Deep Learning-Based Carotid Plaque Segmentation from B-Mode Ultrasound Images

Carotid ultrasound measurement of total plaque area (TPA) provides a method for quantifying carotid plaque burden and monitoring changes in carotid atherosclerosis in response to medical treatment. Plaque boundary segmentation is required to generate the TPA measurement; however, training of observers and manual delineation are time consuming. Thus, our objective was to develop an automated plaque segmentation method to generate TPA from longitudinal carotid ultrasound images. In this study, a deep learning-based method, modified U-Net, was used to train the segmentation model and generate TPA measurement. A total of 510 plaques from 144 patients were used in our study, where the Monte Carlo cross-validation was used by randomly splitting the data set into 2/3 and 1/3 for training and testing. Two observers were trained to manually delineate the 510 plaques separately, which were used as the ground-truth references. Two U-Net models (M1 and M2) were trained using the two different ground-truth data sets from the two observers to evaluate the accuracy, variability and sensitivity on the ground-truth data sets used for training our method. The results of the algorithm segmentations of the two models yielded strong agreement with the two manual segmentations with the Pearson correlation coefficient r = 0.989 (p < 0.0001) and r = 0.987 (p < 0.0001). Comparison of the U-Net and manual segmentations resulted in mean TPA differences of 0.05 ± 7.13 mm² (95% confidence interval: 14.02–13.02 mm²) and 0.8 ± 8.7 mm² (17.85–16.25 mm²) for the two models, which are small compared with the TPA range in our data set from 4.7 to 312.8 mm². Furthermore, the mean time to segment a plaque was only 8.3 ± 3.1 ms. The presented deep learning-based method described has sufficient accuracy with a short computation time and exhibits high agreement between the algorithm and manual TPA measurements, suggesting that the method could be used to measure TPA and to monitor the progression and regression of carotid atherosclerosis.     

Three-dimensional Ultrasound Quantification of Intensive Statin Treatment of Carotid Atherosclerosis

This study was designed to evaluate 3-D ultrasound (3DUS)–derived vessel wall volume (VWV), a 3-D measurement of the carotid artery intima and media, including atherosclerotic plaque, in patients enrolled in a randomized placebo-controlled three-month study of intensive atorvastatin treatment. Thirty-five subjects with carotid stenosis >60% who provided written informed consent and completed a randomized, double-blind, placebo-controlled study were evaluated at baseline and at three months after receiving either 80 mg atorvastatin (16 subjects, nine male, mean age 68 ± 8.6 y) or placebo (19 subjects, 15 male, mean age 70 ± 9.4 y) daily. 3DUS images were acquired and 3DUS VWV was manually segmented by a single observer. Individual lumen and wall segmentation contours were also used to generate carotid atherosclerosis thickness difference maps by establishing correspondence between points along the vessel wall and lumen segmentation contour surfaces, and digitally subtracting registered baseline and follow-up thickness maps. 3DUS VWV increased by 70 ± 140 mm³ (+4.9 ± 10.3%) in the placebo group and decreased by 30 ± 110 mm³ (-1.4 ± 7.7%) in the atorvastatin group (p < 0.05). Two-dimensional maps generated from the VWV measurements show localized heterogeneity and vessel wall thickness changes for all subjects, mainly in the common carotid artery. Carotid 3DUS VWV is a quantitative measure of atherosclerosis burden including the intima, media and plaque, with sensitivity to detect changes over short periods of time. Quantitative VWV thickness difference maps provide visual evidence of the spatial and temporal dynamics of carotid artery changes. (E-mail: gep@imaging.robarts.ca)     

Automated Skin Segmentation in Ultrasonic Evaluation of Skin Toxicity in Breast Cancer Radiotherapy

