The goal of this work is to model analytic variability in fMRI pipelines as a Feature Model (FM), and to use such FM to explore and sample the fMRI analytic space.
The goal of an FM is to represent the allowed combination of feature in a given software system. A feature is an increment in a system functionality, that offer a configuration option. Some combination of feature can be allowed or forbidden, and such constraints are included in FMs.
For this work, we choose to use the Universal Variability Language (UVL) (Sundermann et al., 2021).