Greetings,
I have have thinking about using Glasser parcellation on ABIDE-I dataset. However, there are few hurdles.
There are few atlases already provided by nilearn but the Glasser 360 atlas, is there any script or .nii Glasser atlas which i could use to achieve the task.
Thank you very much.
Regards
Hi @Saqib_Mamoon,
Does this work for you? GitHub - brainspaces/glasser360: 360-node volumetric atlas derived from the multimodal parcellation from Glasser and colleagues
Best,
Steven
Thank you sharing, Yes, I think will work for me.
from nilearn.input_data import NiftiLabelsMasker
from nilearn.datasets import fetch_icbm152_2009
mnitemp = fetch_icbm152_2009()
glassermasker = NiftiLabelsMasker(labels_img='Glasser_masker.nii.gz',mask_img=mnitemp['mask'])
glassermasker.fit()
# In[3]:
import numpy as np
from nilearn.datasets import fetch_abide_pcp
import pandas
import os
df = pandas.read_csv('Phenotypic_V1_0b_preprocessed1.csv', sep='\t')
site_id=np.unique(df['SITE_ID'])
print(site_id)
for S in (site_id):
data_dir='test'
#DX_GROUP: 1 is autism, 2 is control
abidedata_func_normal=fetch_abide_pcp(data_dir=data_dir, n_subjects= None, pipeline='cpac', band_pass_filtering=True, global_signal_regression=True, derivatives=['func_preproc'], quality_checked=True,DX_GROUP=2,SITE_ID=S)
abidedata_func_autism=fetch_abide_pcp(data_dir=data_dir, n_subjects= None, pipeline='cpac', band_pass_filtering=True, global_signal_regression=True, derivatives=['func_preproc'], quality_checked=True,DX_GROUP=1,SITE_ID=S)
print(len(abidedata_func_normal['func_preproc']))
print(len(abidedata_func_autism['func_preproc']))
## Extaction of time series for normal
abide_ts=[]
abide_ts_normal=[]
for i in range(len(abidedata_func_normal['func_preproc'])):
ts_nor = glassermasker.transform(imgs=abidedata_func_normal['func_preproc'][i])
ts_normal = ts_nor
abide_ts_normal.append(ts_normal)
abide_ts.append(ts_normal)
## Extaction of time series for autism
abide_ts_autism=[]
for i in range(len(abidedata_func_autism['func_preproc'])):
ts_aut = glassermasker.transform(imgs=abidedata_func_autism['func_preproc'][i])
ts_autism = ts_aut
abide_ts_autism.append(ts_autism)
abide_ts.append(ts_autism)
print(len(abide_ts))
labels=np.concatenate((np.zeros(len(abide_ts_normal)),np.ones(len(abide_ts_autism))))
print(labels.shape)
np.savez_compressed("site_"+str(S)+"time_series.npz",
labels=labels,
abide_ts_normal=abide_ts_normal,
abide_ts_autism=abide_ts_autism,
abide_ts=abide_ts
)
shutil.rmtree('test/ABIDE_pcp')
I have this section of the code, and I am running it, the results are not out-yet, the atlas is used to parcellate the data after it is downloaded. Do you think this is the right to do this?
Also I kept the band_pass_filtering=True, global_signal_regression=True, is this good in-general?
Thanks!
This isn’t the right solution, it didn’t work. is there any other script which will do the work?
You are welcome to use the MNI 2009c asymmetric volume space version of the Glasser atlas we developed for AFNI. It is distributed by default with AFNI. Some earlier versions had poor alignment with any MNI space.
Hi @dglen Thank you for your valuable suggestion. Could you please give an example or how it could be done in python language. Thanks in advance.
