J. Stoitsis

Comparison of Multiresolution Features for Texture Classification of Carotid Atherosclerosis From B-Mode Ultrasound

In this paper, a multiresolution approach is suggested for texture classification of atherosclerotic tissue from B-mode ultrasound. Four decomposition schemes, namely, the discrete wavelet transform, the stationary wavelet transform, wavelet packets (WP), and Gabor transform (GT), as well as several basis functions, were investigated in terms of their ability to discriminate between symptomatic and asymptomatic cases. The mean and standard deviation of the detail subimages produced for each decomposition scheme were used as texture features. Feature selection included 1) ranking the features in terms of their divergence values and 2) appropriately thresholding by a nonlinear correlation coefficient. The selected features were subsequently input into two classifiers using support vector machines (SVM) and probabilistic neural networks. WP analysis and the coiflet 1 produced the highest overall classification performance (90% for diastole and 75% for systole) using SVM. This might reflect WPs ability to reveal differences in different frequency bands, and therefore, characterize efficiently the atheromatous tissue. An interesting finding was that the dominant texture features exhibited horizontal directionality, suggesting that texture analysis may be affected by biomechanical factors (plaque strains).

Comparison of Block Matching and Differential Methods for Motion Analysis of the Carotid Artery Wall From Ultrasound Images

Motion of the carotid artery wall is important for the quantification of arterial elasticity and contractility and can be estimated with a number of techniques. In this paper, a framework for quantitative evaluation of motion analysis techniques from B-mode ultrasound images is introduced. Six synthetic sequences were produced using 1) a real image corrupted by Gaussian and speckle noise of 25 and 15 dB, and 2) the ultrasound simulation package Field II. In both cases, a mathematical model was used, which simulated the motion of the arterial wall layers and the surrounding tissue, in the radial and longitudinal directions. The performance of four techniques, namely optical flow (OFHS), weighted least-squares optical flow (OFLK(WLS)), block matching (BM), and affine block motion model (ABMM), was investigated in the context of this framework. The average warping indices were lowest for OFLK(WLS) (1.75 pixels), slightly higher for ABMM (2.01 pixels), and highest for BM (6.57 pixels) and OFHS (11.57 pixels). Due to its superior performance, OFLK(WLS) was used to quantify motion of selected regions of the arterial wall in real ultrasound image sequences of the carotid artery. Preliminary results indicate that OFLK(WLS) is promising, because it efficiently quantified radial, longitudinal, and shear strains in healthy adults and diseased subjects.

A Modular Software System to Assist Interpretation of Medical Images—Application to Vascular Ultrasound Images

ANALYSIS, a modular software system that can assist interpretation of medical images, is presented in This work. ANALYSIS was designed to meet the needs of the general medical image analysis procedure and can be used for the pre-processing and analysis of medical images of various modalities and formats, in an attempt to assist diagnosis. The system includes two main modules for texture and motion analysis of selected regions of interest (ROIs). Motion can be estimated using region tracking and block matching. Texture features can be estimated using first order statistics, second order statistics, laws texture energy, and fractal dimension. Moreover, ANALYSIS includes modules for image preprocessing, statistical analysis, classification and clustering using fuzzy c-means. It provides a friendly user interface, control fields, tool buttons and menus. The system was used to assist diagnosis of carotid atherosclerosis from B-mode ultrasound images. Motion and texture characteristics where estimated for 10 symptomatic and 9 asymptomatic subjects. After dimensionality reduction of the extracted texture features, 23 features were considered most significant. The estimation of motion vectors on the luminal surface of the plaque can identify patterns of inherent plaque motion, which may be responsible for clinical symptoms. ANALYSIS can assist interpretation of medical images through lesion detection, characterization and assessment. Further studies for different medical diagnostic tasks using different imaging modalities proves the possibility of the system to be used in clinical practice.

