Carotid artery intima-media thickness measurement in children with normal and increased body mass index: a comparison of three techniques

Carotid artery intima-media thickness measurement in children with normal and increased body mass... Background Common carotid artery intima-media thickness is a marker of subclinical atherosclerosis. In children, increased intima-media thickness is associated with obesity and the risk of cardiovascular events in adulthood. Objective To compare intima-media thickness measurements using B-mode ultrasound, radiofrequency (RF) echo tracking, and RF speckle probability distribution in children with normal and increased body mass index (BMI). Materials and methods We prospectively measured intima-media thickness in 120 children randomly selected from two groups of a longitudinal cohort: normal BMI and increased BMI, defined by BMI ≥85th percentile for age and gender. We followed Mannheim recommendations. We used M’Ath-Std for automated B-mode imaging, M-line processing of RF signal amplitude for RF echo tracking, and RF signal segmentation and averaging using probability distributions defining image speckle. Statistical analysis included Wilcoxon and Mann-Whitney tests, and Pearson correlation coefficient and intra-class correlation coefficient (ICC). Results Children were 10–13 years old (mean: 11.7 years); 61% were boys. The mean age was 11.4 years (range: 10.0–13.1 years) for the normal BMI group and 12.0 years (range: 10.1–13.5 years) for the increased BMI group. The normal BMI group included 58% boys and the increased BMI group 63% boys. RF echo tracking method was successful in 79 children as opposed to 114 for the B-mode method and all 120 for the probability distribution method. Techniques were weakly correlated: ICC=0.34 (95% confidence interval [CI]: 0.27–0.39). Intima-media thickness was significantly higher in the increased BMI than normal BMI group using the RF techniques and borderline for the B-mode technique. Mean differences between weight groups were: B-mode, 0.02 mm (95% CI: 0.00 to 0.04), P=0.05; RF echo tracking, 0.03 mm (95% CI: 0.01 to 0.05), P=0.01; and RF speckle probability distribution, 0.03 mm (95% CI: 0.01 to 0.05), P=0.002. Conclusion Though techniques are not interchangeable, all showed increased intima-media thickness in children with increased BMI. RF echo tracking method had the lowest success rate at calculating intima-media thickness. For patient follow-up and cohort comparisons, the same technique should be used throughout. . . . . . . Keywords Atherosclerosis B-mode ultrasound Carotid artery Children Intima-media thickness Obesity Radiofrequency echo tracking * Ramy El Jalbout Laboratory of Biorheology and Medical Ultrasonics, ramy.el–jalbout.hsj@ssss.gouv.qc.ca University of Montreal Hospital Research Center (CRCHUM), Montreal, Quebec, Canada 1 Department of Pediatrics, University of Montreal, Department of Medical Imaging, University of Montreal, Sainte-Justine University Health Center, Sainte-Justine University Health Center, Montreal, Quebec, Canada 3175 Cote-Sainte-Catherine Road, Montreal, Quebec H3T 1C5, Canada Department of Radiology, University of Montreal Hospital Center (CHUM), Montreal, Quebec, Canada 1074 Pediatr Radiol (2018) 48:1073–1079 Introduction (>102 cm in males and >88 cm in females) [9]. Exclusion criteria included being a pregnant or breastfeeding mother, Common carotid artery intima-media thickness is a reliable families planning to move, and children on a restricted diet marker of subclinical atherosclerosis [1]. In children, increased or taking anti-hypertensive drugs or corticosteroids [10, 11]. intima-media thickness is associatedwithobesityandthe risk The independent variables included BMI, age and gender. of cardiovascular events in adulthood [1, 2]. The use of intima- All information was collected during the same visit. Weight media thickness in risk stratification, however, is limited by two was measured twice on an electronic scale to the nearest factors. The first is a lack of standardization in the acquisition 0.1 kg. Height was measured twice to the nearest 0.1 cm. and analysis of intima-media thickness measures. The second is BMI was calculated according to the formula weight (kg)/ a lack of reproducibility and validated age- and gender-specific height (m) . We defined overweight and obesity according normative values against which to compare patient results. The to the Centers for Disease Control age- and gender-adjusted Association for European Paediatric Cardiology (AEPC) rec- z-score for weight and height [12]. ognized the need for standardized normative data in its recom- mendations for the measurement, analysis and interpretation of Image acquisition intima-media thickness in children [3]. There are two main methods for measuring common carot- Primary end points were intima-media thickness measured by id artery intima-media thickness: ultrasound (US) B-mode B-mode, RF echo tracking and RF speckle probability distri- images (manual, semiautomated or automated) and radiofre- bution. Images were acquired according to the Mannheim quency (RF) multiple M-line analysis (using echo tracking of recommendations [13]. Participants were in the supine posi- RF amplitude). In adults, the two methods correlate well [4]. tion with their head tilted 45° to the opposite side. We mea- Automated B-mode imaging with dedicated software (M’Ath- sured the far wall of the common carotid artery at 1 cm from Std) was validated for intima-media thickness measurement the carotid bifurcation. A high frequency linear US probe was and showed excellent reproducibility in the PARC (Paroi positioned longitudinally perpendicular to the vessel wall. The Artérielle et Risque Cardiovasculaire) study [5]. In pediatrics, software, whether for the B-mode measurement or for the RF studies used either US [6] or echo tracking [7]. To our knowl- echo tracking measurement, calculates the intima-media edge, there are no studies comparing B-mode and RF in chil- thickness automatically along the indicated segment. dren and few published normative values using either tech- nique [3, 6, 7]. B-mode US has nonetheless demonstrated B-mode intima-media thickness increased intima-media thickness in obese children [8]. The objective of this study was therefore to compare com- B-mode images were obtained using an ATL 5000 HDI US mon carotid artery intima-media thickness using B-mode US, unit (Advanced Technology Laboratories, Bothell, WA) RF echo tracking, and segmentation of the vessel wall using equipped with a 5.5–12.5 MHz probe. Focus depth was set RF speckle probability distribution in children with normal around 25 mm, with a frame rate of 25 per s and gain settings and increased body mass index (BMI). adjusted optimally to facilitate edge detection. A US technol- ogist with 10 years of experience scanned the child and saved a static longitudinal image of the common carotid artery that Materials and methods shows best the intima-media-lumen interface on gray-scale image. The 2-D image is instantaneously transferred to a com- Study design and patient population puter with M’Ath-Std software (Metris, Argenteuil, France). The technologist then drew a 1-cm-long line next to the We randomly selected 120 children from two groups of the intima-media-lumen interface (Fig. 1). Automated computer- Quebec Adipose and Lifestyle Investigation in Youth ized intima-media thickness is then generated by the software (QUALITY) prospective cohort of 564 children: a normal [14, 15]. We sought a quality index >0.5 for each measure BMI group and an increased BMI group, as