JP2021053411A - Ultrasonic diagnosis of cardiac performance by single degree of freedom chamber segmentation - Google Patents

Abstract

To provide users with a simple automated tool for delineating the location of an acceptable heart chamber border.SOLUTION: An ultrasonic diagnostic imaging system has a user control by which a user positions the user's selection of a heart chamber border in relation to two machine-drawn heart chamber tracings. The user-selected border is positioned by a single degree of freedom control which positions the border as a function of a single user-determined value. This overcomes the vagaries of machine-drawn borders and their acceptance by clinicians, and the clinicians can create repeatedly-drawn borders and exchange the control value for use by others to obtain the same results.SELECTED DRAWING: Figure 5b

Description

The present invention relates to medical ultrasound diagnostic systems, in particular to the use of single degree of freedom cardiac segmentation.

Ultrasound imaging is routinely used to diagnose cardiac performance by measuring parameters such as ejection fraction and cardiac output. Such measurements require that the volume of blood in the heart chambers at various phases of the cardiac cycle be contoured in a two- or three-dimensional image of the heart chambers. In general, measurements of heart chamber volume, such as the left ventricle, are generated by tracing the endocardial boundary of the heart chamber with the user's hand. Such tracing is subject to great variability due to differences in the criteria used by different users to determine where to position the tracing. Automatic methods have been developed that attempt to automate this boundary tracing, such as the automated boundary tracing procedure described in US Pat. No. 6,491,636 (Chenal et al.).

In this technique, anatomical landmarks of the heart chamber, including the corners of the mitral valve plane and the apex of the ventricle, are positioned. One of several standard endocardial shapes verified by experts fits these landmarks. The automatically drawn boundaries can be manually adjusted by rubber banding, which allows the user to move control points on the boundaries to adjust their final position on the endocardium. This process is performed on images acquired in the end systolic and end diastolic cardiac phases. The two boundaries can be compared or subtracted to calculate ejection fraction or cardiac output. For example, U.S. Pat. No. 2009/0136109 should compare the identified endocardial boundaries (which define the medial surface of the myocardium) with the identified epicardial boundaries (which define the outer surface of the myocardium). Discloses that it produces a myocardial thickness volume. The myocardial thickness volume of US Pat. No. 2009/0136109 is hollow, and inside the hollow space is the volume of the heart chamber. A similar approach is described in European Patent Application Publication No. 2434454A2, in which the volume of the left ventricle (LV) is calculated based on the detected LV boundary. The system of EPO No. 2434454A2 uses a variety of machine learning to locate and track the heart wall. These training techniques are based on training datasets, which use a database of known cases collected for machine learning. Ejection fraction is well calculated by established methods such as the automated Simpson algorithm (rule of disks) to measure the percentage of cavity volume expelled by each contraction of the heart. be able to.

However, automated image analysis does not necessarily provide acceptable delineation of the heart chambers for all users. This lack of success is largely due to the inability of automated methods to consistently locate boundaries that any given user thinks should be placed. Most of this poor performance is due to variations between individual different users about what the anatomical landmarks that specify where the true boundaries are located.

An object of the present invention is to provide the user with a simple automated tool for delineating the location of an acceptable heart chamber boundary. Another object of the present invention is that such delineation is standardized, comparable, and capable of producing repeatable results between different users. Such results allow one clinician to approach a single value, which single value is understood by other clinicians with the same tools and communicated to other clinicians. Can be done.

According to the principles of the present invention, an ultrasonic diagnostic system and method for diagnosing cardiac performance are described. An image of the heart chamber is acquired and the image is segmented to delineate the medial and lateral boundaries of the myocardium, or, as an alternative, segmented to delineate some (3+) boundaries of the myocardium. Will be done. Preferably, this is done by an automated image analysis processor, such as using cardiac model data. User controls are provided to allow the user to define the chamber boundary location preferred by the user for one, both, or some of the segmented boundaries. User control provides a single value with a single degree of freedom, for example the percentage of distance to a segmented boundary, and the user modifies the single value to position the true chamber boundary. To do. A single value can be shared with other users of the same tool, allowing other users to get the same results with respect to other images and thus obtain standardization of ventricular boundary identification.

