JPH06215108A - Positioning method for radiation picture - Google Patents

Abstract

PURPOSE:To rapidly and highly precisely attain positioning without recording markers or the like for positioning with an object, in the positioning method of a radiation picture. CONSTITUTION:The binarization of a picture signal is operated by using a certain threshold value, and two closed areas are set at each X-ray picture 4a and 4b based it. That is, closed areas 8a and 8b are set at the X-ray picture 4a, and closed areas 8c and 8d are set at the X-ray picture 4b. Then, the centers of gravity 9a, 9b, 9c and 9d of each closed area 8a, 8b, 8c and 8d are found, and the X-ray picture 4b is rotated, enlarged or reduced, and moved in parallel by using those centers of gravity so that the X-ray picture 4b can be matched with the X-ray picture 4a. At first, the X-ray picture 4b is moved in parallel so that the center of gravity 9c of the X-ray picture 4b can be matched with the center of gravity 9a of the X-ray picture 4a, and then the X-ray picture 4b is rotated and enlarged or reduced by using the center of gravity 9d as an axis until the center of gravity 9d is matched with the center of gravity 9b so that the positioning of the two X-ray pictures 4a and 4b can be attained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of aligning images by correcting the positional deviation of a plurality of images which are subjected to a radiation image superimposing process or a subtraction process, and more specifically, to aligning each radiation image. The present invention relates to a method for aligning an image without using a marker for the image.

[0002]

2. Description of the Related Art Recently, using a stimulable phosphor, radiation image information of a subject such as a human body is once recorded on a stimulable phosphor sheet (hereinafter referred to as a stimulable phosphor sheet) and excited. There has been proposed a method of scanning with light to cause stimulated emission, photoelectrically reading the stimulated emission to obtain an image signal, and processing the image signal to obtain a radiation image of a subject having good diagnostic suitability ( For example, JP-A-55-12429 and JP-A-55-116340
No. 55-163472, No. 56-11395, No. 56-104645, etc.). This final image can be played as a hard copy or on a CRT.

On the other hand, overlay processing of radiation images has been conventionally known (see, for example, JP-A-56-11399).
Generally, radiation images are used for diagnostic purposes and other purposes, but in using them, it is required to satisfactorily detect minute radiation absorption differences of a subject. The degree of this detection in the radiographic image is called contrast detectability or simply detectability. The higher the detectability, the higher the diagnostic performance.
It can be said that the radiographic image has high practical value. Therefore, in order to improve the diagnostic performance, it is desired to increase this detectability, but the biggest obstacle factor is various noises. The superposition processing is a method of significantly reducing the various noises, making it possible to clearly observe even a slight radiation absorption difference of the subject in the final image, and greatly improving the detectability. That is, a radiation image is captured (accumulated and recorded) on a plurality of stacked stimulable phosphor sheets, and a plurality of image signals obtained by subjecting the plurality of sheets to a reading process are subjected to an addition process. To reduce.

Conventionally, in order to actually perform this addition processing, for example, two stimulable phosphor sheets are placed in a cassette in an overlapping manner to photograph a subject, and the two stimulable phosphor sheets are normally used. The reading process similar to the reading process is sequentially performed to obtain two sets of image signals, and the addition process is performed on the two sets of image signals.

On the other hand, subtraction processing of radiation images has been conventionally known. The subtraction of the radiographic image means that two radiographic images captured under different conditions are photoelectrically read to obtain digital image signals, and then these digital image signals are subjected to subtraction processing by associating each pixel of both images with each other. A method of obtaining a difference signal for extracting a specific structure in a radiation image, and using the difference signal thus obtained, a radiation image in which only the specific structure is extracted can be reproduced.

There are basically the following two methods for this subtraction processing. That is, (1) a specific structure is extracted by subtracting (subtracting) the image signal of the radiation image in which the contrast agent is not injected from the image signal of the radiation image in which the specific structure is emphasized by the injection of the contrast agent. The so-called temporal subtraction processing, and (2) irradiating the same subject with radiation having different energy distributions, or irradiating the two radiation detecting means with radiation after passing through the subject by changing the energy distribution state, As a result, images having different specific structures are made to exist between the two radiation images, and then the image signals of the two radiation images are appropriately weighted and then subtracted (subtract) to obtain the specific structures. This is a so-called energy subtraction process for extracting the image.

