Disease Mapping
NC SIDS Disease Mapping
Spatial analysis of sudden infant death syndrome across North Carolina counties using BYM and Poisson models with demographic covariates.
Public Health & Environment
Disease mapping, spatial epidemiology, and environmental exposure modeling using Bayesian hierarchical models with spatial random effects.
Disease Mapping
Scottish Lip Cancer Disease Mapping
BYM2 spatial model for mapping relative risk across 56 Scottish districts. Separates spatial and unstructured variation, quantifies the effect of outdoor work on lip cancer incidence, and validates against MCMC with r = 0.9999.
Explore this example
Disease Mapping
Ohio Lung Cancer
Spatio-temporal disease mapping of lung cancer incidence in Ohio counties with temporal trends and spatial random effects.
Spatial Epidemiology
Gambia Malaria
Geostatistical modeling of malaria prevalence in The Gambia using SPDE-based spatial random fields with environmental covariates.
Environmental Health
Spain PM2.5 Exposure
Spatial modeling of fine particulate matter concentrations across Spain using SPDE mesh-based Gaussian random fields.
Classification
Pima Diabetes Classification
Bayesian logistic regression for diabetes classification using the Pima Indians dataset with uncertainty-aware predictions.
Why Bayesian for Public Health?
Public health data is inherently spatial, often sparse, and demands principled uncertainty quantification for policy decisions:
- Spatial smoothing stabilises disease rates in areas with small populations by borrowing strength from neighbors
- Exceedance probabilities identify disease hotspots with quantified confidence, supporting targeted interventions
- Covariate adjustment separates known risk factors from residual spatial patterns
- INLA speed enables routine surveillance and real-time disease mapping where MCMC is impractical