neurotools

Creative and original name? Maybe not. Lots of cool and helpful tools for working with neuroimaging data? Yes.

What's Included?

neurotools is split into a number of different directories / sub-packages, each including different functionality (though key pieces from different sub-directories are shared allowing for some cool functionality!). Be low are some brief descriptions of the different sub-packages.

• loading : This directory encompasses a set of utilities for loading generic neuroimaging data, as well as some ABCD study specific loading utilities.

• misc : A set of misc. functions that don't fit in nicely anywhere else.

• plotting : Tools and functions for plotting surface representations of the data. Also smart functions for easy quick plots and collages of surface representations. Still a big work in progress as plotting can be messy, but in general these tools are designed to help make it easier and more stream-lined.

• random : Different from misc... this includes functions that are based on randomness. These include utilities for generating constrained permutations based on block structures and also for generating random / null surface parcellations.

• rely : Included here are utilities for running 'reliability' like tests. These were featured in https://github.com/sahahn/ABCD_Consortium_Analysis

• stats : Featured here are wrappers and other code for running statical tools, from basic functions (i.e., calculating cohen's d effect size with or without NaN's) to running mixed linear models.

• transform : Tools for transforming data - that is to say, tools for extracting ROIs from surface level data are provided, with utility for inverse_transforming them as well. Also included, but still more as a work in progress, are some utilities for converting between different common surfaces, e.g., from fs LR 32k surface parcellations to a version without medial walls and with the sub-cortical regions included.

Besides these main functionalities, another main folder is included, called "slurm". This folder will store examples for using the neurotools package on a SLURM cluster is a massively paralell way. Included are, and in the future will be, examples with extra scripts and pieces responsible for submitting and collecting jobs.

Install

The easiest way to install is through pypip via

pip install bp-neurotools

You can also install the latest version via github directly, by cloning (git clone https://github.com/sahahn/neurotools) and then pip installing (navigate into the directory then run pip install .) the repository.

This will install for you the latest version. Note that on running it for the first time, it will install for you automatically the required external data that the library leverages.