Since this subtraction processing is an extremely effective method especially for medical diagnosis, it has received a great deal of attention in recent years, and its research and development have been actively promoted by making full use of electronic technology.

However, the following problems occur in the radiation image superposition processing method and subtraction processing method using the above-described stimulable phosphor sheet.

That is, in each of the above-described processing methods using the stimulable phosphor sheet, two (or sometimes three or more) stimulable phosphor sheets are sequentially or simultaneously inserted into the photographing table and overlapped with each other. The radiation image to be subtracted is taken, and then the stimulable phosphor sheet is individually inserted into the reading device, and each time the stimulable luminescent light emitted by irradiating the stimulable phosphor sheet with excitation light is detected. As a result, in the process of reading out the radiation image, even if the mechanical accuracy of all devices involved in imaging and reading is increased, positional deviation and rotational deviation occur between the images to be superposed or subtracted. As a result, although various noises are averaged and reduced in this superposition process,
Blurring occurs in the entire image including the edge of the structure in the image, the image to be observed becomes unsuitable for observation, and the image to be erased in the subtraction process is not erased, or conversely the image to be extracted Is erased, a false image is generated, and an accurate subtraction image cannot be obtained. As described above, it has been found that the positional deviation and the rotational deviation described above cause serious trouble in diagnosis.

When such a deviation occurs between the radiation image information stored and recorded on the stimulable phosphor sheet, the radiation image is stored and recorded in the stimulable phosphor as a latent image.
Unlike the case of an X-ray photo film capable of capturing an X-ray image as a visible image, it is impossible to visually match two X-ray images, and it is extremely difficult to correct the deviation.

Further, even if the positional deviation and the rotational deviation which occur between the two radiation images can be detected by some means, if the conventionally known arithmetic processing is carried out to correct the data of the read radiation image, in particular, the rotational deviation will occur. A great amount of time is spent in the correction, which is a very serious problem in practical use.

The applicant of the present invention has proposed a subtraction processing method for a radiographic image using a marker having a shape that provides a reference point or a reference line in Japanese Patent Laid-Open No. 58-163338. This method records the marker on two stimulable phosphor sheets at a fixed position with respect to the radiographic image, detects the marker when reading the radiographic image, and calculates the positional deviation and rotational deviation. Either one of the radiation images to be subtracted is rotated and / or moved on the digital data, and the image data is subtracted between the corresponding pixels of the radiation image. The alignment step in the subtraction processing method for a radiation image using this marker can also be applied to the above-described overlay processing method. In that case, after the alignment is performed, the addition processing of the image data may be performed between the corresponding pixels of the radiation image.

[0013]

However, in this method, the marker as described above must be accumulated and recorded together with the subject in the stimulable phosphor sheet each time a radiographic image is taken. Then, there is a problem that the image information of the subject cannot be obtained from the portion of the accumulated and recorded radiation image that overlaps the position of the marker.

In view of the above circumstances, the present invention provides a method of aligning a radiographic image that enables rapid and highly accurate alignment without recording a marker or the like with an object for alignment. The purpose is.

[0015]

A first method of aligning a radiation image according to the present invention is a method of aligning a plurality of radiation images for overlay processing or subtraction processing of radiation images. At least 2 having substantially the same shape on the plurality of images
A closed area is set at each location, each centroid of each closed area is obtained, and each coordinate of each centroid of the reference image of the plurality of images matches each coordinate of each centroid of the other images. As described above, the other image is rotated, enlarged or reduced, and moved in parallel.

The above-mentioned closed region is a region closed at a certain threshold value. For example, when the chest of a human body is imaged, the lung field can be made a closed region. Corresponding closed areas on the plurality of images can be considered to be substantially the same shape, even if the images are tilted relative to each other. Therefore, this corresponding closed region can be used as an index for alignment. This closed region can be set by binarizing with a certain threshold value.

As described above, the shape of the corresponding closed region is substantially the same among a plurality of images, and although the information (image signal) is slightly different depending on the photographing conditions, the positional relationship of the respective centroids is different. They are substantially equal in the closed region. Therefore, two common points can be set in each image, which allows alignment to be performed.

