Umeå universitets logga

Identification of risk markers based on quantitative ultrasound texture analysis of carotid plaques has the ability to define vulnerable components that correlate with increased cardiovascular risk. However, data describing factors with the potential to influence the measurement variability of risk markers are limited. The aim of this study was to evaluate the influence of electrocardiogram-guided image selection, plaque echogenicity and area on carotid plaque risk markers and their variability in asymptomatic carotid plaques. Plaque risk markers were measured in 57 plaques during three consecutive heartbeats at two cardiac cycle time instants corresponding to the electrocardiogram R-wave (end diastole) and end of T-wave (end systole), resulting in six measurements for each plaque. Risk marker variability was quantified by computing the coefficient of variation (CV) across the three heartbeats. The CV was significantly higher for small plaques (area <15 mm2, 10%) than for large plaques (area >15 mm2, 6%) (p <0.001) in measurements of area, and the CV for measurements of gray-scale median were higher for echolucent plaques (<40, 15%) than for echogenic plaques (>40, 9%) (p <0.001). No significant differences were found between systole and diastole for the mean of any risk marker or the corresponding CV value. However, in a sub-analysis, the echolucent plaques were found to have a higher CV during systole compared with diastole. The variability also caused plaque type reclassification in 16% to 25% of the plaques depending on cutoff value. The results of this study indicate that echolucent and small plaques each contribute to increased risk marker variability. Based on these results, we recommend that measurements in diastole arc preferred to reduce variation, although we found that it may not be possible to characterize small plaques accurately using contemporary applied risk markers. 

Aims: To determine the inter-sonographer reproducibility of carotid ultrasound plaque detection using Mannheim consensus in a subclinical population and evaluate associations related to the reproducibility.

Methods and results: Bilateral ultrasound screening for carotid plaques defined by Mannheim consensus was performed on 106 subclinical participants. Two different sonographers scanned the same participant, and reproducibility of plaque detection was measured by Cohens kappa. Associations with reproducibility were evaluated by comparing wall, and plaque characteristics between subjects with plaques identified in one and both scans. In general, the inter-sonographer reproducibility of plaque detection was substantial with a kappa value of 0 center dot 70 (95% CI 0 center dot 60-0 center dot 80). Plaques detected in only one scan had significantly lower plaque area and plaque thickness (6 center dot 82 mm(2) and 1 center dot 45 mm) as compared to plaques detected in both scans (11 center dot 65 mm(2) and 1 center dot 96 mm, P<0 center dot 001).

Conclusion: Minor carotid plaques contribute to decreased reproducibility as compared to large plaques when screening for subclinical atherosclerosis using Mannheim consensus. Using an alternative plaque definition based on plaque thickness >1.5 mm and plaque area >10 mm(2) could increase the reproducibility of plaque detection in subclinical atherosclerosis.