Skin toxicity is the most common side effect of breast cancer radiotherapy and impairs the quality of life of many breast cancer survivors. We, along with other researchers, have recently found quantitative ultrasound to be effective as a skin toxicity assessment tool. Although more reliable than standard clinical evaluations (visual observation and palpation), the current procedure for ultrasound-based skin toxicity measurements requires manual delineation of the skin layers (i.e., epidermis-dermis and dermis-hypodermis interfaces) on each ultrasound B-mode image. Manual skin segmentation is time consuming and subjective. Moreover, radiation-induced skin injury may decrease image contrast between the dermis and hypodermis, which increases the difficulty of delineation. Therefore, we have developed an automatic skin segmentation tool (ASST) based on the active contour model with two significant modifications: (i) The proposed algorithm introduces a novel dual-curve scheme for the double skin layer extraction, as opposed to the original single active contour method. (ii) The proposed algorithm is based on a geometric contour framework as opposed to the previous parametric algorithm. This ASST algorithm was tested on a breast cancer image database of 730 ultrasound breast images (73 ultrasound studies of 23 patients). We compared skin segmentation results obtained with the ASST with manual contours performed by two physicians. The average percentage differences in skin thickness between the ASST measurement and that of each physician were less than 5% (4.8 ± 17.8% and −3.8 ± 21.1%, respectively). In summary, we have developed an automatic skin segmentation method that ensures objective assessment of radiation-induced changes in skin thickness. Our ultrasound technology offers a unique opportunity to quantify tissue injury in a more meaningful and reproducible manner than the subjective assessments currently employed in the clinic.     

Inter-Scan Reproducibility of Carotid Plaque Volume Measurements by 3-D Ultrasound

We tested a novel 3-D matrix transducer with respect to inter-scan reproducibility of carotid maximum plaque thickness (MPT) and volume measurements. To improve reproducibility while focusing on the largest plaque/most diseased part of the carotid artery, we introduced a new partial plaque volume (PPV) measure centered on MPT. Total plaque volume (TPV), PPV from a 10-mm segment and MPT were measured using dedicated semi-automated software on 38 plaques from 26 patients. Inter-scan reproducibility was assessed using the t-test, Bland–Altman plots and Pearson's correlation coefficient. There was a mean difference of 0.01?mm in MPT (limits of agreement: ?0.45 to 0.42?mm, Pearson's correlation coefficient: 0.96). Both volume measurements exhibited high reproducibility, with PPV being superior (limits of agreement: ?35.3?mm³ to 33.5?mm³, Pearson's correlation coefficient: 0.96) to TPV (limits of agreement: ?88.2 to 61.5?mm³, Pearson's correlation coefficient: 0.91). The good reproducibility revealed by the present results encourages future studies on establishing plaque quantification as part of cardiovascular risk assessment and for follow-up of disease progression over time.     

Method for Carotid Artery 3-D Ultrasound Image Segmentation Based on CSWin Transformer

Precise segmentation of carotid artery (CA) structure is an important prerequisite for the medical assessment and detection of carotid plaques. For automatic segmentation of the media–adventitia boundary (MAB) and lumen–intima boundary (LIB) in 3-D ultrasound images of the CA, a U-shaped CSWin transformer (U-CSWT) is proposed. Both the encoder and decoder of the U-CSWT are composed of hierarchical CSWT modules, which can capture rich global context information in the 3-D image. Experiments were performed on a 3-D ultrasound image data set of the CA, and the results indicate that the U-CSWT performs better than other convolutional neural network (CNN)-based and CNN–transformer hybrid methods. The model yields Dice coefficients of 94.6 ± 3.0% and 90.8 ± 5.1% for the MAB and LIB in the common carotid artery (CCA) and 92.9 ± 4.9% and 89.6 ± 6.2% for MAB and LIB in the bifurcation, respectively. Our U-CSWT is expected to become an effective method for automatic segmentation of 3-D ultrasound images of CA.     

Reproducibility of Two 3-D Ultrasound Carotid Plaque Quantification Methods

Compared with single 2-D images, emerging 3-D ultrasound technologies hold the promise of reducing variability and increasing sensitivity in the quantification of carotid plaques for individual cardiovascular risk stratification. Inter- and intra-observer agreement between a manual, cross-sectional, 2-D freehand sweep and a mechanical 3-D ultrasound investigation of 62 carotid artery plaques is reported with intra-class correlation coefficients (with 95% confidence intervals). Inter-observer agreement was 0.60 (0.29–0.77) for the freehand method and 0.89 (0.83–0.93) for the mechanical 3-D acquisition. The use of semi-automated computerized planimetric measurements of plaque burden has high intra-observer repeatability, but is vulnerable to systematic inter-observer differences. For the 2-D freehand sweep, a considerable contribution to variation is introduced by the scanning procedure itself, that is, the lack of controlled motion along the third dimension. Future implementation of 3-D ultrasound quantification in large-scale studies of inter-individual cardiovascular risk assessment seems justified using the methods described.     