As one of the methods for obtaining the center of gravity of the set closed region, there is a method of determining image points in a subject image described in Japanese Patent Laid-Open No. 2-28782. In this method, based on an image signal obtained from each pixel on a recording sheet on which a radiation image including a subject image is recorded, an image signal value corresponding to each pixel or a reciprocal of the image signal value is used. The center of gravity is obtained by weighting each pixel. In the present invention, the image signal is not from the whole radiation image, but only the image signal in the closed region surrounded by a certain threshold value may be used to obtain the center of gravity. The center of gravity in the present invention may be obtained by any method other than this method.

The above rotational movement, enlargement or reduction, and parallel movement may be performed simultaneously or separately. For example, when two closed regions are set in each radiographic image, the images are translated so that one position of the center of gravity of the closed region matches, and then the other center of gravity is matched. It can be rotated and enlarged or reduced, and a line segment that connects two centroids in each image is set, and the image is rotated and moved so that the inclinations of the line segments are equal. It can also be translated and scaled up or down to match.

A second method of aligning radiation images of the present invention is a method of aligning a plurality of radiation images for overlay processing or subtraction processing of radiation images. , At least two closed regions having substantially the same shape among the plurality of images are set, orthogonal coordinates are defined for each of the plurality of images, and the centers of gravity of all the closed regions are respectively obtained, An expression for converting the coordinate values of the other image into the coordinate values of the reference image so that the coordinates of the center of gravity of the reference image of the images match the coordinates of the center of gravity of the other image. ,

[0021]

[Equation 2]

(Where u and v are the coordinates of the reference closed area, x and y
Is a coordinate of the general closed region, a, b, c, d are coefficients for rotational movement correction and enlargement or reduction correction, and e, f are coefficients for affine transformation represented by A first affine transformation for performing at least rotational movement correction and enlargement or reduction ratio correction on the other image using a coefficient is performed, and a plurality of images subjected to the first affine transformation are substantially At least two common regions of interest are set, each region of a reference image of the plurality of images is set as a reference region, and the regions of other images are set as template regions, and the template region is set as the reference region. Template matching is performed to obtain coordinate values of at least two points corresponding to each other in the plurality of images, and the coordinate points are matched so that the corresponding points match. The second affine transformation is performed to obtain a coefficient of the affine transformation represented by ## EQU1 ## and perform a rotational movement correction, an enlargement or reduction ratio correction, and a parallel movement correction on a plurality of images including the template region. It is what

Here, the template matching is a process of finding the best matching place by moving the template region on the image including the reference region when the template region and the reference region are set as described above. The point representing the place gives the coordinates of the corresponding point.

In such template matching, the correlation method and SSDA (Sequential Similarity Detection Algo)
lithms).

The correlation method is a method in which a product is calculated for each corresponding pixel and a value obtained by standardizing the sum of the products (hereinafter referred to as a standardized value) is used as a measure of superposition. In this standardization, the sum of the products (squares) of the pixels themselves in each area is calculated, the product of each sum is further calculated, and the square root of this product is calculated as the sum of the products of the corresponding pixels. This is done by using the denominator. When the superposition is perfect, the product of each pixel that is a numerator does not become the square of the pixel itself due to noise, etc. Therefore, even if the standardized value does not become 1, the maximum value that is the closest to 1 is obtained. It is believed that Therefore, it is considered that the template region is moved variously on the image including the standard region, and the superposition is achieved with the above-described movement that maximizes the standardized value. However, the movement having the maximum standardized value cannot be determined unless all the movements are completed. For details on this method, see, for example, Smith's "Automated cloud tracking usi.
ngprecisely aligned digital ATS pictures "ibi
d. , July 1972, volume c-21, pages 715-729.

