Ebook: Plaque Imaging: Pixel to Molecular Level

Different classifications have been proposed in the literature for the characterization of atherosclerotic plaque morphology, resulting in considerable confusion. For example plaques containing medium of high level uniform echoes were classified as homogeneous by others and correspond closely to dense and calcified plaques, other types. This survey is to understand different types of plaque when imaged using ultrasound and MR.

The increasing amount of medical images produced and stored daily in hospitals needs a datrabase management system that organizes them in a meaningful way, without the necessity of time‐consuming textual annotations for each image. One of the basic ways to organize medical images in taxonomies consists of clustering them depending of plaque appearance (for example, intravascular ultrasound images). Although lately, there has been a lot of research in the field of Content‐Based Image Retrieval systems, mostly these systems are designed for dealing a wide range of images but not medical images. Medical image retrieval by content is still an emerging field, and few works are presented in spite of the obvious applications and the complexity of the images demanding research studies. In this chapter, we overview the work on medical image retrieval and present a general framework of medical image retrieval based on plaque appearance. We stress on two basic features of medical image retrieval based on plaque appearance: plaque medical images contain complex information requiring not only local and global descriptors but also context determined by image features and their spatial relations. Additionally, given that most objects in medical images usually have high intra‐ and inter‐patient shape variance, retrieval based on plaque should be invariant to a family of transformations predetermined by the application domain. To illustrate the medical image retrieval based on plaque appearance, we consider a specific image modality: intravascular ultrasound images and present extensive results on the retrieval performance.

This chapter reviews current MRI techniques to differentiate stable versus high risk atherosclerosis and discusses the development of non‐invasive MR imaging techniques to characterize atherosclerotic plaques. Tissue specific MR signal features will be described according to histo‐pathological evaluation standards and comprehensive imaging protocol for the identification of different lesion types will be introduced.

The material composition and morphology of the atherosclerotic plaque components are considered to be more important determinants of acute coronary ischemic syndromes than the degree of stenosis. When a vulnerable plaque ruptures it causes an acute thrombotic reaction. Rupture prone plaques contain a large lipid pool covered by a thin fibrous cap. The stress in these caps increases with decreasing thickness. Additionally, the cap may be weakened by macrophage infiltration. IntraVascular UltraSound (IVUS) elastography might be an ideal technique to assess the presence of lipid pools and to identify high stress regions. Elastography is a technique that assesses the local elasticity (strain and modulus) of tissue. It is based on the principle that the deformation of tissue by a mechanical excitation is a function of its material properties. The deformation of the tissue is determined using ultrasound. For intravascular purposes, the intraluminal pressure is used as the excitation force. The radial strain in the tissue is obtained by cross‐correlation techniques on the radio frequency signals. The strain is color‐coded and plotted as a complimentary image to the IVUS echogram. IVUS elastography, and IVUS palpography (which uses the same principle but is faster and more robust), have been extensively validated using simulations and by performing experiments in vitro and in vivo with diseased arteries from animals and humans. Strain was shown to be significantly different in various plaque types (absent, fatty, fibrous or calcified). A high strain region with adjacent low strain at the lumen vessel‐wall boundary has 88% sensitivity and 89% specificity for detecting vulnerable plaques. High strain regions at the lumen plaque‐surface have 92% sensitivity and 92% specificity for identifying macrophages. Furthermore, the incidence of vulnerable‐plaque‐specific strain patterns in humans has been related to clinical presentation (stable angina, unstable angina or acute myocardial infarction) and the level of C‐reactive protein. In conclusion, the results obtained with IVUS (strain and modulus) elastography/palpography, show the potential of the technique to become a unique tool for clinicians to assess the vulnerability and material composition of plaques.

Numerical simulations, which are based on reliable biomechanical models of blood vessels, can help to get a better understanding of cardiovascular diseases such as atherosclerosis, and can be used to develop optimal medical treatment strategies.

The adventitia is the outer most layer of blood vessels and its mechanical properties are essentially determined by the three‐dimensional, structural arrangement of collagen fibre bundles embedded in the tissue. Global information such as the orientation statistics of the fibre bundles as well as detailed information as the crimp of the single fibres within the bundles is of particular interest in biomechanical modeling.