Fully Automated Carotid Plaque Segmentation in Combined Contrast-Enhanced and B-Mode Ultrasound

Carotid plaque segmentation in B-mode ultrasound (BMUS) and contrast-enhanced ultrasound (CEUS) is crucial to the assessment of plaque morphology and composition, which are linked to plaque vulnerability. Segmentation in BMUS is challenging because of noise, artifacts and echo-lucent plaques. CEUS allows better delineation of the lumen but contains artifacts and lacks tissue information. We describe a method that exploits the combined information from simultaneously acquired BMUS and CEUS images. Our method consists of non-rigid motion estimation, vessel detection, lumen–intima segmentation and media–adventitia segmentation. The evaluation was performed in training (n = 20 carotids) and test (n = 28) data sets by comparison with manually obtained ground truth. The average root-mean-square errors in the training and test data sets were comparable for media–adventitia (411 ± 224 and 393 ± 239 μm) and for lumen–intima (362 ± 192 and 388 ± 200 μm), and were comparable to inter-observer variability. To the best of our knowledge, this is the first method to perform fully automatic carotid plaque segmentation using combined BMUS and CEUS.     

Three-Dimensional Carotid Ultrasound Segmentation Variability Dependence on Signal Difference and Boundary Orientation

Quantitative measurements of the progression (or regression) of carotid plaque burden are important in monitoring patients and evaluating new treatment options. We previously developed a quantitative metric to analyze changes in carotid plaque morphology from 3-D ultrasound (US) on a point-by-point basis. This method requires multiple segmentations of the arterial wall and lumen boundaries to obtain the local standard deviation (SD) of vessel-wall-plus-plaque thickness (VWT) so that t-tests could be used to determine whether a change in VWT is statistically significant. However, the requirement for multiple segmentations makes clinical trials laborious and time-consuming. Therefore, this study was designed to establish the relationship between local segmentation SD and local signal difference on the arterial wall and lumen boundaries. We propose metrics to quantify segmentation SD and signal difference on a point-by-point basis, and studied whether the signal difference at arterial wall or lumen boundaries could be used to predict local segmentation SD. The ability to predict the local segmentation SD could eliminate the need of repeated segmentations of a 2-D transverse image to obtain the local segmentation standard deviation, thereby making clinical trials less laborious and saving time. Six subjects involved in this study were associated with different degrees of atherosclerosis: three carotid stenosis subjects with mean plaque area >3 cm² and >60% carotid stenosis were involved in a clinical study evaluating the effect of atorvastatin, a cholesterol-lowering and plaque-stabilizing drug; and three subjects with carotid plaque area >0.5 cm² were subjects with moderate atherosclerosis. Our results suggest that when local signal difference is higher than 8 greyscale value (GSV), the local segmentation SD stabilizes at 0.05 mm and is thus predictable. This information provides a target value of local signal difference on the arterial boundaries that should be achieved to obtain an accurate prediction of local segmentation SD. (E-mail: bcychiu@alumni.uwo.ca)     

Endovascular Shear Strain Elastography for the Detection and Characterization of the Severity of Atherosclerotic Plaques: In Vitro Validation and In Vivo Evaluation

