JPH05167927A - Image processor - Google Patents

Abstract

PURPOSE:To obtain a transportation outline for blood vessels and a core image from a digital subtraction angiographic image even when the blood vessels are complecatedly branched by extracting the maximum point of a gradient value profile relating to a vertical direction, i.e., a gradient direction, to a blood vessel direction. CONSTITUTION:A profile forming part 7 forms a change in a gradient value, i.e., a gadient value profile, around each picture element point by using gradient values and gradient directions of all picture element points in an angiographic image inputted from a gradient value/gradient direction calculating part 6. The formed profile is sent to a maximum point extracting part 8, which extracts the maximum points P0, P1 of the profile to use them as the outline points of a blood vessel. The threshold range of a density value or the gradient value may be previously experimentally and empirically set up or properly set up by an operator and the obtained outline points are displayed on an external image display device 9.

Description

Detailed Description of the Invention

[0001]

The present invention relates to a DSA (Digital Subt
raction angiography) An image processing apparatus for extracting a contour point or a core point of a blood vessel from an image.

[0002]

2. Description of the Related Art A DSA image, which is one of blood vessel-enhanced images, is obtained by performing a difference process between an X-ray transmission image before injection of a contrast agent and an X-ray transmission image after injection of a contrast agent. The appearance of this DSA image has significantly improved the efficiency and diagnostic ability of blood vessel diagnosis. At present, attempts are being made to obtain various blood vessel images using this DSA image. One of them is to try to capture a blood vessel wall (blood vessel contour line) or a blood vessel core line (center line). The main method is a binarization method or a tracking method or a tracking method.

In the binarization method, an appropriate threshold value is set for a DSA image, the threshold value is compared with the density value of each pixel, and the pixel value is polarized before and after the threshold value to create a binary image. , A method of extracting a boundary portion between a blood vessel image portion and a background portion of the binary image to obtain a contour line image of the blood vessel.

Further, in the case of the tracking method, the point of the maximum density value regarding each point existing on the circumference of a predetermined radius centering on a certain designated point on the DSA image is set as the center point of the blood vessel, and the center point is set first. In this method, the core line image is obtained by repeatedly performing this operation by regarding it as the designated point of (1) and adding continuity to each obtained center point. The zero crossing method (Zero Crossing) is applied to the density profile of the DSA image in the direction perpendicular to the core line direction of the core line image. In the zero crossing method, the sign of the second derivative is the blood vessel part. This is a method of utilizing the inversion at the boundary between the background portion and the background portion and smoothing the DSA image with the weight of the Gaussian distribution function, and applying the Laplacian to obtain the zero crossing.

[0005]

However, in the binarization method, when a binary image is obtained, the setting of a threshold value for distinguishing a blood vessel portion from other portions is very delicate, and depending on the level of this threshold value. As a result, the range of the blood vessel image on the binary image varies, and as a result, the obtained blood vessel contour image is not accurate. In addition, regarding the threshold value setting, an operator often intervenes to determine an accurate value, but since this needs to be determined by trial and error, it requires a great deal of effort and is not practical. Absent.

On the other hand, in the tracking method or the tracking method, when the core line image is obtained, if the tracking of the next designated point fails in the middle of the tracking of the center point, the subsequent center point is not extracted and an accurate core line image is obtained. I can't get it. Furthermore, if you do not track the specified point for every blood vessel on the image,
There will be blood vessels that cannot be extracted. Therefore, as a natural result, the obtained blood vessel contour image is not accurate. Therefore, an object of the present invention is to provide an image processing apparatus capable of quickly obtaining an accurate blood vessel contour line image and a blood vessel core line image from a DSA image.

[0007]

An image processing apparatus according to the present invention comprises means for calculating a horizontal density gradient value and a vertical density gradient value for each pixel point of a blood vessel emphasized image, the horizontal density gradient value and the vertical density gradient value. Means for obtaining a gradient value for each pixel point and a gradient line extending from each pixel point in a direction perpendicular to the blood vessel direction using a density gradient value; and a gradient value for the pixel point on the gradient line or the blood vessel-enhanced image And a means for extracting the maximum point or the maximum point of the profile as a contour point on the contour line of the blood vessel or a core point on the core line of the blood vessel. ..

