Pilot Deployment of Viz–Intracranial Hemorrhage for Intracranial Hemorrhage Detection: Real-World Performance in a Stroke Code Cohort

ntracranial hemorrhage (ICH) is associated with a high disease burden and poor outcomes. During the workup of stroke codes, it is essential that ICH is excluded. In this study, we report the real-life performance of Viz ICH, in a prospective consecutive series of stroke code patients. The Figure depicts the main processing steps of the algorithm. and September 2020. The NCCT impressions served as the clinical reference standard test, while Viz ICH served as the index test. Although Viz ICH offers a volumetric segmentation result, in instances where ICH was detected, this output was translated into a qualita-tive, binary (positive or negative) characterization. Exclusion criteria included contrast contamination (eg, if NCCT was performed following CT angiography), nondiagnostic NCCTs (due to motion artifact), and metal artifact. A patient workflow is provided (Figure S1).