SSDA uses the sum of the absolute values of the differences (residual error) for each pixel as a measure of superposition.
When the superposition is perfect, it is considered that the residual is minimized even if it is not zero due to noise or the like. Therefore, it is considered that the template region is moved variously on the image including the reference region, and the registration is achieved with the movement that minimizes the residual error. At this time, if the overlay is deviated, the residual error rapidly increases when the pixels are sequentially added. Therefore, SSDA is a method in which if the residual exceeds a certain threshold during the addition, the addition is immediately terminated and the next movement is performed. Since the calculation used is only addition, and in many cases the calculation is aborted halfway, the calculation time is greatly shortened. For details of this method, see, for example, Barnea et al. “A class of algorithms for fast digital image reg.
istration ”IEEE. Trans. , February 1972 Volume c-21, 179-
It is described on page 186.

In the present invention, the coordinate value of the other image is set to the reference so that the coordinates of the center of gravity of the image serving as the reference of the plurality of images match the coordinates of the center of gravity of the other images. The affine transformation represented by the above equation for converting into the coordinate value of the image is performed, and at least the rotational movement correction and the enlargement or reduction ratio correction are performed on the other images. That is, the affine transformation is performed using the coefficients a, b, c, and d indicating the rotational movement correction and the enlargement or reduction ratio correction.

Next, the template matching described above is performed on the image subjected to this affine transformation.
Since this template matching is performed on the image on which the rotational movement correction and the enlargement or reduction ratio correction have been performed, more accurate matching is performed.

By the affine transformation based on the corresponding points obtained by the template matching, the rotational movement correction, the enlargement or reduction ratio correction and the parallel movement correction are performed on the image including the template area. That is, the affine transformation is performed using the coefficients a, b, c and d indicating the rotational movement correction and the enlargement or reduction ratio correction and the coefficients e and f indicating the parallel movement correction.

[0030]

In the radiation image registration method of the present invention, at least two closed regions are set in the radiation image instead of recording a marker for alignment together with the subject, and the centers of gravity of these closed regions are set. Since the area is determined to be rotationally moved, enlarged or reduced, and moved in parallel based on the center of gravity of the area, it is possible to obtain the radiation image information of the portion that overlaps with the conventional marker. Also,
Another alignment method of the present invention obtains the centers of gravity of these closed regions, performs rotational movement correction, enlargement or reduction correction based on the centers of gravity, and then based on at least two regions of interest. Since the template matching is performed and the rotational movement correction, the enlargement or reduction correction, and the parallel movement correction are performed, the alignment accuracy can be further improved.

[0031]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on the embodiments shown in the drawings.

FIG. 1 is a schematic view of an X-ray imaging apparatus which is an embodiment of an apparatus for recording a radiation image used in the present invention. The X-ray image obtained by this imaging is used for energy subtraction processing.

The stimulable phosphor sheets 5 and 7 are stacked with the filter 6 sandwiched therebetween with the sheet 7 facing downward. An X-ray tube 2 that emits X-rays 3 through a subject 4 is arranged on this. The X-ray imaging apparatus 1 is configured as described above.

The subject 4 is illuminated by the X-rays 3 emitted from the X-ray tube 2. The X-rays 3a transmitted through the subject 4 are irradiated on the first stimulable phosphor sheet 5, and a part of the energy of the X-rays 3a is recorded on the first stimulable phosphor sheet 5, whereby the sheet 5 is stored. An X-ray image of the subject 4 is stored and recorded in the. The X-rays 3b that have passed through the sheet 5 further pass through the filter 6, and the X-rays 3c that have passed through the filter 6 are applied to the second stimulable phosphor sheet 7. As a result, the X-ray image of the subject 4 is also stored and recorded on the sheet 7.

FIG. 2 shows each stimulable phosphor sheet 5 and 7.
FIG. 3 is a diagram schematically showing an X-ray image accumulated and recorded in FIG. The X-ray images 4a and 4b of the subject 4 are accumulated and recorded on substantially the entire surfaces of the stimulable phosphor sheets 5 and 7, respectively.

FIG. 3 shows an embodiment of an X-ray image reading apparatus which is an embodiment of a reading unit for reading a radiation image used in the present invention and an embodiment of an arithmetic unit which carries out the alignment method of the present invention and performs subtraction processing. It is a perspective view of an image processing display device.

After radiography is performed by the X-ray radiographing apparatus 1 shown in FIG. 1, the first and second stimulable phosphor sheets 5 and 7 are set one by one at a predetermined position of the X-ray image reading apparatus 10. It Here, the case of reading the first X-ray image accumulated and recorded in the first stimulable phosphor sheet 5 will be described.