[0008]

According to the present invention, the density gradient value in the scanning line direction and the density gradient value in the direction perpendicular to the direction are calculated for each pixel point of the blood vessel emphasized image, and the horizontal gradient value and the vertical gradient value are calculated. And obtain a gradient value for each pixel or a gradient line extending from the pixel point in the direction perpendicular to the blood vessel direction, and obtain a profile of the gradient value or the density value of the blood vessel-enhanced image with respect to the pixel point on the gradient line. Creating a maximum point or maximum point of the profile, and obtaining the maximum point or the maximum point is the contour point image on the contour line of the blood vessel or the core point on the core line of the blood vessel to obtain the blood vessel contour line image or the blood vessel core line image You can

[0009]

Embodiments will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the arrangement of an image processing apparatus according to the first embodiment of the present invention, and FIG. 2 is an example of a DSA image output from the DSA image creating section shown in FIG. FIG.

The DSA image producing apparatus 1 makes a difference between an X-ray transmission image obtained before injection of a contrast agent and an X-ray transmission image obtained after injection of a contrast agent into a blood vessel from an X-ray imaging apparatus (not shown). DSA (Digital Subtractio) with emphasis on blood vessels
n Angiography) image is created and the DSA image I0 is supplied to the image processing apparatus according to the present invention which is downstream. It should be noted that this part does not have to be a creation device as long as it supplies the DSA image I0 to the image processing device, and may be, for example, a storage device that stores the already obtained DSA image I0. ..

The image processing apparatus 2 is an apparatus for obtaining a blood vessel contour image which is the result of the apparatus of this embodiment. Here, the terms used below will be briefly described. That is,
The "horizontal direction" means a direction parallel to the scanning line direction of the DSA image I0, and the "vertical direction" means the DSA image I0.
Means a direction perpendicular to the scanning line direction, and the "gradient value at the point P" means the rate of change of the density value in the DSA image I0 with respect to the point P, in other words, the degree of inclination of the change of the density value. The "gradient direction at the point P" means a direction from the point P toward the center of the blood vessel, in other words, a direction perpendicular to the blood vessel direction regarding the point P,
The “gradient line regarding the point P” is a line parallel to the gradient direction and passing through the line point P.

The image processing apparatus 2 includes a gradient processing section 3 for receiving the DSA image I0 from the DSA image creating section 1, a profile creating section 7 for receiving the output of the gradient processing section 3, and an output of the profile creating section 7. And a maximum value extraction unit 8 for receiving the maximum value.

The gradient processing unit 3 is a processing unit which inputs the DSA image I0 and individually obtains the gradient value and the gradient direction for all the pixel points of the DSA image I0. Gradient processing unit 3
Is a vertical direction differential processing unit 4 and a horizontal direction differential processing unit 5 for calculating a vertical direction gradient value and a horizontal direction gradient value for each pixel point, and a vertical direction gradient value and a horizontal direction obtained by these processing units 4 and 5. And a gradient value / gradient direction calculation unit 6 that calculates a gradient value and a gradient direction for each pixel point using the gradient value. The gradient value is the rate of change of the density value as described above, that is, it is the same as the differential value based on the horizontal direction (scanning line direction) or the vertical direction. Figure 3
FIG. 6 is a diagram showing a vertical gradient value, a horizontal gradient value, a gradient value obtained from them, and a gradient direction regarding a certain pixel point P0 of the DSA image I0. The vertical direction differentiation processing unit 4 uses the vertical direction DS passing through the pixel point P0 (x, y).
A pixel point P0 of the pixel value (density value) column of the A image I0 (x,
The pixel point P0 (x, x,
Compute the vertical gradient value hy (x, y) for y). The horizontal direction differential processing section 5 performs a differential process on the pixel point P0 (x, y) of the pixel value (density value) column of the horizontal DSA image I0 passing through the pixel point P0 (x, y) to obtain the pixel point P0. Horizontal gradient value hx with respect to (x, y)
Calculate (x, y). Then, the gradient value / gradient direction calculation unit 6 expresses the vertical gradient value hy (x, y) and the horizontal gradient value hx (x, y) regarding the pixel point P0 (x, y) by the following equation (1). ), The gradient value G (x, y) regarding the pixel point P0 (x, y) is obtained.
y) is calculated, and the gradient direction θ (x, y) is calculated using the equation (2). G (x, y) = {hx (x, y) ² ＋ hy (x, y) ² } ^1/2 … (1) θ (x, y) = tan ^-1 (hx (x, y) / hy (x, y))… (2)

The gradient direction θ (x, y) relating to the pixel point P0 (x, y) obtained here is always the point P0 (x, y).
To the center of the blood vessel, that is, the direction perpendicular to the long axis direction of the blood vessel B. This principle will be briefly described with reference to FIGS. Note that FIG. 4 shows the horizontal direction D1, the vertical direction D2, and the gradient direction θ with respect to the pixel point P0.
FIG. 5 is a diagram showing (x, y), FIG. 5 is a diagram showing a density profile in the horizontal direction D1, and FIG. 6 is a diagram showing a density profile in the vertical direction D2. As shown in FIG. 4, it is assumed that the blood vessel direction BD has an inclination of the angle θ'at the point P0 with respect to the horizontal direction D1. At this time, as shown in FIGS. 5 and 6, the maximum density value Nmax of the density profile in the horizontal direction D1 (usually at the blood vessel center) and the maximum density value Nmax of the density profile in the vertical direction D2 are both Since the density values are in the vicinity of the same blood vessel, the maximum diameters in each direction are substantially the same, and thus are substantially the same. That is, the difference amount between the density value of the contour portion of the blood vessel and the density value of the central portion of the blood vessel is substantially the same.
However, the range included in the blood vessel B in the horizontal direction D1 is P0 to P1 while the blood vessel B is in the vertical direction D2.
The range P0 to P2 included in the above becomes shorter than P0 to P1 depending on the angle θ '. Therefore, the magnitude relationship between the gradient value in the vertical direction and the gradient value in the horizontal direction with respect to the pixel point P0 (x, y) conforms to the angle θ ', and a composite vector obtained from each gradient value having each direction component. , The gradient direction θ (x, y) is always substantially the same as the direction from the point P0 (x, y) to the center of the blood vessel.

The profile creating unit 7 uses the gradient values and the gradient directions for all the pixel points of the DSA image input from the gradient value / gradient direction calculating unit 6 to determine the gradient value in the gradient direction around each pixel point. Create a change, or gradient value profile. FIG. 7 is a diagram showing the gradient value profile for the gradient line in the gradient direction D12 passing through the pixel point P0 shown in FIG. As described above, this gradient value profile is obtained for all pixel points, and these gradient value profiles are output to the maximum value extraction unit 8.

The local maximum value extraction unit 8 extracts local maximum points of those gradient value profiles. The obtained maximum point P0 P1 becomes the contour point of the blood vessel B. This is because the density value change rate is large near the blood vessel contour part, that is, the gradient value is larger than that inside the blood vessel or outside the blood vessel. Furthermore, the change in the gradient value is greatest when the direction is the gradient direction θ, which is the direction perpendicular to the blood vessel direction,
Along with that, the maximum points can be easily extracted. In extracting the maximum point, a predetermined condition may be imposed and only the maximum value satisfying the condition may be extracted. In this case, maximum points other than the blood vessel contour points can be excluded. The predetermined condition is, for example, that a threshold range is set for the density value or gradient value of the point, and the contour point is extracted only when included in the threshold range. The threshold value range for the density value or the gradient value in this case may be set experimentally or empirically in advance, or may be set appropriately by the operator. The obtained maximum points, that is, contour points, are output to the external image display device 9, the image storage device, or the like.

As described above, according to this embodiment, by extracting the maximum point of the gradient value profile in the direction perpendicular to the blood vessel direction (gradient direction), the accurate contour point of the blood vessel can be obtained. Furthermore, by extracting the gradient value profile and the maximum point for all the pixel points, an unextracted blood vessel contour line does not occur even when the blood vessel has a complicated branch, unlike the tracking method.

Next, a second embodiment will be described. In this embodiment, a blood vessel core line image is obtained by using the blood vessel contour line image obtained in the first embodiment. FIG. 8 is a block diagram showing the configuration of the apparatus of this embodiment, and FIG. 9 is a diagram showing a method of obtaining points (core points) on the vascular core line according to this embodiment.

As shown in FIG. 8, the core point extraction processing unit 10
Is for inputting a contour line image from the image processing unit 2 and outputting a blood vessel core line image to the image display device 9 or the like. The core point extraction processing unit 10 extracts the core points of blood vessels by the method shown in FIG.
That is, for each contour point on the contour line input from the image processing unit 2, the profile of the value (density value) of each pixel point of the DSA image on the connection line from the point to the contour point opposite to the point is image processed. It is created by using the gradient direction of the point obtained in the part 2, and the point having the maximum density value in the profile is extracted. The point having the maximum density value is the point on the blood vessel core line. In this way, by extracting core points for all contour points on the contour line image, a blood vessel core line image can be obtained. If continuity of each core point is required, it can be easily obtained by applying a generally used tracking method to this blood vessel core line image.

