JP2011235010A - X-ray diagnosis apparatus, image processor and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve a lengthy image joined with a plurality of images.SOLUTION: This X-ray diagnosis device 10 comprises: an X-ray irradiation unit 1; an X-ray detection unit 2; a joining position-setting unit 32 which sets a position at which the plurality of images are joined by being overlapped with one another at a part of each image; a pixel position correction unit 33 which corrects a pixel position of an overlapped region included in at least one image; and a lengthy-image creation unit 34 which creates the lengthy image which is joined in the overlapped region at which the correction is performed. The lengthy image is outputted to an output device 4 which is composed of a monitor or the like.

Description

  The present invention relates to an X-ray diagnostic apparatus, an image processing apparatus, and a program, and particularly to an image joining technique for long imaging.

  Long-length imaging refers to an imaging technique in which a continuous subject exceeding the detection region of one X-ray detector is imaged a plurality of times by a plurality of detectors or one detector and joined (or also called connection). Of these imaging methods, when performing long imaging using a single detector, the detector is moved multiple times, imaging is performed, a plurality of captured X-ray images are stored, and the stored plurality of images are joined. And create one long image. Here, as a technique for joining a plurality of images into a single image, for example, Patent Document 1 extracts a region where a spine is imaged by image processing, searches for a highly correlated position in the spine region, and joins the images. A medical image processing apparatus for deriving a position is disclosed.

JP 2008-067916 A

  In the above-mentioned Patent Document 1, when there is no movement of the body of the subject, alignment can be performed, but there is an image distortion caused by, for example, body movement of the subject in the overlapping region of a plurality of consecutive images. Could not align the position of the image area that was displaced due to body movement. Therefore, there is a problem that a boundary (edge) such as a bone or a diaphragm is displayed twice in the overlap region due to the movement of the subject's breathing or the like, and as a result, a long image (which is difficult to diagnose the overlap region) There is also a problem that it is also referred to as a bonded image).

  The present invention has been made in view of the above problems, and even when there is a movement of the subject's body in the overlapping region of a plurality of continuous images, the length in which the images are not displayed twice in the overlapping region. An X-ray diagnostic apparatus, an image processing apparatus, and a program for creating a scale image are provided.

  In order to solve the above problems, an X-ray diagnostic apparatus according to the present invention includes an X-ray irradiation unit that irradiates X-rays, and an X-ray that detects X-rays emitted from the X-ray irradiation units and converts them into electrical signals. Detection means, joining position setting means for setting positions where a plurality of images based on the electrical signals obtained from the X-ray detection means are joined by overlapping a part of each image, and overlapping included in at least one image A pixel position correcting unit that corrects a pixel position of the alignment region; and a long image generating unit that generates a long image obtained by joining the plurality of images in the corrected overlapping region. And

  The image processing apparatus according to the present invention includes a plurality of images obtained by the medical image pickup apparatus and at least one of the joining position setting means for setting a joining position by overlapping a part of each image. Pixel position correcting means for correcting the pixel position of the superimposed area, and a long image creating means for creating a long image obtained by joining the plurality of images in the corrected overlapping area. It is characterized by.

  In addition, the image processing program according to the present invention includes a step of setting a position where a plurality of images obtained by the medical imaging apparatus are joined by overlapping a part of each image, and a superposition included in at least one image A step of correcting the pixel position of the region and a step of creating a long image obtained by joining the plurality of images in the superimposed region where the correction has been performed are performed.

  According to the present invention, in order to connect a plurality of images after correcting the pixel position in the overlap region, the subject, for example, an edge portion such as a bone is matched in the overlap region. Thus, the image quality of the long image can be improved.

