Value of Image Biomarkers Based on Dual-Energy Computed Tomography Angiography Material Separation Technique in Predicting Early Hematoma Expansion in Spontaneous Intracerebral Hemorrhage

Abstract

ObjectiveThis study aims to investigate the connection between the leakage sign (LS) and hematoma expansion (HE) in cases of spontaneous intracerebral hemorrhage (ICH). The investigation employs dual-energy computed tomography angiography (DECTA). MethodsA prospective cohort study was conducted, in which clinical and DECTA imaging data were collected from ICH patients within 6 hours of onset between January 2021 and June 2023. Exposure factors included DE-LS and traditional imaging biomarkers. The occurrence of HE on CT rescanned within 24 hours was the observed outcome. Exposed and confounding factors were considered in both univariate and multivariate regression analyses based on the results. Logistic and adjusted Poisson regressions were employed, and odds ratios (OR) and relative risks (RR) were calculated with 95% confidence intervals (CI). ResultsThe study enrolled a total of 90 patients, of whom 32 cases manifested HE, while 58 cases did not exhibit HE. Univariate analysis revealed statistically significant differences in parameters such as admission SBP, CRP, GCS, baseline hematoma volume, and imaging biomarkers likeDE-SS, and DE-LS. The OR value of DE-LS was determined as 48.21, with an RR value of 7.51. Multivariate adjusted Poisson regression analysis demonstrated that DE-LS was a robust independent predictor(RR = 4.11, 95% CI: 1.49 - 11.35; P < 0.001). ConclusionsDECTA-based DE-LS stands out as an independent predictor of HE. The utilization of RR values over OR values is endorsed when assessing the risk of HE prediction.