Lyman: an fMRI Analysis Ecosystem¶
Note
These are the docs for version 1.0 of lyman. Documentation for the 2+ series, a nearly complete rewrite of the library, are available here.
Lyman is a high-level ecosystem for analyzing neuroimaging data using open-source software. It aims to support an analysis workflow that is powerful, flexible, and reproducible, while automating as much of the processing as possible.
Lyman offers a command-line based interface to a set of pipelines, where FSL, Freesurfer, and Python-based tools are integrated using Nipype. These pipelines will take raw Nifti files and process them all the way through a basic group analysis with minimal manual intervention. Important intermediate files that might be useful for later analysis are saved in predictable locations at the completion of the pipelines.
Because the processing is heavily automated, lyman also generates a number of static plots and images that are useful for understanding the results of the analyses and diagnosing any problems that might arise. These files are stored alongside the data they correspond with in the output directories. Although it is possible to manually browse them, a much better approach is to use the companion zielger webapp, which is tightly integrated with the lyman results and makes it very easy to understand what has happened with your data.
Lyman is provided freely and with open source in the hope that it might be useful. However, there is no guarantee of support or stability. There has been some effort put into documentation, but not every aspect of using the tools will be obvious. Lyman supports a specific approach to analyzing data and may not work for every experiment. Finally, the code may change between releases in a way that is not backwards compatible.