In order to obtain a sufficiently large amount of data for biomechanical modeling, a fully automatic method for the structural analysis of the soft tissue is required. In this contribution we present methods based on computer vision to fulfill this task. We start by discussing proper tissue preparation and imaging techniques that have to be used to obtain data, which reliably represents the real three‐dimensional tissue structure. The next step is concerned with algorithms that robustly segment the collagen fibre bundles and cope with various kinds of artifacts. Novel segmentation techniques for robust segmentation of individual fibril bundles and methods for estimation of their parameters, such as location, shape, mean fibril orientation, crimp of fibrils, etc, is discussed. The proposed algorithms are based on novel perceptual grouping methods operating on the extracted orientation data of fibrils.

Finally, we demonstrate the results obtained by our fully automatic method on real data. In addition, for a more quantitative assessment, we introduce a generative structural model that enables the synthesis of three‐dimensional fibre bundles with well‐defined characteristics.

In this chapter we present recent developments in the modeling of coronary artery biomechanics. We first introduce the pathology and localization of lesions in the circulatory system. Recent fluid and structural modeling of CAD is presented and discussed. At the end of the chapter, we present recent effort in coupling these two modeling domains using fluid‐structure interaction (FSI).

The chapter presents the Cardiac CT for the assessment of cardiovascular pathology with an emphasis on the detection of coronary atherosclerosis. Cardiac CT is a robust technology for the non‐invasive assessment for a spectrum of cardiovascular disease processes. This imaging modality can provide assessment of atherosclerotic plaque burden and coronary artery disease risk through coronary calcium scoring. Advances in spatial and temporal resolution, electrocardiographic triggering methodology, and image reconstruction software have helped in the evaluation of coronary artery anatomy and vessel patency, providing the ability to noninvasively diagnose or rule out significant epicardial coronary artery disease. This technique also allows the 3‐Dimensional simultaneous imaging of additional cardiac structures including coronary veins, pulmonary veins, atria, ventricles, aorta and thoracic arterial and venous structures, with definition of their spatial relationships for the comprehensive assessment of a variety of cardiovascular disease processes.

X‐ray computed tomography (CT) is widely available in the world and has the ability to provide high definition, thin section imaging of any body part. In particular, CT over the past decade has been shown in numerous publications to allow for quantitation of coronary calcification, a proven surrogate for coronary artery atheromatous plaque. Electron beam tomography (EBT) and multi‐detector CT (MDCT) have been studied for these purposes; although the majority of the data has been derived from EBT studies.

This chapter details the patho‐biology of atherosclerotic disease, the basis of using EBT (and/or CT in general) to define atherosclerotic plaque including the technical and engineering pitfalls and promises, and details issues of clinical application.

Studies done on carotid arteries suggest that the morphology and composition of atherosclerotic plaque are predictive of stroke risk. The goal of this investigation has been to demonstrate that the true acoustic integrated backscatter (IBS) from plaque regions can be measured non‐invasively, based on which plaque composition may be inferred and thus become a tool to estimate the likelihood of a lesion or plaque being stable or vulnerable, i.e. having a risk of causing a stroke.

To obtain the true IBS non‐invasively, the scattering and aberrating effect of the intervening tissue layers must be overcome. This is achieved by using the IBS from arterial blood as a reference backscatter, specifically the backscatter from a blood volume along the same scan line as and adjacent to the region of interest. We have shown that the variance of the IBS estimate of the blood backscatter signal can be quantified and reduced to a specified tolerable level.

The aim of this chapter is to summarise the recent advances in ultrasonic plaque characterisation and to evaluate the efficacy of computer aided diagnosis based on neural and statistical classifiers using as input texture and morphological features. Several classifiers like the K‐Nearest Neighbour (KNN) the Probabilistic Neural Network (PNN) and the Support Vecton Machine (SVM) are evaluated resulting to a diagnostic accuracy up to 71.2%.

Intravascular ultrasound images represent a unique tool to guide interventional coronary procedures, this technique allows to supervise the cross‐sectional locations of the vessel morphology and to provide quantitative and qualitative information about the causes and severity of coronary diseases. At the moment, the automatic extraction of this kind of information is performed without taking into account the basic signal principles that guide the process of image generation. In this chapter, we overview the main physical principles and factors that affect the IVUS generation; we propose a simple physics‐based approach for IVUS image simulation that is defined as a discrete representation of the tissue by individual scatterrers elements with given spatial distribution and backscattering differential cross sections. In order to generate the physical model that allows to construct synthetic IVUS images, we analyze the process of pulse emission, transmission and reception of the ultrasound signal as well as its interaction with the different tissues scatterrers of the simulated artery. In order to obtain the 3D synthetic image sequences we involve the dynamic behavior of the heart/arteries and the catheter movement in the image generation model. Having an image formation model allows to study rhe physics parameters that participate during the image generation and to achieve a better understanding and robust interpreting of IVUS image structures. Moreover, this model allows to comprehend, simulate and solve several limitations of IVUS sequences, to extract important image parameters to be taken into account when developing robust image processing algorithms as well as to construct wide synthetic image sequence databases in order to validate different image processing techniques.