The block diagram which shows the ultrasonic diagnostic imaging system constructed by the principle of this invention. The block diagram which shows the detail of the boundary detection of the QLQB processor of FIG. 1 by the principle of this invention. The figure which shows the landmark of the left ventricle which is useful for boundary detection. The figure which shows the landmark of the left ventricle which is useful for boundary detection. A diagram showing a specialist-derived endocardial boundary shape used for automated boundary detection. A diagram showing a specialist-derived endocardial boundary shape used for automated boundary detection. A diagram showing a specialist-derived endocardial boundary shape used for automated boundary detection. The figure which shows the contour depiction of the epicardial boundary and the endocardial boundary in the image of the left ventricle. The figure which shows the contour depiction of the epicardial boundary and the endocardial boundary in the image of the left ventricle. A flowchart of the operation of a cardiac model that is deformable to find a cardiac boundary according to the present invention. The figure which shows the cardiac image of the left ventricle at the end diastole and the end contraction where the endocardial boundary and the interface between the trabeculae myocardium and the densified myocardium are traced. The figure which shows the cardiac image of the left ventricle at the end diastole and the end contraction where the endocardial boundary and the interface between the trabeculae myocardium and the densified myocardium are traced. FIG. 7b shows an ultrasound image of the end of systole, with a user-defined boundary located at 0% of the distance between the two heart boundaries outlined in FIG. 7b. FIG. 7b shows an ultrasound image of the end of systole, with a user-defined boundary positioned at 100% of the distance between the two heart boundaries outlined in FIG. 7b. A diagram showing a cardiac image showing a user-defined boundary located at 40% of the distance from endocardial tracing to the densified myocardial interface. A diagram showing user control with a single degree of freedom according to an embodiment of the present invention, in which user-defined boundaries are positioned relative to some boundaries of the myocardium. A diagram showing a user-defined heart chamber from a biplane image measured volumemetrically using disk rules. FIG. 5 showing a user-defined heart chamber wireframe model from a 3D ultrasound image prior to volumetric measurements.

The ultrasonic diagnostic imaging system 10 constructed by the principles of the present invention is shown in the form of a block diagram with reference to FIG. 1 first. The ultrasonic probe 12 has an array 14 of ultrasonic transducers that transmit and receive ultrasonic pulses. The array can be a one-dimensional linear array or curved array for two-dimensional imaging, or it can be a two-dimensional matrix of transducer elements for three-dimensional electron beam steering. The ultrasonic transducers in the array 14 transmit ultrasonic energy and receive echoes that are returned in response to this transmission. The transmission frequency control circuit 20 controls the transmission of ultrasonic energy in a desired frequency or band of frequencies through a transmission / reception (“T / R”) switch 22 coupled to the ultrasonic transducer of the array 14. The amount of time the transducer array is activated to transmit a signal can be synchronized to an internal system clock (not shown) or to a body function such as the cardiac cycle. The cardiac cycle waveform is provided by the ECG device 26. When the heartbeat is in the desired phase of the cycle, as determined by the waveform provided by the ECG device 26, the probe is instructed to acquire an ultrasound image. For example, this allows acquisition in end-diastolic and end-systolic cardiac phases. The frequency and bandwidth of the ultrasonic energy generated by the transmission frequency control circuit 20 are controlled by the control signal ftr generated by the central controller 28.

Echoes from the transmitted ultrasonic energy are received by the transducers in the array 14, which generate echo signals, which are analog-digital through the T / R switch 22 if the system uses a digital beamformer. ("A / D") It is coupled to the converter 30 and digitized by the A / D converter 30. Further analog beam formers can be used. The A / D converter 30 samples the received echo signal at a sampling frequency controlled by the signal fs generated by the central controller 28. The desired sampling rate dictated by sampling theory is at least twice the maximum frequency of the received passband and can be on the order of 30-40 MHz. Higher sampling rates than the minimum requirements are even more desirable.