defined by BMI indicating that more than 50% of the measurements along ≥85th percentile for age and gender, 2 years after enrollment the 1-cm-long line were used to calculate the final averaged in the cohort between 2005 and 2011. The study was approved intima-media thickness. An average of two values was obtain- by the Ethics Boards of the Quebec Heart and Lung Institute ed during optimal resolution, independent of cardiac cycle. and Sainte-Justine University Hospital. Written informed as- sent and consent were obtained from all participants and their RF echo tracking intima-media thickness parents, respectively. Inclusion criteria were Caucasian origin, age 8–10 years at time of entry into the original cohort, We used a linear array L10–5 40-mm US transducer (model healthy, and having at least one obese parent (defined by 410,503) on the MyLab 70 platform (ESAOTE) equipped with BMI ≥30 kg/m or an elevated waist circumference an ART.LAB platform (Pie Medical Imaging BV, Maastricht, Pediatr Radiol (2018) 48:1073–1079 1075 calculating RF US signals received along a single line of ob- servation (M-line processing) with time-averaging (using an electrocardiogram [ECG]) and motion compensation algo- rithms [16]. An average of two such intima-media thickness values was obtained (Fig. 2). RF speckle probability distribution intima-media thickness The same RF-digitized data stream (cine-loop acquisition of 2 to 5 cardiac cycles including 250–300 images) was post- processed to obtain an intima-media thickness using a differ- ent semiautomatic algorithm. This algorithm was implement- ed on a dedicated platform (ORS Visual; Object Research Systems Inc., Montreal, Canada). A research assistant (M.- H. R. C.) with 8 years of experience blinded to the results of B mode and RF echo tracking manually initialized the seg- Fig. 1 B-mode longitudinal US image of the common carotid artery of a mentation on one image of the data stream by drawing two 12-year-old boy shows the line drawn by the technologist next to the contours: the lumen/intima-media and the intima-media/ad- intima-media-lumen interface indicating the location of the automatic ventitia interfaces. The software automatically computed measurement by the M’Ath-Std system software. The software automatically gave the intima-media thickness value averaged over the these two interfaces on all subsequent frames (Fig. 3). The 1-cm segment of the vessel and gave the quality index (we aimed an index intima-media thickness was the average of all segmented superior to 0.5 as explained in text) frames of the cine-loop of the mean distance between the two segmented interfaces. The segmentation of the vessel wall The Netherlands). A 2-5 cardiac cycles long RF-digitized data was based on the method described by Destrempes et al. [17, stream is recorded by the technologist who then indicates a 18]. During the segmentation process, the initial contours are region of interest along the intima-media-lumen interface. refined and tracked over the whole video sequence. For this Automated intima-media thickness is then obtained by Fig. 3 A static image of the common carotid artery of a 12-year-old boy demonstrates the radiofrequency speckle probability distribution technique taken from a video sequence on which automatically computed contours of the intima-media borders (green) with the lumen Fig. 2 Radiofrequency echo tracking static image of an 11-year-old girl and adventitia are shown. The segmentation is manually initialized on a shows the location to measure the intima-media thickness (green) on the single frame (not shown). Both contours are then automatically traced on far vessel wall indicated by the technologist. The software automatically all subsequent frames of the 2–5 cardiac cycle-long video sequence. The gave individual intima-media thickness values during the entire cardiac intima-media thickness is the average (over all frames) of the mean cycle depending on the lumen diameter as well as the averaged final value distance between the two segmented contours 1076 Pediatr Radiol (2018) 48:1073–1079 purpose, a Bayesian segmentation model was used. In this model, the echogenicity of the intima-media layer, the lumen and the adventitia are modeled by mixtures of Nakagami dis- tributions and the motion of the vessel wall is estimated with optical flow. An average of systolic and diastolic intima-media thickness of all subsegments is obtained. Statistical analysis We did a descriptive analysis of the study population, checked for normality of the data distribution using the Shapiro-Wilk test, and used the Wilcoxon test for paired samples to compare intima-media thickness obtained by each of the three tech- niques. The intra-class correlation coefficient (ICC) was used to measure correlation. We plotted Bland-Altman and regres- Fig. 4 Box and whisker plot of mean intima-media thickness: B-mode: sion graphs to verify the ranges of agreement. We checked for 0.56+/−0.004 mm (95% confidence interval [CI]: 0.55–0.57), radiofrequency echo tracking: 0.47+/−0.006 mm (95% CI: 0.46–0.48) the correlation between intima-media thickness and age using and radiofrequency probability distribution: 0.34+/−0.005 mm (95% the Pearson correlation. We used the Mann-Whitney U test for CI: 0.33–0.35) the difference between children with normal BMI and those with increased BMI. Analysis was done using MedCalc sta- tistical software version 17.9.7 (Ostend, Belgium). difference of 0.14 mm (95% CI: 0.12–0.16, P<0.001) for the comparison of the RF echo tracking and RF speckle probabil- Results ity distribution techniques. ICC for the comparison of all three techniques was 0.34 (95% CI: 0.27–0.39). Agreement and The age range of the 120 randomly selected children was 10 to regression analysis between any two techniques is shown in 13 years including 73 boys. The mean age was 11.4 years the Bland-Altman scatter plots (Fig. 5). (95% confidence interval [CI]: 11.2–11.6) for the normal The mean B-mode intima-media thickness for normal BMI BMI group and 12.0 years (95% CI: 11.8–12.2) for the in- group (n=59) was 0.55+/−0.006 mm (95% CI: 0.54–0.56). creased BMI group. In the normal BMI group, 58% were For the increased BMI group (n=55), B-mode intima-media boys, and in the increased BMI group 63% were boys. The thickness was 0.57+/−0.006 mm (95% CI: 0.56–0.58). The two groups were significantly different with respect to BMI mean RF echo tracking intima-media thickness for the normal (mean z-score=−0.73 (95% CI: -0.92 to-0.54) for the normal BMI group (n=46) was 0.46+/−0.008 mm (95% CI: 0.45– BMI group and 1.96 (95% CI: 1.83–2.08) for the increased 0.47) and for the increased BMI group (n=33), the mean BMI group (P<0.001). The age was statistically different was 0.49+/−0.008 (95% CI: 0.48–0.50). The mean RF speckle (P<0.001). Gender ratios did not differ significantly between probability distribution intima-media thickness for the normal the groups (P=0.7). Although we measured intima-media BMI group (n=60) was 0.32+/−0.006 mm (95% CI: 0.31– thickness on all children, data for RF echo tracking were often 0.33). For the increased BMI group (n=60), the mean was not available as the software could not automatically detect the 0.36+/−0.006 mm (95% CI: 0.35–0.37). intima-media-lumen interface mainly due to artifacts. We Intima-media thickness was statistically significantly could not call the children back as the present analysis was higher in the increased BMI group for the RF echo tracking started in 2016, at least 5 years after data collection. Final and the RF speckle probability distribution techniques and analysis included 