Because the carotid bifurcation lies close to the surface of the neck, without overlying bony structures or air spaces, trans-cutaneous B-mode ultrasound has been a very effective tool for evaluating carotid disease. Aside from the utility of Doppler ultrasound to measure carotid stenosis, B-mode ultrasound provides detailed information regarding carotid wall morphology (Fig.2).

A leading application of carotid ultrasound is the measurement of intima-media thickness (IMT).

Fig. 1. Cross-section of a type VI plaque specimen from carotid endarterectomy stained with Movat's pentachrome. A rupture (arrow) of the blue fibrous cap has occurred exposing the thrombogenic lipid-rich necrotic core (N) to the lumen (L). Resultant in-traplaque hemorrhage appears red.

As shown in Fig. 2, distinct echoes occur at the boundaries between the lumen and intima and between the media and adventitia. The average distance between these echoes is the IMT, which serves as a measurement of the degree of diffuse intimal thickening. By comparing the distance between these lines to various histological measurements in the aorta and carotid arteries, Pignoli et al. [56] demonstrated that this distance did correspond to the combined thickness of the intima and media. Further evaluations showed that this measurement could be made accurately at the far wall [74] and that IMT does appear to represent atherosclerosis as opposed to a general response to changes in shear and tensile stress [7].

Also evident in Fig. 2 is a significant plaque distal to the region where IMT is measured. Such plaques range in appearance from echolucent to hy-perechoic which has led to efforts to characterize plaque composition based on echogenicity. These studies have divided echogenicity into two [30], four [25], or six [60] subjective categories as well as investigating direct measures of echogenicity [65], [71]. Comparisons of echogenicity to histological samples has generally found that echolucent plaques are

Fig. 2. Longitudinal B-mode ultrasound image of a carotid artery with a large atherosclerotic plaque (arrow) and demonstrating intima-media thickness (IMT) measurement. A region (box) of the common carotid artery proximal to the plaque is selected and echoes corresponding to the intimal and medial boundaries are detected (lines). The average distance between these lines is IMT.

likely to contain intraplaque hemorrhage and other soft plaque materials. Wilhjelm et al. [71] further showed that the mean gray scale value of the plaque by B-mode ultrasound is significantly correlated with decreasing amounts of soft plaque materials (r = -0.42). Hatsukmai et al. [30] demonstrated that the echolucent region could be localized within the plaque to identify the quadrant containing the soft plaque features. This study also found a prevalence of speckled calcification and foam cells in echolucent regions in addition to intraplaque hemorrhage and necrosis.

Hyperechoic regions in carotid plaques are associated with fibrous tissue. In a recent study, the thickness of the hyperechoic plaque region adjacent to the lumen was shown to agree well with histolog-ically measured cap thickness [17].

B-mode ultrasound is attractive because the imaging procedure is relatively simple. However, B-mode ultrasound provides limited view angles for imaging plaques. Furthermore, plaque appearance can vary with the view angle and calcifications can cause shadow artifacts that obscure deeper plaque structures [54]. To some extent, these limitations are addressed by 3D imaging techniques [78], [61].

X-ray computed tomography (CT) is another effective technology for visualizing plaque morphology. In CT angiograms used to measure carotid stenosis clinically, calcifications are clearly visible (Fig. 3). The presence of high density regions within the carotid wall has been definitively linked to calcifications. Furthermore, the amount of calcification visible by CT has been shown to correlate with the percent stenosis in the carotid [48] and with reduced inflammatory content of the vessel wall [62].

In addition to being sensitive to calcified content, CT has also shown promise for identifying other components within carotid plaque. Walker et al. [69] showed that decreasing tissue density was associated with increasing plaque lipid content. Fi

Images Ultrasund Vasculature
Fig. 3. Maximum intensity projection (MIP) X-ray Computed Tomography image of the carotid vasculature showing large areas of calcification (arrows) in the left and right carotid arteries. A significant stenosis of the right carotid artery is indicated by an arrowhead.

brous plaques have been shown to yield densities similar to soft tissues [54].

An important advantage of CT is its ability to yield absolute measures of tissue densities in terms of Hounsfield units. This permits consistent thresholds to be defined for characterizing plaque content. Disadvantages of CT include the use of ionizing radiation and nephrotoxic contrast agents.

Magnetic resonance imaging (MRI) has been most thoroughly investigated for its ability to characterize plaque morphology. MRI is uniquely able to tune the acquisition to increase sensitivity to specific physical characteristics of the tissue. Images can be specifically weighted to highlight regional proton density, magnetic relaxation time constants T1 and T2, flowing blood, or uptake of injected contrast agents. Images with each of these contrast weightings are analyzed together to provide a comprehensive picture of plaque composition (Fig. 4).

The use of black-blood imaging techniques, particularly T1-weighted double inversion recovery techniques, allows clear delineation of the vessel wall boundaries (inner and outer) in cross-sectional images of the carotid. The areas of individual slices can

Carotid Artery Ultrasound

Fig. 4. Axial, multiple-contrast-weighting magnetic resonance images of an atherosclerotic carotid artery distal to the bifurcation with corresponding histology. Images are a T1-weighted, b T2-weighted, c proton-density-weighted, d contrast-enhanced T1 -weighted, e time-of-flight, f histology stained with Movat's pentachrome. The lumen of the internal carotid artery is indicated by an "L"and a region of calcification is indicated by an arrow. The large hemorrhagic core (red) leads to bright signal on Tl-weighted and time-of-flight images.