This work explores the potential of shear strain elastograms to identify vulnerable atherosclerotic plaques. The Lagrangian speckle model estimator (LSME) elasticity imaging method was further developed to estimate shear strain elasticity (SSE). Three polyvinyl alcohol cryogel vessel phantoms were imaged with an intravascular ultrasound (IVUS) scanner. The estimated SSE maps were validated against finite-element results. Atherosclerosis was induced in carotid arteries of eight Sinclair mini-pigs using a combination of surgical techniques, diabetes and a high-fat diet. IVUS images were acquired in vivo in 14 plaques before euthanasia and histology. All plaques were characterized by high magnitudes in SSE maps that correlated with American Heart Association atherosclerosis stage classifications (r = 0.97, p < 0.001): the worse the plaque condition the higher was the absolute value of SSE, i.e. |SSE| (e.g., mean |SSE| was 3.70 ± 0.40% in Type V plaques, whereas it was reduced to 0.11 ± 0.01% in normal walls). This study indicates the feasibility of using SSE to highlight atherosclerotic plaque vulnerability characteristics.     

Comparison of B‐Mode Ultrasound, 3‐Dimensional Ultrasound,and Magnetic Resonance Imaging Measurements of Carotid Atherosclerosis

Objective. We compared the intraobserver and interscan variability of carotid atherosclerosis measured using B‐mode ultrasound for quantifying intima media thickness (IMT), 3‐dimensional ultrasound (3DUS) for quantifying vessel wall volume (VWV) and total plaque volume (TPV), and magnetic resonance imaging (MRI) for measuring VWV. We also evaluated the associations of these measurements and sample sizes required to detect specific changes in patients with moderate atherosclerosis. Methods. Ten patients were evaluated with B‐mode ultrasound, MRI, and 3DUS twice within 14 ± 2 days. Measurements of IMT, MRI VWV, 3DUS VWV, and 3DUS TPV were performed by single observers using manual (VWV and TPV) and semiautomated (IMT) segmentation. Results. Intraobserver coefficients of variation were 3.4% (IMT), 4.7% (3DUS VWV), 6.5% (MRI VWV), and 23.9% (3DUS TPV). Interscan coefficients of variation were 8.1% (MRI VWV), 8.9% (IMT), 13.5% (3DUS VWV), and 46.6% (3DUS TPV). Scan‐rescan linear regressions were significant for 3DUS TPV (R² = 0.57), 3DUS VWV (R² = 0.59), and IMT (R² = 0.75) and significantly different (P < .05) for MRI VWV (R² = 0.87). Conclusions. B‐mode ultrasound‐derived IMT provided the highest intraobserver and interscan reproducibility. Three‐dimensional measurements of VWV derived from 3DUS and MRI provided both high sensitivity and high intraobserver and interscan reliability.     

Non-invasive Vascular Radial/Circumferential Strain Imaging and Wall Shear Rate Estimation Using Video Images of Diagnostic Ultrasound

The aim of this work was to develop a convenient method for radial/circumferential strain imaging and shear rate estimation that could be used as a supplement to the current routine screening for carotid atherosclerosis using video images of diagnostic ultrasound. A reflection model-based correction for gray-scale non-uniform distribution was applied to B-mode video images before strain estimation to improve the accuracy of radial/circumferential strain imaging when applied to vessel transverse cross sections. The incremental and cumulative radial/circumferential strain images can then be calculated based on the displacement field between consecutive B-mode images. Finally, the transverse Doppler spectra acquired at different depths along the vessel diameter were used to construct the spatially matched instantaneous wall shear values in a cardiac cycle. Vessel phantom simulation results revealed that the signal-to-noise ratio and contrast-to-noise ratio of the radial and circumferential strain images were increased by 2.8 and 5.9 dB and by 2.3 and 4.4 dB, respectively, after non-uniform correction. Preliminary results for 17 patients indicated that the accuracy of radial/circumferential strain images was improved in the lateral direction after non-uniform correction. The peak-to-peak value of incremental strain and the maximum cumulative strain for calcified plaques are evidently lower than those for other plaque types, and the echolucent plaques had higher values, on average, than the mixed plaques. Moreover, low oscillating wall shear rate values, found near the plaque and stenosis regions, are closely related to plaque formation. In conclusion, the method described can provide additional valuable results as a supplement to the current routine ultrasound examination for carotid atherosclerosis and, therefore, has significant potential as a feasible screening method for atherosclerosis diagnosis in the future.     