The schematic diagram which shows the structural example of the X-ray diagnostic apparatus which concerns on this embodiment The flowchart which shows the flow of the long image creation process by the X-ray diagnostic apparatus which concerns on this embodiment. The flowchart which shows the flow of the pixel position correction process which concerns on this embodiment. Explanatory drawing which shows the derivation | leading-out process of an overlapping area | region Explanatory drawing which shows the setting process of a point of interest and a region of interest based on a small region Explanatory drawing showing the scanning area (scanning direction of the area of interest) Explanatory drawing which shows pixel position correction processing Conceptual diagram of projective transformation

  Hereinafter, embodiments of the X-ray diagnostic apparatus of the present invention will be described with reference to the drawings. In the following description, the same components are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a schematic diagram illustrating a configuration example of an X-ray diagnostic apparatus 10 according to the present embodiment. The X-ray diagnostic apparatus 10 includes an X-ray generator 1 that emits X-rays, an X-ray detector 2 that detects X-rays emitted from the X-ray generator 1 and converts them into electrical signals, and an X-ray detector 2. The image processing device 3 for creating a long photographic image based on the electrical signal indicating the intensity of the X-rays obtained from the above, and the output device 4 for outputting (displaying) the long photographic image. The image processing device 3 receives an image signal from the X-ray detector 2 and stores it, and a joining position setting unit 32 that detects and sets joining positions of a plurality of images stored in the image storage device 31. A pixel position correcting unit 33 that corrects a pixel position of a region where a plurality of images to be joined overlap (hereinafter referred to as an “overlapping region”) at the joint position set by the joint position setting unit 32; A long image creation unit 34 for creating a long image obtained by joining the plurality of images.

  The image storage device 31 is configured by a device that temporarily or fixedly stores an electrical signal corresponding to the intensity of transmitted X-rays of a subject, such as a magnetic disk, nonvolatile, or volatile memory.

  The joint position setting unit 32, the pixel position correcting unit 33, and the long image creating unit 34 are an image processing program that executes these functions, an arithmetic / control device (for example, a CPU), a storage device (for example, a memory), and an input / output The functions of the image processing program are realized by cooperating with hardware composed of devices.

  Next, based on FIG. 2 thru | or FIG. 8, the flow of the process which produces a long image with the X-ray diagnostic apparatus which concerns on this embodiment is demonstrated. FIG. 2 is a flowchart showing a flow of long image creation processing by the X-ray diagnostic apparatus according to the present embodiment. FIG. 3 is a flowchart showing the flow of pixel position correction processing according to the present embodiment. FIG. 4 is an explanatory diagram showing the process for deriving the overlapping area. FIG. 5 is an explanatory diagram showing a process for setting a point of interest and a region of interest based on a small region. FIG. 6 is an explanatory diagram showing a scanning region (scanning direction of the region of interest). FIG. 7 is an explanatory diagram illustrating pixel position correction processing. FIG. 8 is a conceptual diagram of projective transformation.

  In the following description, the X-ray diagnostic apparatus 10 captures two images (image A and image B in FIG. 4) by shifting the X-ray detector 2 to different positions while having an overlapping portion with respect to the subject. Then, a process for joining these two images to create a long image will be described as an example. Image A and image B are images taken with overlapping portions along the body axis direction of the subject. The lower region (foot side) of image A and the upper region (head side) of image B are , Each includes a region where the same part of the subject is imaged. Overlapping and joining the regions where the same part is imaged creates a long image. Hereinafter, it demonstrates along the step of FIG.

(Step S1)
X-ray imaging is performed (S1). The X-ray diagnostic apparatus 10 overlaps the same part of a subject (not shown) located between the X-ray generator 1 and the X-ray detector 2 to overlap the X-ray generator 1 and the X-ray detector 2. Are moved relative to each other, and continuous shooting is performed, a plurality of images (corresponding to the images A and B in FIG. 4 in this embodiment) are captured, an electric signal is output, and transmitted to the image storage device 31. The image storage device 31 stores images A and B based on the electrical signal.

