Document resource
Background Acute ischemic stroke (AIS) and large vessel occlusion (LVO) are neurological emergencies requiring rapid diagnosis and intervention. Electroencephalography (EEG) is a non-invasive tool that may aid early diagnosis. While few quantitative EEG parameters have been evaluated, many of these are not routinely used in clinical practice, and the utility of qualitative EEG remains unclear.Methods 100 patients with clinically suspected stroke were recruited from the emergency department (ED) of a comprehensive stroke centre by FAST criteria. Patients underwent EEG in addition to their stroke work-up. EEGs were assessed qualitatively by two blinded senior neurologists and stratified into normal and abnormal waveforms (focal or generalised slowing, epileptiform activity and periodic discharges).Results Performance of EEG in predicting AIS and LVO yielded a sensitivity of 47.9% and 63.2%, specificity of 65.5% and 60.5%, positive predictive value (PPV) of 85.3% and 27.3%, and negative predictive value (NPV) of 33.9% and 87.5%, respectively. When EEG was combined with clinical variables in a logistic regression to predict AIS and LVO, sensitivity was 91.3% and 52.6%, specificity 60.7% and 96.2%, PPV 85.1% and 76.9%, and NPV 73.9% and 89.3%, respectively. Asymmetrical slowing on EEG was significantly associated with LVO ( p = 0.007), with 42.1% of LVO demonstrating this on EEG.Conclusions The data highlights the potential for integrating EEG with clinical variables to improve diagnostic accuracy in AlS and LVO. While EEG alone had variable utility, its combination with clinical variables enhanced predictive performance. Further studies with larger cohorts are needed to validate these findings.