting import plot_stat_map, show
display = plot_stat_map(weight_img, mean_img, title='Milkshake vs. H2O')
display.savefig(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/milkshake_vs_h2O/images/all_p1_SVM_nimask.png'
)
# Saving the results as a Nifti file may also be important
weight_img.to_filename(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/milkshake_vs_h2O/images/all_p1_SVM_nimask.nii'
)
# Let's now retrieve a motor contrast from a Neurovault repository motor_images = datasets.fetch_neurovault_motor_task() print(motor_images.images) ############################################################################### # motor_images is a list of filenames. We need to take the first one tmap_filename = motor_images.images[0] ############################################################################### # Visualizing a 3D file # ---------------------- # # The file contains a 3D volume, we can easily visualize it as a # statistical map: from nilearn import plotting plotting.plot_stat_map(tmap_filename) ############################################################################### # Visualizing works better with a threshold plotting.plot_stat_map(tmap_filename, threshold=3) ############################################################################### # Visualizing one volume in a 4D file # ----------------------------------- # # We can download resting-state networks from the Smith 2009 study on # correspondance between rest and task rsn = datasets.fetch_atlas_smith_2009()['rsn10'] print(rsn) ###############################################################################
# For this, we first define the contrasts as we would do for a single session
n_columns = design_matrices[0].shape[1]
contrast_val = np.hstack(([-1, -1, 1, 1], np.zeros(n_columns - 4)))
#########################################################################
# Statistics for the first session
from nilearn import plotting
cut_coords = [-129, -126, 49]
contrast_id = 'DSt_minus_SSt'
fmri_glm = fmri_glm.fit(fmri_img[0], design_matrices=design_matrices[0])
summary_statistics_session1 = fmri_glm.compute_contrast(contrast_val,
output_type='all')
plotting.plot_stat_map(summary_statistics_session1['z_score'],
bg_img=mean_img_,
threshold=3.0,
cut_coords=cut_coords,
title='{0}, first session'.format(contrast_id))
#########################################################################
# Statistics for the second session
fmri_glm = fmri_glm.fit(fmri_img[1], design_matrices=design_matrices[1])
summary_statistics_session2 = fmri_glm.compute_contrast(contrast_val,
output_type='all')
plotting.plot_stat_map(summary_statistics_session2['z_score'],
bg_img=mean_img_,
threshold=3.0,
cut_coords=cut_coords,
title='{0}, second session'.format(contrast_id))
print("Nested CV score: %.4f" % np.mean(nested_cv_scores))
# In[ ]:
# Here is the image
coef = svc.coef_
# reverse feature selection
coef = feature_selection.inverse_transform(coef)
# reverse masking
weight_img = masker.inverse_transform(coef)
# Use the mean image as a background to avoid relying on anatomical data
from nilearn import image
mean_img = image.mean_img(dataset)
mean_img.to_filename(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/LF_HS_vs_h2O/images/4wp2_k2_mean_nimask.nii'
)
# Create the figure
from nilearn.plotting import plot_stat_map, show
display = plot_stat_map(weight_img,
mean_img,
title='SVM weights LF_HS vs h2O 4 waves')
display.savefig(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/LF_HS_vs_h2O/images/4wp2_k2_nimask.png'
)
# Saving the results as a Nifti file may also be important
weight_img.to_filename(
'/projects/niblab/bids_projects/Experiments/ChocoData/derivatives/code/decoding/LF_HS_vs_h2O/images/4wp2_k2_nimask.nii'
)
resampled_affine = resampled_stat_img.affine
template_img = load_img(template)
template_shape = template_img.shape
template_affine = template_img.affine
print("""Shape comparison:
- Original t-map image shape : {0}
- Resampled t-map image shape: {1}
- Template image shape : {2}
""".format(original_shape, resampled_shape, template_shape))
print("""Affine comparison:
- Original t-map image affine :\n {0}
- Resampled t-map image affine:\n {1}
- Template image affine :\n {2}
""".format(original_affine, resampled_affine, template_affine))
from nilearn import plotting
plotting.plot_stat_map(stat_img,
bg_img=template,
cut_coords=(36, -27, 66),
threshold=3,
title="t-map in original resolution")
plotting.plot_stat_map(resampled_stat_img,
bg_img=template,
cut_coords=(36, -27, 66),
threshold=3,
title="Resampled t-map")
plotting.show()
