A glossary of concepts

Seeing how the parts of lyman fit together requires understanding several hierarchies of concepts that structure the codebase, interactions with the library, and the organization of raw and processed data on disk.

Dataset organization

The first hierarchy of concepts pertains to the granularity of data:

subject

An individual who participates in the project.

session

A continuous block of time that a subject was in the scanner.

run

A continuous acquisition of data from the scanner corresponding to a time series image.

Analysis parameterization

The second hierarchy of concepts pertains to the granularity of analysis parameterization:

project

Corresponds to the whole data set that any invocation of lyman will interact with.

experiment

Corresponds to a subset of the data that was collected with consistent acquisition parameters.

model

Corresponds to the set of signal-processing parameters and elements of a linear model that are used to analyze an experiment’s data in a particular way.

Directory structure

lyman directory

The location where files that define analysis parameters are stored. Should be identified with the LYMAN_DIR environment variable when executing the workflows.

data directory

The location where data that will be input into the lyman workflows are stored. Should also correspond to the Freesurfer SUBJECTS_DIR.

proc directory

The location where lyman will write processed output files for persistent storage after workflow execution.

cache directory

The location where lyman will write intermediate files for temporary storage during workflow execution.