Segmentation of intravascular ultrasound images: a knowledge-based approach

Intravascular ultrasound imaging of coronary arteries provides important information about coronary lumen, wall, and plaque characteristics. Quantitative studies of coronary atherosclerosis using intravascular ultrasound and manual identification of wall and plaque borders are limited by the need for observers with substantial experience and the tedious nature of manual border detection. We have developed a method for segmentation of intravascular ultrasound images that identifies the internal and external elastic laminae and the plaque-lumen interface. The border detection algorithm was evaluated in a set of 38 intravascular ultrasound images acquired from fresh cadaveric hearts using a 30 MHz imaging catheter. To assess the performance of our border detection method we compared five quantitative measures of arterial anatomy derived from computer-detected borders with measures derived from borders manually defined by expert observers. Computer-detected and observer-defined lumen areas correlated very well (r=0.96, y=1.02x+0.52), as did plaque areas (r=0.95, y=1.07x-0.48), and percent area stenosis (r=0.93, y=0.99x-1.34.) Computer-derived segmental plaque thickness measurements were highly accurate. Our knowledge-based intravascular ultrasound segmentation method shows substantial promise for the quantitative analysis of in vivo intravascular ultrasound image data.     

Robust simultaneous detection of coronary borders in complex images

Visual estimation of coronary obstruction severity from angiograms suffers from poor inter- and intraobserver reproducibility and is often inaccurate. In spite of the widely recognized limitations of visual analysis, automated methods have not found widespread clinical use, in part because they too frequently fail to accurately identify vessel borders. The authors have developed a robust method for simultaneous detection of left and right coronary borders that is suitable for analysis of complex images with poor contrast, nearby or overlapping structures, or branching vessels. The reliability of the simultaneous border detection method and that of the authors' previously reported conventional border detection method were tested in 130 complex images, selected because conventional automated border detection might be expected to fail. Conventional analysis failed to yield acceptable borders in 65/130 or 50% of images. Simultaneous border detection was much more robust (p<.001) and failed in only 15/130 or 12% of complex images. Simultaneous border detection identified stenosis diameters that correlated significantly better with observer-derived stenosis diameters than did diameters obtained with conventional border detection (p<0.001), Simultaneous detection of left and right coronary borders is highly robust and has substantial promise for enhancing the utility of quantitative coronary angiography in the clinical setting.     

Simultaneous detection of both coronary borders

A method for simultaneous detection of both coronary borders that is based on three-dimensional graph searching principles is presented. The simultaneous method and the authors' previously reported conventional method were applied to 29 coronary images, of which 19 were selected because conventional methods might be expected to have difficulty. Coronary borders identified by the two methods were visually compared. In the 19 difficult images, simultaneous border detection yielded superior results in 7 images and equivalent results in 12 images. Superior or equivalent results were obtained in the remaining 10 typical images. In a set of 43 uncomplicated images, minimal lumen diameters derived using simultaneous border detection correlated well with diameters derived using conventional border detection (r=0.97), diameters obtained from observer-defined borders (r=0.91), and diameters obtained using the Brown-Dodge quantitative coronary arteriography method (r=0.85). Thus simultaneous detection of left and right coronary borders provides improved accuracy in the detection of vessel borders in difficult coronary angiograms.     

Adaptive approach to accurate analysis of small-diameter vessels incineangiograms

In coronary vessels smaller than 1 mm in diameter, it is difficult to accurately identify lumen borders using existing border detection techniques. Computer-detected diameters of small coronary vessels are often severely overestimated due to the influence of the imaging system point spread function and the use of an edge operator designed for a broad range of diameter vessel sizes. Computer-detected diameters may be corrected if a calibration curve for the X-ray system is available. Unfortunately, the performance of this postprocessing diameter correction approach is severely limited by the presence of image noise. The authors report here a new approach that uses a two-stage adaption of edge operator parameters to optimally match the edge operator to the local lumen diameter. In the first stage, approximate lumen diameters are detected using a single edge operator in a half-resolution image. Depending on the approximate lumen size, one of three edge operators is selected for the second full-resolution stage in which left and right coronary borders are simultaneously identified. The method was tested in a set of 72 segments of nine angiographic phantom vessels with diameters ranging from 0.46 to 4.14 mm and in 82 clinical coronary angiograms. Performance of the adaptive simultaneous border detection method was compared to that of a conventional border detection method and to that of a postprocessing diameter correction border detection method. Adaptive border detection yielded significantly improved accuracy in small phantom vessels and across all vessel sizes in comparison to the conventional and postprocessing diameter correction methods (p<0.001 in all cases). Adaptive simultaneous coronary border detection provides both accurate and robust quantitative analysis of coronary vessels of all sizes     

