A two-level multimodality imaging Bayesian network approach for classification of partial epilepsy: Preliminary data

► Two-level multi-modality Bayesian network for classification of focal epilepsy. ► 1st level Bayesian network (GAMMA) identifies regions of GM and FA abnormalities. ► 2nd level Bayesian network uses information to provide group membership probabilities. ► Accuracy to distinguish between 3 different epilepsy types was 64-87%.