As described above, according to this embodiment, by extracting the maximum point of the density value profile in the direction perpendicular to the blood vessel direction (gradient direction), the accurate center point of the blood vessel can be obtained. Furthermore, by extracting the maximum points (core points) corresponding to all the contour points on the contour line image obtained in the first embodiment, even if the blood vessel has a complicated branch, As described above, the unextracted blood vessel core line does not occur. Furthermore, in the case of reconstructing a three-dimensional image of a blood vessel from a DSA stereo image, in the present embodiment, the blood vessel core line images are obtained for every two stereo images for the left and right eyes, and the two blood vessel core line images are obtained. By obtaining the positional correspondence between the core lines, the three-dimensional coordinates of the blood vessel can be obtained.

The present invention is not limited to the above embodiments, but can be modified in various ways. For example, first, second
The original image supplied to the image processing apparatus of the embodiment is DSA
Although it is an image, it may be an image in which blood vessel enhancement processing is performed.

[0023]

As described above, according to the present invention, D
A density gradient value in the scanning line direction and a density gradient value in a direction perpendicular to the direction are calculated for each pixel point of the SA image, and the gradient value for each pixel is calculated using the horizontal gradient value and the vertical gradient value. Alternatively, when a gradient line extending from the pixel point in a direction perpendicular to the blood vessel direction is obtained, a profile of the gradient value or the density value of the blood vessel-enhanced image with respect to the pixel point on the gradient line is created, and the maximum point or the maximum point of the profile is created. An accurate blood vessel contour line is obtained from the DSA image by extracting points and obtaining the blood vessel contour line image or the blood vessel core line image by using the maximum point as the maximum point as the contour point on the contour line of the blood vessel or the core point on the core line of the blood vessel. It is possible to provide an image processing apparatus that can quickly obtain an image and a blood vessel core line image.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of a DSA image output from the DSA image creation unit shown in FIG.

FIG. 3 is a diagram illustrating a gradient value and a gradient direction in the gradient processing unit shown in FIG.

4A and 4B are diagrams showing a horizontal direction and a vertical direction with respect to a pixel point P0 in explaining the principle of obtaining the gradient direction shown in FIG.

FIG. 5 is a diagram showing a horizontal density profile shown in FIG. 4;

FIG. 6 is a diagram showing a vertical density profile shown in FIG. 4;

FIG. 7 is a diagram showing a gradient value profile regarding the gradient line shown in FIG. 3;

FIG. 8 is a block diagram showing the arrangement of an image processing apparatus according to another embodiment of the present invention.

9 is a diagram showing a center point extracting process in the center point extracting device shown in FIG. 8;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... DSA image creation unit, 2 ... Image processing device, 3 ... Gradient processing unit, 4 ... Vertical direction differentiation processing unit, 5 ... Horizontal direction differentiation processing unit, 6 ... Gradient value / gradient direction calculation unit, 7 ... Profile creation unit , 8 ... local maximum value extraction unit, 9 ... image display device.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location // G06F 15/70 335 Z 9071-5L

Claims (2)

[Claims] 1. A means for calculating a horizontal density gradient value and a vertical density gradient value for each pixel point of a blood vessel enhanced image; and a means for calculating each of the pixel points by using the horizontal density gradient value and the vertical density gradient value. Means for obtaining a gradient value and a gradient line extending in the direction perpendicular to the blood vessel direction from each of the pixel points, means for creating a profile of the gradient value relating to the pixel points on the gradient line, and a maximum point of the profile being the contour of the blood vessel. And a means for extracting the contour points on the line. 2. A means for calculating a horizontal density gradient value and a vertical density gradient value for each pixel of a blood vessel emphasized image, and a blood vessel direction of each pixel point using the horizontal density gradient value and the vertical density gradient value. A means for obtaining a gradient line directed in the vertical direction, a means for creating a profile of density values on the blood vessel-enhanced image with respect to pixel points on the gradient line, and a maximum point of the profile as a point on the core line of the blood vessel A device for extracting a blood vessel core line, which comprises: