Ultrasound imaging of carotid artery stenosis: application of the Society of Radiologists in Ultrasound Consensus Criteria to a Single Institution Clinical Practice

Abstract

Carotid duplex Doppler ultrasound (CDDU) is increasingly used for the evaluation of internal carotid artery (ICA) stenosis. In CDDU, velocity measurements are used to estimate the degree of ICA stenosis. Traditionally, radiologists have relied on institutional experience and published research when interpreting CDDU. In 2003, a consensus committee of experts convened as the Society of Radiologists in Ultrasound Consensus Committee and proposed standard criteria for grading ICA stenosis including the use of peak systolic velocity (PSV) of greater than 230 cm/s for assigning ICA stenosis of greater than 70%. The purpose of this study was to evaluate the accuracy of the Society of Radiologists in Ultrasound Consensus Criteria in classifying carotid stenoses. This study shows the following: (1) that the criterion of PSV of greater than 230 cm/s for angiographic stenosis of greater than 70% performs as predicted by the consensus committee, with sensitivity of 95.3% (95% confidence interval [CI], 0.89-0.99) and specificity of 84.4% (95% CI, 0.80-0.88); (2) using Pearson correlations, there is no statistical difference found between the correlation of PSV with angiography (0.825 [95% CI, 0.792-0.853]), end diastolic velocity with angiography (0.762 [95% CI, 0.718-0.799]), and the ICA/common carotid artery (CCA) systolic ratio with angiography (0.766 [95% CI, 0.723-0.802]). The correlation of the ICA/CCA diastolic ratio with angiography (0.643 [95% CI, 0.584-0.696]) is less predictive at a 95% confidence interval than the other 3 velocity-based variables, and (3) when the 4 velocity-based variables are taken in pairs (eg, PSV and end diastolic velocity), there is no pair that shows statistically significant improvement in performance. Peak systolic velocity in combination with other variables does show a slight trend toward superior performance.