The stimulable phosphor sheet 5 in which the first X-ray image is stored and recorded, which is set at a predetermined position, is moved in the direction of the arrow Y by the sheet conveying means 15 such as an endless belt driven by a driving means (not shown). Is transported (sub-scanning). On the other hand, a light beam 17 emitted from a laser light source 16 is reflected and deflected by a rotating polygon mirror 19 driven by a motor 18 and rotating at a high speed in the arrow Z direction, and after passing through a focusing lens 20 such as an fθ lens, an optical path is passed by a mirror 21. Incident on the stimulable phosphor sheet 5 by changing the direction, and the sub-scanning direction (arrow Y direction)
The main scanning is performed in the direction of arrow X, which is substantially vertical to From the place where the light beam 17 of the stimulable phosphor sheet 5 is irradiated, the stimulated emission light 22 of a light amount corresponding to the accumulated and recorded X-ray image information is emitted, and this stimulated emission light 22 is a light guide. It is guided by 23 and is photoelectrically detected by a photomultiplier (photomultiplier tube) 24. The light guide 23 is made by molding a light guide material such as an acrylic plate, and is arranged so that the linear incident end face 23a extends along the main scanning line on the stimulable phosphor sheet 5. , Injection end face 23 formed in an annular shape
The light receiving surface of the photomultiplier 24 is coupled to b. The photostimulated luminescent light 22 that has entered the light guide 23 through the incident end face 23a travels through the inside of the light guide 23 after repeating total reflection, and then exits from the exit end face 23b to exit from the photomultiplier 24.
The photostimulated luminescent light 22 representing the radiation image is converted into an electric signal by the photomultiplier 24. The analog signal S output from the photomultiplier 24 is
After being logarithmically amplified by the log amp 25, the A / D converter 26
Image signal S input to and sampled by
O is obtained. This image signal SO represents the first X-ray image accumulated and recorded on the first stimulable phosphor sheet 5, and is referred to as the first image signal SO ₁ . The first image signal SO ₁ is temporarily recorded in the internal memory in the image processing display device 30.

The image processing display device 30 includes a keyboard 31 for inputting various instructions, a CRT display 32 for displaying a visible image based on auxiliary information for the instructions and an image signal,
A floppy disk drive device 33 in which a floppy disk as an auxiliary storage medium is loaded and driven, and a CPU
A main body 34 having a built-in memory and an internal memory is provided.

Then, in the same manner as described above, the second image signal SO ₂ representing the second X-ray image accumulated and recorded on the second stimulable phosphor sheet 7 is obtained, and this second image signal is obtained. SO
_{2 is} also temporarily stored in the internal memory in the image processing display device 30.

When the two image signals SO ₁ and SO ₂ to be subjected to the subtraction operation are stored in the internal memory in this way, these two image signals SO ₁ and SO ₂ are read out and these two images are read. The images are aligned so that the corresponding subtraction operation is performed between the pixels of the X-ray images carried by the signals SO ₁ and SO ₂ .

Here, in the present embodiment, the image signal SO
_A method of aligning two X-ray images represented by ₁ and SO ₂ will be described.

First, the first radiation image registration method of the present invention will be described.

In this alignment method, an image signal is binarized using a certain threshold value, and at least two closed regions are set based on this. The two closed areas set in this way are shown in FIG. In FIG. 2, these closed regions are the lung fields indicated by the diagonal lines. That is, the closed regions 8a and 8b are set in the X-ray image 4a, and the closed regions 8c and 8d are set in the X-ray image 4b. Furthermore, the center of gravity 9a, 9b, 9c, 9d corresponding to each closed region 8a, 8b, 8c, 8d is shown. These centroids are obtained by the method described in JP-A-2-28782 as described above. However, the method of determining the center of gravity in the present invention is not limited to this method, and may be determined by any method.

The manner in which the X-ray image 4b is rotationally moved, enlarged or reduced, and moved in parallel using these centers of gravity to match the X-ray image 4a will be described with reference to FIGS. In these drawings, the X-ray image 4a is shown by a solid line and the X-ray image 4b is shown by a dotted line.

