Usage

Start by importing EasyHCP.

import easyhcp

easyhcp.hcpscraper.setup_credentials()

Set Up AWS credentials from access keys into a credentials file

Notes

This function will open/create a file ‘~/.aws/credentials’, that will then include a section: [hcp] AWS_ACCESS_KEY_ID=XXXXXXXXXXXXXXXX AWS_SECRET_ACCESS_KEY=XXXXXXXXXXXXXXXX The keys are credentials that you can get from HCP (see https://wiki.humanconnectome.org/display/PublicData/How+To+Connect+to+Connectome+Data+via+AWS)

easyhcp.hcpscraper.get_structural_data(subject_list, scan_type, preprocessed=True, MNISpace=True, out_dir='.')

Gets structural data for a list of subjects, and stores them in BIDS-like format in the specified output directory

Parameters:

subject_list : list

List of subjects to get data for

scan_type: list

List of types of structural scans to get

preprocessed : bool

Gets preprocessed data

MNISpace : bool

Gets data registered in MNI Space

out_dir : str

Path to output directory

Notes

Local filenames are changed to match our expected conventions. .. [R993021282679-1] Gorgolewski et al. (2016). The brain imaging data structure, a format for organizing and describing outputs of neuroimaging experiments. Scientific Data, 3:: 160044. DOI: 10.1038/sdata.2016.44.

easyhcp.hcpscraper.get_resting_data(subject_list, scan_run=['rfMRI_REST1_LR', 'rfMRI_REST2_LR', 'rfMRI_REST1_RL', 'rfMRI_REST2_RL'], preprocessed=True, MNISpace=True, out_dir='.')

Gets resting data for runs for a list of subjects, and stores them in BIDS-like format in the specified output directory

Parameters:

subject_list : list

List of subjects to get data for

scan_run: list

List of types of structural scans to get

preprocessed : bool

Gets preprocessed data

MNISpace : bool

Gets data registered in MNI Space

out_dir : str

Path to output directory

Notes

Local filenames are changed to match our expected conventions. .. [Red044de8b140-1] Gorgolewski et al. (2016). The brain imaging data structure, a format for organizing and describing outputs of neuroimaging experiments. Scientific Data, 3:: 160044. DOI: 10.1038/sdata.2016.44.

easyhcp.hcpscraper.train_test_split(root: str, split_folds: (0.7, 0.2, 0.1), scan_type: list, convert_to_npy: bool = False) → None

splits an hcp dataset into train, test, val and converts the .nii.gz files to .npy for easier processing checks shape to ensure t1 and t2 are same dim

Parameters:

root: str

root directory where raw files are stored

split_folds: tuple(float, float, float)

What fraction to divide data in for train, test, val

scan_type: list

What scans to divive into train - test splits