from nilearn.maskers import NiftiLabelsMasker
from nilearn.datasets import fetch_icbm152_2009
mnitemp = fetch_icbm152_2009()
glassermasker = NiftiLabelsMasker(labels_img='glasser360MNI.nii.gz',mask_img=mnitemp['mask'])
glassermasker.fit()
# In[3]:
import numpy as np
count = []
for i in range(0, 360):
count.append(i)
from nilearn.datasets import fetch_abide_pcp
import pandas
import os
df = pandas.read_csv('Phenotypic_V1_0b_preprocessed1.csv', sep=',')
site_id=np.unique(df['SITE_ID'])
print(site_id)
for S in (site_id):
data_dir='test'
#DX_GROUP: 1 is autism, 2 is control
abidedata_func_normal=fetch_abide_pcp(data_dir=data_dir, n_subjects= None, pipeline='cpac', band_pass_filtering=True, global_signal_regression=True, derivatives=['func_preproc'], quality_checked=True,DX_GROUP=2,SITE_ID=S)
abidedata_func_autism=fetch_abide_pcp(data_dir=data_dir, n_subjects= None, pipeline='cpac', band_pass_filtering=True, global_signal_regression=True, derivatives=['func_preproc'], quality_checked=True,DX_GROUP=1,SITE_ID=S)
##########################################Normal#########################################################################
## Extaction of time series for normal
abide_ts=[]
print(abide_ts)
abide_ts_normal=[]
filename_normal=[]
for i in range(len(abidedata_func_normal['func_preproc'])):
ts_nor = glassermasker.transform(abidedata_func_normal['func_preproc'][i])
print(abidedata_func_normal['func_preproc'][i])
filename = os.path.basename(abidedata_func_normal['func_preproc'][i])
file_name_without_extension = filename[:-7]
print(file_name_without_extension)
ts_normal = ts_nor[:,count]
abide_ts_normal.append(ts_normal)
abide_ts.append(ts_normal)
filename_normal.append(file_name_without_extension)
# Create a directory to store the text files
text_file_dir = 'Normal'
if not os.path.exists(os.path.join(S, text_file_dir)):
os.makedirs(os.path.join(S, text_file_dir))
# Iterate over the file paths in the list
for file, name in zip(abide_ts, filename_normal):
#Convert to dataframe
df = pd.DataFrame(file)
# Create a text file for the DataFrame
text_file_path = os.path.join(S, text_file_dir, '{}.txt'.format(name))
# Save the DataFrame to the text file
df.to_csv(text_file_path, index=False, header=False)
############################################# Autism ########################################################################
## Extaction of time series for normal
abide_ts=[]
print(abide_ts)
abide_ts_autism=[]
filename_autism=[]
for i in range(len(abidedata_func_autism['func_preproc'])):
ts_nor = glassermasker.transform(abidedata_func_autism['func_preproc'][i])
print(abidedata_func_autism['func_preproc'][i])
filename = os.path.basename(abidedata_func_autism['func_preproc'][i])
file_name_without_extension = filename[:-7]
print(file_name_without_extension)
ts_autism = ts_nor[:,count]
abide_ts_autism.append(ts_autism)
abide_ts.append(ts_autism)
filename_autism.append(file_name_without_extension)
# Create a directory to store the text files
text_file_dir = 'Autism'
if not os.path.exists(os.path.join(S, text_file_dir)):
os.makedirs(os.path.join(S, text_file_dir))
# Iterate over the file paths in the list
for file, name in zip(abide_ts, filename_autism):
#Convert to dataframe
df = pd.DataFrame(file)
# Create a text file for the DataFrame
text_file_path = os.path.join(S, text_file_dir, '{}.txt'.format(name))
# Save the DataFrame to the text file
df.to_csv(text_file_path, index=False, header=False)
Could you please guide where to make changes to use “MNI 2009c asymmetric volume space version of the Glasser atlas”, I have this code running.
bundle of thanks!
I expect you would change the NIFTI file that you have been using to MNI_Glasser_HCP_v1.0.nii.gz instead.
Hi @dglen Thank you for your guidance. So far what I understand is, The “MNI_Glasser_HCP_v1.0.nii.gz” is already in MNI space, so don’t have to use “mask_img”, so just have to directly apply the “MNI_Glasser_HCP_v1.0.nii.gz” to the data.
Like:
glassermasker = NiftiLabelsMasker(labels_img='MNI_Glasser_HCP_v1.0.nii.gz',mask_img=None)
glassermasker.fit()
and later in my case:
ts_nor = glassermasker.transform(abidedata_func_normal['func_preproc'][i])
Have i got your point right?
Thanks in advance.
I don’t know anything about the calling code here, but there doesn’t seem anything specific to any particular values, so it seems reasonable to try it with our Glasser parcellation in MNI space. I can’t tell what the glassermasker.transform does. It might be useful to find out what you intend to do with the Glasser atlas.
MANAGED BY INCF