Tissue characterization in intravascular ultrasound images

Intravascular ultrasound (IVUS) imaging permits direct visualization of vascular pathology. It has been used to evaluate lumen and plaque in coronary arteries and its clinical significance for guidance of coronary interventions is increasingly recognized. Conventional manual evaluation is tedious and time-consuming. This paper describes a highly automated approach to segmentation of coronary wall and plaque, and determination of plaque composition in individual IVUS images and pullback image sequences. The determined regions of plaque were classified in one of three classes: soft plaque, hard plaque, or hard plaque shadow. The method's performance was assessed in vitro and in vivo in comparison with observer-defined independent standards. In the analyzed images and image sequences, the mean border positioning error of the wall and plaque borders ranged from 0.13-0.17 mm. Plaque classification correctness was 90%.     

Lumen centerline detection in complex coronary angiograms

The authors have developed a method for lumen centerline detection in individual coronary segments that is based on simultaneous detection of the approximate positions of the left and right coronary borders. This approach emulates that of a clinician who visually identifies the lumen centerline as the midline between the simultaneously-determined left and right borders of the vessel segment of interest. The authors' lumen centerline detection algorithm and 2 conventional centerline detection methods were compared to carefully-defined observer-identified centerlines in 89 complex coronary images. Computer-detected and observer-defined centerlines were objectively compared using 5 indices of center line position and orientation. The quality of centerlines obtained with the new simultaneous border identification approach and the 2 conventional centerline detection methods was also subjectively assessed by an experienced cardiologist who was unaware of the analysis method. The authors' centerline detection method yielded accurate centerlines in the 89 complex images. Moreover, their method outperformed the 2 conventional methods as judged by all 5 objective parameters (p<0.001 for each parameter) and by the subjective assessment of centerline quality (p<0.001). Automated detection of lumen centerlines based on simultaneous detection of both coronary borders provides improved accuracy in complex coronary arteriograms     

Evaluation of three-dimensional segmentation algorithms for the identification of luminal and medial-adventitial borders in intravascular ultrasound images

Intravascular ultrasound (IVUS) provides direct depiction of coronary artery anatomy, including plaque and vessel area, which is important in quantitative studies on the progression or regression of coronary artery disease. Traditionally, these studies have relied on manual evaluation, which is laborious, time consuming, and subject to large interobserver and intraobserver variability. A new technique, called active surface segmentation, alleviates these limitations and makes strides toward routine analyses. However, for three-dimensional (3-D) plaque assessment or 3-D reconstruction to become a clinical reality, methods must be developed which can analyze many images quickly. Presented is a comparison between two active surface techniques for three-dimensional segmentation of luminal and medial-adventitial borders. The force-acceleration technique and the neighborhood-search technique accurately detected both borders in vivo (r2 = 0.95 and 0.99, Williams' index = 0.67 and 0.65, and r2 = 0.95 and 0.99, WI = 0.67 and 0.70, respectively). However, the neighborhood-search technique was significantly faster and required less computation. Volume calculations for both techniques (r2 = 0.99 and r2 = 0.99) also agreed with a known-volume phantom. Active surface segmentation allows 3-D assessment of coronary morphology and further developments with this technology will provide clinical analysis tools.     

Intravascular ultrasound image segmentation: a three-dimensional fast-marching method based on gray level distributions

Intravascular ultrasound (IVUS) is a catheter based medical imaging technique particularly useful for studying atherosclerotic disease. It produces cross-sectional images of blood vessels that provide quantitative assessment of the vascular wall, information about the nature of atherosclerotic lesions as well as plaque shape and size. Automatic processing of large IVUS data sets represents an important challenge due to ultrasound speckle, catheter artifacts or calcification shadows. A new three-dimensional (3-D) IVUS segmentation model, that is based on the fast-marching method and uses gray level probability density functions (PDFs) of the vessel wall structures, was developed. The gray level distribution of the whole IVUS pullback was modeled with a mixture of Rayleigh PDFs. With multiple interface fast-marching segmentation, the lumen, intima plus plaque structure, and media layers of the vessel wall were computed simultaneously. The PDF-based fast-marching was applied to 9 in vivo IVUS pullbacks of superficial femoral arteries and to a simulated IVUS pullback. Accurate results were obtained on simulated data with average point to point distances between detected vessel wall borders and ground truth <0.072 mm. On in vivo IVUS, a good overall performance was obtained with average distance between segmentation results and manually traced contours <0.16 mm. Moreover, the worst point to point variation between detected and manually traced contours stayed low with Hausdorff distances <0.40 mm, indicating a good performance in regions lacking information or containing artifacts. In conclusion, segmentation results demonstrated the potential of gray level PDF and fast-marching methods in 3-D IVUS image processing.     

