Detection of infundibula using three-dimensional time-of-flight magnetic resonance angiography with volume rendering at 3.0 Tesla compared to digital subtraction angiography.

We aimed to prospectively evaluate the effectiveness of unenhanced, three-dimensional time-of-flight magnetic resonance angiography (3D-TOF-MRA) with volume rendering (VR) at 3.0 Tesla in the detection of infundibula (IF). A total of 136 consecutive patients with suspected or known aneurysms detected by MRA were examined using DSA, and IF were identified. A 3D specialist adjusted window width to the appropriate threshold, took level and MR source images and performed MRA post-processing techniques, including VR and the single artery highlighting method. VR-DSA was obtained from the rotational DSA data, and two-dimensional (2D)-DSA and VR-DSA were combined as the gold standard. Three blinded readers evaluated the diagnostic sensitivity and accuracy of 3D-TOF-MRA with VR. The 2D-DSA and VR-DSA revealed 44 IF in 34 of the 136 patients and no cerebral IF in 102 patients. For MRA, sensitivity per patient for all readers ranged from 89.5% to 97.1%, while accuracy per patient ranged from 97.1% to 99.3%. Per IF sensitivity for all readers ranged from 91.7% to 97.7%, while the per IF accuracy ranged from 97.3% to 99.3%. For detection of IF at the internal carotid artery, sensitivity for all readers ranged from 89.5% to 97.1%, and accuracy from 89.5% to 97.4%. For detection of IF at the anterior choroidal arteries, sensitivity and accuracy was 100% in all patients. There were no significant differences in the sensitivity or accuracy of all readers in the detection of single and multiple IF (p>0.05). One false-negative was found for IF by reader 1, two for reader 2 and four for reader 3. We conclude that VR 3D-TOF-MRA at 3.0 T is an ideal, non-invasive imaging exam for the detection of IF and can be used effectively in its diagnosis.