Bayesian Virtual Epileptic Patient (BVEP)

Bayesian Virtual Epileptic Patient (BVEP): a probabilistic framework designed to infer the spatial map of epileptogenicity in a personalized large-scale brain model of epilepsy spread by PPLs using No-U-Turn Sampler (NUTS) and Automatic Differentiation Variational Inference (ADVI), and also now using the state-of-the-art deep learning algorithms for conditional density estimation in simulation-based-inference (SBI) framework.

Installation:

For simulation using TVB:

https://www.thevirtualbrain.org/tvb/zwei

For inference using Stan:

https://mc-stan.org/

For inference using PyMC3:

https://docs.pymc.io/

For inference using SBI:

https://www.mackelab.org/sbi/

Notebooks:

Bayesian inference of 2D Epileptor model:

https://lab.ebrains.eu/user/user-redirect/lab/tree/drive/Shared%20with%20all/Bayesian%20Virtual%20Epileptic%20Patient/BVEP_noncen_pymc3_patient1.ipynb

Cross-validation and hypothesis inference in 2D Epileptor model:

https://lab.ebrains.eu/user/user-redirectm/lab/tree/drive/Shared%20with%20all/Bayesian%20Virtual%20Epileptic%20Patient/EpileptorInferHyposPrior.ipynb

Bayesian inference of VEP whole-brain network model:

shared/Bayesian Virtual Epileptic Patient/BVEP_ode_sbi_sourcelevel_patient1_savesim_v18.ipynb

https://lab.ch.ebrains.eu/hub/user-redirect/lab/tree/shared/Bayesian%20Virtual%20Epileptic%20Patient/BVEP_sde_sbi_maf_seeg_GrExp_patient1.ipynb

SBI-VEP: