PURPOSE: To determine the accuracy of elliptic centric contrast material-en
hanced magnetic resonance (MR) angiography by using conventional angiograph
y as the reference standard.
MATERIALS AND METHODS: Fifty patients were examined prospectively with cont
rast-enhanced MR angiography and conventional angiography. The two examinat
ions were performed within 1 week of each other. Two patients underwent con
ventional angiography of only one carotid artery, which yielded 98 arteries
for comparison.
RESULTS: With conventional angiography as the reference standard and by usi
ng a 70% threshold for internal carotid arterial diameter stenosis, maximum
intensity projection (MIP) images had a sensitivity of 93.3%, specificity
of 85.1%, and accuracy of 87.6%, whereas reformatted transverse source imag
es had a sensitivity of 83.3%, specificity of 97.0%, and accuracy of 92.8%.
Interobserver variability for conventional angiograms was 0.97, for MIP im
ages was 0.91, and for source images was 0.90. The contrast-enhanced MR ang
iographic technique had a sensitivity of 88.9% and specificity of 58.1% for
the presence of irregularity and/or ulceration. All 50 examinations were t
riggered appropriately so that minimal or no venous signal intensity was de
picted.
CONCLUSION: Contrast-enhanced elliptic centric three-dimensional MR angiogr
aphy offers high-spatial-resolution, venous-suppressed images of the caroti
d arteries that appear to be adequate to replace conventional angiography i
n most patients examined prior to carotid endarterectomy.