114 children for B-mode, 79 for RF echo borderline significant for the B-mode technique. The mean tracking and 120 for RF speckle probability distribution. differences using the Mann-Whitney U test were calculated Therefore, for the 74 children who had all three measurements at 0.02 mm+/−0.01 (95% CI: 0.00–0.04, P=0.05) with the B- available, intima-media thickness differed significantly de- mode technique, 0.03 mm +/−0.01 (95% CI: 0.01–0.05, pending on technique. Mean intima-media thicknesses are P=0.02) with the RF echo tracking technique and 0.03 mm shown in Fig. 4. +/−0.01 (95% CI: 0.01–0.05, P=0.002) for the RF speckle The Wilcoxon test showed significant differences of probability distribution technique. Although age was statisti- 0.10 mm (95% CI: 0.09–0.11, P<0.001) for the comparison cally different between the groups, the difference was not of the B-mode and RF echo tracking techniques, a difference clinically significant (11.4 and 12.0 years). Furthermore, the of 0.24 mm (95% CI: 0.23–0.25, P<0.001) for the B-mode Pearson correlation did not show a significant correlation be- and RF speckle probability distribution techniques and a tween intima-media thickness and age ranging between 0.04 Pediatr Radiol (2018) 48:1073–1079 1077 (for B-mode and RF speckle probability distribution) to 0.15 (for RF echo tracking). The AEPC working group and Mannheim consensus stated that there are differences in the left and right intima-media thickness measurements being in- creased on the left side. We herein report an averaged right- and left-side intima-media thickness including one-side data for some participants who did not have both sides. Results remained unchanged when we included only the aver- aged data for the B-mode and RF speckle probability distribution (P=0.032 and P=0.005, respectively). However, the statistical significance was lost for the small sample of RF echo tracking (P=0.25). Discussion Most of the literature on intima-media thickness measurement is on adults. In this study, we compared the various measure- ment techniques in children ages 10–13 years. We found that the three techniques yielded significantly different results. However, regardless of the technique used, the common ca- rotid artery intima-media thickness seemed to increase in chil- dren with increased BMI compared to those with normal BMI. Schreuder et al. [4] found a good correlation between B- mode and RF echo tracking techniques in adults with neuro- logical symptoms related to cardiovascular diseases. They al- so found that B-mode values were higher than those measured by RF echo tracking, similar to our results [4]. Bianchini et al. [19] showed good agreement between B-mode automated system measurement and RF echo tracking techniques in adults with known cardiovascular disease risk factors and in healthy controls. The reported correlation between the two techniques in adults is between 0.76 and 0.86 [4, 20, 21]. These studies state that RF echo tracking-based measurements offer more precision than B-mode measurements. The Mannheim consensus included criteria for obtaining standard- ized intima-media thickness to facilitate data collection, inter- pretation and comparison [13]. The consensus also noted that manual measurements are more observer dependent than semiautomated measurements. In children, manual and semiautomated measurements have been done using B-mode US. Some reports used RF echo tracking techniques [7]. These studies presented a nor- mative range of measurements depending on age and gender [3, 6, 7, 22]. None has compared these techniques. The present Fig. 5 Bland-Altman scatter plots show agreement between the three study showed poor to fair correlation between techniques, but techniques (n=74: dashed lines represent the 95% CI and the regression it also highlighted two important factors. First, it demonstrated line). Solid line represents the mean difference. a Relationship between B- that although very small, there is a statistically significant mode and radiofrequency echo tracking. The range of agreement was −0.01 mm to 0.21 mm; slope: -0.47 (P<0.01). b Relationship between B- lower intima-media thickness in the normal BMI group of mode and radiofrequency probability distribution. The range of agreement children as compared to the increased BMI group using the was0.12mmto0.36mm; slope:-0.41 (P=0.04). c Relationship between RF echo tracking and RF speckle probability distribution tech- radiofrequency echo tracking and radiofrequency probability distribution. niques. Second, the values we obtained for either B-mode or The range of agreement was 0.00 to 0.27 mm; slope: 0.14 (P=0.45). As the mean in (a)and(b) increases, the difference between the techniques decreases RF echo tracking were close to those published in the 1078 Pediatr Radiol (2018) 48:1073–1079 literature. Using semiautomated B-mode measurement, Bohm However, values using R-wave were larger [27]. RF tech- et al. [6] found a mean intima-media thickness of 0.51 mm for niques gave an averaged measure over the entire cardiac cycle. children around 10 years of age, which is close to the 0.56 mm Finally, we did not control for the arterial diameter for observed in our cohort. Engelen et al. [7] found a thickness of both groups and although we averaged left and right 0.4 mm using RF echo tracking for 15 year olds, which is measurements, some participants did not have both sides close to the 0.47 mm in our cohort. RF speckle probability screened. But Loizou et al. [28] did not show a differ- distribution has the smallest values; however, this is close to ence between both sides in healthy asymptomatic sub- the 0.38 mm of the cohort of Jourdan et al. [22] where a jects. This study does not allow us to say which technique is manual B-mode technique was used. The difference between the gold standard. the techniques could be related to the fact that the intima- media thickness is smaller in children than adults and there- fore accurate measurement is challenging. This is in keeping Conclusion with the observation that the regression graphs show a smaller difference with larger measurements. RF-based intima-media thickness in our study was increased Since intima-media thickness in pediatrics is sensitive to in the overweight and obese children compared to normal technique and dependent on age [23] and perhaps gender, the weight children. This difference tends to be borderline signif- AEPC working group set standards for measuring and icant for the B-mode technique. The success rate for the RF reporting it. This is a necessary step before drawing conclu- echo tracking automated technique is inferior to the other two sions about intima-media thickness as a surrogate marker of techniques. The three techniques considered were not inter- subclinical atherosclerosis, especially in obesity [3]. In the changeable suggesting that a single technique should be used same perspective, the American Heart Association published in prospective studies across time for comparability. a scientific statement in 2009 recommending standardized as- sessment of intima-media thickness and arterial stiffness pa- Compliance with ethical standards rameters in children [8]. The cited studies, as ours, demonstrated a relationship between intima-media thick- Conflicts of interest None. ness and BMI in children and between obesity in chil- Open Access This article is distributed under the terms of the Creative dren and intima-media thickness in adulthood [8]. Many Commons Attribution 4.0 International License (http:// of these recommendations are in agreement with those