Die Doppler-Sonographie der extrakraniellen Arteria carotis mit und ohne Doppier-Frequenz-Spektrum-Analyse

Accuracy of the main Doppler methods for evaluating the degree of carotid stenoses (continuous wave, pulsed wave, and color Doppler)

OBJECTIVE

To compare the accuracy of the main Doppler methods for quantifying the degree of carotid stenoses in 133 patients.

METHOD

seven parameters were measured: maximum velocity (V.max) inside the stenosis (by pulsed wave [PW]), grade and index of spectral disturbance (STI) at the outlet of the stenosis (by PW and continuous wave [CW]), ratio of velocities I(IC/CC) in the internal and common carotid (by PW), and ratio of vessel cross-section and residual lumen area (%Color) inside the stenosis (color Doppler). The reference methods were the grades of spectral disturbance and the STI (by CW), already validated against endarterectomy specimens.

CONCLUSION

For poststenosis measurements, a high correlation was found between the grades or STI measured by PW and CW (reference method). The measurement of these parameters was easier in PW mode. For intrastenotic velocity, the increase in V.max was not proportional to the degree of stenosis, and V.max showed large fluctuations for the same degree of stenosis. I(IC/CC) also showed large fluctuations for the same degree of stenosis. The correlation was poor for these two parameters, which could only be used for detecting two groups of stenosis: >75% or >90% in area. Color Doppler (conventional and power) routinely overestimated the degree of stenosis by 10%-15% but correlated better with the reference method and was more accurate and reproducible than V.max. A special procedure can be used to avoid this overestimation and improve the accuracy of the color Doppler.

Reproducibility of carotid artery Doppler frequency measurements

Doppler sonography of 29 extracranial carotid arteries was performed twice within three days or less. Angiography revealed stenoses with from 10 to 95% diameter reduction in 25 internal carotid arteries whereas four vessels were found to be normal. The systolic peak frequency of the internal carotid artery read from the Doppler spectrum (n = 29) could be reproduced very well. This was shown by a linear regression nearest the line of identical values with a coefficient of correlation r = 0.97 (p less than 0.001). Other values derived from the spectral analysis of the Doppler shift signal were not so well reproduced including the peak frequency ratio (systolic peak frequency of the internal carotid artery/systolic peak of the common carotid artery) (n = 22; r = 0.81; p less than 0.001). The mean frequencies read from the zero-crossing detector recordings (n = 20) could not be reproduced as demonstrated by a linear regression far away from the line of identical measurements with a coefficient of correlation r = 0.43 (p less than 0.05).