JPS5991947A - Side edge extraction of ultrasonic diagnostic apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an edge extraction method for an ultrasonic diagnostic apparatus, and particularly to an edge extraction method for an M-mode echocardiogram.

In general, ultrasound diagnostic equipment aims to diagnose the internal tissue structure of the human body by applying ultrasound waves from an ultrasound transducer called a probe to the human body and analyzing the reflected waves from the human body. M-mode ultrasound diagnostic equipment measures the depth from the surface of the human body on the vertical axis.

Time is plotted on the horizontal axis, and the temporal changes in each part of the human body, that is, the state of movement of each part, are displayed on a CRT, etc. This displayed waveform is particularly useful for diagnosing cardiac function as an echocardiogram. Used for observation.

By the way, the person involved in this case does not only perform M-mode display in such an M-mode ultrasonic diagnostic device (,
We have already filed an application for a device that has a margin extraction function that extracts characteristic lines from an M-mode echocardiogram and can also quantify cardiac function parameters such as the size of each part of the heart (Patent Application No. 1982- No. 49828), in the above device, M
When displaying a mode echocardiogram, a method is used in which the echocardiogram is stored in an image memory for digital processing and the reconstructed echocardiogram is displayed on a TV monitor. Therefore, in this case, the screen is composed of pixels each having an address determined by depth and time, and using a TV monitor that can display gray scale, each pixel is displayed as having its own gray scale. Ru.

Here, each characteristic line in the above echocardiogram (this line is not one continuous line on the CRT, but is intermittent and has a large difference in shading, so the characteristic line can be separated from the core echogram). When extracting edges, it is often impossible to draw the edge extraction line accurately.

Therefore, the conventional method for extracting the edges of characteristic lines of an echocardiogram is to extract the edges of characteristic lines in one image memory (out of the pixels of the stored echocardiogram).

Border extraction was performed by extracting all pixels with a density equal to or higher than a certain predetermined density and connecting these pixels on the screen. However, in such conventional edge extraction methods, all pixels with a certain density or higher are extracted regardless of the degree of density, so noise components may enter the extracted pixels or the extracted pixels may may be pixels of other feature lines.

It was difficult to draw the edge extraction line accurately.

This invention was made in view of the above-mentioned problems, and prioritizes each pixel stored in the image haze memory according to its density level.

By specifying each pixel as a component point that constitutes an edge extraction line according to its priority level 1, edge extraction can be performed reliably, and areas with low density can be cut to remove noise components, etc., and sharp images can be obtained. It is an object of the present invention to provide an edge extraction method for an ultrasonic diagnostic apparatus that can obtain edge extraction lines of an echogram.

An embodiment of the present invention will be described below with reference to the drawings.

1 to 4 show a margin extraction method of an ultrasonic diagnostic apparatus according to an embodiment of the present invention, which will be explained below using an echogram of a mitral valve as an example.

First, the maximum and minimum points A1 to Es (black pixels 1 to Et in FIG. 2) of the echo image (1) of the mitral valve shown in FIG.
The operator detects the points on the RT using a light pen, cursor, etc., and designates these points as constituent points of the edge extraction line. Detection of the maximum and @A small points can also be performed automatically by differentiating the echo signal with a differentiating circuit.

Next, the specified maximum and minimum points A1-E are connected with a straight line, and above and below this straight line are surrounded by straight lines a1-a4 and -b4 in FIG. Set a belt range of a predetermined width. The width of this belt range is set because it is thought that the correct characteristic line will exist within this width, and this width is input by the operator while looking at the echogram displayed on the CRT. Alternatively, a preset value may be stored in the ROM with reference to past data and then chanted.

First, set a reference density approximately equal to the maximum density value of each pixel in the image 1 image memory, detect pixels x1 to XIO that are in the belt range and have a density equal to or higher than the reference density,
These points are designated as constituent points of the edge extraction line. Each designated pixel is indicated by an X mark in FIG. Up to this point, the pixels at the maximum and minimum points and each pixel x1 to XIO indicated by the X marks in FIG. That means that.

Furthermore, change the reference density by 1 to a lower density value, and similarly to the above, detect a pixel xtt' = Specify as the constituent points of the edge extraction line.

At this stage, the pixel Xll~! is newly designated as a constituent point of the edge extraction line! 1B is indicated by the X mark in Fig. 4. At this stage, the maximum and minimum points and each of the designated pixels Xl to °x16 are given priority to the edges over other pixels. This means that they are designated as constituent points of the edge extraction line.

Here, the a degree of a pixel in the pixel memory may be O, so the above operation ends after detecting and specifying a pixel with a certain density or higher.

Approximate by connecting nearby designated points with straight lines.

Therefore, according to the above embodiment, each pixel in one image memory is designated as a constituent point of the edge extraction line in descending order of density over other pixels, so edge extraction can be performed exactly once. In addition, by appropriately setting the minimum standard value, it is possible to remove low-density noise components and draw a clear edge extraction line of the echocardiogram on the CKT.

As described above, according to the present invention, the maximum and minimum points of the characteristic line of an echocardiogram are connected with a curved line, a belt range of one belt above and below this straight line is set, and the concentration level of the +mi element in this belt range is set. Prioritize the points and specify the first point from the highest concentration as the constituent point of the edge extraction line, and connect each of the specified constituent points with a straight line to extract the edge. Therefore, in the ultrasonic diagnostic apparatus, there is an effect that noise components can be removed and positive edge extraction can be performed without confusing pixels of other feature lines with pixels of the feature line.

[Brief explanation of the drawing]

FIG. 1 is an echocardiogram of a mitral valve, and FIGS. 2 to 4 are explanatory diagrams for explaining a margin extraction method of an ultrasonic diagnostic apparatus according to an embodiment of the present invention. . A1・A2・B1・B2・C1・C2・Dl・B2・E
l・B2... Maximum/minimum point, Xl-X111... Pixel. Agent Makoto Kuzuno −□V
ni r7 111

Claims (1)

[Claims] (1) An edge extraction method for an ultrasonic diagnostic device that extracts a characteristic line from an M-mode echocardiogram, the edge extraction method being the maximum and minimum points of the characteristic line of the M-mode echocardiogram. The specified maximum and minimum points are connected with a straight line to set a belt range with a predetermined width above and below the straight line, and pixels within the belt range and within the predetermined width are Specify pixels having a density equal to or higher than the reference density as points constituting the edge extraction line,
The feature is that the operation of sequentially lowering the reference density to a lower density and specifying the constituent points of the edge extraction line is repeated a predetermined number of times, and each constituent point specified by the repeated operation is connected with a straight line. A margin extraction method for an ultrasound diagnostic device.