### Fit TV-L1 #################################################################
# Here we're using the regressor object given that the task is to predict a
# continuous variable, the gain of the gamble.
from nilearn.decoding import SpaceNetRegressor
decoder = SpaceNetRegressor(
mask=mask_img,
penalty="tv-l1",
eps=1e-1, # prefer large alphas
memory="cache")
decoder.fit(zmaps, object_category) # fit
# Visualize TV-L1 weights
import matplotlib.pyplot as plt
from nilearn.plotting import plot_stat_map
plot_stat_map(decoder.coef_img_,
title="tv-l1",
display_mode="yz",
cut_coords=[20, -2])
### Fit Graph-Net #############################################################
decoder = SpaceNetRegressor(
mask=mask_img,
penalty="graph-net",
eps=1e-1, # prefer large alphas
memory="cache")
decoder.fit(zmaps, object_category) # fit
# Visualize Graph-Net weights
plot_stat_map(decoder.coef_img_,
title="graph-net",
display_mode="yz",
cut_coords=[20, -2])
axis = fig.subplots(nrows=U.shape[1] * 2, ncols=1)
for k in range(U.shape[1]):
#k = 0
idx = 2 * k
map_img = map_img_l[k]
#ax = fig.add_subplot(111)
#ax.set_title("T-stats T>%.2f" % tstats_thres)
vmax = np.abs(map_arr).max()
axis[idx].set_title("PC%i (EV:%.3f%%)" % (k+1, explained_variance_ratio[k] * 100))
plotting.plot_glass_brain(map_img, colorbar=True, vmax=vmax, figure=fig, axes=axis[idx])
#pdf.savefig()
display = plotting.plot_stat_map(map_img, colorbar=True, draw_cross=True, cmap=cmnl.cold_white_hot, figure=fig, axes=axis[idx+1])#, symmetric_cbar=False)#, cmap=plt.cm.hot_r)#, cut_coords=[16, -4, 0], symmetric_cbar=False, cmap=cold_blue)#, threshold=3,)#, figure=fig, axes=ax)
plt.show()
pdf.savefig()
plt.savefig(prefix+"_components-brain-maps.png")
plt.close(fig)
pdf.close()
########################################################################################################################
cd /neurospin/brainomics/2019_rundmc_wmh/analyses/201909_rundmc_wmh_pca/models/pca_enettv_0.000350_1.000_0.001
fsl5.0-fslsplit components-brain-maps.nii.gz ./components-brain-maps_PC -t
~/git/scripts/brainomics/image_clusters_analysis_nilearn.py /neurospin/brainomics/2019_rundmc_wmh/analyses/201909_rundmc_wmh_pca/models/pca_enettv_0.000350_1.000_0.001/components-brain-maps_PC0000.nii.gz --atlas JHU --thresh_size 10
~/git/scripts/brainomics/image_clusters_analysis_nilearn.py /neurospin/brainomics/2019_rundmc_wmh/analyses/201909_rundmc_wmh_pca/models/pca_enettv_0.000350_1.000_0.001/components-brain-maps_PC0001.nii.gz --atlas JHU --thresh_size 10
~/git/scripts/brainomics/image_clusters_analysis_nilearn.py /neurospin/brainomics/2019_rundmc_wmh/analyses/201909_rundmc_wmh_pca/models/pca_enettv_0.000350_1.000_0.001/components-brain-maps_PC0002.nii.gz --atlas JHU --thresh_size 10
#
def plotting_hrf_stats(v,
t_r,
hrf_ref=None,
stat_type='tp',
display_mode='ortho',
cut_coords=None,
masker=None,
atlas_type='havard',
atlas_kwargs=dict(),
n_scales=122,
plot_dir='.',
fname=None,
save_nifti=False,
verbose=False):
""" Plot, and save as pdf, each stats HRF for each ROIs.
Parameters
----------
v : array, shape (n_hrf_rois, n_times_atom), the initial used HRFs
t_r : float, Time of Repetition, fMRI acquisition parameter, the temporal
resolution
hrf_ref : array or None, shape (n_times_atom, ), (default=None), reference
HRF to plot for comparison
stat_type : str, (default='tp'), statistic to compute on each HRFs possible
choice are ('tp', 'fwhm')
normalized : bool, (default=False), whether or not to normalized by the
l-inf norm each HRFs
display_mode : None or str, coords to cut the plotting, possible value are
None to have x, y, z or 'x', 'y', 'z' for a single cut
cut_coords : tuple or None, MNI coordinate to perform display
masker : Nilearn-Masker like, masker class to perform the inverse Nifti
transformation
atlas_type : str, func, or None, (default=None), atlas type, possible
choice are ['havard', 'basc', given-function]
atlas_kwargs : dict, (default=dict()), additional kwargs for the atlas,
if a function is passed.
n_scales : int, (default=122), number of scale if atlas_type == 'basc'
plot_dir : str, (default='.'), directory under which the pdf is saved
fname : str, (default='v_{fwhm/tp}.pdf'), filename under which the pdf is
saved
save_nifti : bool, (default=False), whether or not to save the image as
Nifti
verbose : bool, (default=False), verbosity level
"""
if stat_type not in ['tp', 'fwhm']:
raise ValueError("stat_type should be in ['tp', 'fwhm'], "
"got {}".format(stat_type))
if atlas_type == 'havard':
_, atlas_rois = fetch_vascular_atlas()
elif atlas_type == 'basc':
n_scales_ = f"scale{int(n_scales)}"
_, atlas_rois = fetch_atlas_basc_2015(n_scales=n_scales_)
elif isinstance(atlas_type, collections.Callable):
_, atlas_rois = atlas_type(**atlas_kwargs)
else:
raise ValueError(f"atlas_type should belong to ['havard', 'basc', "
f"given-function], got {atlas_type}")
hrf_rois = dict()
rois = masker.transform(atlas_rois).astype(int).ravel()
index = np.arange(rois.shape[-1])
for roi_label in np.unique(rois):
hrf_rois[roi_label] = index[roi_label == rois]
_, roi_label_from_hrf_idx, _ = split_atlas(hrf_rois)
raw_atlas_rois = atlas_rois.get_data()
n_hrf_rois, n_times_atom = v.shape
if hrf_ref is not None:
if stat_type == 'tp':
ref_stat = tp(t_r, hrf_ref)
elif stat_type == 'fwhm':
ref_stat = fwhm(t_r, hrf_ref)
for m in range(n_hrf_rois):
v_ = v[m, :]
if stat_type == 'tp':
stat_ = tp(t_r, v_)
stat_name = 'TtP'
elif stat_type == 'fwhm':
stat_ = fwhm(t_r, v_)
stat_name = 'FWHM'
if hrf_ref is not None:
stat_ -= ref_stat
title = "{0}(-{0}-reference) map (s)".format(stat_name)
else:
title = "{} map (s)".format(stat_name)
label = roi_label_from_hrf_idx[m]
raw_atlas_rois[raw_atlas_rois == label] = stat_
stats_map = image.new_img_like(atlas_rois, raw_atlas_rois)
plotting.plot_stat_map(stats_map,
title=title,
colorbar=True,
display_mode=display_mode,
cut_coords=cut_coords,
symmetric_cbar=False)
if save_nifti:
nii_filename = os.path.join(plot_dir, "v_{}.nii".format(stat_type))
stats_map.to_filename(nii_filename)
if fname is None:
fname = "v_{}.pdf".format(stat_type)
fname = os.path.join(plot_dir, fname)
plt.savefig(fname, dpi=150)
if verbose:
print("Saving plot under '{0}'".format(fname))
from nilearn.plotting import plot_stat_map, show
# Use the fmri mean image as a surrogate of anatomical data
from nilearn import image
from nilearn.image import get_data
mean_fmri_img = image.mean_img(func_filename)
threshold = -np.log10(0.1) # 10% corrected
vmax = min(signed_neg_log_pvals.max(),
neg_log_pvals_bonferroni.max())
# Plot thresholded p-values map corresponding to F-scores
display = plot_stat_map(neg_log_pvals_bonferroni_unmasked, mean_fmri_img,
threshold=threshold, cmap=plt.cm.RdBu_r,
display_mode='z', cut_coords=[-1, ],
vmax=vmax)
neg_log_pvals_bonferroni_data = get_data(neg_log_pvals_bonferroni_unmasked)
n_detections = (neg_log_pvals_bonferroni_data > threshold).sum()
title = ('Negative $\\log_{10}$ p-values'
'\n(Parametric two-sided F-test'
'\n+ Bonferroni correction)'
'\n%d detections') % n_detections
display.title(title, y=1.1)
# Plot permutation p-values map
display = plot_stat_map(signed_neg_log_pvals_unmasked, mean_fmri_img,
threshold=threshold, cmap=plt.cm.RdBu_r,
display_mode='z', cut_coords=[-1, ],
# the following line:
canica_components_img.to_filename(
'/home/bk/Desktop/bkrest/canica_resting_state.nii.gz')
from nilearn.plotting import plot_prob_atlas
# Plot all ICA components together
plot_prob_atlas(canica_components_img, title='All ICA components')
from nilearn.image import iter_img
from nilearn.plotting import plot_stat_map
for i, cur_img in enumerate(iter_img(canica_components_img)):
plot_stat_map(cur_img,
display_mode="z",
title="IC %d" % i,
cut_coords=1,
colorbar=False)
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='template')
print('[Example] Fitting dicitonary learning model')
dict_learning.fit(func_filenames)
# taken from Wikipedia
query = "Aphasia is an inability to comprehend or formulate language"
response = encoder(query)
print(response.keys())
######################################################################
# The "z_map" entry of the results is a brain map showing the anatomical
# regions that are most strongly associated with the query in the neuroimaging
# literature. It is a `Nifti1Image` which can be saved, displayed, etc.
from nilearn import plotting
print(type(response["z_map"]))
plotting.plot_stat_map(response["z_map"],
display_mode="z",
title="aphasia",
threshold=3.1)
######################################################################
#
# Display the map on the cortical surface:
view = plotting.view_img_on_surf(response["z_map"], threshold=3.1)
view.open_in_browser()
# (in a Jupyter notebook, we can display an inline view):
view
######################################################################
#
# Or open interactive viewer:
design_matrix=design_matrix.iloc[:, 0:40],
ax=ax[i])
fig.savefig(f'{out_dir}{sub}/figures/{sub}_contrasts_trials.png')
#--- trial effects estimation
zmaps = np.zeros((89, 105, 89, n_trials))
for i in range(n_trials):
zmap = fmri_glm_non_smoothed.compute_contrast(np.asarray(
trial_contrasts[f'Q{i+1:02}']),
output_type='z_score')
zmaps[:, :, :, i] = zmap.get_fdata()
p = plot_stat_map(zmap,
threshold=3,
display_mode='z',
cut_coords=8,
black_bg=False,
title='Q{:0>2d}'.format(i + 1))
p.savefig(f'{out_dir}{sub}/figures/{sub}_Q{i+1:02}_zmap.png')
zmap_img = nib.Nifti1Image(zmaps, zmap.affine, zmap.header)
nib.save(zmap_img, f'{out_dir}{sub}/{sub}_{task}_Q_all_zmaps.nii.gz')
#--- overal effect estimation
fig, ax = plt.subplots(1, 1)
fig.set_size_inches(17, 3)
plot_contrast_matrix(trial_contrasts['overall'],
design_matrix=design_matrix.iloc[:, 0:40],
ax=ax)
fig.savefig(f'{out_dir}{sub}/figures/{sub}_contrast_overall.png')
from gclda.utils import get_resource_path
###############################################################################
# Load model and initialize decoder
# ----------------------------------
model_file = join(get_resource_path(), 'models/Neurosynth2015Filtered2',
'model_200topics_2015Filtered2_10000iters.pklz')
model = Model.load(model_file)
###############################################################################
# Read in image to decode
# --------------------------------------
file_to_decode = '../data/faces_specificity_z.nii.gz'
img_to_decode = nib.load(file_to_decode)
fig = plotting.plot_stat_map(img_to_decode,
display_mode='z',
threshold=3.290527,
cut_coords=[-28, -4, 20, 50])
###############################################################################
# Decode image
# -------------
df, topic_weights = decode_continuous(model, img_to_decode)
###############################################################################
# Get associated terms
# ---------------------
df = df.sort_values(by='Weight', ascending=False)
print(df.head(10))
###############################################################################
# Plot topic weights
def plot_brain(objIn,
how="full",
thr_upper=None,
thr_lower=None,
save=False,
**kwargs):
"""
More complete brain plotting of a Brain_Data instance
Args:
obj: (Brain_Data) object to plot
how: (str) whether to plot a glass brain 'glass', 3 view-multi-slice mni 'mni', or both 'full'
thr_upper: (str/float) thresholding of image. Can be string for percentage, or float for data units (see Brain_Data.threshold()
thr_lower: (str/float) thresholding of image. Can be string for percentage, or float for data units (see Brain_Data.threshold()
save (str): if a string file name or path is provided plots will be saved into this directory appended with the orientation they belong to
kwargs: optionals args to nilearn plot functions (e.g. vmax)
"""
if thr_upper or thr_lower:
obj = objIn.threshold(upper=thr_upper, lower=thr_lower)
else:
obj = objIn.copy()
views = ["x", "y", "z"]
coords = [
range(-50, 51, 8),
range(-80, 50, 10),
range(-40, 71, 9),
] # [-88,-72,-58,-38,-26,8,20,34,46]
cmap = "RdBu_r"
if thr_upper is None and thr_lower is None:
print("Plotting unthresholded image")
else:
if isinstance(thr_upper, str):
print("Plotting top %s of voxels" % thr_upper)
elif isinstance(thr_upper, (float, int)):
print("Plotting voxels with stat value >= %s" % thr_upper)
if isinstance(thr_lower, str):
print("Plotting lower %s of voxels" % thr_lower)
elif isinstance(thr_lower, (float, int)):
print("Plotting voxels with stat value <= %s" % thr_lower)
if save:
path, filename = os.path.split(save)
filename, extension = filename.split(".")