Coronary artery disease is the number one cause of death in the United States and the Western world, and approximately 250,000 affected people die per year without ever being admitted to a hospital. One of the main reasons of such a high death‐rate without any diagnosis is that more than 50 or heart‐attacks) occur in patients with no prior history of known heart disease or symptoms. Coronary artery disease leads to the occlusion of arteries that are vital in providing nutrients to the heart muscles. The disease develops by progressive accumulation or formation of “plaque” within an artery. Certain types of plaques could occlude blood flow and yet might be “stable”. These plaques usually have a high fibrous content, and are known as hard plaques. On the other hand, “unstable” or “soft” plaques might not cause much occlusion but could be vulnerable to rupture. Rupture of such plaques could lead to total or partial occlusion in arteries resulting in sudden cardiac death or heart‐attack. In fact, 68 coronary arteries are less than 50.

Intravascular ultrasound (IVUS) is a minimally invasive imaging modality that provides cross‐section images of arteries in real‐time, allowing visualization of atherosclerotic plaques in vivo. In standard IVUS gray‐scale images, calcified regions of plaque and dense fibrous components generally reflect ultrasound energy well and thus appear bright and homogeneous on IVUS images. Conversely, regions of low echo reflectance in IVUS images are usually labeled “soft” or “mixed” plaque. However, this visual interpretation has been demonstrated to be very inconsistent in accurately determining plaque composition and does not allow real‐time assessment of quantitative plaque constituents.

Spectral analysis of the backscattered radiofrequency (RF) ultrasound signals allows detailed assessment of plaque composition. Advanced mathematical techniques can be employed to extract spectral information from these RF data to determine composition. The spectral content or signature of RF data reflected from tissue depends on density, compressibility, concentration, size, etc. A combination of spectral parameters were used to develop statistical classification schemes for analysis of in vivo IVUS data in real‐time. The clinical data acquisition system is ECG gated and the analysis software developed by our group reconstructs IVUS gray‐scale images from the acquired RF data. A combination of spectral parameters and active contour models is used for real‐time 3D plaque segmentation followed by computation of color‐coded tissue maps for each image cross‐section and longitudinal views of the entire vessel. The “fly‐through” mode allows one to visualize the complete length of the artery internally with the histology components at the lumen surface. In addition, vessel and plaque metrics such as areas and volumes of individual plaque components (collagen, fibro‐lipid, calcium, lipid‐core) are also available.

Imaging of the coronary arteries is usually performed by X‐ray contrast angiography or intravascular ultrasound (IVUS). Angiography provides information about the vessel lumen and its geometry. IVUS offers more detailed information that also includes the vascular wall. The chapter describes these two imaging modalities and their geometrically correct fusion yielding a 3‐D and/or 4‐D representation of the coronary geometry and morphology. The image‐derived information is used for assessment of coronary function and plaque severity, blood flow related indices are determined using computational fluid dynamics. Detailed description of the methodology is followed by validation and clinical studies.

A new horizon in imaging of atherosclerotic plaque is the use of contrast enhancing agents to investigate molecular and cellular markers. This chapter reviews the state of the art in contrast‐enhanced imaging of atherosclerotic plaque. Topics include enhancement characteristics of different plaque tissues, modeling of dynamic contrast‐enhanced MRI, and targeted contrast agents. For each topic, specific techniques for image acquisition and post‐processing are presented. Quantitative results are summarized for experiments conducted with histological validation. Finally, potential applications of these techniques in detecting vulnerable lesions and in drug development are discussed.

