Getting started

graynet currently offers the following streams of processing:

  • Cortical graynet : using vertex-wise ROIs defined on the cortex. This is useful to analyze network-level features based on cortical thickness, curvature, sulcal depth and gyrification. Base-level features can easily obtained from running Freesurfer

  • Volumetric graynet stream : using voxel-wise ROIs defined over the whole brain relying on a volumetric atlas. This is useful to analyze network-level features based on gray matter density dervied from voxel-based morphometry (VBM) or similar approaches. Base-level features can easily obtained from CAT12 toolbox within the SPM ecosystem.

We plan to offer the following soon:

In both streams, in addition to the computation of pair-wise network-level features, graynet will help you compute ROI-wise statistics (individual, not pair-wise) for visualization (median thickness in PCG, or variance in GM density within amygdala), as well as to serve as a baseline for network-level features.

The package offers both Command line interface and an API, to better integrate with your workflow. However, the CLI is the recommended/most-tested gateway.

For uniform processing across subjects, graynet needs:

  • an atlas with pre-defined ROIs

  • each subject be registered to that atlas (vertex- or voxel-wise), so ROIs correspond across all subjects

  • the extracted features are in a format readable by nibabel (Freesurfer formats and Nifti are strongly recommended)

The following steps should help you get started and going quickly:

  • Install graynet using this command in a terminal: pip install -U graynet

  • Refer to Cortical graynet and Volumetric graynet pages for more details on the individual streams of processing and examples.

If you run into any issues, or have a feature wish or suggestions, please let me know here by opening an issue.

Thanks for trying out graynet. I’d appreciate if you can cite it using the details in Citation.