Echo signal samples from the individual transducers in the array 14 are delayed and summed by the beamformer 32 to form a coherent echo signal. For 3D imaging using a two-dimensional array, the microbeam former located on the probe and the mainframe of the system mainframe, as described in US Pat. No. 6,013,032 (Savord) and US Pat. No. 6,375,617 (Fraser). It is preferable to separate the beam former from the former. The digital coherent echo signal is filtered by the digital filter 34. In the illustrated ultrasonic system, the transmit frequency and receiver frequency are individually controlled so that the beamformer 32 can receive a band of frequency different from the transmit band, such as the harmonic band. The digital filter 34 can bandpass filter the signal and further shift the frequency band to a lower or baseband frequency range. The digital filter can be, for example, the type of filter disclosed in US Pat. No. 5,833,613. The filtered echo signal from the tissue is coupled from the digital filter 34 to the B-mode processor 36 in the case of conventional B-mode image processing.

The filtered echo signal of the contrast agent (eg, microbubbles) is coupled to the contrast agent signal processor 38. Contrast media are often used to more clearly delineate the endocardial wall with respect to contrast media in the blood pool of the heart chamber, or as described in US Pat. No. 6,692,438, for example, myocardial microminiatures. Used to perform vascular perfusion studies. The contrast agent signal processor 38 preferably separates the echoes returned from the harmonic contrast agent by a pulse inversion technique, and the echoes resulting from the transmission of the pulses modulated in a plurality of different aspects to the image location are It is synthesized to cancel the fundamental signal component and emphasize the harmonic signal component. For example, a suitable pulse inversion technique is described in US Pat. No. 6,186,950.

The filtered echo signal from the digital filter 34 is further coupled to a conventional Doppler processor 40 for Doppler processing to generate velocity and power Doppler signals. The output signals from these processors can be displayed as planner images, are further coupled to a 3D image processor 42 that renders the 3D image, and the 3D image is stored in the 3D image memory 44. Three-dimensional rendering can be performed as described in US Pat. No. 5,720,291, US Pat. No. 5,474,073 and US Pat. No. 5,485,842, all of which are incorporated herein by reference. ..

The signals from the contrast agent signal processor 38, the B-mode processor 36 and the Doppler processor 40, and the 3D image signal from the 3D image memory 44 are coupled to a cineloop memory 48 that stores image data for each of a large number of ultrasonic images. Will be done. The image data is preferably stored in the cineloop memory 48 in a plurality of sets, and each set of the image data corresponds to the image acquired at each time. Image data within a group can be used to display parametric images showing tissue perfusion at individual times between heartbeats. The group of image data stored in the cineloop memory 48 can also be further stored in a permanent memory device such as a disk drive or digital video recorder for later analysis, for example. In this embodiment, the image is further coupled to the QLAB processor 50, where the image is analyzed to automatically outline the boundaries of the heart, thereby allowing the user to be true of the lumen of the heart. Allows the user to position the boundary when he or she is confident that it will most accurately indicate the boundary. The QLAB processor also makes quantified measurements of various aspects of the anatomical structure in the image, as described in US Patent Application Publication No. 2005/0075567 and International Publication No. 2005/054898. , Delineate tissue boundaries and margins with automated boundary tracing. The data and images generated by the QLAB processor are displayed on the display 52.

