Magnetic resonance angiography is now commercially available for a var
iety of scanners and is being increasingly applied in the diagnosis of
cerebrovascular disorders. Considering the clinical consequences, esp
ecially in intracranial aneurysms, studies to determine the sensitivit
y and specificity of the method are essential. Here we report our expe
rience with a 3D-FISH time-of-flight magnetic resonance angiography pr
otocol in 52 patients who have suffered an acute subarachnoid hemorrha
ge. In 26 of the 52 patients, conventional angiography identified 31 a
neurysms (3-20 mm) that were confirmed during surgery or autopsy. Magn
etic resonance angiography correctly identified 28 of the 31 aneurysms
(sensitivity 90.3 %) and missed one ruptured (3 mm) and two incidenta
l aneurysms (3 mm) in patients with multiple aneurysms. The sensitivit
y for a ruptured aneurysm was 96 %. The 26 patients who suffered subar
achnoid hemorrhage without evidence of an intracranial aneurysm on rep
eated angiography served as a control group. Magnetic resonance angiog
raphy revealed no false-positive findings, resulting in a specificity
of 100 %. In correlation with the literature, we conclude that magneti
c resonance angiography is not sensitive enough for the management of
acute sub arachnoid hemorrhage. However, the method provides important
complementary information for definition of the bleeding site in pati
ents with multiple aneurysms. In addition, the calculation of projecti
ons not possible with conventional angiography can aid surgical planni
ng. Since only very small aneurysms were missed by magnetic resonance
angiography, the sensitivity seems appropriate to screen asymptomatic
patients who are at risk for intracranial aneurysms,