New EEG Technology Facilitates Seizure Screening in Urgent Care Settings
A group of researchers from Infinite Biomedical Technologies (IBT) and The Johns Hopkins University School of Medicine presented findings demonstrating the accuracy of investigational screening technology designed to assess if a patient's symptoms should be classified as a seizure or another condition.
"Access to an EEG reading, the gold standard for identifying and classifying seizures, is limited in most urgent care settings. Often the EEG machine and/or technologists are not readily available to the Emergency Department (ED), and eventual diagnosis by a specialist may be delayed for hours, or in some cases days," said study author Peter W. Kaplan, MB, FRCP, Professor of Neurology, The Johns Hopkins University School of Medicine, Baltimore, MD.
To address this problem, the collaborative team, with grant support from the National Institutes of Health, designed the Seizure Vector (SV) algorithm to express EEG readings as a numeric seizure score. Based on the score, ED staff could quickly classify and triage patients. Potential classifications include:
* Epileptic seizures: refer to a neurologist for further evaluation.
* Non-epileptic events: refer for neurological, medical or psychiatric evaluation.
The team's hope is that this technology, when used in the ED setting, will enable first-response personnel to screen for seizures in patients and make rapid triage decisions, such as timely referral to a specialist for evaluation, diagnosis and treatment.
To validate the algorithm, researchers collected EEGs from 40 adults with a variety of seizure types, and a blinded epileptologist classified them into "normal" or "seizure" categories. A total of 2,035 episodes of seizures and 3,867 episodes of normal data were recorded. When the SV algorithm was applied to the same recordings, it differentiated between "seizure" and "normal" episodes with 95.0% sensitivity and 95.2% specificity.