FIG. 2 shows further details of the operation of the QLAB processor to delineate user-defined heart chamber boundaries according to the principles of the present invention. The echocardiographic image is provided by the source of the cardiac image data 60, which may be one of the cineloop memory 48, 3D image memory 44 or image processor 36, 38 or 40 of FIG. The cardiac image is transmitted to the automatic boundary detection (ABD) processor 62. The ABD processor can be a fully automatic or semi-automatic (user-assisted) image processor that outlines the boundaries of the heart chamber in a cardiac image, some of which are described below. In a typical semi-automatic ABD system, the user is within the cardiac image by a pointing device such as a mouse or trackball, usually located on the ultrasound system control panel 70, or by a workstation keyboard that manipulates the cursor over the image. Specify the first landmark in. In the example of FIG. 3a, for example, the first landmark designated is the medial mitral annulus (MMA) at the bottom of the left ventricle (LV) in the view illustrated. When the user clicks on the MMA in the image, a graphic marker like the white control point indicated by the number "1" in the figure appears. The user then specifies a second landmark, in this example the lateral mitral annulus (LMA), the second landmark being indicated by the number "2" in FIG. 3b. Marked by a second white control point. A line generated by the ABD processor then automatically connects the two control points, and the line represents the mitral valve plane in this longitudinal view of the left ventricle. The user moves the pointer to the apex of the endocardium, which is the top point in the cavity of the left ventricle. When the user moves the pointer to this third landmark in the image, the template shape of the ventricular cavity of the left ventricle is as the user-operated pointer looks for the apex of the LV cavity as shown in FIG. 5a. Dynamically follow the cursor, distorting and expanding. This template, shown as a white line in FIG. 5a, is anchored by first and second control points 1 and 2 and passes through a third control point where the user has the apex of the heart. When the pointer is clicked to position the third control point 3, it is positioned at the apex of the heart. Typical LV cavity boundary templates are shown in FIGS. 4a, 4b and 4c. These templates are determined from many expert tracings of the LV endocardial border of many patients. Template 80 in FIG. 4a is an elongated template specific to many normal patients. Template 82 in FIG. 4b is more bulging in shape, which is characteristic of many patients with congestive heart failure. Template 84 is a third possibility, an even more teardrop. The template that best fits the three user-identified anatomical landmarks is selected by the ABD processor 62 and distorted to fit the three user-specified landmarks. As shown in FIG. 5a, the endocardial templates 80, 82 or 84 provide an approximate tracing of the endocardium of the LV when they are positioned and fitted to the landmarks. In the example of FIG. 5a, the black line that bisects the left ventricle follows the pointer as it approaches the apex and indicates the apex. This black line is anchored between the center of the line indicating the mitral valve plane and the apex of the left ventricle, essentially indicating the center line between the center of the mitral valve and the apex of the cavity.

When the ABD processor 62 finds the endocardial lining of the LV, it then attempts to find the epicardial boundary. This is shown in FIG. 5b, where the user moved the cursor and clicked on the apex 4 outside the dark myocardium in the image. The image of FIG. 5 is a contrast-enhanced harmonic image, where the LV cavity is full of contrast, but the contrast has not yet completely perfused the myocardium, and for that reason, in this image, The LV cavity appears very bright relative to the darker surrounding myocardium. When the user clicks on the epicardial apex, the ABD processor selects an outer or epicardial template similar to the template in FIG. 4, as described above, and performs it as shown in FIG. 5b. Fit to the epicardium. Cardiac images now show its endocardial boundary (the line connecting markers 1, 3 and 2), which is the blood pool-myocardial interface, and its epicardial boundary (markers 1, 4), which is the outermost surface of the heart. And the line connecting 2) and.

These endocardial and epicardial boundaries are contoured in the image by tracing generated by the graphics generator 66.

Instead of semi-automatic operation that requires user interaction, the ABD processor can delineate the boundaries of the LV completely automatically, as described in US Pat. No. 6,491,636 described above. As described in this context, the image processor can be configured to automatically find the mitral valve corner and apex and fit the template to an automatically located landmark. However, as shown in FIG. 6, a preferred technique for automatically contouring myocardial boundaries is the cardiac model. A cardiac model is a spatially defined mathematical description that represents the tissue structure of a typical heart.

