A Novel Method for Seizure Detection In Intracranial Eeg Recordings

Nearly 1 out of every 100 person on the planet is afflicted to Epilepsy, which is characterized by the occurrence of spontaneous seizures[ 8]. A victim maybe be given sufficiently high dose of anticonvulsant medication, in order to prevent seizures, but they may suffer from side effects. In 20-40% of cases with epilepsy, medication is not effective, even after surgical removal of brain tissues that cause epilepsy, many continue to still experience unprompted seizures. In spite of the fact that seizures occur sporadically, the patients suffer from persistent anxiety, due to possibility of a seizure occurring. The potential to help the patients in leading a more normal life can be done with the help of Seizure forecasting systems. If we are able to predict seizures as early as possible then we have the chance of effectively aborting the seizure using responsive neurostimulation. However if we fail to detect the seizure in its early stages then it becomes very hard to abort seizures. This model aims to create a suitable model to detect the seizure in its early stages so that proper actions can be taken.