MRI

To install dependencies for MRI preprocessing, use one of the following:

pip install -r requirements_mri_preproc.txt conda create -n mri_env --file requirements_mri_preproc.txt

download-kaggle-series-folder-wise.ipynb

Run in Kaggle https://www.kaggle.com/competitions/rsna-intracranial-aneurysm-detection to zip and download a specific DICOM series folder along with train.csv metadata. Outputs: {SeriesInstanceUID}.zip and train.csv with download links.

Dicom_Loading_Preprocessing.ipynb

Inspects raw DICOM MRI data for one patient, extracts metadata, orientation, voxel spacing, calculates isotropy/anisotropy and visualizes a slice. Prepares data for downstream NIfTI conversion or deep learning tasks in the RSNA aneurysm challenge.

FMRI

ABIDE .1D Downloader

Python script to download ABIDE preprocessed ROI time series files using the AWS CLI. Download the phenotypic data file as well. Click here to download the phenotypic CSV file

Right-click and choose “Save As” if it opens as text.

Requirements

pip install pandas tqdm wget

Usage

python download_abide.py
--pheno_csv ./data/Phenotypic_V1_0b_preprocessed1.csv
--out_dir ./abide_data/cpac_cc200
--pipe cpac
--roi cc200
--fg filt_noglobal

FCN

ABIDE FCN Generator

Python script to generate Functional Connectivity Networks (FCNs) from ABIDE preprocessed ROI time series .1D files using Nilearn.

Requirements

pip install pandas tqdm nilearn numpy

Usage

python generate_fcn.py
--ts_dir ./Abide/abide_dparsf_cc200
--pheno_csv ./Abide/Phenotypic_V1_0b_preprocessed1.csv
--out_dir ./Abide/outputs/dparsf
--roi cc200
--method correlation