If the heart appears in the diagnostic image, the heart model can be fitted to the heart, thereby defining the particular anatomy of the imaged heart. The process of FIG. 6 begins at 90 with the acquisition of a cardiac image. The location of the heart is located in the cardiac image at 92 by processing the image data with a generalized Hough transform. Here, the posture of the heart is not defined, and in 94, the translation, rotation and scaling misalignment of the heart in the image data is corrected by using one similarity transformation of the entire heart model. At 96, the model is modified and affine transformations are assigned to specific regions of the heart. Deformation constraints are relaxed in 98 by allowing the heart model to transform with respect to piecewise affine transformations, and shape-constrained deformable models are resized and deformed. Each part of the model fits the actual patient's anatomy as shown in the image in the acquired phase of the cardiac cycle. In this way, the model is precisely adapted to the organ boundaries shown in the cardiac image, thereby providing endocardial lining, the interface between trabeculaeted and densified myocardium, and the epicardial boundary. Define individual boundaries, including. In a preferred embodiment of such a cardiac model, the interface between the trabeculae and densified myocardium is first found, which is gently illuminated with brightly illuminated areas in the ultrasound image. It looks like a well-defined gradient to the area. Endocardial boundaries are generally less well defined in cardiac models due to the desire to be able to find variable locations of endothelial linings that are not so well defined when they appear in ultrasound images. Unlike the contrast-enhanced cardiac images of FIGS. 5a and 5b, non-enhanced ultrasound images are generally relative between the relatively high-intensity echo tissue surrounding the myocardium and the moderate-intensity myocardium. It shows a strong intensity gradient and a relatively smaller gradient between the myocardium and the blood pool in the low intensity lumen. This requires first identifying the outer myocardial border and then the inner endocardial border when diagnosing the image in the absence of contrast agent. Once the coordinates of the boundaries are found, they are transmitted to the graphics generator 66, which produces a trace that is placed on the image at the calculated position.

FIG. 7 shows two ultrasound images, the first image having both boundaries of the contoured myocardium at the end of diastole (FIG. 7a) and the second image at the end of systole. It has both traced myocardial boundaries (Fig. 7b). The densified myocardial border is traced in black and the endocardial border is traced in white in these images. Both boundaries are thus drawn and the user controls the user control on the control panel 70 to indicate the location between the two tracings that the user considers to be the true cavity boundary. To control. In one embodiment, the user operates a one-variable controller, which allows the user to already draw between the endocardial boundary already drawn and the trabeculae and densified myocardium. The true endocardial boundary can be positioned at a location that is displaced by a selected percentage of the distance between the drawn interface. FIG. 8a shows the end-systolic image of FIG. 7b when the one-variable controller is set to 0%, FIG. 8b shows the same image set to 100% by user control, in this case a white line boundary. Is located outside the myocardium. User-defined boundary visual tracing is further generated by the graphics generator 66 of the system of FIG. 1 due to overlays on the ultrasound image. The user-configured boundary is the desired location measured orthogonally to the endocardial tracing and is located at the desired percentage of the distance between the two boundaries. In these examples, the two myocardial border tracings are not shown for brevity of description, but are one possible implementation, and if desired, automatically drawn tracings are shown. Can also be done.

FIG. 8c shows a user-adjusted boundary (shown as a white line) positioned by the user such that the boundary is 40% away from the endocardial tracing towards the densified myocardial interface. Shown. This is done by moving the slider 100 to the left or right in its slot 102. As the slider is manipulated by the user, the user-controlled boundary tracing 110 is moved back and forth between the two boundaries of the myocardium. In the example of FIG. 8c, the slider 100 is illustrated as a softkey control on a display screen operated by a mouse or other user interface control, but as an alternative, the slide is a physical slider, knob, and so on. Alternatively, it can be a switch, or a trackball on a conventional ultrasonic system control panel. The user control can also be implemented as a rocker control, a toggle button, a list box or a numerical input box. If the user has a suitable percentage in many cases, this can be saved as the default value. In the example of FIG. 8c, the numerical percentage is displayed on the screen and changes as the slider 100 is moved. In addition, a magnified view 104 of the identified myocardial border portion is shown in FIG. 8d. The user clicks to the left at the point of the myocardium in the image, the part of the myocardium appears in the magnified view 104, and the user-operated boundary 110 is the myocardial boundary 106 (outer boundary) and 108 (inner boundary). Shown in between. As the user operates the slider 100, the boundary 110 is moved between the two boundaries 106 and 108 drawn by the system.

