Skip to content

GriffithsLab/kernel-flow-tools

Repository files navigation

Kernel Flow Tools

Summary

KFTools is a Python library for Python-based analyses of fNIRS and EEG data from the Kernel Flow system.

Please note that the code here is still very preliminary, under active development, and subject to substantial change.

Rationale

As one of the first scientific groups undertaking new research projects with the Kernel Flow system, it became quickly apparent that establishing a wider user community will be one of the best routes to rapid and effective progress for all concerned.

As open source and open science advocates, we have elected to pursue a ‘fully open’ and public development approach here. A chief motivation behind this is to bring in contributors and collaborators who are interested in working together to move things forward more quickly that any of us would be able to individually. So, if you are interested in getting involved, don’t hesitate to reach out to John ( j dot davidgriffiths at gmail dot com ), or just introduce yourself via an issue.

Structure

The objective of KFTools is to act as a set of thin wrappers on actual analysis software. It is Python-based, with a sprinkling of Matlab here and there. The wrapper functions do a few useful data organization things, and have some useful expectiations / knowledge about file structures, experiment types, event coding conventions, etc.

The KFTools functions are mostly based on two established and best-in-class neuroimaging anaysis libraries: MNE (and especially mne-nirs), and Nilearn. There is also some Homer3 analysis functionality.

There are three main components to the code base:

doc folder - Documentation pages text and organization
kftools folder - The importable python library
examples folder - Example usage scripts that become the gallery items in the CI-managed sphinx gallery site

About

Collaborative open-source Kernel Flow TD-fNIRS analysis development

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published