creativecommons.org/licenses/by/4.0/), which permits unrestricted use, of the AEPC working group in terms of technique used. distribution, and reproduction in any medium, provided you give appro- Our results showed clear differences in intima-media priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. thickness between normal BMI and increased BMI group for the RF techniques and a similar trend for B- mode. This is in agreement with the study published by Ozcetin et al. [24] using RF echo tracking. Intima- media thickness is therefore a potential marker of early References vascular changes in at-risk children. However, it should be used in combination with other markers, such as pulse wave 1. McGill HC Jr, McMahan CA, Zieske AW et al (2000) Associations velocity or elastography in order to increase specificity. This of coronary heart disease risk factors with the intermediate lesion of observation applies to children [8, 25] and adults alike [26]. atherosclerosis in youth. The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) research group. Arterioscler Our study has several limitations, including the small num- Thromb Vasc Biol 20:1998–2004 ber of measurements due to the low success rate of the RF 2. Baker JL, Olsen LW, Sorensen TI (2007) Childhood body-mass echo tracking technique. Since the measurements are index and the risk of coronary heart disease in adulthood. N Engl semiautomated, interobserver variability becomes less signif- J Med 357:2329–2337 icant. We did not study the interobserver variability due to the 3. Dalla Pozza R, Ehringer-Schetitska D, Fritsch P et al (2015) Intima media thickness measurement in children: a statement from the small error of estimating the 1-cm distance from the carotid Association for European Paediatric Cardiology (AEPC) working bifurcation. Our measurements did not include different an- group on cardiovascular prevention endorsed by the Association for gles when acquiring the images, contrary to recommendations European Paediatric Cardiology. Atherosclerosis 238:380–387 in the Mannheim consensus. Although we used the 4. Schreuder FH, Graf M, Hameleers JM et al (2009) Measurement of common carotid artery intima-media thickness in clinical practice: Mannheim consensus recommendations for image acquisi- comparison of B-mode and RF-based technique. Ultraschall Med tion, we did not have ECG monitoring for B-mode based 30:459–465 measurements. Selamet Tierney et al. [27]showed good re- 5. Touboul PJ, Vicaut E, Labreuche J et al (2005) Baseline character- producibility in intima-media thickness measures using either istics and carotid intima-media thickness reproducibility in the ECG-gated R-wave or subjective optimal visualization. PARC study. Cerebrovasc Dis 19:57–63 Pediatr Radiol (2018) 48:1073–1079 1079 6. Bohm B, Hartmann K, Buck M et al (2009) Sex differences of 17. Destrempes F, Meunier J, Giroux MF et al (2009) Segmentation in ultrasonic B-mode images of healthy carotid arteries using mixtures carotid intima-media thickness in healthy children and adolescents. Atherosclerosis 206:458–463 of Nakagami distributions and stochastic optimization. IEEE Trans 7. Engelen L, Ferreira I, Stehouwer CD et al (2013) Reference inter- Med Imaging 28:215–229 vals for common carotid intima-media thickness measured with 18. Destrempes F, Meunier J, Giroux MF et al (2011) Segmentation of echotracking: relation with risk factors. Eur Heart J 34:2368–2380 plaques in sequences of ultrasonic B-mode images of carotid arter- 8. Urbina EM, Williams RV, Alpert BS et al (2009) Noninvasive as- ies based on motion estimation and a Bayesian model. IEEE Trans sessment of subclinical atherosclerosis in children and adolescents: Biomed Eng 58:2202–2211 recommendations for standard assessment for clinical research: a 19. Bianchini E, Bozec E, Gemignani V et al (2010) Assessment of scientific statement from the American Heart Association. carotid stiffness and intima-media thickness from ultrasound data. Hypertension 54:919–950 J Ultrasound Med 29:1169–1175 9. Grundy SM, Cleeman JI, Daniels SR et al (2005) Diagnosis and 20. Willekes C, Hoeks AP, Bots ML et al (1999) Evaluation of off-line management of the metabolic syndrome: an American Heart automated intima media thickness detection of the common carotid Association/National Heart, lung and blood institute scientific state- artery based on M-line signal processing. Ultrasound Med Biol 25: ment. Circulation 112:2735–2752 57–64 10. Lambert M, Van Hulst A, O’Loughlin J et al (2012) Cohort profile: 21. Van Bortel LM, Vanmolkot FH, Van der Heijden-Spek JJ et al the Quebec adipose and lifestyle investigation in youth cohort. Int J (2001) Does B-mode common carotid artery intima-media thick- Epidemiol 41:1533–1544 ness differ from M-model? Ultrasound Med Biol 27:1333–1336 11. Paradis G, Lambert M, O'Loughlin J et al (2003) The Québec child 22. Jourdan C, Wuhl E, Litwin M et al (2005) Normative values for and adolescent health and social survey: design and methods of a intima-media thickness and distensibility of large arteries in healthy cardiovascular risk factor survey for youth. Can J Cardiol 19:523– adolescents. J Hypertens 23:1707–1715 23. Buljan K, Butkovic-Soldo S, Janculjak D et al (2015) Relationship 12. Centers for Disease Control and Prevention Growth Charts. https:// between age and thickness of carotid arteries in a population with- cdc.gov/growthcharts/.2000. Accessed 11 Sept 2017 out risk factors for atherosclerosis. Coll Antropol 39:779–784 13. Touboul PJ, Hennerici MG, Meairs S et al (2012) Mannheim carot- 24. Ozcetin M, Celikyay ZR, Celik A et al (2012) The importance of id intima-media thickness and plaque consensus (2004-2006-2011): carotid artery stiffness and increased intima-media thickness in An update on behalf of the advisory board of the 3rd and 4th obese children. S Afr Med J 102:295–299 watching the risk symposium 13th and 15th European Stroke 25. Weberruß H, Pirzer R, Böhm B et al (2016) Intima-media thickness Conferences, Mannheim, Germany, 2004, and Brussels, Belgium, and arterial function in obese and non-obese children. BMC Obes 3: 2011. Cerebrovasc Dis 34:290–296 14. Touboul PJ, Hennerici MG, Meairs S et al (2007) Mannheim carot- 26. Novo G, Di Miceli R, Novo S (2013) Is local stiffness, as measured id intima-media thickness consensus (2004-2006): An update on by radio frequency, more sensitive than intima-media thickness? Int behalf of the advisory board of the 3rd and 4th watching the risk Angiol 32:575–580 symposium 13th and 15th European Stroke Conferences, 27. Selamet Tierney ES, Gauvreau K, Jaff MR et al (2015) Carotid Mannheim, Germany, 2004, and Brussels, Belgium, 2006. artery intima-media thickness measurements in the youth: repro- Cerebrovasc Dis 23:75–80 ducibility and technical considerations. J Am Soc Echocardiogr 15. Touboul JP, Prati P, Scarabin PY et al (1992) Use of monitoring 28:309–316 software to improve the measurement of carotid wall thickness by B-mode imaging. J Hypertens 10:S37–S41 28. Loizou CP, Nicolaides A, Kyriacou E et al (2015) A comparison of 16. Hoeks AP, Willekes C, Boutouyrie P et al (1997) Automated detec- ultrasound intima-media thickness measurements of the left and tion of local artery wall thickness based on M-line signal process- right common carotid artery. IEEE J Transl Eng Health Med 3: ing. Ultrasound Med Biol 23:1017–1023 1900410 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Pediatric Radiology Springer Journals

Carotid artery intima-media thickness measurement in children with normal and increased body mass index: a comparison of three techniques

Free
7 pages

Loading next page...
