Estimating central blood pressure from aortic flow: development and assessment of algorithms.

Central blood pressure (cBP) is a highly prognostic cardiovascular (CV) risk factor whose accurate, invasive assessment is costly and carries risks to patients. We developed and assessed novel algorithms for estimating cBP from noninvasive aortic hemodynamic data and a peripheral blood pressure measurement. These algorithms were created using three blood flow models: the two- and three-element Windkessel (0-D) models and a one-dimensional (1-D) model of the thoracic aorta. We tested new and existing methods for estimating CV parameters (left ventricular ejection time, outflow BP, arterial resistance and compliance, pulse wave velocity, and characteristic impedance) required for the cBP algorithms, using virtual (simulated) subjects (n = 19,646) for which reference CV parameters were known exactly. We then tested the cBP algorithms using virtual subjects (n = 4,064), for which reference cBP were available free of measurement error, and clinical datasets containing invasive (n = 10) and noninvasive (n = 171) reference cBP waves across a wide range of CV conditions. The 1-D algorithm outperformed the 0-D algorithms when the aortic vascular geometry was available, achieving central systolic blood pressure (cSBP) errors ≤ 2.1 ± 9.7 mmHg and root-mean-square errors (RMSEs) ≤ 6.4 ± 2.8 mmHg against invasive reference cBP waves (n = 10). When the aortic geometry was unavailable, the three-element 0-D algorithm achieved cSBP errors ≤ 6.0 ± 4.7 mmHg and RMSEs ≤ 5.9 ± 2.4 mmHg against noninvasive reference cBP waves (n = 171), outperforming the two-element 0-D algorithm. All CV parameters were estimated with mean percentage errors ≤ 8.2%, except for the aortic characteristic impedance (≤13.4%), which affected the three-element 0-D algorithm's performance. The freely available algorithms developed in this work enable fast and accurate calculation of the cBP wave and CV parameters in datasets containing noninvasive ultrasound or magnetic resonance imaging data.NEW & NOTEWORTHY First, our proposed methods for CV parameter estimation and a comprehensive set of methods from the literature were tested using in silico and clinical datasets. Second, optimized algorithms for estimating cBP from aortic flow were developed and tested for a wide range of cBP morphologies, including catheter cBP data. Third, a dataset of simulated cBP waves was created using a three-element Windkessel model. Fourth, the Windkessel model dataset and optimized algorithms are freely available.

Comparison of CT and CMR for detection and quantification of carotid artery calcification: the Rotterdam Study.

BACKGROUND: Carotid artery atherosclerosis is an important risk factor for stroke. As such, quantitative imaging of carotid artery calcification, as a proxy of atherosclerosis, has become a cornerstone of current stroke research. Yet, population-based data comparing the computed tomography (CT) and cardiovascular magnetic resonance (CMR) for the detection and quantification of calcification remain scarce. METHODS: A total of 684 participants from the population-based Rotterdam Study underwent both a CT and CMR of the carotid artery bifurcation to quantify the amount of carotid artery calcification (mean interscan interval: 4.9 ± 1.2 years). We investigated the correlation between the amount of calcification measured on CT and CMR using Spearman's correlation coefficient, Bland-Altman plots, and linear regression. In addition, using logistic regression modeling, we assessed the association of CT and CMR based calcification volumes with a history of stroke. RESULTS: We found a strong correlation between CT and CMR based calcification volumes (Spearman's correlation coefficient:0.86, p-value ≤0.01). Bland-Altman analyses showed a good agreement, though CT based calcification volumes were systematically larger. Finally, calcification volume assessed with either imaging modality was associated with a history of stroke with similar effect estimates (odds ratio (OR) per 1-SD increase in calcification volume: 1.52 (95% CI:1.00;2.30) for CT, and 1.47 (95% CI:1.01;2.14) for CMR. CONCLUSION: CT based and CMR based volumes of carotid artery calcification are highly correlated, but CMR based calcification is systematically smaller than those obtained with CT. Despite this difference, both provide comparable information with regard to a history of stroke.

Aortic length measurements for pulse wave velocity calculation: manual 2D vs automated 3D centreline extraction.

