Automated detection of the CT-angiography spot sign in acute intracerebral hemorrhage

Early hematoma expansion is a major determinant of outcome after intracerebral hemorrhage (ICH). The CTA spot sign predicts hematoma expansion and poor outcome but requires specialized expertise for interpretation. Automated detection facilitates imaging-based selection strategies for hemostatic therapies. We evaluated the diagnostic performance of an automated algorithm for spot sign detection using multiphase CTA. We analyzed patients presenting with spontaneous ICH within 4.5 h of symptom onset who underwent multiphase CTA. Fifty scans were randomly selected from a prospective registry. The automated algorithm independently classified each scan as spot sign positive or negative. An expert neuroradiologist served as reference standard. Sensitivity for spot sign detection was 95.7%, specificity was 74.1%. In this pilot study, automated detection of the CTA spot sign on dynamic CTA was feasible and demonstrated high sensitivity but moderate specificity, supporting further independent validation of CARL. Such tools may standardize spot sign identification, support clinical trial enrollment, and have the potential to facilitate rapid selection of patients at high risk of hematoma expansion.