In-vivo quantitative ultrasound evaluation of carotid plaque

Carotid atherosclerotic plaque is a major contributor to cerebrovascular disease. Noninvasive quantitative ultrasound (QUS) evaluation of plaque morphology is requisite for comprehensive risk assessment. System and setting dependencies, however, limit current grayscale analyses. Here, integrated backscatter (IB) and attenuation coefficient (AC) were calculated in-vivo and compared with pathology scores of excised plaques. Radiofrequency data acquired prior to a carotid endarterectomy procedure were processed on 6 subjects. Average percent stenosis was 71.67±11.69. QUS parameters were calculated in-vivo with 70% axial overlap at end-diastole using the reference phantom method (AC, IB) and optimum power-spectral shift estimator (AC). Parametric maps were created to visualize QUS parameter variation in plaque. Histopathology examination was reported for a representative section of fixed plaque. Plaque was scored for hemorrhage, hemosiderin, and inflammation on an ordinal scale and was assessed for percent cholesterol/fibrinoid and calcium. Relationships between QUS parameters and histopathology were examined. Our results indicate that plaques with more calcium on histopathology (n=2, >20% calcium) have high IB (approaching 0 dB) and elevated AC (approaching 1.5 dB/cm-MHz). Plaques with high percent cholesterol/fibrinoid (n=4, >60% cholesterol/fibrinoid) have low IB (approaching -30 dB) and moderate AC (approaching 1 dB/cm-MHz). We are also able to distinguish calcified and soft plaque regions on parametric maps. These findings suggest that IB and AC can be used to noninvasively evaluate plaques based on tissue composition. Thus, these QUS parameters may provide important information for a risk-assessment model of carotid plaque. Future work will compare QUS parameters in-vivo with high frequency characterizations of plaque ex-vivo and 3D reconstructions of histopathology sections.