In the example of FIG. 4, the center of gravity 9c of the X-ray image 4b is moved in parallel so as to coincide with the center of gravity 9a of the X-ray image 4a. After that, the two X-ray images 4a and 4b can be aligned by rotating and enlarging or reducing the center of gravity 9d about the center of gravity 9c until the center of gravity 9d coincides with the center of gravity 9b. The rotation angle at that time is the line segment I connecting the centers of gravity 9a and 9b.
Is the angle formed by the line segment II connecting the centers of gravity 9c and 9d.

In the example of FIG. 5, the X-ray image 4b is rotationally moved so that the line segment I and the line segment II described above are parallel to each other. This rotational movement is due to the center of gravity 4c, 4d, one point on the line segment II,
Any point may be used as an axis. After that, the two X-ray images 4a and 4b can be aligned by translating and enlarging or reducing the X-ray image 4b so that the corresponding centroids match.

Next, a second radiation image registration method of the present invention will be described.

First, the centroids of all closed regions are obtained in the same manner as the first radiation image registration method of the present invention described above (see FIG. 2). At the same time, as the common orthogonal coordinates, the x-axis and y-axis coordinate systems are set for the sheet 5, and the u-axis and v-axis coordinate systems are set for the sheet 7. This x-axis, u
The axis is the left-right direction on the paper surface of FIG. 1, and the y-axis and the v-axis are the directions perpendicular to the paper surface of FIG. Also, at this time, an X-ray image
The closed regions 8a and 8b of 4a will be referred to as template regions, and the closed regions 8c and 8d of X-ray image 4b will be referred to as reference regions. Then, using the coordinates of the center of gravity of the closed region, the X-ray image 4a is made so that the coordinates of the template region and the coordinates of the reference region are the same.
Rotate, enlarge or reduce, and translate.
In this embodiment, the X-ray image 4a is aligned with the X-ray image 4b.

The first X carried by the first image signal SO ₁
The coordinates of at least two barycenters on the line image and the sampling points on the line segment connecting the barycenters are (X ₁ ,
Y ₁ ), at least two centroids on the second X-ray image carried by the second image signal SO ₂ and sampling points corresponding to (X ₁ , Y ₁ ) above on the line segment connecting the centroids. With coordinates (X ₂ , Y ₂ ), a, b, c, d,
Affine transformation when e and f are coefficients

[0051]

[Equation 3]

By converting the coordinates of the first X-ray image 4a according to the above, the first X-ray image 4a and the second X-ray image 4b are superimposed. Here, in coordinate transformation based on equation (1),
Enlarging or reducing the entire first X-ray image in the X and Y directions independently of each other, rotating and moving the entire X-ray image 4a, and parallelizing the X-ray image 4a in the X and Y directions. Moving is all done at the same time. However, in the second radiographic image registration method of the present invention, the registration is not performed by only one affine transformation, so here it is assumed that rotational movement and conversion of enlargement or reduction ratio are performed. .

Next, the coefficients a, b, c included in the equation (1),
A method of obtaining d, e, f will be described.

Equation (1) is

[0055]

[Equation 4]

[0056]

[Equation 5]

It is divided into Here, the coordinates of the centers of gravity 9a and 9b on the X-ray image 4a and one point on the line segment I are (X ₁₁ , Y ₁₁ ), (X ₁₂ , Y ₁₂ ), and (X ₁₃ , Y ₁₃ ), respectively. , The barycenters 9c and 9d on the X-ray image 4b, and the coordinates of one point on the line segment II are (X ₂₁ , Y ₂₁ ), (X ₂₂ , Y ₂₂ ),
(X ₂₃ , Y ₂₃ ). At this time, from equations (2) and (3),

[0058]

[Equation 6]

[0059]

[Equation 7]

[0060]

[Equation 8]

[0061]

[Equation 9]

[0062]

[Equation 10]

[0063]

[Equation 11]

It becomes The coefficients to be calculated here are a, b,
Since there are 6 of c, d, e, and f, the above (2a), (2b), (2
c); can be obtained based on the six formulas (3a), (3b), and (3c),

[0065]

[Equation 12]

[0066]

[Equation 13]

[0067]

[Equation 14]

[0068]

[Equation 15]

[0069]

[Equation 16]

[0070]

[Equation 17]

It becomes

However, as described above, since only the rotational movement and the conversion of enlargement or reduction need be performed in the first affine transformation, it is not necessary to obtain e and f here, but it is necessary in the final affine transformation. So e,
The method of obtaining f is described here.

