This retrospective validation study aimed to assess the accuracy of qER–NCCT, a fully automated software by Qure.ai, for intracerebral hemorrhage (ICH) volume quantification on non-contrast computed tomography (NCCT). The study population was derived from the Swedish Stroke Register (RS) and included 1,649 ICH patients diagnosed between 2016 and 2019 in Region Skåne. ICH volumes were measured using qER–NCCT, manual segmentation (Sectra IDS7), and the ABC/2 method, with 500 cases independently reviewed by two radiologists. qER–NCCT demonstrated a sensitivity of 97% in ICH detection and showed excellent agreement with manual segmentation (ICC = 0.96) and good agreement with the ABC/2 method (ICC = 0.86). The software tended to underestimate volumes, particularly in large (>30 ml), heterogeneous hemorrhages. Interrater agreement for manual segmentation was ICC = 0.996 (95% CI: 0.99–1.00). The findings suggest that qER–NCCT is a highly reliable tool for automated ICH volume quantification, aiding early diagnostics, prognostication, and population-wide volumetric analysis, though further refinements are needed for irregular, large hemorrhages.
