We performed a preliminary study comparing three-dimensional time-of-f
light (3 D TOF) magnetic resonance angiography (MRA) and spiral CT ang
iography (SCTA) in the detection and assessment of internal carotid ar
tery stenosis. Digital subtraction angiography (DSA) was the reference
examination. We examined 20 patients with signs of cerebrovascular in
sufficiency, who underwent MRA, SCTA and DSA within a 3 day period. Bo
th internal carotid arteries were assessed by three blinded readers fo
r degree of stenosis at two different levels (bulb and remaining secti
on) giving a total of 80 assessments. Interobserver variability, sensi
tivity, specificity, diagnostic accuracy, concordance, overestimation
and underestimation were assessed. Interobserver variability was not s
tatistically significant. MRA showed higher sensitivity, specificity,
diagnostic accuracy and concordance than SCTA (92.0% vs 80.8 %, 98.2 %
vs 96.4 %, 96.3 % vs 91.3 % and 96.0 % vs 88.0%,respectively). MRA ga
ve rise to a 5.0 % overestimation rate, whereas SCTA occasioned a 7.5
% underestimation rate. These differences are not statistically signif
icant. These results suggest that MRA is a more useful, noninvasive mo
dality for assessment of the internal carotid artery with a more than
70% stenosis.