(Step S2)
The joining position setting unit 32 sets the joining position of the images A and B stored in the image storage device 31 (S2). The joint position setting process may obtain a joint position input from the user, or may be derived from a plurality of captured images A and B by image processing using the image processing device 30. As an example of this image processing, for example, as shown in FIG. 4, a region 50 consisting of a partial region below the image A, and a search range 51 above the image B and having a larger area than the region 50. And define. Then, while shifting the region 50 within the search range 51, search for a position with the highest density correlation value in the search range 51, and secure the position of the region 50 with respect to the search range 51 at that time, Is set as the joint position, and the overlapping areas in the images A and B at this time are obtained as the overlapping area 50 and the overlapping area 52 of each image. The algorithm for detecting the joint position is not limited to the above, and any algorithm may be used as long as the joint position and the overlapping region between the image A and the image B can be detected.

(Step S3)
The image position correction unit 33 corrects the pixel positions of the overlapping region 50 of the image A and the overlapping region 52 of the image B at the bonding position set by the bonding region detection unit 32 (S3). In the following pixel position correction process, the overlapping area 50 is divided into small areas. On the other hand, in the overlapping area 52, a scanning area having the highest correlation with the small area is obtained for each small area. Then, the pixel position is corrected by projective transformation of both the small area and the scanning area, and the overlapping area 50 and the overlapping area 52 are matched. Next, the pixel position correcting process by the pixel position correcting unit 33 will be described in the order of FIG.

(Step S301)
The pixel position correcting unit 33 performs a small area dividing process for dividing one of the two image overlapping areas, for example, the overlapping area 50 of the image A, into small areas (S301). This small area is a unit area for performing projection projection in step S303 to be described later.

First, as shown in FIG. 4, the pixel position correcting unit 33 extracts overlapping regions 50 and 52 from two overlapping images A and B, respectively. Next, as shown in FIG. 5 (a), one of the image regions, for example the overlay area 50, the lower right diagonal hatched square small areas 57 (FIG. 5 the length of one side is composed of d ₁ (Corresponding to the area). In FIG. 5A, a plurality of horizontal lines 53 h and vertical lines 53 v depicted by dotted lines in the overlapping area 50 are set at a pitch d ₁ , and the overlapping area 50 is divided into a plurality of small areas 57. Note that the length d ₁ of one side at this time may be changed in proportion to the image size of the overlapping region 50.

(Step S302)
The image position correction unit 33 searches for the pixel having the highest relevance among the pixels of the other image area, that is, the overlapping area 52 of the image B, for all the vertex pixels of each small area divided in step S301. The correlation value deriving process is performed (S302).

Image position correction unit 33, the vertices of the small area, i.e. 5 dotted (a) 53h, sets all intersections of 53v as interest point 54, a square shape having a side length of its periphery is d ₂ The region of interest 55 is set. The size d ₂ may be a size of 2 × d _{1 or} may be smaller than that in order to reduce the calculation cost.

  The pixel position correcting unit 33 sets a region 56 having the same shape as the region of interest 55 (hereinafter, this region is referred to as a “scanning region”) as the other image region, that is, the overlapping region 52. As shown in FIG. 6, the pixel position correcting unit 33 compares the region of interest 55 and the scanning region 56 at each position while scanning the scanning region 56 in the horizontal direction and the vertical direction within the overlapping region 52. Find the correlation. The reference for the correlation height at this time may be the deviation of the pixel values of the region of interest 55 and the scanning region 56, or the difference between the feature values such as the average value or median value of the pixel values of each region is small. May be used.

FIG. 7 is an explanatory diagram showing a relationship between a point of interest and a region of interest, a scanning region having the highest correlation with the region of interest, and a point of interest in the scanning region. The point of interest in the overlapping area 50 of the image A is denoted as P ₁ (i _n , j _n ), and its coordinates are defined as (x1i _n , y1j _n ). Note that (i _n , j _n ) represents the pixel array number. The processing in this step will be described by taking four interest points P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), and P ₁ (2,2) in FIG. 7 as an example. In the following description, only four points of interest will be described, but the same processing is performed on the vertices of all small regions.

The pixel position correcting unit 33 performs a region of interest for each of the four points of interest P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), and P ₁ (2,2). 55a, 55b, 55c, and 55d are set. Then, the scanning region having the highest correlation is obtained for each region of interest. In FIG. 7, 56a, 56b, 56c, and 56d are obtained as scanning regions having the highest correlation with each of the regions of interest 55a, 55b, 55c, and 55d.