Characterization of carotid atherosclerosis based on motion and texture features and clustering using fuzzy c-means

Analysis of B-mode ultrasound images of the carotid atheromatous plaque includes the estimation of texture from static images and the estimation of motion from image sequences. The combination of these two types of information may be valuable for accurate diagnosis of vascular disease. The purpose of this paper was to study texture and motion patterns of carotid atherosclerosis and select the optimal combination of features that can characterize plaque. B-mode ultrasound images of 10 symptomatic and 9 asymptomatic plaques were interrogated. A total of 99 texture features were estimated using first-order statistics, second-order statistics, Laws texture energy and the fractal dimension. Only five texture features were significantly different between the two groups. In the same subjects, the motion of selected plaque regions was estimated using region tracking and block-matching and expressed through: a/maximal surface velocity (MSV), and b/maximal relative surface velocity (MRSV). MSV and MRSV were significantly lower in asymptomatic plaques suggesting more homogeneous motion patterns. Clustering using fuzzy c-means correctly classified 74% of plaques based on texture features only, and 79% of plaques based on motion features only. Classification performance reached 84% when a combination of motion and texture features was used.

Comparison of B-mode, M-mode and Hough transform methods for measurement of arterial diastolic and systolic diameters

Measurements of arterial diameter during the cardiac cycle are increasingly used to study the mechanical properties of the arterial wall and changes associated with disease. In this paper, diastolic and systolic diameters of the carotid arteries were estimated from ultrasound imaging using the following three different procedures: a/ B-mode imaging with region tracking and block-matching, b/ M-mode imaging with automated edge detection and c/ automatic segmentation of the arterial lumen at diastole and systole using the Hough transform. Transverse images of the carotid artery were used, in which the arterial lumen has an almost circular appearance. The values for systolic and diastolic diameters estimated with the Hough transform, 0.69plusmn0.04 and 0.61plusmn0.06, respectively, were closer to those estimated with B-mode and motion tracking, 0.75plusmn0.07 and 0.67plusmn0.09. A large difference was found for a subject with an atherosclerotic vessel wall. It is concluded that the Hough transform can be efficiently used to automatically segment healthy arterial wall lumen from B-mode ultrasound images of the carotid artery, assuming a circular shape. In atherosclerotic vessel walls the assumption for circular shape may no longer be valid, and thus the use of an elliptical shape may be more appropriate

Analysis and quantification of arterial wall motion from B-mode ultrasound images - comparison of block-matching and optical flow

Motion of the carotid atheromatous plaque may be responsible for plaque rupture and cerebrovascular symptoms. B-mode ultrasound allows non-invasive recording of arterial wall and plaque motion. Our aim was to analyze quantitatively patterns of arterial wall motion with different techniques. Temporal sequences of digitized B-mode ultrasound images of the carotid arteries of 10 young healthy subjects were interrogated. Arterial wall motion was analyzed using: a/ block-matching, and b/ optical flow. The motion of selected regions of the luminal surface of the arterial wall was estimated using region tracking and block-matching. The motion of areas of the arterial wall was estimated using optical flow. Waveforms showing radial and axial displacements, as well as radial and axial velocities were produced for the selected ROIs using both techniques. Both techniques produced waveforms with peaks, corresponding to cardiac cycle events, that occurred at similar time points. To study the similarity of the waveforms obtained from the two techniques, a cross-correlation coefficient was calculated. Cross-correlation coefficients were 0.72plusmn0.22 and 0.70plusmn0.19 for displacements and velocities, respectively, in the radial direction. In the axial direction, the coefficients were 0.32plusmn0.39 and 0.24plusmn0.22 for displacements and velocities, respectively. On the basis of this relative comparison of methods, we conclude that significant observations can be made for each motion analysis technique in terms of characterization of the mechanical properties of the tissue