Patients with Unstable Atherosclerosis Have More Echolucent Carotid Plaques Compared with Stable Atherosclerotic Patients: A 3-D Ultrasound Study

Using a novel 3-D ultrasound system, we aimed to determine differences in carotid plaque size and echogenicity in two atherosclerotic groups. Seventy patients admitted with acute myocardial infarction (aMI) and 69 patients known with chronic peripheral arterial disease (cPAD) were included. The cPAD group had larger plaque volumes (median: 70.24 mm³, interquartile range [40.12–135.61] vs. 55.41 mm³ [4.24–84.31], p = 0.004), thicker plaques (2.45 mm [1.85–3.25] vs. 1.99 mm [1.55 – 2.64], p = 0.005) and higher gray-scale medians (GSMs) (mean: 71.75, standard deviation: 21.55 vs. 60.99 [24.09], p = 0.006) than the aMI group. After adjustment for traditional risk factors, the difference persisted for thickness and volume. The difference in GSM persisted after adjustment for volume only. Patients with stable atherosclerotic disease had larger and brighter carotid plaques compared with unstable atherosclerotic patients. 3-D ultrasound may prove useful in identifying thromboembolic risk.     

Mapping spatial and temporal changes in carotid atherosclerosis from three-dimensional ultrasound images

This study was designed to evaluate changes in carotid atherosclerosis using plaque and wall thickness maps derived from three-dimensional ultrasound (3DUS) images. Five subjects with carotid stenosis were scanned at baseline and 3 mo as part of a placebo-controlled intensive statin treatment study and three subjects with moderate atherosclerosis were scanned at baseline and again within 14 +/- 2 d. 3DUS-derived vessel wall volume (VWV) was measured using manual segmentation to provide segmentation contours that were used to generate scan and rescan carotid atherosclerosis thickness maps and thickness difference maps. There was no significant difference in VWV between scan and rescan for the three subjects scanned twice in 2 wk or the single subject treated with placebo. There was a significant difference between scan and rescan VWV for carotid stenosis subjects treated with atorvastatin (p < 0.001). Carotid atherosclerosis thickness difference maps showed visual qualitative evidence of thickness changes in vessel wall and plaque thickness in the common carotid artery for all statin-treated subjects and no change in a placebo-treated subject and subjects scanned twice in 2 wk. Carotid atherosclerosis thickness difference maps generated from 3DUS images provide evidence of vessel wall and plaque thickness changes for all subjects assessed.     

Echo-Computed Tomography Strain Imaging of Healthy and Diseased Carotid Specimens

To improve our understanding of the mechanical behavior of human atherosclerotic plaque tissue, fully 3-D geometrical, morphological and dynamical information is essential. For this purpose, four-dimensional (3-D+t) strain imaging using an ultrasound tomography approach (echo-computed tomography) was performed in carotid arteries in vitro. The method was applied to a carotid phantom (CPh), a porcine carotid artery (PC) and human carotid atherosclerotic plaque samples (HC, n = 5). Each sample was subjected to an intraluminal pressure, after which 2-D longitudinal ultrasound images were obtained for 36 angles along the circumferential direction. Local deformations were estimated using a 2-D strain algorithm, and 3-D radial strain data were reconstructed. At systole, median luminal strains of 15% (CPh) and 18% (PC) were found, which is in agreement with the stiffness of the material and applied pressure pulse. The elastographic signal-to-noise ratio was consistent in all directions and ranged from 16 to 36 dB. Furthermore, realistic but more complex strain patterns were found for the HC, with 99th percentile systolic strain values ranging from 0.1% to 18%.     