glass_save = os.path.join(path, filename + "_glass." + extension)
x_save = os.path.join(path, filename + "_x." + extension)
y_save = os.path.join(path, filename + "_y." + extension)
z_save = os.path.join(path, filename + "_z." + extension)
else:
glass_save, x_save, y_save, z_save = None, None, None, None
saves = [x_save, y_save, z_save]
if how == "full":
plot_glass_brain(obj.to_nifti(),
display_mode="lzry",
colorbar=True,
cmap=cmap,
plot_abs=False,
**kwargs)
if save:
plt.savefig(glass_save, bbox_inches="tight")
for v, c, savefile in zip(views, coords, saves):
plot_stat_map(obj.to_nifti(),
cut_coords=c,
display_mode=v,
cmap=cmap,
bg_img=resolve_mni_path(MNI_Template)["brain"],
**kwargs)
if save:
plt.savefig(savefile, bbox_inches="tight")
elif how == "glass":
plot_glass_brain(obj.to_nifti(),
display_mode="lzry",
colorbar=True,
cmap=cmap,
plot_abs=False,
**kwargs)
if save:
plt.savefig(glass_save, bbox_inches="tight")
elif how == "mni":
for v, c, savefile in zip(views, coords, saves):
plot_stat_map(obj.to_nifti(),
cut_coords=c,
display_mode=v,
cmap=cmap,
bg_img=resolve_mni_path(MNI_Template)["brain"],
**kwargs)
if save:
plt.savefig(savefile, bbox_inches="tight")
del obj # save memory
return
masker = estimator.masker_
# Drop output maps to a Nifti file
components_img = masker.inverse_transform(estimator.components_)
components_img.to_filename('%s_resting_state.nii.gz' % names[estimator])
components_imgs.append(components_img)
###############################################################################
# Visualize the results
from nilearn.plotting import (plot_prob_atlas, find_xyz_cut_coords, show,
plot_stat_map)
from nilearn.image import index_img
# Selecting specific maps to display: maps were manually chosen to be similar
indices = {dict_learning: 1, canica: 31}
# We select relevant cut coordinates for displaying
cut_component = index_img(components_imgs[0], indices[dict_learning])
cut_coords = find_xyz_cut_coords(cut_component)
for estimator, components in zip(estimators, components_imgs):
# 4D plotting
plot_prob_atlas(components,
view_type="filled_contours",
title="%s" % names[estimator],
cut_coords=cut_coords,
colorbar=False)
# 3D plotting
plot_stat_map(index_img(components, indices[estimator]),
title="%s" % names[estimator],
cut_coords=cut_coords,
colorbar=False)
show()
def plotting_spatial_comp(u,
variances,
masker,
plot_dir='.',
fname=None,
display_mode='ortho',
perc_voxels_to_retain=0.1,
bg_img=None,
save_nifti=False,
verbose=False):
""" Plot, and save as pdf, each spatial estimated component.
