A quantitative EEG index for the recognition of arterial ischemic stroke in children

Abstract

ObjectiveTo describe and assess performance of the Correlate Of Injury to the Nervous system (COIN) index, a quantitative electroencephalography (EEG) metric designed to identify areas of cerebral dysfunction concerning for stroke. MethodsCase-control study comparing continuous EEG data from children with acute ischemic stroke to children without stroke, with or without encephalopathy. COIN is calculated continuously and compares EEG power between cerebral hemispheres. Stroke relative infarct volume (RIV) was calculated from quantitative neuroimaging analysis. Significance was determined using a two-sample t-test. Sensitivity, specificity, and accuracy were measured using logistic regression. ResultsAverage COIN values were −34.7 in the stroke cohort compared to −9.5 in controls without encephalopathy (p = 0.003) and −10.5 in controls with encephalopathy (p = 0.006). The optimal COIN cutoff to discriminate stroke from controls was −15 in non-encephalopathic and −18 in encephalopathic controls with >92% accuracy in strokes with RIV > 5%. A COIN cutoff of −20 allowed discrimination between strokes with <5% and >5% RIV (p = 0.027). ConclusionsWe demonstrate that COIN can identify children with acute ischemic stroke. SignificanceCOIN may be a valuable tool for stroke identification in children. Additional studies are needed to determine utility as a monitoring technique for children at risk for stroke.