BACKGROUND: Pulse wave velocity (PWV) is a biomarker for the intrinsic stiffness of the aortic wall, and has been shown to be predictive for cardiovascular events. It can be assessed using cardiovascular magnetic resonance (CMR) from the delay between phase-contrast flow waveforms at two or more locations in the aorta, and the distance on CMR images between those locations. This study aimed to investigate the impact of different distance measurement methods on PWV. We present and evaluate an algorithm for automated centreline tracking in 3D images, and compare PWV calculations using distances derived from 3D images to those obtained from a conventional 2D oblique-sagittal image of the aorta. METHODS: We included 35 patients from a twin cohort, and 20 post-coarctation repair patients. Phase-contrast flow was acquired in the ascending, descending and diaphragmatic aorta. A 3D centreline tracking algorithm is presented and evaluated on a subset of 30 subjects, on three CMR sequences: balanced steady-state free precession (SSFP), black-blood double inversion recovery turbo spin echo, and contrast-enhanced CMR angiography. Aortic lengths are subsequently compared between measurements from a 2D oblique-sagittal plane, and a 3D geometry. RESULTS: The error in length of automated 3D centreline tracking compared with manual annotations ranged from 2.4 [1.8-4.3] mm (mean [IQR], black-blood) to 6.4 [4.7-8.9] mm (SSFP). The impact on PWV was below 0.5m/s (<5%). Differences between 2D and 3D centreline length were significant for the majority of our experiments (p < 0.05). Individual differences in PWV were larger than 0.5m/s in 15% of all cases (thoracic aorta) and 37% when studying the aortic arch only. Finally, the difference between end-diastolic and end-systolic 2D centreline lengths was statistically significant (p < 0.01), but resulted in small differences in PWV (0.08 [0.04 - 0.10]m/s). CONCLUSIONS: Automatic aortic centreline tracking in three commonly used CMR sequences is possible with good accuracy. The 3D length obtained from such sequences can differ considerably from lengths obtained from a 2D oblique-sagittal plane, depending on aortic curvature, adequate planning of the oblique-sagittal plane, and patient motion between acquisitions. For accurate PWV measurements we recommend using 3D centrelines.

Three-dimensional carotid ultrasound plaque texture predicts vascular events.

BACKGROUND AND PURPOSE: Carotid ultrasound atherosclerosis measurements, including those of the arterial wall and plaque, provide a way to monitor patients at risk of vascular events. Our objective was to examine carotid ultrasound plaque texture measurements and the change in carotid plaque texture during 1 year in patients at risk of events and to compare these with measurements of plaque volume and other risk factors as predictors of vascular events. METHODS: We evaluated 298 patients with carotid atherosclerosis using 3-dimensional (3D) ultrasound at baseline and after 1 year and measured carotid plaque volume and 376 measures of plaque texture. Patients were followed up to 5 years (median [range], 3.12 [0.77-4.66]) for myocardial infarction, transient ischemic attack, and stroke. Sparse Cox regression was used to select the most predictive plaque texture measurements in independent training sets using a 10-fold cross-validation, repeated 5×, to ensure unbiased results. RESULTS: Receiver operator curves and Kaplan-Meier analysis showed that changes in texture and total plaque volume combined provided the best predictor of vascular events. In multivariate Cox regression, changes in plaque texture (median hazard ratio, 1.4; P<0.001) and total plaque volume (median hazard ratio, 1.5 per 100 mm(3); P<0.001) were both significant predictors, whereas the Framingham risk score was not. CONCLUSIONS: Changes in both plaque texture and volume are strongly predictive of vascular events. In high-risk patients, 3D ultrasound plaque measurements should be considered for vascular event risk prediction.

A reliable, non-invasive measurement tool for anisotropy in normal skin and scar tissue.

BACKGROUND: Anisotropy of the skin varies depending on different locations and pathological conditions. Currently, no reliable non-invasive measurement tool is available for tissue anisotropy. The Reviscometer is an anisotropy measurement tool that measures the resonance running time (RRT) of a shock wave. This study was initiated to establish the reliability of the Reviscometer on normal skin and scars, and to provide basic information on tissue alignment in normal skin and scars. METHODS: Fifty volunteers and 50 patients underwent measurements on normal skin and scars, respectively. All measurements were performed by the same two observers. Measurements on normal skin were performed on the forearm, upper arm, and abdomen. RESULTS: The results showed that the intraclass correlation coefficient of the inter-observer reliability was > or =0.79 on normal skin and > or =0.86 on scars. In normal skin, the highest mean RRT was found on the abdomen (156.4+/-48.8), followed by the upper arm (123.2+/-33.6) and the forearm (112.5+/-24.3). A significantly lower mean RRT was found in scars (52.3+/-21.9) compared with normal skin (91.6+/-37.7). CONCLUSION: Reviscometer measurements were reliable for normal skin and scars. In addition, clear differences between scars and normal skin but also within different locations on normal skin were identified. The Reviscometer can be considered for the evaluation of the efficacy of different treatments.