Comparison of grey scale median (GSM) measurement in ultrasound images of human carotid plaques using two different softwares

Grey scale median (GSM) measured on ultrasound images of carotid plaques has been used for several years now in research to find the vulnerable plaque. Centres have used different software and also different methods for GSM measurement. This has resulted in a wide range of GSM values and cut‐off values for the detection of the vulnerable plaque. The aim of this study was to compare the values obtained with two different softwares, using different standardization methods, for the measurement of GSM on ultrasound images of carotid human plaques. GSM was measured with Adobe Photoshop^® and with Artery Measurement System (AMS) on duplex ultrasound images of 100 consecutive medium‐ to large‐sized carotid plaques of the Beta‐blocker Cholesterol‐lowering Asymptomatic Plaque Study (BCAPS). The mean values of GSM were 35·2 ± 19·3 and 55·8 ± 22·5 for Adobe Photoshop^® and AMS, respectively. Mean difference was 20·45 (95% CI: 19·17–21·73). Although the absolute values of GSM differed, the agreement between the two measurements was good, correlation coefficient 0·95. A chi‐square test revealed a kappa value of 0·68 when studying quartiles of GSM. The intra‐observer variability was 1·9% for AMS and 2·5% for Adobe Photoshop. The difference between softwares and standardization methods must be taken into consideration when comparing studies. To avoid these problems, researcher should come to a consensus regarding software and standardization method for GSM measurement on ultrasound images of plaque in the arteries.     

User-Guided Segmentation of Preterm Neonate Ventricular System from 3-D Ultrasound Images Using Convex Optimization

A three-dimensional (3-D) ultrasound (US) system has been developed to monitor the intracranial ventricular system of preterm neonates with intraventricular hemorrhage (IVH) and the resultant dilation of the ventricles (ventriculomegaly). To measure ventricular volume from 3-D US images, a semi-automatic convex optimization-based approach is proposed for segmentation of the cerebral ventricular system in preterm neonates with IVH from 3-D US images. The proposed semi-automatic segmentation method makes use of the convex optimization technique supervised by user-initialized information. Experiments using 58 patient 3-D US images reveal that our proposed approach yielded a mean Dice similarity coefficient of 78.2% compared with the surfaces that were manually contoured, suggesting good agreement between these two segmentations. Additional metrics, the mean absolute distance of 0.65 mm and the maximum absolute distance of 3.2 mm, indicated small distance errors for a voxel spacing of 0.22 × 0.22 × 0.22 mm³. The Pearson correlation coefficient (r = 0.97, p < 0.001) indicated a significant correlation of algorithm-generated ventricular system volume (VSV) with the manually generated VSV. The calculated minimal detectable difference in ventricular volume change indicated that the proposed segmentation approach with 3-D US images is capable of detecting a VSV difference of 6.5 cm³ with 95% confidence, suggesting that this approach might be used for monitoring IVH patients' ventricular changes using 3-D US imaging. The mean segmentation times of the graphics processing unit (GPU)- and central processing unit-implemented algorithms were 50 ± 2 and 205 ± 5 s for one 3-D US image, respectively, in addition to 120 ± 10 s for initialization, less than the approximately 35 min required by manual segmentation. In addition, repeatability experiments indicated that the intra-observer variability ranges from 6.5% to 7.5%, and the inter-observer variability is 8.5% in terms of the coefficient of variation of the Dice similarity coefficient. The intra-class correlation coefficient for ventricular system volume measurements for each independent observer ranged from 0.988 to 0.996 and was 0.945 for three different observers. The coefficient of variation and intra-class correlation coefficient revealed that the intra- and inter-observer variability of the proposed approach introduced by the user initialization was small, indicating good reproducibility, independent of different users.     

Enhanced characterization of calcified areas in intravascular ultrasound virtual histology images by quantification of the acoustic shadow: validation against computed tomography coronary angiography

We enhance intravascular ultrasound virtual histology (VH) tissue characterization by fully automatic quantification of the acoustic shadow behind calcified plaque. VH is unable to characterize atherosclerosis located behind calcifications. In this study, the quantified acoustic shadows are considered calcified to approximate the real dense calcium (DC) plaque volume. In total, 57 patients with 108 coronary lesions were included. A novel post-processing step is applied on the VH images to quantify the acoustic shadow and enhance the VH results. The VH and enhanced VH results are compared to quantitative computed tomography angiography (QTA) plaque characterization as reference standard. The correlation of the plaque types between enhanced VH and QTA differs significantly from the correlation with unenhanced VH. For DC, the correlation improved from 0.733 to 0.818. Instead of an underestimation of DC in VH with a bias of 8.5 mm³, there was a smaller overestimation of 1.1 mm³ in the enhanced VH. Although tissue characterization within the acoustic shadow in VH is difficult, the novel algorithm improved the DC tissue characterization. This algorithm contributes to accurate assessment of calcium on VH and could be applied in clinical studies.     