Parameters
----------
u : array, shape (n_atoms, n_voxels), the spatial maps
variances : array, shape (n_atoms, ) the order variances for each
components
masker : Nilearn-Masker like, masker class to perform the inverse Nifti
transformation
plot_dir : str, (default='.'), directory under which the pdf is saved
fname : str, (default='u.pdf'), filename under which the pdf is saved
display_mode : None or str, coords to cut the plotting, possible value are
None to have x, y, z or 'x', 'y', 'z' for a single cut
perc_voxels_to_retain : float, (default=0.1), percentage of voxels to
retain when plotting the spatial maps
bg_img : Nifti-like or None, (default=None), background image, None means
no image
save_nifti : bool, (default=False), whether or not to save the image as
Nifti
verbose : bool, (default=False), verbosity level
"""
if display_mode in ['x', 'y', 'z']:
cut_coords = 1
colorbar = False
compress_plot = True
else:
display_mode = 'ortho'
cut_coords = None
colorbar = True
compress_plot = False
n_atoms, n_voxels = u.shape
img_u = []
for k in range(1, n_atoms + 1):
u_k = u[k - 1]
last_retained_voxel_idx = int(perc_voxels_to_retain * n_voxels)
th = np.sort(u_k)[-last_retained_voxel_idx]
expl_var = variances[k - 1]
if compress_plot:
title = "Map-{}".format(k)
else:
title = "Map-{} (explained variance = {:.2e})".format(k, expl_var)
img_u_k = masker.inverse_transform(u_k)
img_u.append(img_u_k)
if bg_img is not None:
plotting.plot_stat_map(img_u_k,
title=title,
colorbar=colorbar,
display_mode=display_mode,
cut_coords=cut_coords,
threshold=th,
bg_img=bg_img)
else:
plotting.plot_stat_map(img_u_k,
title=title,
colorbar=colorbar,
display_mode=display_mode,
cut_coords=cut_coords,
threshold=th)
if save_nifti:
nii_filename = os.path.join(plot_dir, "u_{0:03d}.nii".format(k))
img_u_k.to_filename(nii_filename)
plt.savefig(os.path.join(plot_dir, "u_{0:03d}.pdf".format(k)), dpi=150)
pdf_files = os.path.join(plot_dir, 'u_*.pdf')
if fname is None:
fname = 'u.pdf'
pdf_file = os.path.join(plot_dir, fname)
if compress_plot:
cmd_cat = ("pdfjam --suffix nup --nup 8x5 --no-landscape {} "
"--outfile {}".format(pdf_files, pdf_file))
subprocess.call(cmd_cat, shell=True)
cmd_crop = "pdfcrop {0} {0}".format(pdf_file)
subprocess.call(cmd_crop, shell=True)
else:
cmd = "pdftk {} cat output {}".format(pdf_files, pdf_file)
subprocess.call(cmd, shell=True)
subprocess.call("rm -f {}".format(pdf_files), shell=True)
if verbose:
print("Saving plot under '{0}'".format(pdf_file))
# 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.
components_img = estimator.components_img_
components_img.to_filename('%s_resting_state.nii.gz' %
names[estimator])
components_imgs.append(components_img)
###############################################################################
# Visualize the results
# ----------------------
from nilearn.plotting import (plot_prob_atlas, find_xyz_cut_coords, show,
plot_stat_map)
from nilearn.image import index_img
# Selecting specific maps to display: maps were manually chosen to be similar
indices = {dict_learning: 25, canica: 33}
# We select relevant cut coordinates for displaying
cut_component = index_img(components_imgs[0], indices[dict_learning])
cut_coords = find_xyz_cut_coords(cut_component)
for estimator, components in zip(estimators, components_imgs):
# 4D plotting
plot_prob_atlas(components, view_type="filled_contours",
title="%s" % names[estimator],
cut_coords=cut_coords, colorbar=False)
# 3D plotting
plot_stat_map(index_img(components, indices[estimator]),
title="%s" % names[estimator],
cut_coords=cut_coords, colorbar=False)
show()
design_matrix = pd.DataFrame(np.hstack((tested_var, np.ones_like(tested_var))),
columns=['fluency', 'intercept'])
###########################################################################
# Fit of the second-level model
from nistats.second_level_model import SecondLevelModel
model = SecondLevelModel(smoothing_fwhm=5.0)
model.fit(contrast_map_filenames, design_matrix=design_matrix)
##########################################################################
# To estimate the contrast is very simple. We can just provide the column
# name of the design matrix.
z_map = model.compute_contrast('fluency', output_type='z_score')
###########################################################################
# We compute the fdr-corrected p = 0.05 threshold for these data
from nistats.thresholding import map_threshold
_, threshold = map_threshold(z_map, alpha=.05, height_control='fdr')
###########################################################################
#Let us plot the second level contrast at the computed thresholds
from nilearn import plotting
plotting.plot_stat_map(
z_map,
threshold=threshold,
colorbar=True,
title='Group-level association between motor activity \n'
'and reading fluency (fdr<0.05')
plotting.show()