Interactive virtual endoscopy in coronary arteries based on multimodality fusion

A novel approach for platform-independent virtual endoscopy in human coronary arteries is presented in this paper. It incorporates previously developed and validated methodology for multimodality fusion of two X-ray angiographic images with pullback data from intravascular ultrasound (IVUS). These modalities pose inherently different challenges than those present in many tomographic modalities that provide parallel slices. The fusion process results in a three- or four-dimensional (3-D/4-D) model of a coronary artery, specifically of its lumen/plaque and media/adventitia surfaces. The model is used for comprehensive quantitative hemodynamic, morphologic, and functional analyses. The resulting quantitative indexes are then used to supplement the model. Platform-independent visualization is achieved through the use of the ISO/IEC-standardized Virtual Reality Modeling Language (VRML). The visualization includes an endoscopic fly-through animation that enables the user to interactively select vessel location and fly-through speed, as well as to display image pixel data or quantification results in 3-D. The presented VRML virtual-endoscopy system is used in research studies of coronary atherosclerosis development, quantitative assessment of coronary morphology and function, and vascular interventions.     

Ovarian ultrasound image analysis: follicle segmentation

Ovarian ultrasound is an effective tool in infertility treatment. Repeated measurements of the size and shape of follicles over several days are the primary means of evaluation by physicians. Currently, follicle wall segmentation is achieved by manual tracing which is time consuming and susceptible to inter-operator variation. An automated method for follicle wall segmentation is reported that uses a four-step process based on watershed segmentation and knowledge-based graph search algorithm which utilizes priori information about follicle structure for inner and outer wall detection. The automated technique was tested on 36 ultrasonographic images of women's ovaries. Validation against manually traced borders has shown good correlation of manually defined and computer-determined area measurements (R2 = 0.85 - 0.96). The border positioning errors were small: 0.63+/-0.36 mm for inner border and 0.67+/-0.41 mm for outer border detection. The use of watershed segmentation and graph search methods facilitates fast, accurate inner and outer border detection with minimal user-interaction.     

Effects of vessel geometry and catheter position on dose delivery in intracoronary brachytherapy

In-stent restenosis is commonly observed in coronary arteries after intervention. Intravascular brachytherapy has been found effective in reducing the recurrence of restenosis after stent placement. Conventional dosing models for brachytherapy with beta (beta) radiation neglect vessel geometry as well as the position of the delivery catheter. This paper demonstrates in computer simulations on phantoms and on in vivo patient data that the estimated dose distribution varies substantially in curved vessels. In simulated phantoms of 50-mm length with a shape corresponding to a 60 degrees - 180 degrees segment of a respectively sized torus, the average dose in 2-mm depth was decreased by 2.70%-7.48% at the outer curvature and increased by 2.95%-9.70% at the inner curvature as compared with a straight phantom. In vivo data were represented in a geometrically correct three-dimensional model that was derived by fusion of intravascular ultrasound (IVUS) and biplane angiography. These data were compared with a simplified tubular model reflecting common assumptions of conventional dosing schemes. The simplified model yielded significantly lower estimates of the delivered radiation and the dose variability as compared with a geometrically correct model (p < 0.001). The estimated dose in ten vessel segments of eight patients was on average 8.76% lower at the lumen/plaque and 6.52% lower at the media/adventitia interfaces (simplified tubular model relative to geometrically correct model). The differences in dose estimates between the two models were significantly higher in the right coronary artery as compared with the left coronary artery (p < 0.001).     

