Purpose: The authors investigated the effectiveness of high-resolution magnetic resonance angiography (MRA) using a single-artery highlighting technique for the accurate diagnosis of intracranial aneurysms at 3.0 T in a large cohort of patients with suspected intracranial aneurysms against the current gold standard, i.e. cerebral angiography.
Methods: A total of 307 patients with suspected aneurysms were referred for three-dimensional time-of-flight MRA (3D-TOF-MRA) prior to digital subtraction angiography (DSA). We summarised patient-based, aneurysm-based and vessel-based diagnostic performance parameters, namely, accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for detecting intracranial aneurysms with 3D-TOF-MRA. Interobserver agreement was calculated with the kappa (κ) statistic.
Results: Patient-based accuracy was 98.1%, sensitivity 99.5%, specificity 95.3%, PPV 97.6% and NPV 99% according to observer A; and 97.7%, 99.5%, 94.3%, 97.1% and 99%, respectively, according to observer B. Aneurysmbased accuracy was 98.3%, sensitivity 99.6%, specificity 95.2%, PPV 98.1% and NPV 99% according to observer A, and 98.0%, 99.6%, 94.3%, 97.7% and 99%, respectively, according to observer B. Vessel-based accuracy was 99.0%, sensitivity 99.6%, specificity 98.7%, PPV 97.8%, and NPV 99.7% according to observer A, and 98.9%, 99.6%, 98.5%, 97.3%, and 99.7%, respectively, according to observer B. Interobserver reading differences were not significant. The interobserver agreement was good, with a kappa (κ) value of 0.931 for patient-based evaluation, 0.934 for aneurysmbased evaluation and 0.953 for vessel-based evaluation.
Conclusions: Our results demonstrate the effectiveness of high-resolution MRA using a single-artery-highlighting technique for accurate diagnosis of intracranial aneurysms in a large cohort of patients with suspected aneurysms.