FEMA Analysis
Fast and Efficient Mixed-effects Algorithm
Fast and Efficient Mixed-effects Algorithm (FEMA) is a MATLAB-based package for running mass-univariate linear mixed-effects (LME) and generalized additive mixed models (GAMMS) analyses on large-scale data. This container packages FEMA together with its runtime dependencies so it can be reproduced within DEAP. More details on FEMA are available on GitHub.
The Fast and Efficient Mixed-effects Algorithm (FEMA) fits linear mixed-effects models (LME) to large-scale neuroscience data (Parekh et al. 2024, 2026). FEMA makes possible LME analyses of whole-brain vertex-wise, voxel-wise, connectome-wide, region-of-interest (ROI) imaging and non-imaging data in large samples, accounting for complex designs and dependencies among observations. The FEMA implementation in DEAP allows you to specify complex statistical models and visualize the results using a custom UI.
FEMA Container
Users can pull the container using the command docker pull deapscience/deap_fema:<tag>.
The <tag> can be specified as latest, which refers to the most recent version of the FEMA release. Alternatively, it can be a specific version of FEMA along with the container version. The FEMA version consists of the three numbers following the v, while the container version is represented by a single number that comes after the c.
For example, in FEMA_v4_0_2_c0, the FEMA version is 4.0.2, and the container version is 0. The FEMA version corresponds to the version listed on the cmig_tools [FEMA release page[(https://github.com/cmig-research-group/cmig_tools/releases). For the majority of FEMA releases, the container version is typically 0.
The container uses Ubuntu 24.04 as the OS. It contains MATLAB r2024b (no GUI), R 4.5.1, and a few R data science packages such as dplyr and arrow.
The container uses the same license as FEMA.