Fig. 4. Axial, multiple-contrast-weighting magnetic resonance images of an atherosclerotic carotid artery distal to the bifurcation with corresponding histology. Images are a T1-weighted, b T2-weighted, c proton-density-weighted, d contrast-enhanced T1 -weighted, e time-of-flight, f histology stained with Movat's pentachrome. The lumen of the internal carotid artery is indicated by an "L"and a region of calcification is indicated by an arrow. The large hemorrhagic core (red) leads to bright signal on Tl-weighted and time-of-flight images.

be combined into a total wall volume using Simpson's rule. This wall volume has been shown to correlate with plaque volume measured ex vivo with a correlation coefficient of 0.92 [44]. Additionally, this measurement is highly reproducible, with an error standard deviation under 6% [35].

Besides the ability of MRI to depict vessel boundaries, it is also able to depict substructures of the plaque. Using a bright-blood, 3D-time-of-flight imaging technique, Hatsukami et al. [31] showed that a hypointense band adjacent to the lumen indicates a thick (>0.25 mm) fibrous cap and that communication of a bright plaque interior with the lumen is indicative of a fibrous cap rupture. Agreement between the MR findings and histological state of the fibrous cap showed a Kappa value of 0.83. The addition of information from black-blood sequences to the fibrous cap evaluation was further shown to aid in discriminating juxtaluminal calcifi cation and flow artifacts [49]. By comparing MRI findings to histology, this study showed a sensitivity of 0.81 and a specificity of 0.90 for identifying an unstable cap in vivo

MRI is also able to detect the soft plaque components underlying the fibrous cap by their unique combinations of intensities under different contrast weightings. Especially apparent by MRI is a high signal on T1-weighted imaging corresponding to soft plaque components (necoritc core or hemorrhage), which can be identified with a sensitivity of 85% and a specificity of 92% [80]. Hyperintensity is especially apparent in the event of intraplaque hemorrhage, which led to the development of direct thrombus imaging [50]. Chu et al. [12] showed that subsequent breakdown of the hemorrhage components leads to characteristic changes in signal intensity that allow hemorrhage to be classified as fresh, recent, or old. Comparison with histology showed that these stages could be identified with a Cohen's Kappa equal to 0.7. In another study, hemorrhage was subdivided into cases where the hemorrhage did, or did not directly communicate with the lumen [34]. Direct communication implies a disruption of the fibrous cap and was differentiated from intrapla-que hemorrhage with an accuracy of 96%.

MRI is also capable of quantifying components of the plaque. In a recent study, carotid plaques were divided into regions of necrotic core, calcification, loose matrix, and dense fibrous tissue [58]. Volumes of each component correlated closely with histo-logically measured volumes (correlation coefficients ranging from 0.55 to 0.74). The measurements were statistically equivalent with the exception of calcification, which was overestimated by MRI relative to histology, but was still closely correlated.

The ability of MRI to depict plaque composition in vivo also enables classification of human carotid atherosclerotic plaque according to American Heart Association (AHA) classifications. In a recent study, cross sectional images were classified into types I/II, III, IV/V, VI, VII, and VIII. With the exception of types I/II and VIII, which were under-represented in the study, all classifications predicted histological classifications with sensitivities and specificities exceeding 80%. For all classifications, MRI and histology showed good agreement, with Cohen's k= 0.74 [10].

Further delineation of plaque composition may be facilitated by contrast enhanced (CE) MRI using gadolinium-based contrast agents. In parallel investigations, Yuan et al. and Wasserman et al. showed that comparison of pre- and post-CE MRI improves differentiation of necrotic core from fibrous tissue [81], [70]. In the latter study, CE MRI helped discriminate fibrous cap from lipid core with a contrast-to-noise ratio as good as or better than that with T2-weighted images but with approximately twice the SNR. These findings provide quantitative evidence that CE MRI is a viable tool for in vivo study of atherosclerosis and can be used in combination with other contrast weightings to identify plaque composition.

Gadolinium-enhanced MRI also provides information regarding plaque inflammatory characteristics, including in-growth of neovasculature and macro-

phages. In a kinetic analysis of dynamic CE MRI from 16 subjects with carotid atherosclerosis, Kerwin et al. [40] compared the fractional blood volume measured by MRI to the area of neovasculature in corresponding histological specimens. A correlation coefficient of 0.80 was found suggesting that dynamic CE MRI provides a means for prospectively studying the link between neovasculature, inflammation, and plaque vulnerability. In a follow-on study, plaque macrophage content was shown to be associated with the transfer constant, describing the rate of uptake by the plaque. A correlation coefficient of 0.76 was observed [41].

Although MRI has tremendous capabilities for characterizing carotid plaque, most information is derived by combining several images, with different contrast weightings or from time-series images. The complexity of the data is the biggest limitation for MRI, and it requires specialized analysis tools and streamlined procedures for image review. To assist in the analysis, specialized algorithms for registration of carotid MRI have been proposed [38], [39]. Additionally, methods for automated segmentation of the plaque into constituent components are under development. For the most part, these studies have focused on ex vivo imaging experiments, with established segmentation procedures [63], [13], [33]. Recent experiments in vivo show that more specialized segmentation procedures that overcome the limitations of in vivo imaging are able to identify many of the principle plaque components [23], [43]. Delineation of the wall itself has been facilitated by active contour techniques for boundary detection [79], [28], [29]. Finally, these features have been combined into integrated processing packages, such as CASCADE (Fig. 5 [75]).

Was this article helpful?

0 0
Relaxation Audio Sounds Autumn In The Forest

Relaxation Audio Sounds Autumn In The Forest

This is an audio all about guiding you to relaxation. This is a Relaxation Audio Sounds with sounds from Autumn In The Forest.

Get My Free MP3 Audio

Post a comment