 
/lp/springer_journal/carotid-artery-intima-media-thickness-measurement-in-children-with-k0wTqL1pN4
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Imaging / Radiology; Pediatrics; Neuroradiology; Nuclear Medicine; Ultrasound; Oncology
ISSN
0301-0449
eISSN
1432-1998
D.O.I.
10.1007/s00247-018-4144-6
Publisher site
See Article on Publisher Site

Abstract

Background Common carotid artery intima-media thickness is a marker of subclinical atherosclerosis. In children, increased intima-media thickness is associated with obesity and the risk of cardiovascular events in adulthood. Objective To compare intima-media thickness measurements using B-mode ultrasound, radiofrequency (RF) echo tracking, and RF speckle probability distribution in children with normal and increased body mass index (BMI). Materials and methods We prospectively measured intima-media thickness in 120 children randomly selected from two groups of a longitudinal cohort: normal BMI and increased BMI, defined by BMI ≥85th percentile for age and gender. We followed Mannheim recommendations. We used M’Ath-Std for automated B-mode imaging, M-line processing of RF signal amplitude for RF echo tracking, and RF signal segmentation and averaging using probability distributions defining image speckle. Statistical analysis included Wilcoxon and Mann-Whitney tests, and Pearson correlation coefficient and intra-class correlation coefficient (ICC). Results Children were 10–13 years old (mean: 11.7 years); 61% were boys. The mean age was 11.4 years (range: 10.0–13.1 years) for the normal BMI group and 12.0 years (range: 10.1–13.5 years) for the increased BMI group. The normal BMI group included 58% boys and the increased BMI group 63% boys. RF echo tracking method was successful in 79 children as opposed to 114 for the B-mode method and all 120 for the probability distribution method. Techniques were weakly correlated: ICC=0.34 (95% confidence interval [CI]: 0.27–0.39). Intima-media thickness was significantly higher in the increased BMI than normal BMI group using the RF techniques and borderline for the B-mode technique. Mean differences between weight groups were: B-mode, 0.02 mm (95% CI: 0.00 to 0.04), P=0.05; RF echo tracking, 0.03 mm (95% CI: 0.01 to 0.05), P=0.01; and RF speckle probability distribution, 0.03 mm (95% CI: 0.01 to 0.05), P=0.002. Conclusion Though techniques are not interchangeable, all showed increased intima-media thickness in children with increased BMI. RF echo tracking method had the lowest success rate at calculating intima-media thickness. For patient follow-up and cohort comparisons, the same technique should be used throughout. . . . . . . Keywords Atherosclerosis B-mode ultrasound Carotid artery Children Intima-media thickness Obesity Radiofrequency echo tracking * Ramy El Jalbout Laboratory of Biorheology and Medical Ultrasonics, ramy.el–jalbout.hsj@ssss.gouv.qc.ca University of Montreal Hospital Research Center (CRCHUM), Montreal, Quebec, Canada 1 Department of Pediatrics, University of Montreal, Department of Medical Imaging, University of Montreal, Sainte-Justine University Health Center, Sainte-Justine University Health Center, Montreal, Quebec, Canada 3175 Cote-Sainte-Catherine Road, Montreal, Quebec H3T 1C5, Canada Department of Radiology, University of Montreal Hospital Center (CHUM), Montreal, Quebec, Canada 1074 Pediatr Radiol (2018) 48:1073–1079 Introduction (>102 cm in males and >88 cm in females) [9]. Exclusion criteria included being a pregnant or breastfeeding mother, Common carotid artery intima-media thickness is a reliable families planning to move, and children on a restricted diet marker of subclinical atherosclerosis [1]. In children, increased or taking anti-hypertensive drugs or corticosteroids [10, 11]. intima-media thickness is associatedwithobesityandthe risk The independent variables included BMI, age and gender. of cardiovascular events in adulthood [1, 2]. The use of intima- All information was collected during the same visit. Weight media thickness in risk stratification, however, is limited by two was measured twice on an electronic scale to the nearest factors. The first is a lack of standardization in the acquisition 0.1 kg. Height was measured twice to the nearest 0.1 cm. and analysis of intima-media thickness measures. The second is BMI was calculated according to the formula weight (kg)/ a lack of reproducibility and validated age- and gender-specific height (m) . We defined overweight and obesity according normative values against which to compare patient results. The to the Centers for Disease Control age- and gender-adjusted Association for European Paediatric Cardiology (AEPC) rec- z-score for weight and height [12]. ognized the need for standardized normative data in its recom- mendations for the measurement, analysis and interpretation of Image acquisition intima-media thickness in children [3]. There are two main methods for measuring common carot- Primary end points were intima-media thickness measured by id artery intima-media thickness: ultrasound (US) B-mode B-mode, RF echo tracking and RF speckle probability distri- images (manual, semiautomated or automated) and radiofre- bution. Images were acquired according to the Mannheim quency (RF) multiple M-line analysis (using echo tracking of recommendations [13]. Participants were in the supine posi- RF amplitude). In adults, the two methods correlate well [4]. tion with their head tilted 45° to the opposite side. We mea- Automated B-mode imaging with dedicated software (M’Ath- sured the far wall of the common carotid artery at 1 cm from Std) was validated for intima-media thickness measurement the carotid bifurcation. A high frequency linear US probe was and showed excellent reproducibility in the PARC (Paroi positioned longitudinally perpendicular to the vessel wall. The Artérielle et Risque Cardiovasculaire) study [5]. In pediatrics, software, whether for the B-mode measurement or for the RF studies used either US [6] or echo tracking [7]. To our knowl- echo tracking measurement, calculates the intima-media edge, there are no studies comparing B-mode and RF in chil- thickness automatically along the indicated segment. dren and few published normative values using either tech- nique [3, 6, 7]. B-mode US has nonetheless demonstrated B-mode intima-media thickness increased intima-media thickness in obese children [8]. The objective of this study was therefore to compare com- B-mode images were obtained using an ATL 5000 HDI US mon carotid artery intima-media thickness using B-mode US, unit (Advanced Technology Laboratories, Bothell, WA) RF echo tracking, and segmentation of the vessel wall using equipped with a 5.5–12.5 MHz probe. Focus depth was set RF speckle probability distribution in children with normal around 25 mm, with a frame rate of 25 per s and gain settings and increased body mass index (BMI). adjusted optimally to facilitate edge detection. A US technol- ogist with 10 years of experience scanned the child and saved a static longitudinal image of the common carotid artery that Materials and methods shows best the intima-media-lumen interface on gray-scale image. The 2-D image is instantaneously transferred to a com- Study design and patient population puter with M’Ath-Std software (Metris, Argenteuil, France). The technologist then drew a 1-cm-long line next to the We randomly selected 120 children from two groups of the intima-media-lumen interface (Fig. 1). Automated computer- Quebec Adipose and Lifestyle Investigation in Youth ized intima-media thickness is then generated by the software (QUALITY) prospective cohort