htwangtw/hcp-surface-format

HCP surface format from preprocessed volume/freesurface data

Volume to HCP CIFTI

Python functions (and bash scripts) to project freesurfer/MNI volume space data to HCP CIFTI files and downsample to resolutions < 32k.

Data input

Structural - Freesurfer recon-all results (white matter boarder visually inspected)
Functional - Basic preprocessed resting state timeseries and average volumes in MNI space.

fMRI preprocessing:

• Motion correction
• Noise regression
• Band-pass filtered
• No smoothing

Optional:

• Global signal regression

Environment setup

Python:

• Python 3.6 and above is required

The following softwares and databases are required:

• HCP pipeline (minimum: files under HCPpipeline/global/)
• BALSA_database
• Freesurfer
• FSL
• Connectome workbench

After installation, the following environment variables are also needed by Python functions:

• WB_DIR = /path/to/workbench/binary/directory
• HCP_PIPELINES_DIR = /path/to/hcp/pipelines/repository
• HCP_STANDARD_MESH_ATLASES_DIR = /path/to/hcp/standard/meshes (under $HCP_PIPELINES_DIR/global/templates/standard_mesh_atlases by default)
• HCP_BALSA_DIR = /path/to/balsa/database

Python [WIP]

An example of how to use the Python functions is included in playground.py

Bash

The pipeline can be run in the following order

common space template

Only need to run these files once

1. CreateNewResTemplate.sh
2. DownsampleGroupTemplate.sh

subject preprocessing

1. GiftiReady.sh
2. GoodvoxelsRibbon.sh
3. NeocorticalResampler.sh
4. SubcorticalResampler.sh