Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
2002-05-06
2004-03-16
Getzow, Scott M. (Department: 3762)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
Reexamination Certificate
active
06708055
ABSTRACT:
1.0 BACKGROUND OF THE INVENTION
1.1. Field of the Invention
The invention concerns systems for automatically analyzing echocardiographic digital images of the heart, especially two-dimensional images acquired using apical four-chamber view of the heart. The systems preferably employ a collection of matched filters on the images to automatically locate and measure features of the heart.
1.2. Description of Related Art
Two-dimensional ultrasonic imaging is used as an important non-invasive technique in the comprehensive characterization of a number of body organs. In ultrasonic imaging, a sound pulse is sent along a ray from a transducer towards the organ that is being imaged. The pulse is attenuated and reflected when it hits a medium with an acoustic impedance different from that of the medium in which the pulse is traveling. The time the sound pulse takes in transit is a measure of the distance of the medium interface from the transducer. The amount of energy that is reflected is a measure of the difference in acoustic impedance across the interface. Since the energy of the pulse diminishes as it travels, post-processing of the reflected signal includes time/gain control that compensates for the attenuation of the signal over time. Assuming the pulse travels at a single speed in the body, and by taking rays uniformly distributed across a given plane, a two-dimensional record of the received energy in spatial (Cartesian, polar) coordinates can be used to present a cross-sectional view of the imaged organ.
Echocardiography is the application of ultrasonic imaging to the heart. Echocardiography has experienced widespread acceptance in the evaluation of cardiac disease, structure, and function of the heart. This acceptance is in large part due to its non-invasive nature, and to its real-time capability for observing both cardiac structure and motion. Using echocardiography, quantitative information may be obtained concerning cardiac anatomy, chamber diameter and volume, wall thickness, valvular structure, ejection fraction, etc. (Weyman, 1994).
The real-time capability of echocardiography may be used to measure variations in the shape of heart structures throughout the cardiac cycle (Weyman et al., 1984). These analyses require the complete determination of inner (endocardial) and outer (epicardial) boundaries of the heart wall, particularly of the left ventricle. Present evidence indicates that sensitive detection of ischemic disease with two-dimensional echocardiography requires knowledge of the endocardial border on echocardiographic frames throughout the cardiac cycle (Weyman et al., 1984).
Because both global and regional left ventricular function are major variables used to determine prognosis in cardiac disease, there is considerable interest in the ability to quantify function indexes from echocardiographic images. Presently, such indexes, e.g., left ventricular chamber volume and left ventricular ejection fraction, are calculated from observer-defined cardiac boundaries traced with either a light pen, a digitizing tablet, or either a mouse, trackball, or other suitable computer input device. Tracing of endocardial borders on two-dimensional echocardiograms, however, is tedious and the selected borders may be highly subjective. Indeed, in most systematic studies, substantial intra-observer and inter-observer variability has been found in such observer-defined cardiac boundaries (Weyman et al., 1984). An echocardiogram is a generic term for an image formed using ultrasound. It contrasts with images produced by x-ray, magnetic resonance, or other techniques (which are typically referred to as “radiologic medical tomographic images”). An echocardiogram may also be referred to as a “sonogram,” and is an image that is formed using ultrasound as the type of wave producing the images.
One advantage of ultrasound (sonograms) as an image formation technique, is that images can be formed very rapidly, meaning typically 10 to possibly 80 times per second, therefore, since the heart has one beat per cycle, it is feasible to catch up to typically 30 to 50 images in this series through one heart cycle. The registration point moves, or at least can move on every frame and therefore, in both the short axis and the long axis, the inventors have elected to locate that point on each frame. It can not be assumed to stay fixed. Measurements may be made on each frame (i.e. a single image from a series of images) and adjusted to the size of the heart by taking into account the scale factor of the image; in other words, how many pixels represent a centimeter can be changed by the operator. The scale factor relating the number of pixels in one centimeter in the image must be known so that the measurements can be adjusted to the true size of the heart. The “registration point” is the origin of the system of coordinates.
Manually defining such boundaries becomes increasingly labor intensive when the analysis of a complete cardiac cycle is needed to provide a description of the systolic and diastolic wall motion pattern, or when a number of echocardiographic frames have to be processed in order to obtain an extended time-history of cardiac function. It is therefore desirable to automate as much as possible the determination of boundaries of echocardiographic images, as well as other structural features. Automated definition of the boundaries and features would improve the reliability of analyses by eliminating the subjectivity of manual tracing.
In the past several years, advances in computer data processing technology have allowed the application of several different automatic boundary detection methods to echocardiographic images (Conetta et al., 1985). However, most researchers have had difficulties with image enhancement and boundary detection with echocardiographic images because of the low signal-to-noise ratio and large discontinuities in such images. Thus, automated border detection has been reported in two-dimensional echocardiographic images, but only when the images have been of good quality and certain smoothing techniques have been employed prior to edge detection in order to render the endocardial edge more continuous. An overview of the field is set forth in Kerber (1988).
U.S. Pat. No. 5,797,396 to Geiser and Wilson (1998) (specifically incorporated herein by reference in its entirety) describes a method for the automated analysis of short-axis views of a heart.
1.3 Deficiencies in the Prior Art
What is lacking in the prior art is a method for automatically determining quantitative characteristics of ultrasonic images, especially long-axis echocardiographic images that provide details of the apical four-chamber view of a heart. In particular, there is also a need for a method that can automatically determine key regions of an imaged structure and approximate the borders of such a structure.
2.0 SUMMARY OF THE INVENTION
The present invention is directed at making automated measurements on echocardiographic 2-D images acquired using the apical 4-chamber view. In a broad aspect, the invention comprises a system for processing 2-D digital images of the heart to analyze the structure and functioning of the heart. These features include the right and left ventricles, the septum, the mitral valve apparatus and the myocardium.
The invention provides a method for automatically analyzing a long-axis image of a heart, the method comprising the steps of: (a) generating an image frame of the myocardium, interventricular septum, and mitral valve annulus of the heart, the image frame comprising a plurality of rows and a plurality of columns of pixels in a digital format; (b) determining an approximate position of the interventricular septum from the diagnostic image comprising: (i) passing a filter through the image to determine a maximum mean pixel intensity, wherein the maximum is a first approximate position of the interventricular septum; (ii) defining a second approximate position of the interventricular septum with a series of straight line filters; (iii) obtaining a best fit line through th
Geiser Edward A.
Wilson David C.
Clarke Dennis P.
Getzow Scott M.
Miles & Stockbridge P.C.
University of Florida
LandOfFree
Method for automated analysis of apical four-chamber images... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for automated analysis of apical four-chamber images..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automated analysis of apical four-chamber images... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3243874