Correlation of high-resolution, B-mode and continuous-wave Doppler sonography with arteriography in the diagnosis of carotid stenosis.

One hundred thirty-one carotid bifurcations in 73 symptomatic patients were prospectively studied with high-resolution, B-mode sonography (HRS) and continuous-wave Doppler ultrasound (CWD). Twenty-three bifurcations were studied because of asymptomatic bruits, 30 for transient ischemic attacks, and 36 for focal or nonfocal symptoms or signs suggesting occlusive disease. The remaining 42 bifurcations were asymptomatic, with symptoms or signs confined to the contralateral bifurcation. Findings in the common, external, and internal carotid arteries (total, 393 vessels) were tabulated separately for plaque severity (HRS), ulceration (HRS), and degree of occlusion (HRS and CWD). HRS correctly estimated plaque severity in 67% of vessels, with 13% overestimated and 20% underestimated. HRS was only 44% sensitive for detection of 81 angiographically demonstrated ulcers. Among 187 occlusive lesions (ranges, less than 50%, greater than or equal to 50%, less than 70%, greater than or equal to 70% decrease in diameter or complete occlusion) HRS correctly estimated luminal narrowing in 62%, CWD in 30%, and HRS plus CWD in 70%. Among greater than 50% occlusive lesions, HRS was correct in 48%, CWD in 59%, and HRS plus CWD in 69%. These results suggest that (a) HRS is not accurate for diagnosis of ulceration, (b) Doppler and B-mode combined are more accurate than separate use of either procedure for evaluation of occlusive lesions, and (c) the level of accuracy in evaluating plaque and stenosis provides a general estimation of severity of disease but does not allow for precise diagnosis.