User-operated boundaries 110 can also be positioned using a range of 100% or greater, as shown in magnified view 104'. The user-defined position A (110'A) is positioned beyond the densified myocardial boundary 106'and is represented in a range of 100% or more, and the user-defined position C (110'C) is the endocardial region. It is positioned within the range of Racing 108'and is represented by a negative percentage range. The user-defined position B (110'B), located between the endocardium and the densified myocardial boundary, can be represented in the range between 0% and 100%.

Alternatively, for the convenience of the user, the user-operated (defined) boundaries can be moved for three contours, as shown in FIG. 8d (lower right view 104'). And the three contours are the same as in the example above, the densified myocardial boundary 106'and the epicardial boundary 108' corresponding to the percentages 100% and 0%, respectively, and the additional boundary 109' corresponding to the percentage value of 200%. , And the additional boundary 109'is located at the epicardial boundary. In the illustrated example, the slider 100 is adjusted to a value of 150% corresponding to the user-defined boundary 110'position between the densified myocardial boundary 106'and the epicardial boundary 109'. An alternative to the third boundary addresses the fact that beyond the densified myocardial boundary (106') there is another boundary, the epicardial boundary (shown in this example as an additional boundary). However, this epicardial boundary can also be characterized by a prominent detectable gradient within the ultrasound image, which can be used to suppress the single degree of freedom slider. This epicardial boundary can be further identified by a boundary image processor. The slider is a controller with a single degree of freedom. The user sets the position of the boundary 110 around the entire perimeter of the cavity by simply setting a single value controlled by the slider. The value can be passed on to other users, who can get the same result using the same number.

9 and 10 illustrate how the user-defined heart chamber boundaries of the present invention can be used to measure parameters (eg, cardiac output and ejection fraction). In the perspective view of FIG. 9, two user-defined boundaries 210 and 212 of the simultaneously acquired biplane image of the LV are shown on the base 220 representing the mitral valve plane. The apex marker at the two boundaries is shown at 230. In this example, the image planes of the two biplane image boundaries are orthogonal to each other. The volume within the two boundaries 210 and 212 is mathematically divided into a plurality of spaced planes 222 parallel to the basal plane 220. These planes intersect the left side of the boundary 210 as shown by a and a and the right side of the boundary 210 as shown by c and c. The plane intersects the front side of the tracing 212 as shown in b and b.

The ellipse is mathematically fitted to the four intersection points a, b, c, d of each plane 222 as shown in FIG. Curves or splines other than ellipses can be used, including arcs and irregular shapes, but ellipses provide the advantage that Simpson's rule is clinically valid when performed by ellipses. To do. The volume of the disc defined by the plane 222 and the ellipse can be calculated by the disc rules to calculate the volume of the LV.

FIG. 10 shows a wireframe model composed of user-defined boundaries of a three-dimensional heart chamber image. Horizontal sections 232, 234, and 236 of the wireframe are boundaries that intersect vertical boundary sections 210, 212 at intersection points a, b, c and d. The horizontal section is parallel to the mitral valve plane base 220. The volume in the wireframe can be determined by modified disk rule calculations or other volumetric calculation techniques. When the volume calculated for the end-systolic phase image shown in FIG. 9 or FIG. 10 is subtracted from the calculated volume for the end-expansion image and divided by the same end-expansion volume, the result is the ejection fraction. It is a calculated value of.