Longitudinal assessment of carotid plaque texture in three-dimensional ultrasound images based on semi-supervised graph-based dimensionality reduction and feature selection.

With continuous development of therapeutic options for atherosclerosis, image-based biomarkers sensitive to the effect of new interventions are required to be developed for cost-effective clinical evaluation. Although 3D ultrasound measurement of total plaque volume (TPV) showed the efficacy of high-dose statin, more sensitive biomarkers are needed to establish the efficacy of dietary supplements expected to confer a smaller beneficial effect. This study involved 171 subjects who participated in a one-year placebo-controlled trial evaluating the effect of pomegranate. A framework involving a feature selection technique known as discriminative feature selection (DFS) and a semi-supervised graph-based regression (SSGBR) technique was proposed for sensitive detection of plaque textural changes over the trial. 376 textual features of plaques were extracted from 3D ultrasound images acquired at baseline and a follow-up session. A scalar biomarker for each subject were generated by SSGBR based on prominent textural features selected by DFS. The ability of this biomarker for discriminating pomegranate from placebo subjects was quantified by the p-values obtained in Mann-Whitney U test. The discriminative power of SSGBR was compared with global and local dimensionality reduction techniques, including linear discriminant analysis (LDA), maximum margin criterion (MMC) and Laplacian Eigenmap (LE). Only SSGBR (p=4.12×10-6) and normalized LE (p=0.002) detected a difference between the two groups at the 5% significance level. As compared with ΔTPV, SSGBR reduced the sample size required to establish a significant difference by a factor of 60. The application of this framework will substantially reduce the cost incurred in clinical trials.

Three-dimensional ultrasound evaluation of the effects of pomegranate therapy on carotid plaque texture using locality preserving projection.

BACKGROUND AND OBJECTIVE: Dietary supplements are expected to confer a smaller beneficial effect than medical treatments. Therefore, there is a need to develop cost-effective biomarkers that can demonstrate the efficacy of such supplements for carotid atherosclerosis. The aim of this study is to develop such a biomarker based on the changes of 376 plaque textural features measured from 3D ultrasound images. METHODS: Since the number of features (376) was greater than the number of subjects (171) in this study, principal component analysis (PCA) was applied to reduce the dimensionality of feature vectors. To generate a scalar biomarker for each subject, elements in the reduced feature vectors produced by PCA were weighted using locality preserving projections (LPP) to capture essential patterns exhibited locally in the feature space. 96 subjects treated by pomegranate juice and tablets, and 75 subjects receiving placebo-matching juice and tablets were evaluated in this study. The discriminative power of the proposed biomarker was evaluated and compared with existing biomarkers using t-tests. As the cost of a clinical trial is directly related to the number of subjects enrolled, the cost-effectiveness of the proposed biomarker was evaluated by sample size estimation. RESULTS: The proposed biomarker was more able to discriminate plaque changes exhibited by the pomegranate and placebo groups than total plaque volume (TPV) according to the result of t-tests (TPV: p=0.34, Proposed biomarker: p=1.5×10-5). The sample size required by the new biomarker to detect a significant effect was 20 times smaller than that required by TPV. CONCLUSION: With the increase in cost-effectiveness afforded by the proposed biomarker, more proof-of-principle studies for novel treatment options could be performed.

Pediatric heterozygous familial hypercholesterolemia patients have locally increased aortic pulse wave velocity and wall thickness at the aortic root.

Familial hypercholesterolemia (FH) is an autosomal dominant disorder that affects 1 in 250 people. Aortic stiffness, measured by pulse wave velocity (PWV), is an independent predictor for cardiovascular events. Young FH patients are a unique group with early vessel wall disease that may serve to elucidate the determinants of aortic stiffness. We hypothesized that young FH patients would have early changes in aortic stiffness compared to healthy, age- and sex-matched reference values. Thirty-three FH patients ( ≥ 7 years age; mean age 14.6 ± 3.3 years; 26/33 on statin therapy) underwent cardiac MRI. PWV was determined using propagation of flow waveform from aortic arch phase contrast images. Distensibility and aortic wall thickness (AWT) were measured at the ascending, proximal descending, and diaphragmatic aorta. Ventricular volumes and left ventricular (LV) myocardial mass were measured from 2D cine images. These parameters were compared to age- and sex-matched reference values. FH patients had significantly higher PWV (4.5 ± 0.8 vs. 3.5 ± 0.3 m/s; p < 0.001), aortic distensibility, and ascending aortic wall thickness (1.37 ± 0.18 vs. 1.30 ± 0.02 mm; p < 0.05) compared to reference. There was no difference in aortic area or descending aortic wall thickness between groups. Young FH patients had aortic changes with increased aortic pulse wave velocity in the setting of increased aortic distensibility, accompanied by increased thickness of the ascending aortic wall. Presence of these early findings in young patients despite the majority being on statin therapy support enhanced screening and aggressive treatment of familial hypercholesterolemia to prevent potential future cardiovascular events.

