Marek Belohlavek

Three- and Four-Dimensional Cardiovascular Ultrasound Imaging: A New Era for Echocardiography

Three-dimensional and four-dimensional ultrasonography were pioneered in the 1960s yet have been used little clinically. Only recently have advances in cardiovascular ultrasound equipment and in digital image storage, manipulation, and display techniques made three- and four-dimensional imaging clinically feasible. In this report, we review the historical development of these technologies during 3 decades to their culmination in current state-of-the-art technology. Examples of such multidimensional images are presented, with special emphasis on clinical applications. Although several limitations persist, three-dimensional cardiovascular ultrasonography seems likely to enhance imaging of the heart and vessels in a manner similar to the advent of two-dimensional echocardiography in the M-mode era. Clinician-scientists will soon be able to extract an object, such as the heart, from the body electronically for the purpose of anatomic, functional, and histologic analysis without adverse effect on the patient.

Multidimensional Visualization in Echocardiography: An Introduction

X-ray films depict three-dimensional objects as shadows in a two-dimensional plane; thus, objects become superimposed. Computed tomography and other types of tomographic imaging, such as ultrasonography, acquire two-dimensional images of a material property within a thin slice. Sequential adjacent two-dimensional tomograms can be used to construct three-dimensional displays of objects. Visualization, a field of computer science, enables scientists to measure image attributes (extraction of features), identify features (classification), separate objects from one another (segmentation), and produce comprehensible, information-dense images from three-dimensional data sets (rendering). A three-dimensional rendering of the heart can be used to represent only one component of the heart, such as the atrial septum or the ventricular chamber, and can be shaded or colored to enhance comprehension. Three-dimensional images rendered sequentially over time result in a dynamic four-dimensional display. This report describes multidimensional visualization of objects and tissues and specifically discusses examples from echocardiography.

Toroidal geometry: Novel three-dimensional intracardiac imaging with a phased-array transducer

Recent advances in small, linear-array transducers have opened new avenues for three-dimensional image acquisition from an intracardiac approach. The purpose of this study was to introduce a novel method of image acquisition using toroidal geometry, explore its fidelity of reproduction of three-dimensional cardiac anatomy, and determine whether a whole-heart scan is achievable. Acquisition was accomplished through 360-degree incremental rotation of a rigid endoscope with a side-mounted ultrasound transducer. The procedure was first tested with the use of a gelatin model to define far-field slice resolution with 1.8-degree rotational increments. Comparison of three-dimensional scans of cardiac specimens with corresponding photographs confirmed that toroidal geometry can provide a high-quality display of structures from all sides. We conclude that whole-heart three-dimensional scanning from within the cardiac chambers is possible with toroidal geometry. The quality of depicted anatomy depends on transducer location within the heart, distance from the transducer, density of slices, and image resolution. The potential of intracardiac three-dimensional ultrasound imaging includes detailed spatial evaluation of cardiac morphology, determination of appropriate placement of investigative or therapeutic devices (catheters, closure devices, etc.), and assessment of cardiac function.

Extraction of Endocardial Boundary From Echocardiographic Images by Means of the Kohonen Self-Organizing Map

There is considerable clinical interest in development of algorithms for reproducible determination of endocardial boundaries in echocardiographic images.1-9 This is feasible in the era of computer-aided analysis of cardiac morphology and function. However, ultrasound images are notoriously difficult to process because they are typically incomplete (dropouts, noise, etc.). Thus, automatic endocardial detection techniques require image enhancement to deal with discontinuous border definition. Our initial experiences with self-organizing maps (SOM) for the delineation of endocardial echoes is very encouraging and discussed in this manuscript. The objective was to determine whether this form of neural network combined with algorithms for edge detection can perform reproducible automated endocardial boundary delineation in artifact-prone echocardiographic images. The SOM has been preferred because: 1) no external operator is necessary to oversee the learning process of the unsupervised neural net, 2) it can be initialized with certain target-relevant shapes, 3) topological relationships are maintained between the neural net lattice nodes (the nodes define the vertices of surface tiles which may be useful for curved distances, surface area, and volume calculation), and 4) similar SOM algorithms have been successfully applied to other complex images.10

APPLICATION OF ORDER STATISTICS FILTERS TO CARDIAC BOUNDARY DETERMINATION IN ULTRASOUND IMAGES

Nonlinear order statistics filters are useful for impulsive noise removal and edge preservation. Speckle is a source of impulsive noise in ultrasound images. Edge preservation is important for determination of cardiac boundaries in echocardiographic images. Additionally, order statistics filters also suppress Gaussian noise which is often present in echocardiograms. Pitas and Venetsanopoulos reviewed order statistics filters2 and described a class of edge detectors1 based on nonlinear means and medians. They showed that a linear combination of these filters, resulting in a ranked-order filter (that operates with a selected range of ordered maximum and minimum gray level values), will further improve the noise removal capabilities.