Subarachnoid Hemorrhage in Patients with Good Clinical Grade: Accuracy of 3.0-T MR Angiography for Detection and Characterization

Purpose To evaluate the diagnostic accuracy of aneurysm detection and the fidelity of morphologic characterization of three-dimensional (3D) time-of-flight (TOF) magnetic resonance (MR) angiography at 3.0 T in patients with a Glasgow coma score of 15 and noncontrast material–enhanced computed tomography (CT) findings that showed acute nontraumatic subarachnoid hemorrhage. Materials and Methods The institutional review board approved this prospective study, and patients provided informed consent. A total of 277 patients who had not experienced trauma but in whom nonenhanced CT showed subarachnoid hemorrhage, who had a Glasgow coma score of 15, and who underwent both 3D TOF MR angiography and digital subtraction angiography (DSA) (the reference standard) were included. Three observers who were blinded to clinical and DSA results independently analyzed all 3D TOF MR angiographic data sets. The receiver operating characteristic curve was applied to analysis of the detection of aneurysms with 3D TOF MR angiography by using patient- and aneurysm-based evaluations. Multivariate logistic regression analysis was performed to identify aneurysm-specific variables, including size, shape (daughter sac/lobulation/margin), neck width (wide if > 4 mm or if fundus-to-neck ratio was < 2), and relation to adjacent artery, that significantly affected morphologic assessment with 3D TOF MR angiography. Results Aneurysms were depicted with DSA in 225 patients. In patient- and aneurysm-based evaluations, respectively, 3D TOF MR angiography yielded accuracies of 96.8% (268 of 277) and 96.6% (309 of 320), sensitivities of 98.2% (219 of 223) and 98.1% (260 of 265), specificities of 91% (49 of 54) and 89% (49 of 55), positive predictive values of 97.8% (219 of 224) and 97.7% (260 of 266), and negative predictive values of 92% (49 of 53) and 91% (49 of 54). Accuracy of display of morphologic features was 92.5% (236 of 255) for size, 86.3% (220 of 255) for neck width, 94.5% for shape (241 of 255), and 96.9% (247 of 255) for relationship to adjacent vessel. Width of aneurysm neck was the only variable that significantly affected the morphologic assessment of 3D TOF MR angiography (odds ratio, 0.378; 95% confidence interval: 0.337, 8.347; P = .004). Conclusion Three-dimensional TOF MR angiography can accurately depict cerebral aneurysms and accurately display their morphologic features in stable patients with subarachnoid hemorrhage and a Glasgow coma score of 15. © RSNA, 2017