Maximization of regional probabilities using Optimal Surface Graphs: Application to carotid artery segmentation in MRI.

PURPOSE: We present a segmentation method that maximizes regional probabilities enclosed by coupled surfaces using an Optimal Surface Graph (OSG) cut approach. This OSG cut determines the globally optimal solution given a graph constructed around an initial surface. While most methods for vessel wall segmentation only use edge information, we show that maximizing regional probabilities using an OSG improves the segmentation results. We applied this to automatically segment the vessel wall of the carotid artery in magnetic resonance images. METHODS: First, voxel-wise regional probability maps were obtained using a Support Vector Machine classifier trained on local image features. Then, the OSG segments the regions which maximizes the regional probabilities considering smoothness and topological constraints. RESULTS: The method was evaluated on 49 carotid arteries from 30 subjects. The proposed method shows good accuracy with a Dice wall overlap of 74.1 ± 4.3%, and significantly outperforms a published method based on an OSG using only surface information, the obtained segmentations using voxel-wise classification alone, and another published artery wall segmentation method based on a deformable surface model. Intraclass correlations (ICC) with manually measured lumen and wall volumes were similar to those obtained between observers. Finally, we show a good reproducibility of the method with ICC = 0.86 between the volumes measured in scans repeated within a short time interval. CONCLUSIONS: In this work, a new segmentation method that uses both an OSG and regional probabilities is presented. The method shows good segmentations of the carotid artery in MRI and outperformed another segmentation method that uses OSG and edge information and the voxel-wise segmentation using the probability maps.

Carotid Artery Wall Segmentation in Multispectral MRI by Coupled Optimal Surface Graph Cuts.

We present a new three-dimensional coupled optimal surface graph-cut algorithm to segment the wall of the carotid artery bifurcation from Magnetic Resonance (MR) images. The method combines the search for both inner and outer borders into a single graph cut and uses cost functions that integrate information from multiple sequences. Our approach requires manual localization of only three seed points indicating the start and end points of the segmentation in the internal, external, and common carotid artery. We performed a quantitative validation using images of 57 carotid arteries. Dice overlap of 0.86 ± 0.06 for the complete vessel and 0.89 ± 0.05 for the lumen compared to manual annotation were obtained. Reproducibility tests were performed in 60 scans acquired with an interval of 15 ± 9 days, showing good agreement between baseline and follow-up segmentations with intraclass correlations of 0.96 and 0.74 for the lumen and complete vessel volumes respectively.

Multi-Center MRI Carotid Plaque Component Segmentation Using Feature Normalization and Transfer Learning.

Automated segmentation of plaque components in carotid artery magnetic resonance imaging (MRI) is important to enable large studies on plaque vulnerability, and for incorporating plaque composition as an imaging biomarker in clinical practice. Especially supervised classification techniques, which learn from labeled examples, have shown good performance. However, a disadvantage of supervised methods is their reduced performance on data different from the training data, for example on images acquired with different scanners. Reducing the amount of manual annotations required for each new dataset will facilitate widespread implementation of supervised methods. In this paper we segment carotid plaque components of clinical interest (fibrous tissue, lipid tissue, calcification and intraplaque hemorrhage) in a multi-center MRI study. We perform voxelwise tissue classification by traditional same-center training, and compare results with two approaches that use little or no annotated same-center data. These approaches additionally use an annotated set of different-center data. We evaluate 1) a nonlinear feature normalization approach, and 2) two transfer-learning algorithms that use same and different-center data with different weights. Results showed that the best results were obtained for a combination of feature normalization and transfer learning. While for the other approaches significant differences in voxelwise or mean volume errors were found compared with the reference same-center training, the proposed approach did not yield significant differences from that reference. We conclude that both extensive feature normalization and transfer learning can be valuable for the development of supervised methods that perform well on different types of datasets.

Atherosclerotic plaque component segmentation in combined carotid MRI and CTA data incorporating class label uncertainty.

