We investigated the agreement of multislice computed tomography angiography (CTA) and magnetic resonance angiography (MRA) in the quantitative measurement of carotid artery stenosis. The dependency of the agreement of the chosen postprocessing procedures was also investigated.Methods:
Fifty consecutive symptomatic patients were included in this study. In all patients, a CTA was performed with a 16-slice CT scanner. Within 30 days, the extracranial vessels were examined using a combined time-of-flight and contrast-enhanced MRA. The CT data sets were used to calculate the degree of stenosis according to the North American Symptomatic Carotid Endarterectomy Trial, European Carotid Surgery Trial, and Common Carotid methods by means of the 1-mm thick, transverse raw data (RD), a sagittal maximum-intensity projection reconstruction, and sagittal multiplanar reconstruction. In addition, a semiautomated analysis was done using a specialized postprocessing software. For all combinations of postprocessing procedures and methods of calculating the degree of stenosis, the correlation coefficient and the agreement based on Bland/Altman plots were calculated.Results:
Eleven of the 100 primarily included carotid arteries could not be evaluated. The correlation coefficients for all combinations were comparable and lied in the interval between 0.932 and 0.787. The best correlation was found for the combination of RD/sagittal multiplanar reconstruction and ECST method. The evaluation of the agreement gave a systematic overestimation of CTA between 1.9% and 10.7% with a 95% confidence interval between ±26.7% and ±43.3%. With the semiautomated postprocessing software, additional 33 vessels could not be evaluated. The agreement of the calculated degrees of stenoses was worse than that of the planar procedures.Conclusions:
CTA and MRA had a feasible agreement in measuring the degree of stenosis of the carotid arteries. The best result could be obtained for the evaluation of the RD and the NASCET method. In this case one has to take into account a systematic overestimation of CTA of 1.9%. The combination with an additional reconstructive postprocessing procedure did not improve the result but might be useful for the radiologist to identify the location of the closest narrowing.