SozRank: A new approach for localizing the epileptic seizure onset zone

摘要: Epilepsy is one of the most common neurological disorders affecting about 1% world population. For patients with focal seizures that cannot be treated antiepileptic drugs, treatment a surgical procedure for removal seizure onset zone (SOZ). In this work we introduce an algorithm automatic localization (SOZ) in epileptic based on electrocorticography (ECoG) recordings. The proposed builds upon hypothesis abnormal excessive (or synchronous) neuronal activity brain leading to starts SOZ and then spreads other areas brain. Thus, when starts, signals recorded at electrodes close should have relatively large causal influence rest signals. executed two steps. First, represents set using directed graph which nodes correspond recording edges’ weights quantify pair-wise between Then, infers from estimated variant PageRank followed by novel post-processing phase. Inference results 19 show match inferred approach expert neurologists (success rate 17 out 19).