Atherosclerotic plaque composition can indicate plaque vulnerability. We segment atherosclerotic plaque components from the carotid artery on a combination of in vivo MRI and CT-angiography (CTA) data using supervised voxelwise classification. In contrast to previous studies the ground truth for training is directly obtained from 3D registration with histology for fibrous and lipid-rich necrotic tissue, and with µCT for calcification. This registration does, however, not provide accurate voxelwise correspondence. We therefore evaluate three approaches that incorporate uncertainty in the ground truth used for training: I) soft labels are created by Gaussian blurring of the original binary histology segmentations to reduce weights at the boundaries between components, and are weighted by the estimated registration accuracy of the histology and in vivo imaging data (measured by overlap), II) samples are weighted by the local contour distance of the lumen and outer wall between histology and in vivo data, and III) 10% of each class is rejected by Gaussian outlier rejection. Classification was evaluated on the relative volumes (% of tissue type in the vessel wall) for calcified, fibrous and lipid-rich necrotic tissue, using linear discriminant (LDC) and support vector machine (SVM) classification. In addition, the combination of MRI and CTA data was compared to using only one imaging modality. Best results were obtained by LDC and outlier rejection: the volume error per vessel was 0.9±1.0% for calcification, 12.7±7.6% for fibrous and 12.1±8.1% for necrotic tissue, with Spearman rank correlation coefficients of 0.91 (calcification), 0.80 (fibrous) and 0.81 (necrotic). While segmentation using only MRI features yielded low accuracy for calcification, and segmentation using only CTA features yielded low accuracy for necrotic tissue, the combination of features from MRI and CTA gave good results for all studied components.

Automated segmentation of atherosclerotic histology based on pattern classification.

BACKGROUND: Histology sections provide accurate information on atherosclerotic plaque composition, and are used in various applications. To our knowledge, no automated systems for plaque component segmentation in histology sections currently exist. MATERIALS AND METHODS: We perform pixel-wise classification of fibrous, lipid, and necrotic tissue in Elastica Von Gieson-stained histology sections, using features based on color channel intensity and local image texture and structure. We compare an approach where we train on independent data to an approach where we train on one or two sections per specimen in order to segment the remaining sections. We evaluate the results on segmentation accuracy in histology, and we use the obtained histology segmentations to train plaque component classification methods in ex vivo Magnetic resonance imaging (MRI) and in vivo MRI and computed tomography (CT). RESULTS: In leave-one-specimen-out experiments on 176 histology slices of 13 plaques, a pixel-wise accuracy of 75.7 ± 6.8% was obtained. This increased to 77.6 ± 6.5% when two manually annotated slices of the specimen to be segmented were used for training. Rank correlations of relative component volumes with manually annotated volumes were high in this situation (P = 0.82-0.98). Using the obtained histology segmentations to train plaque component classification methods in ex vivo MRI and in vivo MRI and CT resulted in similar image segmentations for training on the automated histology segmentations as for training on a fully manual ground truth. The size of the lipid-rich necrotic core was significantly smaller when training on fully automated histology segmentations than when manually annotated histology sections were used. This difference was reduced and not statistically significant when one or two slices per section were manually annotated for histology segmentation. CONCLUSIONS: Good histology segmentations can be obtained by automated segmentation, which show good correlations with ground truth volumes. In addition, these can be used to develop segmentation methods in other imaging modalities. Accuracy increases when one or two sections of the same specimen are used for training, which requires a limited amount of user interaction in practice.

Multi-feature-based plaque characterization in ex vivo MRI trained by registration to 3D histology.

We present a new method for automated characterization of atherosclerotic plaque composition in ex vivo MRI. It uses MRI intensities as well as four other types of features: smoothed, gradient magnitude and Laplacian images at several scales, and the distances to the lumen and outer vessel wall. The ground truth for fibrous, necrotic and calcified tissue was provided by histology and µCT in 12 carotid plaque specimens. Semi-automatic registration of a 3D stack of histological slices and µCT images to MRI allowed for 3D rotations and in-plane deformations of histology. By basing voxelwise classification on different combinations of features, we evaluated their relative importance. To establish whether training by 3D registration yields different results than training by 2D registration, we determined plaque composition using (1) a 2D slice-based registration approach for three manually selected MRI and histology slices per specimen, and (2) an approach that uses only the three corresponding MRI slices from the 3D-registered volumes. Voxelwise classification accuracy was best when all features were used (73.3 ± 6.3%) and was significantly better than when only original intensities and distance features were used (Friedman, p < 0.05). Although 2D registration or selection of three slices from the 3D set slightly decreased accuracy, these differences were non-significant.