Simulating dynamic B-mode ultrasound image data of the common carotid artery

Motion of the carotid artery wall from sequences of ultrasound images has been promising in providing useful indices characterizing elasticity of healthy and diseased arterial wall. The validation of motion analysis algorithms remains, however, a challenging task mainly because the actual tissue motion field is not readily available. In this paper, a methodology is suggested for the generation of simulated dynamic B-mode ultrasound image data of the common carotid artery. The approach consists of three main stages: (a) the generation of a scattering map using a real ultrasound image as a template, (b) the assumption of a mathematical model of arterial wall deformation in the radial and axial directions and (c) the generation of simulated sequential data by displacing the scattering map according to the deformation model. The procedure is based on the use of FIELD II, an ultrasound simulation package incorporating realistic transducer features. A total of 29 frames with a temporal separation of 0.04s were generated covering a cardiac cycle of approximately 1.16s duration. The simulation procedure is computationally expensive, but the resulting dynamic image data can be used efficiently in tasks involving evaluation of motion analysis algorithms.

A mathematical model of the mechanical deformation of the carotid artery wall and its application to clinical data

The study of arterial wall mechanics, including the study of stresses and strains experienced by the vascular wall, is pivotal in our understanding of arterial physiology. In this paper, a mathematical model is provided describing the deformation of the arterial wall in terms of 6 parameters. Actual deformation waveforms were also obtained from the analysis of B-mode ultrasound image sequences of the carotid artery using block-matching. The mathematical model was fitted to the clinical data using nonlinear least squares to determine the 6 parameters for 6 different locations along the posterior and 6 along the anterior walls, on the interface between the lumen and the intima-media complex (L-IM). On the posterior wall, 6 locations were also investigated at the interface between the intima-media complex and the adventitia (IM-A) as well as at the adventitia-surrounding tissue (A-T) boundary. The root mean square error was low for all locations indicating a good fit of the proposed model to the clinical data. The amplitude of the deformation, expressed through parameter alpha, was significantly lower in the A-T interface compared to the other two interfaces. The time when the systolic peak occurs, expressed through parameter t1, was significantly lower in the L-IM interface compared to the other two interfaces. Preliminary findings from a small group of diseased wall locations suggested that the parameters a, b and t1 were significantly different than healthy cases. This probably reflects alterations of arterial wall mechanics due to disease. This study showed that the proposed mathematical model is a satisfactory representation of the mechanical deformation of the carotid artery wall in the radial direction and can provide valuable information in the understanding of the mechanical behavior of the arterial wall.

A Communication Platform for Tele-monitoring and Tele-management of Type 1 Diabetes

Type 1 diabetes mellitus is a chronic disease characterized by blood glucose levels out of normal range due to inability of insulin production. This dysfunction leads to many short- and long-term complications. In this paper, a system for tele-monitoring and tele-management of type 1 diabetes patients is proposed, aiming at reducing the risk of diabetes complications and improving quality of life. The system integrates wireless personal area networks (WPAN), mobile infrastructure, and Internet technology along with commercially available and novel glucose measurement devices, advanced modeling techniques, and tools for the intelligent processing of the available diabetes patients information. The integration of the above technologies enables intensive monitoring of blood glucose levels, treatment optimisation, continuous medical care, and improvement of quality of life for type 1 diabetes patients, without restrictions in everyday life activities

Development of an integrated breast tissue density classification software system

The current work aims at the classification of breast tissue according to Breast Imaging Reporting and Data System (BIRADS), based on texture features from mammographic images. To this end an integrated software system was developed in visual C++ using the .NET 2.0 Framework. The system takes as inputs pictures in most of the popular bitmap formats as well as DICOM and provides as output a specific breast density category according to the BIRADS system. The functionality of the system is provided by three modules: (a) the pre-processing module, where a set of tools for image manipulation (rotation, crop, gray level adjustment) are available accompanied by the ability to perform anisotropic filtering to the input image, (b) the breast segmentation module where the breast region is separated from the image background and pectoral muscle using characteristics of monogenic signals and Gabor wavelets respectively and (c) the breast tissue density classification module where the breast tissue is categorized according to the BIRADS, using texture characteristics and probabilistic Latent Semantic Analysis (pLSA). Special emphasis has been given to the development of a functional and user-friendly interface.