Abstract

Background and Purpose— The aim of the study was to investigate the usefulness of the computed tomography (CT) island sign for predicting early hematoma growth and poor functional outcome. Methods— We included patients with spontaneous intracerebral hemorrhage (ICH) who had undergone baseline CT within 6 hours after ICH symptom onset in our hospital between July 2011 and September 2016. Two readers independently assessed the presence of the island sign on the admission noncontrast CT scan. Multivariable logistic regression analysis was used to analyze the association between the presence of the island sign on noncontrast admission CT and early hematoma growth and functional outcome. Results— A total of 252 patients who met the inclusion criteria were analyzed. Among them, 41 (16.3%) patients had the island sign on baseline noncontrast CT scans. In addition, the island sign was observed in 38 of 85 patients (44.7%) with hematoma growth. Multivariate logistic regression analysis demonstrated that the time to baseline CT scan, initial hematoma volume, and the presence of the island sign on baseline CT scan independently predicted early hematoma growth. The sensitivity of the island sign for predicting hematoma expansion was 44.7%, specificity 98.2%, positive predictive value 92.7%, and negative predictive value 77.7%. After adjusting for the patients’ age, baseline Glasgow Coma Scale score, presence of intraventricular hemorrhage, presence of subarachnoid hemorrhage, admission systolic blood pressure, baseline ICH volume, and infratentorial location, the presence of the island sign (odds ratio, 3.51; 95% confidence interval, 1.26–9.81; P =0.017) remained an independent predictor of poor outcome in patients with ICH. Conclusions— The island sign is a reliable CT imaging marker that independently predicts hematoma expansion and poor outcome in patients with ICH. The noncontrast CT island sign may serve as a potential marker for therapeutic intervention.