(Step S303)
The pixel position correcting unit 33 obtains the center point of each scanning region, that is, the point corresponding to the point of interest for each region of interest as the pixel P ₂ (i _n , j _n ) having the highest correlation with the point of interest ( S303). In FIG. 7, the pixels P ₂ (1,1), P ₂ (1,2), P ₂ (2,1), and P ₂ (2,2) that are the center points of the scanning regions 56a, 56b, 56c, and 56d. ) Is determined as the pixel having the highest correlation with respect to each interest point P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), P ₁ (2,2). Hereinafter, representative pixel _{P 2 (i n, j n} ) the coordinates of (x2i _n, y2j _n).

(Step S304)
In order to perform pixel correction processing for moving each interest point and a pixel having the highest correlation with each interest point, first, the pixel position correcting unit 33 first performs each interest point and its interest point. And the corrected coordinates for matching the coordinates of the pixel having the highest correlation (S304).

Pixel position correcting unit 33, each of the interest point _{P 1 (i n, j n} ) coordinate (x1i _n, y1j _n) with the coordinates P ₂ of high correlation value coordinates (center point obtained in step S302) ( i _n, the coordinates (x2i _n of j _n), and Y2j _n), the difference SubtractionPoint (i _n, the j _n), determined by the following equation (1).
SubtractionPoint (i _n , j _n ) = (x1i _n -x2i _n , y1j _n -y2j _n ) ... (1)
Where (x1i _n , y1j _n ): coordinates of the point of interest P ₁ (i _n , j _n )
(x2i _n , y2j _n ): coordinates of pixel P ₂ (i _n , j _n ) with the highest correlation

Next, the pixel position correcting unit 33 weights the difference of the derived coordinates with respect to the coordinates (x1i _n , y1j _n ) of the interest point P ₁ (i _n , j _n ) and adds the weighted difference to the coordinates of the interest point. The coordinates (x′1i _n , y′1j _n ) of the pixel P ′ ₁ (i _n , j _n ) after coordinate correction are obtained. For other image (overlay area 52), with respect to interest point _{P 1 (i n, j n} ) and highest correlation pixels _{P 2 (i n, j n} ) of the coordinate (x2i _n, Y2j _n), performs weighting the difference coordinates for negating the value, the pixel P ₂ (i _n, j _n) of the coordinate (x2i _n, y2j _n) pixels P _'2 after the coordinate modification made to the (i _n, j _n) Find the coordinates of.

The weighting at this time is an image in which the point of interest P ₁ (i _n , j _n ) is provided around the half of the vertical matrix of the overlapping region 50 (and the overlapping region 52) (the overlapping region 50 of the image A). ) Is weighted from 1 to 0, and the other image P ₂ (i _n , j _n ) (superimposed region 52 of image B) is weighted from 0 to 1 to perform linear weighting. Perform these weights, the pixel _{P 1 (i n, j n} ), P 2 (i n, j n) coordinate (x2i _n, Y2j _n) after coordinate modification when the expression of equation (2) ( It becomes like 3).

(x′1i _n , y′1j _n ) = (x1 i _n , y1j _n ) + A (y) · SubtractionPoint (i _n , j _n ) (2)
(x'2i _n , y'2j _n ) = (x2i _n , y2j _n ) + (-(1- A (y))) ・ SubtractionPoint (i _n , j _n ) (3)
However, (x'1i _n , y'1j _n ): coordinates of pixel P ' ₁ (i _n , j _n ) after coordinate correction
(x'2i _n , y'2j _n ): Coordinate of pixel P ' ₂ (i _n , j _n ) with the highest correlation after coordinate correction
A (y): Weighting factor

In Expressions (2) and (3), P ₁ is a pixel when the overlapping region 50 is at the bottom of the image as in image A in FIG. 1, and P ₂ is superimposed as in image B in FIG. The case where the pixel when the area | region 52 exists in the upper part of an image is used is shown. When the positional relationship between the overlapping regions 50 and 52 is reversed, the coefficients A (y) and-(1-A (y)) applied to SubtractionPoint (i _n , j _n ) are _expressed by equations (2) and (3). Vice versa. Further, the weighting coefficient is obtained by the following equation (4).
A (y) =-1 / Height + 1 (4)
However, Height: y matrix size of the overlapping region 50 (or 52) y: The maximum value of the y component of the overlapping region 50 (or 52) P (i, j) is substituted.

(Step S305)
The pixel position correcting unit 33 matches the small area defined by the point of interest and the area defined by the pixel having the highest correlation with the vertex of the small area (hereinafter referred to as “comparison area”). In order to perform pixel correction processing for projective transformation of the comparison region, first, corrected coordinates for matching the respective interest points and the coordinates of the pixel having the highest correlation with the interest points are obtained (S305).

Pixel position correcting unit _{33, P 1 (i n, j} n) an area surrounded by the same corresponding sequences surrounded by a number _{P '1 (i n, j} n) in, performs projective transformation. In FIG. 7, four points of interest P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), and P ₁ (2,2) in the overlapping area 50 of the image A are represented. Coordinate correction of each of these interest points P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), P ₁ (2,2) is performed on the enclosed square-shaped small region 57. Projective transformation is performed on a region 58 surrounded by the subsequent pixels P ′ ₁ (1,1), P ′ ₁ (1,2), P ′ ₁ (2,1), and P ′ ₁ (2,2). By this projective transformation, the small region 57 is projectively transformed into a region 58 as shown in FIG.

The projective transformation uses general formulas (5) and (6) for projective transformation.
x '= (a1x + b1y + c1) / (a0x + b0y + 1) (5)
y '= (a2x + b2y + c2) / (a0x + b0y + 1) (6)
Where x: x-axis coordinate y: y-axis coordinate
x ′: x coordinate after projective transformation y ′: y coordinate after projective transformation

Eight variables a0, a1, a2, b0, b1, b2, c1, c2, which are included in equations (5) and (6), are expressed by four points of interest P ₁ (x and y in equations (5) and (6)). 1,1), P ₁ (1,2), P ₁ (2,1), P ₁ (2,2) coordinates (x1i _n , y1j _n ) are substituted and coordinates are corrected to x ', y' Coordinates of four points P ' ₁ (1,1), P' ₁ (1,2), P ' ₁ (2,1), P' ₁ (2,2) (x'1i _n , y '1j _n ) is derived by simultaneous equations consisting of eight equations obtained by substituting.

Similarly, the pixel position correcting unit 33 has the highest correlation corresponding to the points of interest P ₁ (1,1), P ₁ (1,2), P ₁ (2,1), P ₁ (2,2). Similarly to the small region 57, the comparison region 59 surrounded by the four pixels P ₂ (1,1), P ₂ (1,2), P ₂ (2,1), and P ₂ (2,2) Projective transformation to a region 58 ′ surrounded by pixels P ′ ₂ (1,1), P ′ ₂ (1,2), P ′ ₂ (2,1), and P ′ ₂ (2,2) after coordinate correction To do. This area 58 ′ coincides with the area 58 obtained by projective transformation of the small area 57. By this projective transformation, the overlapping area 50 of the image A is converted into the overlapping area 50 ′ in units of the small area 57, and the overlapping area 52 of the image B is the same coordinate as the area 50 ′ in the comparison area 59 units. It is converted into a superposition region 52 'consisting of a system.

(Step S4)
The long image creation processing unit 34 aligns the images A and B so that the overlapping region 50 ′ and the overlapping region 52 ′ that have undergone the projective conversion overlap, and superimposes the density value of the long image. Similar to the calculation of the average pixel value of the matching pixels in the matching region 50 ′ and the overlapping region 52 ′ and the coordinate correction amount calculation, weighting is performed using Expression (4) and the sum is used.

  Then, a plurality of images (image A and image B) whose density and contrast are corrected are joined (S4). The joining may be corrected by giving a weight to the overlapping joining area as in a known method, or one image having the overlapping area may be used.

(Step S5)
Finally, the long image is D / A converted and displayed (S5). The long image output device may be a monitor or a film printer, or may be stored as data on a server or a recording medium.

  According to the present embodiment, when a long image is created by connecting a plurality of images, an image included in the overlapping area is determined with respect to the overlapping area of each image whose approximate overlapping position is known. By correcting the distortion in each image and then connecting, it is possible to prevent the occurrence of a problem that the subject is displayed twice. Thereby, the image quality of the overlapping region in the long image can be improved.

  Further, as in this embodiment, by correcting the pixel position of the overlapping region of each image to be joined, compared to the case where the pixel position of only the overlapping region of one image is corrected, Since the shift amount of the pixel position from the subsequent area is reduced, the overlapping area (joining area) and the subsequent area are smoothed in the long image. Therefore, the image quality of the long image can be improved.

  In the present embodiment, the pixel position is corrected for both of the plurality of images to be joined. However, even if the pixel position is corrected only for the overlapping region of one image, the pixel position shift is small. When the overlapping area and the subsequent image area are so smooth that they do not interfere with the diagnosis, the pixel position may be corrected only for the overlapping area of either one of the images.

  In the above-described embodiment, weighting is performed using the matrix size in the y direction of the overlapping region. However, linear weighting may be performed in the x direction using the matrix size in the x direction. The coordinate correction amount may be a linear weight in the x direction for the x coordinate and a linear weight in the y direction for the y coordinate.

1: X-ray generator, 2: X-ray detector, 3: Image processing device, 4: Output device, 10: X-ray diagnostic device, 31: Image storage device, 32: Junction position setting unit, 33: Pixel position correction Section, 34: long image creation section

Claims (6)

X-ray irradiation means for irradiating X-rays;
X-ray detection means for detecting X-rays emitted from the X-ray irradiation means and converting them into electrical signals;
A plurality of images based on the electrical signal obtained from the X-ray detection means, a joining position setting means for setting a position where a part of each image is superimposed and joined;
Pixel position correcting means for correcting the pixel position of the overlapping region included in at least one image;
A long image creating means for creating a long image in which the plurality of images are joined in the corrected overlapping region;
An X-ray diagnostic apparatus comprising: The pixel position correcting means includes
Means for dividing an overlapping region included in one of the plurality of images into a plurality of small regions;
Correlation value deriving means for searching which region of the overlapping region included in the other image of the plurality of images is most relevant to the divided small region;
Coordinate conversion means for converting the coordinates of the small area so that the coordinates of the most relevant area and the coordinates of the small area match.
The X-ray diagnostic apparatus according to claim 1. The coordinate conversion means performs coordinate conversion on both the small area included in the one image and the most relevant area included in the other image, and coordinates of the small area after the coordinate conversion are performed. And the coordinates of the most relevant area after coordinate transformation,
The X-ray diagnostic apparatus according to claim 2. The coordinate conversion means uses the coordinate correction amount obtained by multiplying the difference between the coordinates of the small area and the coordinates of the most relevant area by a weighting coefficient based on the size of the overlapping area. And the coordinates of the most relevant area are corrected, and coordinate conversion processing is performed using the corrected coordinates.
The X-ray diagnostic apparatus according to claim 3. A joining position setting means for setting a position for joining a plurality of images obtained by the medical imaging apparatus by overlapping a part of each image;
Pixel position correcting means for correcting the pixel position of the overlapping region included in at least one image;
A long image creating means for creating a long image in which the plurality of images are joined in the corrected overlapping region;
An image processing apparatus comprising: Setting a position where a plurality of images obtained by the medical imaging device are joined by overlapping a part of each image; and
Correcting the pixel position of the overlapping region included in at least one of the images;
Creating a long image in which the plurality of images are joined in the corrected overlapping region;
An image processing program for causing a computer to execute.