A comparison of 4D time-resolved MRA with keyhole and 3D time-of-flight MRA at 3.0 T for the evaluation of cerebral aneurysms

3D-TOF-MRA, 4D-TRAK and 3D-DSA were performed sequentially in 52 patients with suspected CAs. 4D-TRAK was acquired using a combination of sensitivity encoding (SENSE) and CE timing robust angiography (CENTRA) k-space sampling techniques at a contrast dose of 10 ml at 3 T. Accuracy, sensitivity, specificity of 4D-TRAK and 3D-TOF-MRA were calculated and compared for the detection of CAs on patient-based and aneurysm-based evaluation using 3D-DSA as a reference.The overall image quality of 4D-TRAK with a contrast dose of 10 ml was in the diagnostic range but still cannot be compared with that of 3D-TOF-MRA. In 52 patients with suspected CAs, fifty-eight CAs were confirmed on 3D-DSA finally. Fifty-one (with 2 false-positives and 9 false-negatives) and 58 (with 1 false-positive and 1 false-negative) CAs were visualized on 4D-TRAK and 3D-TOF-MRA, respectively. Accuracy, sensitivity and specificity on patient-based evaluation of 4D-TRAK and 3D-TOF-MRA were 92.31%, 93.33%, 85.71% and 98.08%, 100%, 85.71%, respectively, and 74.07%, 75.00%, 66.67% and 96.30%, 95.83%, 100% on aneurysm-based evaluation in patients with multiple CAs, respectively. Subgroup analysis revealed that for 19 very small CAs (maximal diameter <3 mm, measured on 3D-DSA), 9 were missed on 4D-TRAK and 1 on 3D-TOF-MRA (P？=？0.008). However, for 39 CAs with maximal diameter？≥？3 mm, the diagnostic accuracy is equally (39 on 4D-TRAK vs. 39 on 3D-TOF-MRA) (P？=？1). In four larger CAs with maximal diameter？≥？10 mm, 4D-TRAK provided a better characterization of morphology than 3D-TOF-MRA.4D-TRAK at a lower contrast dose of 10 ml with a combination of SENSE and CENTRA at 3 T could provide similar diagnostic accuracy rate for CAs with maximal diameter？≥？3 mm, and a better characterization of morphology for larger CAs with maximal diameter？≥？10 mm compared to 3D-TOF-MRA. However, further study is still needed to improve the “vascular edge” artifact and the compromise in spatial resolution in depiction of CAs with maxi