Overview of FEMA analysis in DEAP
Introduction
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.
Before you start
- Before you can start an analysis you will need to save a dataset which includes the covariates you wish to include in your model. See the DEAP documentation for more information on how to save datasets in DEAP.
Contents
| Document | Description |
|---|---|
| Model specification | A step-by-step guide to specifying a LME model using FEMA in DEAP and running the analysis. |
| Results visualization | Using the DEAP surface volumetric viewers to visualize the results of the analysis. |
| Splines | A guide to using splines in FEMA. |
References
Parekh, Pravesh, Chun Chieh Fan, Oleksandr Frei, et al. 2024. “FEMA: Fast and Efficient Mixed-Effects Algorithm for Large Sample Whole-Brain Imaging Data.” Human Brain Mapping 45 (2): e26579. https://doi.org/10.1002/hbm.26579.
Parekh, Pravesh, Nadine Parker, Diliana Pecheva, et al. 2026. FEMA-Long: Modeling Unstructured Covariances for Discovery of Time-Dependent Effects in Large-Scale Longitudinal Datasets. March, 2025.05.09.653146. https://doi.org/10.1101/2025.05.09.653146.