The chapter presents the research to test the hypotheses that (1) vessel wall volume measurements from dark blood MR images with multiple contrast‐weightings (T1W, T2W and PDW) are highly reproducible, and that (2) the intra‐observer and inter‐observer variability of carotid wall volume measurements will be less than those obtained with maximum wall area (MaxWA) measurements. Methods: Sixteen patients (aged 72 ± 7years) with carotid stenosis documented by duplex ultrasound were recruited for the study. Dark blood T1W, PDW and T2W MR images were used to measure carotid wall volume and MaxWA by two independent observers for inter‐observer and intra‐observer variability assessment. Results: The intra‐observer absolute difference of carotid wall volume for T1W, T2W and PDW images were 67.3 ± 47.5 mm³ (2.3 ± 1.8%), 63.2 ± 52.2 mm³ (2.0 ± 1.3%), and 69.8 ± 45.2 mm³ (2.4 ± 1.7%) respectively. The inter‐observer absolute difference of carotid wall volume for T1W, T2W and PDW images were 103.5 ± 141.8 mm³ (3.0 ± 3.1%), 95.9 ± 102.1 mm³ (3.1 ± 2.6%), and 132.1 ± 87.8 mm³ (4.3 ± 2.7%) respectively. The intra‐observer absolute difference of carotid MaxWA for T1W, T2W and PDW images were 6.9 ± 5.0 mm² (4.2 ± 2.9%), 5.1 ± 4.2 mm² (3.1 ± 2.3%) and 7.5 ± 4.7 mm² (4.2 ± 2.7%) respectively. The inter‐observer absolute difference of carotid MaxWA for T1W, T2W and PDW images were 9.5 ± 4.2 mm² (5.8 ± 2.3%), 6.4 ± 6.1 mm² (3.8 ± 3.1%) and 10.8 ± 7.3 mm² (6.1 ± 3.7%) respectively. Both intra‐ and inter‐observer variability in carotid volume measurement tend to be smaller than that in carotid MaxWA measurement with intraclass correlation coefficients ranged 0.932 to 0.987 for volume measurement and 0.822 to 0.946 for MaxWA measurement.

This chapter describes automatic three‐dimensional registration techniques for magnetic resonance images of carotid vessels. The immediate applications include atherosclerotic plaque characterization and plaque burden quantification vector‐based segmentation using dark blood MR images having multiple contrast weightings (proton density (PD), T1, and T2). Another application is the measurement of disease progression and regression with drug trials. A normalized mutual information registration algorithm is applied to compensate movements between image acquisitions. PD, T1, and T2 images were acquired from patients and volunteers and then matched for image analysis. Visualization methods such as contour overlap showed that vessels well aligned after registration. Distance measurements from the landmarks indicated that the registration method worked well with an error of less than 1‐mm.

Considerable evidence has emerged that adverse blood flow patterns are a major factor in the onset of atherosclerotic disease and may play a role in disease progression. This chapter reviews a technique, referred to as vascular computational fluid dynamics (CFD), for characterizing blood flow patterns in large arteries from magnetic resonance angiography (MRA) and velocity‐encoded phase‐contrast magnetic resonance (PC MR) imaging. In vascular CFD, hemodynamic conditions are modeled by the finite‐element method with flow is governed by the incompressible Navier‐Stokes equations. Construction of the vascular CFD models is a multi‐step process. Critical aspects of the methodology are described in detail including surface reconstruction, construction of the volumetric mesh, imposition of boundary conditions and solution of the finite element model. In vitro and in vivo experimentation is discussed that demonstrates, in a preliminary manner, the validity of the methodology. Flow models are presented for carotid arteries with a wide range of atherosclerotic disease. Considerable evidence has emerged that disturbed blood flow patterns are a major factor in the onset of atherosclerotic disease and may play a role in disease progression. The proposed chapter will review a technique for characterizing blood flow patterns in large arteries from magnetic resonance angiography (MRA) and velocity‐encoded phase‐contrast magnetic resonance imaging. This technique, known as vascular computational fluid dynamics (CFD), has been applied extensively to the bifurcation region of the carotid artery, a common site of plaque formation. Common hemodynamic features in this region will be presented based on imaging of a series of normal subjects. Hemodynamic features in the vicinity of the carotid bifurcation will also be presented for a series of subjects with advanced atherosclerotic disease.

One of the principal therapies considered for the control of in‐stent restenosis is the use of drug loaded polymer‐coated stents for local delivery. We present two‐dimensional and three‐dimensional numerical models to study local delivery of drug eluting stents. The impact of various stent and flow parameters on the concentration distribution in the wall are investigated including the effect of the strut size, coating thickness, strut inter‐distance and strut embedment in the vascular wall, blood flowing speed and the respective diffusion coefficients in the blood, wall and polymer. We also present criteria to assess the drug delivery efficiency based of the concept of the therapeutic window which aims at an spatial homogeneous concentration distribution and we introduce the variables to assess the amount of drug delivered in the wall. The results suggest that advection have a much stronger effect compared to diffusion in the blood media and that drug diffusivity in the arterial wall and in the polymer coating significantly affects the drug distribution. It is also shown that fully‐embedded struts provide better spatial drug concentration uniformity after a short period of time and the half‐embedded struts have a better temporal uniformity.