Three-Dimensional Cardiac Strain Imaging in Healthy Children Using RF-Data

In this study, a new radio-frequency (RF)-based, three-dimensional (3-D) strain imaging technique is introduced and applied to 3-D full volume ultrasound data of the heart of healthy children. Continuing advances in performance of transducers for 3-D ultrasound imaging have boosted research on 3-D strain imaging. In general, speckle tracking techniques are used for strain imaging. RF-based strain imaging has the potential to yield better performance than speckle- based methods because of the availability of phase information but such a system output is commercially not available. Furthermore, the relatively low frame rate of 3-D ultrasound data has limited broad application of RF-based cardiac strain imaging. In this study, the previously reported two-dimensional (2-D) strain methodology was extended to the third dimension. Three-dimensional RF-data were acquired in 13 healthy children, in the age range of 6-15 years, at a relatively low frame rate of 38-51 Hz. A 3-D, free-shape, coarse-to-fine displacement and strain estimation algorithm was applied to the RF-data. The heart was segmented using 3-D ellipsoid fitting. Strain was estimated in the radial (R), circumferential (C) and longitudinal directions (L). Our preliminary results reveal the applicability of the 3-D strain estimation technique on full volume 3-D RF-data. The technique enabled 3-D strain imaging of all three strain components. The average strains for all children were in the lateral wall R = 37 ± 10% (infero-lateral) and R = 32% ± 10% (antero-lateral), C = −9% ± 4% (antero-lateral) and C = −9% ± 4% (infero-lateral), L = −18% ± 6 % (antero-lateral) and L = −15% ± 4% (infero-lateral). In the septum, strains were found to be R = 24% ± 10% (antero-septal) and R = 13% ± 5% (infero-septal), C = −13% ± 5% (antero-septal) and −13% ± 5% (infero-septal) and L = −13% ± 3% (antero-septal) and L = −16% ± 5% (infero-septal). Strain in the anterior and inferior walls seemed underestimated, probably caused by the low (in-plane) resolution and poor image quality. The field-of-view as well as image quality were not always sufficient to image the entire left ventricle. It is concluded that 3-D strain imaging using RF-data is feasible, but validation with other modalities and with conventional 3-D speckle tracking techniques will be necessary.     

Validation of 3D ultrasound vessel wall volume: an imaging phenotype of carotid atherosclerosis

Carotid atherosclerotic lesions are a major cause of stroke and the identification and quantification of such lesions in patients is important for the development of a better understanding of atherogenesis in high risk populations and for the design of studies to assess treatment efficacy. Our objective was to develop and validate a new three-dimensional ultrasound (3DUS) measurement or phenotype of carotid atherosclerosis, vessel wall volume (VWV), which is a three-dimensional measurement of vessel wall thickness and plaque within the carotid arteries measured in 3DUS images. To assess both intraobserver and interscan variability, 3DUS images were acquired from the right and left carotid arteries of ten subjects with carotid atherosclerosis scanned twice within a period of 2 wk. For both VWV and total plaque volume (TPV), an expert observer performed five measurement trials of all images acquired at baseline scan and 2-wk rescan with a 5-d period between measurement trials for images. Images were re-randomized for each measurement trial and both TPV and VWV were measured by observers who were blinded to subject identification for each time-point measurement. Coefficients of variation (COV) and intraclass correlation coefficients (ICC), for VWV measurements indicated higher intraobserver (scan COV = 4.6% ICC = 0.95, rescan COV = 3.4%, ICC = 0.96) and interscan reproducibility (COV = 5.7%, ICC = 0.85) than TPV measurements (intraobserver variability scan COV = 22.7% ICC = 0.85, rescan COV = 21.1% ICC = 0.88 and interscan variability, COV = 31.1%, ICC = 0.83), although absolute variances for both phenotypes were very similar (VWV = 90 mm3, TPV = 80 mm3).