In this way, by using the coefficients a, b, c and d obtained according to the equations (4) to (7), coordinate transformation is performed according to the equation (1), so that the inclination of the X-ray image 4a can be calculated. It can be made almost equal to the X-ray image 4b.

In the second alignment method of the present invention,
In order to further improve the accuracy, another affine transformation is performed after this first affine transformation. The second affine transformation is to perform template matching in which the template region in the first X-ray image subjected to the first affine transformation matches the reference region in the second X-ray image.

Here, template matching is performed by using the correlation method or SSDA to match the template area with the reference area as described above. In the correlation method, the point having the maximum standardized value gives the coordinates of the sampling point (corresponding point) as described above. Alternatively, also in SSDA, the point at which the sum of residuals is the minimum gives the coordinates of the sampling point, as described above. The affine transformation described above is performed using the coordinates of the sampling points obtained by these methods. That is, each coefficient of the affine transformation shown in equation (1) is obtained. This is (4) ~
Substituting the above sampling points into equation (9), the coefficients a, b,
It suffices to find six of c, d, e, and f.

Here, the principle that the alignment accuracy is improved by repeating the affine transformation a plurality of times will be described with reference to FIGS. 6 and 7.

FIG. 6 is a graph showing the relationship between the relative inclination between a plurality of images and the position error of the corresponding points obtained by template matching and the position error of the true corresponding points. This graph shows that the smaller the inclination of the image, the smaller the positional deviation of the corresponding points. Also,
For example, when the relative inclination of the image is 2 °, it indicates that the error between the corresponding points obtained by the template matching as described above and the true corresponding points is about 1 pixel. The actual displacement between the images is within ± 3 ° because it occurs inside the cassette, the photographing device, and the reading device, but it is confirmed that the relationship shown in FIG. 6 is effective over a range of approximately ± 5 °. ing.

In FIG. 7, the corresponding points are 100 in the X and Y directions.
FIG. 9 is a diagram showing a true corresponding point in the case where the pixels are separated from each other and a position of the corresponding point obtained by template matching. This FIG. 7A shows the true corresponding points, and FIG.
Indicates the corresponding points obtained. The corresponding points shown in FIG. 7B show a state in which the pixel is displaced by one pixel from the true corresponding point when the relative inclination of the image is 2 ° as described above.

Here, in FIG. 7A, the inclination of the straight line connecting the true corresponding points is θr = tan ⁻¹ (100/100) = 45 °. On the other hand, in FIG. 7B, the slope of the straight line connecting the corresponding points estimated by the template matching is θm = tan ⁻¹ (101/99) = 45.6 ° even in the worst case, which is true. It shows that the maximum deviation from the slope of the line connecting the corresponding points is 0.6 °. That is, it is shown that the relative inclination of the image was originally 2 ° but was improved to 0.6 ° by one affine transformation. Then, when template matching is performed by the method of the present invention and second affine transformation is further performed, it can be seen from FIG. 6 that the shift of corresponding points can be suppressed within 0.3 pixels. In this case, the slope of the straight line connecting the corresponding points estimated by the template matching is θm ′ = tan ⁻¹ (100.3 / 99.7) = 45.2 ° even in the worst case, which is the straight line connecting the true corresponding points. The maximum deviation from the tilt is 0.2 °. Ie 2
It is shown that the degree is improved to 0.2 by affine transformation. Therefore, it will be understood that by repeating the affine transformation a plurality of times as described above, the relative shift related to the rotational movement of the image is gradually reduced, that is, the alignment accuracy is gradually improved.