of 564 children: a normal [14, 15]. We sought a quality index >0.5 for each measure BMI group and an increased BMI group, as defined by BMI indicating that more than 50% of the measurements along ≥85th percentile for age and gender, 2 years after enrollment the 1-cm-long line were used to calculate the final averaged in the cohort between 2005 and 2011. The study was approved intima-media thickness. An average of two values was obtain- by the Ethics Boards of the Quebec Heart and Lung Institute ed during optimal resolution, independent of cardiac cycle. and Sainte-Justine University Hospital. Written informed as- sent and consent were obtained from all participants and their RF echo tracking intima-media thickness parents, respectively. Inclusion criteria were Caucasian origin, age 8–10 years at time of entry into the original cohort, We used a linear array L10–5 40-mm US transducer (model healthy, and having at least one obese parent (defined by 410,503) on the MyLab 70 platform (ESAOTE) equipped with BMI ≥30 kg/m or an elevated waist circumference an ART.LAB platform (Pie Medical Imaging BV, Maastricht, Pediatr Radiol (2018) 48:1073–1079 1075 calculating RF US signals received along a single line of ob- servation (M-line processing) with time-averaging (using an electrocardiogram [ECG]) and motion compensation algo- rithms [16]. An average of two such intima-media thickness values was obtained (Fig. 2). RF speckle probability distribution intima-media thickness The same RF-digitized data stream (cine-loop acquisition of 2 to 5 cardiac cycles including 250–300 images) was post- processed to obtain an intima-media thickness using a differ- ent semiautomatic algorithm. This algorithm was implement- ed on a dedicated platform (ORS Visual; Object Research Systems Inc., Montreal, Canada). A research assistant (M.- H. R. C.) with 8 years of experience blinded to the results of B mode and RF echo tracking manually initialized the seg- Fig. 1 B-mode longitudinal US image of the common carotid artery of a mentation on one image of the data stream by drawing two 12-year-old boy shows the line drawn by the technologist next to the contours: the lumen/intima-media and the intima-media/ad- intima-media-lumen interface indicating the location of the automatic ventitia interfaces. The software automatically computed measurement by the M’Ath-Std system software. The software automatically gave the intima-media thickness value averaged over the these two interfaces on all subsequent frames (Fig. 3). The 1-cm segment of the vessel and gave the quality index (we aimed an index intima-media thickness was the average of all segmented superior to 0.5 as explained in text) frames of the cine-loop of the mean distance between the two segmented interfaces. The segmentation of the vessel wall The Netherlands). A 2-5 cardiac cycles long RF-digitized data was based on the method described by Destrempes et al. [17, stream is recorded by the technologist who then indicates a 18]. During the segmentation process, the initial contours are region of interest along the intima-media-lumen interface. refined and tracked over the whole video sequence. For this Automated intima-media thickness is then obtained by Fig. 3 A static image of the common carotid artery of a 12-year-old boy demonstrates the radiofrequency speckle probability distribution technique taken from a video sequence on which automatically computed contours of the intima-media borders (green) with the lumen Fig. 2 Radiofrequency echo tracking static image of an 11-year-old girl and adventitia are shown. The segmentation is manually initialized on a shows the location to measure the intima-media thickness (green) on the single frame (not shown). Both contours are then automatically traced on far vessel wall indicated by the technologist. The software automatically all subsequent frames of the 2–5 cardiac cycle-long video sequence. The gave individual intima-media thickness values during the entire cardiac intima-media thickness is the average (over all frames) of the mean cycle depending on the lumen diameter as well as the averaged final value distance between the two segmented contours 1076 Pediatr Radiol (2018) 48:1073–1079 purpose, a Bayesian segmentation model was used. In this model, the echogenicity of the intima-media layer, the lumen and the adventitia are modeled by mixtures of Nakagami dis- tributions and the motion of the vessel wall is estimated with optical flow. An average of systolic and diastolic intima-media thickness of all subsegments is obtained. Statistical analysis We did a descriptive analysis of the study population, checked for normality of the data distribution using the Shapiro-Wilk test, and used the Wilcoxon test for paired samples to compare intima-media thickness obtained by each of the three tech- niques. The intra-class correlation coefficient (ICC) was used to measure correlation. We plotted Bland-Altman and regres- Fig. 4 Box and whisker plot of mean intima-media thickness: B-mode: sion graphs to verify the ranges of agreement. We checked for 0.56+/−0.004 mm (95% confidence interval [CI]: 0.55–0.57), radiofrequency echo tracking: 0.47+/−0.006 mm (95% CI: 0.46–0.48) the correlation between intima-media thickness and age using and radiofrequency probability distribution: 0.34+/−0.005 mm (95% the Pearson correlation. We used the Mann-Whitney U test for CI: 0.33–0.35) the difference between children with normal BMI and those with increased BMI. Analysis was done using MedCalc sta- tistical software version 17.9.7 (Ostend, Belgium). difference of 0.14 mm (95% CI: 0.12–0.16, P<0.001) for the comparison of the RF echo tracking and RF speckle probabil- Results ity distribution techniques. ICC for the comparison of all three techniques was 0.34 (95% CI: 0.27–0.39). Agreement and The age range of the 120 randomly selected children was 10 to regression analysis between any two techniques is shown in 13 years including 73 boys. The mean age was 11.4 years the Bland-Altman scatter plots (Fig. 5). (95% confidence interval [CI]: 11.2–11.6) for the normal The mean B-mode intima-media thickness for normal BMI BMI group and 12.0 years (95% CI: 11.8–12.2) for the in- group (n=59) was 0.55+/−0.006 mm (95% CI: 0.54–0.56). creased BMI group. In the normal BMI group, 58% were For the increased BMI group (n=55), B-mode intima-media boys, and in the increased BMI group 63% were boys. The thickness was 0.57+/−0.006 mm (95% CI: 0.56–0.58). The two groups were significantly different with respect to BMI mean RF echo tracking intima-media thickness for the normal (mean z-score=−0.73 (95% CI: -0.92 to-0.54) for the normal BMI group (n=46) was 0.46+/−0.008 mm (95% CI: 0.45– BMI group and 1.96 (95% CI: 1.83–2.08) for the increased 0.47) and for the increased BMI group (n=33), the mean BMI group (P<0.001). The age was statistically different was 0.49+/−0.008 (95% CI: 0.48–0.50). The mean RF speckle (P<0.001). Gender ratios did not differ significantly between probability distribution intima-media thickness for the normal the groups (P=0.7). Although we measured intima-media BMI group (n=60) was 0.32+/−0.006 mm (95% CI: 0.31– thickness on all children, data for RF echo tracking were often 0.33). For the increased BMI group (n=60), the mean was not available as the software could not automatically detect the 0.36+/−0.006 mm (95% CI: 0.35–0.37). intima-media-lumen interface mainly due to artifacts. We Intima-media thickness was statistically significantly could not call the children back as the present analysis was higher in the increased BMI group for the RF echo tracking