基于局部形状结构分类的心血管内超声图像中-外膜边界检测

本文提出了一种基于局部形状结构分类的心血管内超声（Intravascular Ultrasound,IVUS）图像中-外膜边界检测方法.首先利用k-均值（k-means）聚类方法,确定局部形状结构类别;其次通过类别标号索引图像块,并对其进行积分通道特征和自相似性特征提取,构建多分类随机决策森林模型;最后由分类模型寻找IVUS图像的关键点,采用曲线拟合方法,实现IVUS图像中-外膜边界检测.实验结果表明,本文方法能够有效地解决IVUS图像中斑块、伪影和血管分支等造成边缘难以准确检测的问题,与已有算法相比,其JM （Jaccard Measure,JM）达到了88.9%,PAD （Percentage of Area Difference,PAD）降低了19.1%,HD （Hausdorff Distance,HD）减少了9.7%,更准确地识别目标边界的关键点,成功地检测出完整的中-外膜边界.     

Reconstruction and quantification of the carotid artery bifurcation from 3-D ultrasound images

Three-dimensional (3-D) ultrasound is a relatively new technique, which is well suited to imaging superficial blood vessels, and potentially provides a useful, noninvasive method for generating anatomically realistic 3-D models of the peripheral vasculature. Such models are essential for accurate simulation of blood flow using computational fluid dynamics (CFD), but may also be used to quantify atherosclerotic plaque more comprehensively than routine clinical methods. In this paper, we present a spline-based method for reconstructing the normal and diseased carotid artery bifurcation from images acquired using a freehand 3-D ultrasound system. The vessel wall (intima-media interface) and lumen surfaces are represented by a geometric model defined using smoothing splines. Using this coupled wall-lumen model, we demonstrate how plaque may be analyzed automatically to provide a comprehensive set of quantitative measures of size and shape, including established clinical measures, such as degree of (diameter) stenosis. The geometric accuracy of 3-D ultrasound reconstruction is assessed using pulsatile phantoms of the carotid bifurcation, and we conclude by demonstrating the in vivo application of the algorithms outlined to 3-D ultrasound scans from a series of patient carotid arteries.     

Noninvasive vascular elastography: theoretical framework

Changes in vessel wall elasticity may be indicative of vessel pathologies. It is known, for example, that the presence of plaque stiffens the vascular wall, and that the heterogeneity of its composition may lead to plaque rupture and thrombosis. Another domain of application where ultrasound elastography may be of interest is the study of vascular wall elasticity to predict the risk of aneurysmal tissue rupture. In this paper, this technology is introduced as an approach to noninvasively characterize superficial arteries. In such a case, a linear array ultrasound transducer is applied on the skin over the region of interest, and the arterial tissue is dilated by the normal cardiac pulsation. The elastograms, the equivalent elasticity images, are computed from the assessment of the vascular tissue motion. Investigating the forward problem, it is shown that motion parameters might be difficult to interpret; that is because tissue motion occurs radially within the vessel wall while the ultrasound beam propagates axially. As a consequence of that, the elastograms are subjected to hardening and softening artefacts, which are to be counteracted. In this paper, the Von Mises (VM) coefficient is proposed as a new parameter to circumvent such mechanical artefacts and to appropriately characterize the vessel wall. Regarding the motion assessment, the Lagrangian estimator was used; that is because it provides the full two-dimensional strain tensor necessary to compute the VM coefficient. The theoretical model was validated with biomechanical simulations of the vascular wall properties. The results allow believing in the potential of the method to differentiate hard plaques and lipid pools from normal vascular tissue. Potential in vivo implementation of noninvasive vascular elastography to characterize abdominal aneurysms and superficial arteries such as the femoral and the carotid is discussed.     

Semiautomated border tracking of cine echocardiographic ventnrcular images

Two-dimensional ultrasound sector scans of the left ventricle (LV) are commonly used to diagnose cardiac mechanical function. Present quantification procedures of wall motion by this technique entail inaccuracies, mainly due to relatively poor image quality and the absence of a definition of the relative position of the probe and the heart. The poor quality dictates subjective determination of the myocardial edges, while the absence of a position vector increases the errors in the calculations of wall displacement, LV blood volume, and ejection fraction. An improved procedure is proposed here for automatic myocardial border tracking (AMBT) of the endocardial and epicardial edges in a sequence of video images. The procedure includes nonlinear filtering of whole images, debiasing of gray levels, and location-dependent contrast stretching. The AMBT algorithm is based upon tracking movement of a small number of predefined set of points, which are manually defined on the two myocardial borders. Information from one image is used, by utilizing predetermined statistical criteria to iteratively search and detect the border points on the next one. Border contours are reconstructed by Spline interpolation of the border points. The AMBT procedure is tested by comparing processed sequences of cine echocardiographic scan images to manual tracings by an objective observer and to results from previously published data.     