Here, the corresponding points are in the X and Y directions.
An example is shown in which the pixels are separated by 100 pixels, but the present invention is not limited to this. As another example, a case will be described in which corresponding points are located at positions separated by 1000 pixels in the X and Y directions and the relative inclination of the image is 2 °. The slope of the straight line connecting the true corresponding points is the same
45 °. On the other hand, the slope of the straight line connecting the corresponding points estimated by template matching is θm = tan ⁻¹ (1001/999) = 45.06 ° even in the worst case, which is the same as the straight line connecting the true corresponding points. It shows that the maximum deviation is 0.06 °. That is, it shows that the relative inclination of the image was 2 °, but was improved to 0.06 ° by one affine transformation. From this, it can be seen that the accuracy of alignment improves as the distance between corresponding points increases.

[Brief description of drawings]

1 is an example of an apparatus for recording a radiation image X
Schematic diagram of X-ray equipment

FIG. 2 is a diagram schematically showing an X-ray image accumulated and recorded on each stimulable phosphor sheet.

FIG. 3 is an example of an X-ray image reading apparatus that is an example of a reading unit that reads a radiation image used in the present invention and an image processing that is an example of an arithmetic unit that performs the subtraction process by performing the alignment method of the present invention. Perspective view of display device

FIG. 4 is a diagram schematically showing how the X-ray image 4b is rotated and translated using the center of gravity to match the X-ray image 4a.

FIG. 5 is a diagram schematically showing how the X-ray image 4b is rotated and translated using the center of gravity to match the X-ray image 4a.

FIG. 6 is a graph showing the relationship between the relative inclination between a plurality of images and the error between the positions of corresponding points obtained by template matching and the positions of true corresponding points.

FIG. 7 is a diagram showing a position of a true corresponding point and a position of a corresponding point obtained by template matching when the corresponding points are separated by 100 pixels in the X direction and the Y direction.

[Explanation of symbols]

 1 X-ray imaging apparatus 2 X-ray tube 3 X-ray 4 Subject 5 First stimulable phosphor sheet 6 Filter 7 Second stimulable phosphor sheet 8 Closed area 9 Center of gravity 10 X-ray image reading device 15 Sheet conveying means 16 Laser light source 17 Light beam 18 Motor 19 Rotating polygonal mirror 20 Focusing lens 21 Mirror 22 Photostimulated emission light 23 Optical guide 24 Photomultiplier 25 Log amplifier 26 A / D converter 30 Image processing display device 31 Keyboard 32 CRT display 33 Floppy disk Drive unit 34 Main body

Claims (2)

[Claims] 1. A method of aligning a plurality of radiation images for overlaying or subtracting the radiation images, wherein the plurality of images to be aligned have substantially the same shape. Two closed areas are set, each centroid of each closed area is obtained, and each coordinate of each centroid of the reference image among the plurality of images is set to each coordinate of each centroid of the other images. A method for aligning a radiation image, which comprises rotating, enlarging or reducing, and translating the other image so as to match. 2. A method of aligning a plurality of radiation images for overlay processing or subtraction processing of radiation images, wherein at least a plurality of images to be aligned have substantially the same shape. Two closed areas are set, orthogonal coordinates are defined for each of the plurality of images, the center of gravity of each of the closed areas is obtained, and each coordinate of each center of gravity of the reference image of the plurality of images is determined. An equation for converting the coordinate value of the other image into the coordinate value of the reference image so that the coordinates of the centroids of the other image are matched with, (Where u and v are the coordinates of the reference closed area, x and y are the coordinates of the general closed area, a, b, c and d are coefficients that indicate the rotational movement correction and the enlargement or reduction ratio correction, and e and f are the parallel movement corrections. The coefficient of the affine transformation represented by the first affine transformation is performed, and the first affine transformation for performing at least the rotational movement correction and the enlargement or reduction ratio correction on the other image is performed using the coefficient. For multiple images,
At least two regions of interest that are substantially common between the plurality of images are set, each region of the reference image of the plurality of images is set as a reference region, and the regions of other images are set as template regions, Template matching is performed to match the template region with the reference region, the coordinate values of at least two points corresponding to each other in the plurality of images are obtained, and the affine transformation represented by the above equation is performed so that the corresponding points match. A position of a radiographic image characterized by obtaining a coefficient and performing a second affine transformation for performing rotational movement correction, enlargement or reduction ratio correction, and parallel movement correction on a plurality of images including the template region using the coefficient. Matching method.