51
EVENT FORMAT
Lecture
2020
Januar y
12-2 3
Using Smart EEG Analytics to Personalize Treatment for
Medically Refractory Epilepsy
In this talk, Sridevi Sarma will describe a product, EZTrack, which uses personalized brain network models estimated
from EEG data to accurately localize where seizures start in the brain - the epileptogenic zone (EZ). Medications can
suppress the EZ, however, over 30% of epilepsy patients are drug-resistant and account for 80% of the total costs
to treat epilepsy. For these patients, clinicians can surgically remove the EZ to stop seizures altogether, or electrically
stimulate the EZ to suppress seizures. Unfortunately, treatment outcomes average 50% success. The problem is that
EZ can be very difficult to find. In clear cases, a lesion can be found on an MRI scan and the EZ is determined to be
in the vicinity of the lesion. In over 50% of cases, however, MRI scans look normal, forcing clinicians to localize the
EZ by recording EEG activity from several brain regions in the patient.
A team of clinicians then visually inspect streams of EEG data both on the scalp and inside the brain to localize the
EZ. EZTrack is an analysis package that processes EEG data in minutes and delivers heat maps that highlight regions
highly correlated to the EZ. EZTrack has been validated on a retrospective study consisting of 75 patients and 366
seizure events and demonstrated an increase of 20% over clinicians in predicting treatment outcomes. Importantly,
had clinicians been using EZTrack, they would have been warned about misdiagnosis of the EZ and/or misplacement
of electrodes and could have changed their treatment plan.
SPEAKER Sridevi Sarma
CREDITS 1
Associate Professor of Biomedical Engineering. Whiting School of Engineering,
Johns Hopkins School of Medicine
DESCRIPTION