started in 2016, at least 5 years after data collection. Final and the RF speckle probability distribution techniques and analysis included 114 children for B-mode, 79 for RF echo borderline significant for the B-mode technique. The mean tracking and 120 for RF speckle probability distribution. differences using the Mann-Whitney U test were calculated Therefore, for the 74 children who had all three measurements at 0.02 mm+/−0.01 (95% CI: 0.00–0.04, P=0.05) with the B- available, intima-media thickness differed significantly de- mode technique, 0.03 mm +/−0.01 (95% CI: 0.01–0.05, pending on technique. Mean intima-media thicknesses are P=0.02) with the RF echo tracking technique and 0.03 mm shown in Fig. 4. +/−0.01 (95% CI: 0.01–0.05, P=0.002) for the RF speckle The Wilcoxon test showed significant differences of probability distribution technique. Although age was statisti- 0.10 mm (95% CI: 0.09–0.11, P<0.001) for the comparison cally different between the groups, the difference was not of the B-mode and RF echo tracking techniques, a difference clinically significant (11.4 and 12.0 years). Furthermore, the of 0.24 mm (95% CI: 0.23–0.25, P<0.001) for the B-mode Pearson correlation did not show a significant correlation be- and RF speckle probability distribution techniques and a tween intima-media thickness and age ranging between 0.04 Pediatr Radiol (2018) 48:1073–1079 1077 (for B-mode and RF speckle probability distribution) to 0.15 (for RF echo tracking). The AEPC working group and Mannheim consensus stated that there are differences in the left and right intima-media thickness measurements being in- creased on the left side. We herein report an averaged right- and left-side intima-media thickness including one-side data for some participants who did not have both sides. Results remained unchanged when we included only the aver- aged data for the B-mode and RF speckle probability distribution (P=0.032 and P=0.005, respectively). However, the statistical significance was lost for the small sample of RF echo tracking (P=0.25). Discussion Most of the literature on intima-media thickness measurement is on adults. In this study, we compared the various measure- ment techniques in children ages 10–13 years. We found that the three techniques yielded significantly different results. However, regardless of the technique used, the common ca- rotid artery intima-media thickness seemed to increase in chil- dren with increased BMI compared to those with normal BMI. Schreuder et al. [4] found a good correlation between B- mode and RF echo tracking techniques in adults with neuro- logical symptoms related to cardiovascular diseases. They al- so found that B-mode values were higher than those measured by RF echo tracking, similar to our results [4]. Bianchini et al. [19] showed good agreement between B-mode automated system measurement and RF echo tracking techniques in adults with known cardiovascular disease risk factors and in healthy controls. The reported correlation between the two techniques in adults is between 0.76 and 0.86 [4, 20, 21]. These studies state that RF echo tracking-based measurements offer more precision than B-mode measurements. The Mannheim consensus included criteria for obtaining standard- ized intima-media thickness to facilitate data collection, inter- pretation and comparison [13]. The consensus also noted that manual measurements are more observer dependent than semiautomated measurements. In children, manual and semiautomated measurements have been done using B-mode US. Some reports used RF echo tracking techniques [7]. These studies presented a nor- mative range of measurements depending on age and gender [3, 6, 7, 22]. None has compared these techniques. The present Fig. 5 Bland-Altman scatter plots show agreement between the three study showed poor to fair correlation between techniques, but techniques (n=74: dashed lines represent the 95% CI and the regression it also highlighted two important factors. First, it demonstrated line). Solid line represents the mean difference. a Relationship between B- that although very small, there is a statistically significant mode and radiofrequency echo tracking. The range of agreement was −0.01 mm to 0.21 mm; slope: -0.47 (P<0.01). b Relationship between B- lower intima-media thickness in the normal BMI group of mode and radiofrequency probability distribution. The range of agreement children as compared to the increased BMI group using the was0.12mmto0.36mm; slope:-0.41 (P=0.04). c Relationship between RF echo tracking and RF speckle probability distribution tech- radiofrequency echo tracking and radiofrequency probability distribution. niques. Second, the values we obtained for either B-mode or The range of agreement was 0.00 to 0.27 mm; slope: 0.14 (P=0.45). As the mean in (a)and(b) increases, the difference between the techniques decreases RF echo tracking were close to those published in the 1078 Pediatr Radiol (2018) 48:1073–1079 literature. Using semiautomated B-mode measurement, Bohm However, values using R-wave were larger [27]. RF tech- et al. [6] found a mean intima-media thickness of 0.51 mm for niques gave an averaged measure over the entire cardiac cycle. children around 10 years of age, which is close to the 0.56 mm Finally, we did not control for the arterial diameter for observed in our cohort. Engelen et al. [7] found a thickness of both groups and although we averaged left and right 0.4 mm using RF echo tracking for 15 year olds, which is measurements, some participants did not have both sides close to the 0.47 mm in our cohort. RF speckle probability screened. But Loizou et al. [28] did not show a differ- distribution has the smallest values; however, this is close to ence between both sides in healthy asymptomatic sub- the 0.38 mm of the cohort of Jourdan et al. [22] where a jects. This study does not allow us to say which technique is manual B-mode technique was used. The difference between the gold standard. the techniques could be related to the fact that the intima- media thickness is smaller in children than adults and there- fore accurate measurement is challenging. This is in keeping Conclusion with the observation that the regression graphs show a smaller difference with larger measurements. RF-based intima-media thickness in our study was increased Since intima-media thickness in pediatrics is sensitive to in the overweight and obese children compared to normal technique and dependent on age [23] and perhaps gender, the weight children. This difference tends to be borderline signif- AEPC working group set standards for measuring and icant for the B-mode technique. The success rate for the RF reporting it. This is a necessary step before drawing conclu- echo tracking automated technique is inferior to the other two sions about intima-media thickness as a surrogate marker of techniques. The three techniques considered were not inter- subclinical atherosclerosis, especially in obesity [3]. In the changeable suggesting that a single technique should be used same perspective, the American Heart Association published in prospective studies across time for comparability. a scientific statement in 2009 recommending standardized as- sessment of intima-media thickness and arterial stiffness pa- Compliance with ethical standards rameters in children [8]. The cited studies, as ours, demonstrated a relationship between intima-media thick- Conflicts of interest None. ness and BMI in children and between obesity in chil- Open Access This article is distributed under the terms of the Creative dren and intima-media thickness in adulthood [8]. Many Commons Attribution 4.0 International License (http:// of these recommendations are in agreement with those creativecommons.org/licenses/by/4.0/), which permits unrestricted use, of the AEPC working group in terms of technique used. distribution, and reproduction in any medium, provided you give appro- Our results showed clear differences in intima-media priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. thickness between normal BMI and increased BMI group for the RF techniques and a similar trend for B- mode. This is in agreement with the study published by Ozcetin et al. [24] using RF echo tracking. Intima- media thickness is therefore a potential marker of early References vascular changes in at-risk children. However, it should be used in combination with other markers, such as pulse wave 1. McGill HC Jr, McMahan CA, Zieske AW et al (2000) Associations velocity or elastography in order to increase specificity. This of coronary heart disease risk factors with the intermediate lesion of observation applies to children [8, 25] and adults alike [26]. atherosclerosis in youth. The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) research group. Arterioscler Our study has several limitations, including the small num- Thromb Vasc Biol 20:1998–2004 ber of measurements due to the low success rate of the RF 2. Baker JL, Olsen LW, Sorensen TI (2007) Childhood body-mass echo tracking technique. Since the measurements are index and the risk of coronary heart disease in adulthood. N Engl semiautomated, interobserver variability becomes less signif- J Med 357:2329–2337 icant. We did not study the interobserver variability due to the 3. Dalla Pozza R, Ehringer-Schetitska D, Fritsch P et al (2015) Intima media thickness measurement in children: a statement from the small error of estimating the 1-cm distance from the carotid Association for European Paediatric Cardiology (AEPC) working bifurcation. Our measurements did not include different an- group on cardiovascular prevention endorsed by the Association for gles when acquiring the images, contrary to recommendations European Paediatric Cardiology. Atherosclerosis 238:380–387 in the Mannheim consensus. Although we used the 4. Schreuder FH, Graf M, Hameleers JM et al (2009) Measurement of common carotid artery intima-media thickness in clinical practice: Mannheim consensus recommendations for image acquisi- comparison of B-mode and RF-based technique. Ultraschall Med tion, we did not have ECG monitoring for B-mode based 30:459–465 measurements. Selamet Tierney et al. [27]showed good re- 5. Touboul PJ, Vicaut E, Labreuche J et al (2005) Baseline character- producibility in intima-media thickness measures using either istics and carotid intima-media thickness reproducibility in the ECG-gated R-wave or subjective optimal visualization. PARC study. Cerebrovasc Dis 19:57–63 Pediatr Radiol (2018) 48:1073–1079 1079 6. Bohm B, Hartmann K, Buck M et al (2009) Sex differences of 17. Destrempes F, Meunier J, Giroux MF et al (2009) Segmentation in ultrasonic B-mode images of healthy carotid arteries using mixtures carotid intima-media thickness in healthy children and adolescents. Atherosclerosis 206:458–463 of Nakagami distributions and stochastic optimization. IEEE Trans 7. Engelen L, Ferreira I, Stehouwer CD et al (2013) Reference inter- Med Imaging 28:215–229 vals for common carotid intima-media thickness measured with 18. Destrempes F, Meunier J, Giroux MF et al (2011) Segmentation of echotracking: relation with risk factors. Eur Heart J 34:2368–2380 plaques in sequences of ultrasonic B-mode images of carotid arter- 8. Urbina EM, Williams RV, Alpert BS et al (2009) Noninvasive as- ies based on motion estimation and a Bayesian model. IEEE Trans sessment of subclinical atherosclerosis in children and adolescents: Biomed Eng 58:2202–2211 recommendations for standard assessment for clinical research: a 19. Bianchini E, Bozec E, Gemignani V et al (2010) Assessment of scientific statement from the American Heart Association. carotid stiffness and intima-media thickness from ultrasound data. Hypertension 54:919–950 J Ultrasound Med 29:1169–1175 9. Grundy SM, Cleeman JI, Daniels SR et al (2005) Diagnosis and 20. Willekes C, Hoeks AP, Bots ML et al (1999) Evaluation of off-line management of the metabolic syndrome: an American Heart automated intima media thickness detection of the common carotid Association/National Heart, lung and blood institute scientific state- artery based on M-line signal processing. Ultrasound Med Biol 25: ment. Circulation 112:2735–2752 57–64 10. Lambert M, Van Hulst A, O’Loughlin J et al (2012) Cohort profile: 21. Van Bortel LM, Vanmolkot FH, Van der Heijden-Spek JJ et al the Quebec adipose and lifestyle investigation in youth cohort. Int J (2001) Does B-mode common carotid artery intima-media thick- Epidemiol 41:1533–1544 ness differ from M-model? Ultrasound Med Biol 27:1333–1336 11. Paradis G, Lambert M, O'Loughlin J et al (2003) The Québec child 22. Jourdan C, Wuhl E, Litwin M et al (2005) Normative values for and adolescent health and social survey: design and methods of a intima-media thickness and distensibility of large arteries in healthy cardiovascular risk factor survey for youth. Can J Cardiol 19:523– adolescents. J Hypertens 23:1707–1715 23. Buljan K, Butkovic-Soldo S, Janculjak D et al (2015) Relationship 12. Centers for Disease Control and Prevention Growth Charts. https:// between age and thickness of carotid arteries in a population with- cdc.gov/growthcharts/.2000. Accessed 11 Sept 2017 out risk factors for atherosclerosis. Coll Antropol 39:779–784 13. Touboul PJ, Hennerici MG, Meairs S et al (2012) Mannheim carot- 24. Ozcetin M, Celikyay ZR, Celik A et al (2012) The importance of id intima-media thickness and plaque consensus (2004-2006-2011): carotid artery stiffness and increased intima-media thickness in An update on behalf of the advisory board of the 3rd and 4th obese children. S Afr Med J 102:295–299 watching the risk symposium 13th and 15th European Stroke 25. Weberruß H, Pirzer R, Böhm B et al (2016) Intima-media thickness Conferences, Mannheim, Germany, 2004, and Brussels, Belgium, and arterial function in obese and non-obese children. BMC Obes 3: 2011. Cerebrovasc Dis 34:290–296 14. Touboul PJ, Hennerici MG, Meairs S et al (2007) Mannheim carot- 26. Novo G, Di Miceli R, Novo S (2013) Is local stiffness, as measured id intima-media thickness consensus (2004-2006): An update on by radio frequency, more sensitive than intima-media thickness? Int behalf of the advisory board of the 3rd and 4th watching the risk Angiol 32:575–580 symposium 13th and 15th European Stroke Conferences, 27. Selamet Tierney ES, Gauvreau K, Jaff MR et al (2015) Carotid Mannheim, Germany, 2004, and Brussels, Belgium, 2006. artery intima-media thickness measurements in the youth: repro- Cerebrovasc Dis 23:75–80 ducibility and technical considerations. J Am Soc Echocardiogr 15. Touboul JP, Prati P, Scarabin PY et al (1992) Use of monitoring 28:309–316 software to improve the measurement of carotid wall thickness by B-mode imaging. J Hypertens 10:S37–S41 28. Loizou CP, Nicolaides A, Kyriacou E et al (2015) A comparison of 16. Hoeks AP, Willekes C, Boutouyrie P et al (1997) Automated detec- ultrasound intima-media thickness measurements of the left and tion of local artery wall thickness based on M-line signal process- right common carotid artery. IEEE J Transl Eng Health Med 3: ing. Ultrasound Med Biol 23:1017–1023 1900410

Journal

Pediatric RadiologySpringer Journals

Published: May 9, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off