Accuracy of High-Resolution Ultrasound Imaging for Quantitative Assessment of Early Carotid Atherosclerosis

Abstract

Since the introduction of ultrasound technology for the evaluation of carotid atherosclerosis, a continuous controversy exists about the accuracy of this method for quantitative assessment of atherosclerotic plaques. Potential pitfalls and limitations of ultrasound imaging for a correct classification of significant carotid stenosis (>50% lumen narrowing) and occlusion could be compensated by the additional combined use of Doppler sonography when compared with conventional angiographic techniques, which may no longer be used as the only reference method (''gold standard'') despite their traditionally widespread application in major clinical trials [1–3]. In contrast, the validity of ultrasound echotomography for the assessment of minor carotid artery irregularities producing <50% lumen encroachment was substantiated by many studies in comparison to angiography and pathology [4, 5], although characterization of plaque morphology and surface integrity still represents a challenge. The changing view of atherosclerosis as a highly variable disease with potential capacity to recover either by regression or healing rather than to continuously or step-wise progress [6, 7] stimulated efforts to demonstrate the reliability of high-resolution ultrasound imaging for the investigation of the prevalence of initial stages of carotid atherosclerosis in epidemiological studies as well as in clinical trials investigating the efficacy of risk factor reduction. For this particular purpose, three-dimensional plaque visualization [7] and measurement of the intima-media thickness became an increasingly used criterion for the evaluation of atherogenesis [8–11]. Riley et al. [8] reported that standardized ultrasound techniques yielded highly reproducible results for the evaluation of small differences on the order of 0.03 mm, which exceeds at present by far the best available calculated axial resolution for 10 MHz ultrasound emission frequency with a wave length of 0.15 mm. In order to answer the question whether indeed measurement for carotid wall thickness of this order is realistic even if statistical analysis for the detection of group differences is applied, we asked several experts for their authoritative views.