fbesag is a non-stationary extension of the classical Besag model, developed to model spatial variation across heterogeneous sub-regions. It supports structured random effects with varying precision and is fully compatible with the INLA framework.
📄 Paper
The method is described in the article:
Non-stationary Bayesian spatial model for disease mapping based on sub-regions
Published in Statistical Methods in Medical Research, 2024.
🔍 Applications
fbesag has been successfully applied to disease mapping in Brazil and is ideal for:
- Public health risk modeling
- Non-stationary spatial smoothing
- Bayesian inference over irregular areal domains
🤝 Collaborations
We are looking for collaborators interested in:
- Extending fbesag to BYM2 and spatio-temporal models
- Applying it in new medical, environmental, or social science contexts
- Contributing to the development of flexible spatial priors
📬 Stay Connected
If you’d like updates on fbesag or if you're interested in collaborating—please reach out:
📧 esmail.abdulfattah@kaust.edu.sa