Other modifications described above will be readily apparent to those skilled in the art. Instead of percentage quantification, user-defined boundaries can be positioned at incremented distances from manually or automatically traced boundaries. For example, the slider can be calibrated so that the position of the user-defined boundary is offset from the reference boundary by a user-determined distance in millimeters. Instead of using two traced boundaries, a user-defined boundary can be positioned with respect to a single boundary tracing, or from the two boundaries at an interpolated offset. Can be done. User-defined boundaries can also be positioned using a range of 100% or greater, and user-defined boundaries can be within endocardial tracing, or epicardial or densified myocardial boundaries. It can be positioned beyond.

Claims (15)

An ultrasound diagnostic imaging system that determines the boundaries of the lumen of the heart in ultrasound images.
Source of cardiac image data and
A boundary detection processor that responds to the cardiac image data and that identifies at least the inner and outer boundaries of the myocardium in the cardiac image data.
A user control that allows the user to indicate a user-defined heart chamber boundary with respect to the medial boundary and the lateral boundary.
A heart chamber boundary visualizer coupled to the user controller and the boundary detection processor, with respect to at least one of the inner boundary and the outer boundary identified by the boundary detection processor in the cardiac image data. A heart chamber boundary visualizer that positions the user-specified heart chamber boundary,
An ultrasonic diagnostic imaging system that has. The ultrasonic diagnostic imaging system of claim 1, wherein the user controller is further configured to adjust a single degree of freedom that is variable for the user-defined location of the chamber boundary. The medial boundary further comprises an endocardium, or myocardial-blood pool interface.
The ultrasonic diagnostic imaging system according to claim 2, wherein the outer boundary further comprises an epicardium or an interface between the trabeculae myocardium and the densified myocardium. The ultrasonic diagnostic imaging system according to claim 2, wherein the user controller further includes a slider, a knob, a switch, a trackball, a rocker control, a toggle button, a list box or a numerical input box. The ultrasonic diagnostic imaging system according to claim 4, wherein the user controller further has softkey control or physical control. The ultrasonic diagnostic imaging system of claim 2, wherein the variable single degree of freedom is calibrated on the one hand of a percentage or millimeter, and the percentage is related to the distance to the inner and outer boundaries. The ultrasonic diagnostic imaging system according to claim 1, wherein the source of the cardiac image data further includes a memory device having a two-dimensional cardiac image. The ultrasonic diagnostic imaging system according to claim 7, wherein the cardiac image data includes a view of the left ventricle. The ultrasonic diagnostic imaging system according to claim 1, wherein the boundary detection processor further includes a semi-automatic cardiac boundary image processor. The ultrasound diagnostic imaging system of claim 9, wherein the semi-automatic cardiac boundary image processor is coupled to the user controller and further responds to user input defining landmarks in the cardiac image. The ultrasonic diagnostic imaging system according to claim 1, wherein the boundary detection processor further comprises an automatic cardiac boundary image processor. The ultrasonic diagnostic imaging system according to claim 1, wherein the automatic cardiac boundary image processor can operate to identify the boundary of the myocardium in the cardiac image data without user input. A graphics generator that is coupled to the boundary detection processor to generate visible traces of the inner and outer myocardial boundaries.
A display coupled to the source of the cardiac image data and to the graphics generator to display the cardiac image along with the generated display trace of the inner and outer myocardial boundaries.
The ultrasonic diagnostic imaging system according to claim 1, further comprising. A graphics generator that is coupled to the chamber boundary visualizer to generate a user-defined visual trace of the chamber boundary.
A display that is coupled to the source of the cardiac image data and to the graphics generator to display the cardiac image along with a display trace of the cardiac boundary.
The ultrasonic diagnostic imaging system according to claim 1, further comprising. The ultrasonic diagnostic imaging system according to claim 6, wherein the range of percentages of single degrees of freedom is less than 0%, greater than 100%, or both.

Images