NeuroGraph Preprocessing Functionalities¶
- class Age_Dataset(root, dataset_name, dataset, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Brain_Connectome_Rest(root, name, n_rois, threshold, path_to_data, n_jobs, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- construct_adj_postive_perc(corr)[source]¶
construct adjacency matrix from the given correlation matrix and threshold
- extract_from_3d_no(volume, fmri)[source]¶
Extract time-series data from a 3d atlas with non-overlapping ROIs.
- Inputs:
path_to_atlas = ‘/path/to/atlas.nii.gz’ path_to_fMRI = ‘/path/to/fmri.nii.gz’
- Output:
returns extracted time series # volumes x # ROIs
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Brain_Connectome_Rest_Download(root, name, n_rois, threshold, path_to_data, n_jobs, s3, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- extract_from_3d_no(volume, fmri)[source]¶
Extract time-series data from a 3d atlas with non-overlapping ROIs.
- Inputs:
path_to_atlas = ‘/path/to/atlas.nii.gz’ path_to_fMRI = ‘/path/to/fmri.nii.gz’
- Output:
returns extracted time series # volumes x # ROIs
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Brain_Connectome_Task(root, dataset_name, n_rois, threshold, path_to_data, n_jobs, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- construct_adj_postive_perc(corr)[source]¶
construct adjacency matrix from the given correlation matrix and threshold
- extract_from_3d_no(volume, fmri)[source]¶
Extract time-series data from a 3d atlas with non-overlapping ROIs.
- Inputs:
path_to_atlas = ‘/path/to/atlas.nii.gz’ path_to_fMRI = ‘/path/to/fmri.nii.gz’
- Output:
returns extracted time series # volumes x # ROIs
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Brain_Connectome_Task_Download(root, dataset_name, n_rois, threshold, path_to_data, n_jobs, s3, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- extract_from_3d_no(volume, fmri)[source]¶
Extract time-series data from a 3d atlas with non-overlapping ROIs.
- Inputs:
path_to_atlas = ‘/path/to/atlas.nii.gz’ path_to_fMRI = ‘/path/to/fmri.nii.gz’
- Output:
returns extracted time series # volumes x # ROIs
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Dyn_Down_Prep(root, name, s3, n_rois=100, threshold=10, window_size=50, stride=3, dynamic_length=150)[source]¶
Bases:
object
- Dyn_Prep(fmri, regs, n_rois=100, window_size=50, stride=3, dynamic_length=None)[source]¶
Preprocess fMRI data using NeuroGraph preprocessing pipeline and construct dynamic functional connectome matrices
Args:
fmri (numpy array): fmri image regs (numpy array): regressor array rois (int): {100, 200, 300, 400, 500, 600, 700, 800, 900, 1000}, optional, Number of regions of interest. Default=100. window_size (int) : the length of the window, default = 50 stride (int): default: 3 dynamic_length (int) : length of the timeseries to be considered for dynamic graphs. For memory and compution efficiency, we set dynamic length = 50, default = None, if None, consider the whole timeseries object
- class FI_Dataset(root, dataset_name, dataset, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.
- class Gender_Dataset(root, dataset_name, dataset, transform=None, pre_transform=None, pre_filter=None)[source]¶
Bases:
InMemoryDataset
- property processed_file_names¶
The name of the files in the
self.processed_dir
folder that must be present in order to skip processing.