Automated analysis of coronary arterial morphology in cineangiograms: geometric and physiologic validation in humans

A method of coronary border identification is discussed that is based on graph searching principles and is applicable to the broad spectrum of angiographic image quality observed clinically. Cine frames from clinical coronary angiograms were optically magnified, digitized, and graded for image quality. Minimal lumen diameters, referenced to catheter size, were derived from automatically identified coronary borders and compared to those defined using quantitative coronary arteriography and to observer-traced borders. computer-derived minimal lumen diameters were also compared to intracoronary measurements of coronary vasodilator reserve, a measure of the functional significance of a coronary obstruction. To test the robustness of the present border detection method, computer-derived coronary borders were compared to independent standards separately for good and poor angiographic images. The accuracy of computer-identified borders was similar in the two cases.     

Automated analysis of brachial ultrasound image sequences: early detection of cardiovascular disease via surrogates of endothelial function

Early detection of cardiovascular disease would allow timely institution of preventive measures. Arterial endothelium play a primary role in processes leading to the development of atherosclerotic plaque and cardiovascular disease in general. Determination of flow-mediated dilatation (FMD) of brachial arteries from B-mode ultrasound image sequences offers a noninvasive surrogate index of endothelial function. A highly automated method for analysis of brachial ultrasound image sequences is reported and its performance assessed. The method overcomes the variability of brachial ultrasound images across subjects by incorporating machine learning and quality control steps. The automated method outperformed conventional manual analysis by providing a decreased analysis bias, increased reproducibility, and improved measurement accuracy. Consequently, it decreases inter- and intraobserver as well interinstitution variability. The method has been employed in a number of population studies with thousands of subjects analyzed.     

Quantitative analysis of ultrasound B-mode images of carotid atherosclerotic plaque: correlation with visual classification and histological examination

This paper presents a quantitative comparison of three types of information available for 52 patients scheduled for carotid endarterectomy: subjective classification of the ultrasound images obtained during scanning before operation, first-and second-order statistical features extracted from regions of the plaque in still ultrasound images from three orthogonal scan planes and finally a histological analysis of the surgically removed plaque. The quantitative comparison was made with the linear model and with separation of the available data into training and test sets. The comparison of subjective classification with features from still ultrasound images revealed an overall agreement of 60% for classification of echogenicity and 70% for classification of structure. Comparison of the histologically determined relative volume of soft materials with features from the still images revealed a correlation coefficient of r = -0.42(p = 0.002). for mean echogenicity of the plaque region. The best performing feature was of second order and denoted Contrast (r = -0.5). Though significant, the latter correlation is probably not strong enough to be useful for clinical prediction of relative volume of soft materials for individual patients. Reasons for this is discussed in the paper, together with suggestions for improvements.     

In Vivo IVUS-Based 3-D Fluid–Structure Interaction Models With Cyclic Bending and Anisotropic Vessel Properties for Human Atherosclerotic Coronary Plaque Mechanical Analysis

In this paper, a modeling approach combining in vivo intravascular ultrasound (IVUS) imaging, computational modeling, angiography, and mechanical testing is proposed to perform mechanical analysis for human coronary atherosclerotic plaques for potential more accurate plaque vulnerability assessment. A 44-slice in vivo IVUS dataset of a coronary plaque was acquired from one patient, and four 3-D models with fluid-structure interactions (FSIs) based on the data were constructed to quantify effects of anisotropic vessel properties and cyclic bending of the coronary plaque on flow and plaque stress/strain conditions. Compared to the isotropic model (model 1, no bending, no axial stretch), maximum stress-P₁ (maximum principal stress) values on the cut surface with maximum bending (where applicable) from model 2 (anisotropic, no bending, no stretch), model 3 (anisotropic, with bending, no stretch), and model 4 (anisotropic with bending and stretch) were, respectively, 63%, 126%, and 345% higher than that from model 1. Effects of cyclic bending on flow behaviors were modest (5%-15%). Our preliminary results indicated that in vivo IVUS-based FSI models with cyclic bending and anisotropic material properties could improve the accuracies of plaque stress/strain predictions and plaque vulnerability assessment. Large-scale patient studies are needed to further validate our findings.