Using graph theory measurements acquired from resting state fMRI data combine with machine learning methods to investigate abnormalities in temporal lobe epilepsy and classification.

Many studies has shown structural damage in TLE caused by seizure propagation. We use graph theoretical approach to look at network differences in TLE's brain in order to find abnormalities that may cause seizure. We find out that subcortical regions such as thalamus and hippocampus are abnormally more connected together and with cerebellar regions and these regions are generally less involved in transferring information to other part of the brain from graph theoretical respect of view. In other word, any information pulse that generated in these regions, will circulate faster within these regions which might be the reason for seizure.