JP2014006100A - Image processing system, image processing method, and image processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing system, an image processing method, and an image processing program for specifying the area of a small solid object included in a solid object in a slice image photographing the section of the solid object.SOLUTION: The control unit 21 of an image processing device 20 executes recording processing of a slice image and input processing of a slice interval and magnification. Then, the control unit 21 executes alignment processing in two continuous slice images. The control unit 21 executes, for a specified organic contour in the first slice image, contour alignment processing for the other slice image. Then, the control unit 21 executes two-dimensional evaluation processing for a cell included in each slice image. The control unit 21 executes three-dimensional evaluation processing for identifying similar or different cells among the specified cells of the respective images to count the number of cells, and outputs the number of cells.

Description

  The present invention relates to an image processing system, an image processing method, and an image processing program for specifying a region of a small three-dimensional object included in a three-dimensional object in slice images obtained by photographing the three-dimensional object at a plurality of cross-sectional positions.

  Conventionally, in order to confirm a pathological cell in a specimen tissue, the state of the cell is sometimes evaluated using an image obtained by photographing a cross section of the specimen tissue. Therefore, an image search system for searching cell images from a morphological viewpoint has been studied (for example, see Patent Document 1). In the image database of the image search system described in this document, cell images and cell morphological feature information are stored in association with each other. Then, the image database is searched based on the morphological characteristics of the cell input as the search condition, and the cell image that matches the search condition is displayed on the display unit.

  Furthermore, in the examination of cells, abnormal regions may be counted, and an image filing system for improving the work efficiency of blood cell examinations has been studied (for example, see Patent Document 2). In the image filing system described in this document, a CCD camera that images blood cells in a blood smear and outputs an image signal, a table that associates a specimen number with patient information, and a display device that generates an image And a computer to be displayed. Then, by inputting the sample number, the corresponding patient information is searched from the table. Then, the patient information, the blood cell image, and the classification count value obtained by counting the blood cell classification data are displayed on the display device, and the blood cell image is stored in the database.

  In addition, a system for acquiring and analyzing an image pair obtained as a section of a sample has been studied (for example, see Patent Document 3). In the technique described in this document, an image of each section of a sample is registered. A mathematical transformation rule is then used for each image field identified in one image so that the corresponding image field in the other image can be identified as well. The section is then evaluated by identifying counting events for the objects on the image field in the acquired corresponding image pair using automatic means.

Japanese Patent Laying-Open No. 2003-30202 (first page, FIG. 2) Japanese Patent Laying-Open No. 2005-100452 (first page, FIG. 1) Japanese translation of PCT publication 2010-511213 (first page, FIG. 2)

  A slice image obtained by photographing a cross section of an evaluation target object (for example, a specimen tissue) that is a large three-dimensional object may be created, and an inclusion (for example, a pathological cell) that is a small three-dimensional object included in the slice image may be observed. . In this case, the number of inclusions (number of cells) at the cross-sectional position of the inclusion in the evaluation object can be calculated. Here, when a plurality of slice images are used, the number of cells can be grasped for each cross-sectional position.

  However, if the distance between the cross-sectional positions is more than the size of the inclusion, the inclusions existing between the cross-sections may not be counted. On the other hand, when the distance between the cross-sectional positions is shorter than the size of the inclusion, it cannot be determined whether the inclusion is the same inclusion or another inclusion. For this reason, there is a possibility that the same inclusion is counted twice in a plurality of cross sections. Furthermore, there are cases where the slice image includes distortion. In this case, the same inclusion cannot be determined based on the slice image, and accurate counting becomes difficult.

Thus, when counting the number of inclusions using a plurality of images taken at discrete cross-sectional positions, it is difficult to calculate an accurate number of inclusions.
In addition, when calculating the number of inclusions using image processing of slice images, there is a possibility that the number of slice images is large, the load of image processing increases, and calculation time is required. It is therefore desirable to be able to calculate as efficiently and quickly as possible.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to specify a region of a small three-dimensional object included in a three-dimensional object in a slice image taken at a plurality of cross-sectional positions of the three-dimensional object. An image processing system, an image processing method, and an image processing program are provided.

  In order to solve the above problem, the invention described in claim 1 is included in the first three-dimensional object, and a slice image storage unit that registers slice images taken at a plurality of cross-sectional positions for the first three-dimensional object. An image processing system including a control unit that identifies a region of a second three-dimensional object, wherein the control unit calculates a difference between slice images at successive slice positions recorded in the slice image storage unit. , Means for calculating deformation information for generating a deformed image in which the region of the second three-dimensional object included in the slice image is aligned based on the difference, and a second included in each deformed image The gist is provided with means for specifying a region of a three-dimensional object and means for specifying an image of the same second three-dimensional object included in the continuous deformed image based on the specified region.

  According to a second aspect of the present invention, in the image processing system according to the first aspect, the second solid object region is aligned in the continuous deformed image, and the same second solid object image is specified. The gist is to do.

  According to a third aspect of the present invention, in the image processing system according to the first or second aspect of the present invention, the calculation is based on the interval between the cross-sectional positions where the slice image is taken and the statistical value of the size of the second three-dimensional object. In addition, the gist is to specify an image of the same second three-dimensional object by using the number of slice images including one second three-dimensional object.

  The invention according to claim 4 is the image processing system according to any one of claims 1 to 3, using a continuous area change of the region of the second three-dimensional object included in each deformation image, The gist is to specify an image of the same second three-dimensional object.

  According to a fifth aspect of the present invention, in the image processing system according to any one of the first to fourth aspects, the first second three-dimensional object image identified in the continuous deformed image is used. The gist is to calculate the quantity of the second three-dimensional object included in the three-dimensional object.

  According to a sixth aspect of the present invention, in the image processing system according to any one of the first to fifth aspects, the control means is designated for a range for specifying the second three-dimensional object in one slice image. The present invention further includes means for acquiring a contour and means for deforming the contour based on deformation information of each slice image.

  A seventh aspect of the present invention is the image processing system according to any one of the first to sixth aspects, wherein a low-resolution image is created for each slice image, and the continuous low-resolution image is used to generate the low-resolution image. The gist is to perform alignment of the region of the second three-dimensional object included in the slice image, and to perform alignment of consecutive slice images using the alignment result.

  According to an eighth aspect of the present invention, the slice image storage means for registering slice images taken at a plurality of cross-sectional positions for the first three-dimensional object, and the region of the second three-dimensional object included in the first three-dimensional object are specified. An image processing system comprising a control means for performing image processing, wherein the control means is recorded in the slice image storage means and calculates a difference between slice images at successive cross-sectional positions. Then, based on the difference, a step of calculating deformation information for generating a deformed image in which the region of the second three-dimensional object included in the slice image is aligned, and a second included in each deformed image And a step of specifying the same second three-dimensional object image included in the continuous deformed image based on the specified region.

  The invention according to claim 9 specifies slice image storage means in which slice images taken at a plurality of cross-sectional positions for the first three-dimensional object are registered, and specifies a region of the second three-dimensional object included in the first three-dimensional object A program for executing image processing using an image processing system comprising a control means for performing the control, wherein the control means is recorded in the slice image storage means and is a difference between slice images at successive cross-sectional positions. And means for calculating deformation information for generating a deformed image in which the region of the second three-dimensional object included in the slice image is aligned based on the difference, and the first included in each deformed image And a means for specifying a region of the two-dimensional object, and a function for specifying an image of the same second three-dimensional object included in the continuous deformed image based on the specified region. That.

(Function)
According to the invention described in claim 1, 8 or 9, the control means is recorded in the slice image storage means, calculates a difference between slice images at successive cross-sectional positions, and is included in the slice image based on the difference. Deformation information for generating a deformed image obtained by aligning the region of the second three-dimensional object is calculated. Next, the region of the second three-dimensional object included in each deformed image is specified, and the image of the same second three-dimensional object included in the continuous deformed image is specified based on the specified region. Accordingly, since the second three-dimensional object can be grasped in the continuous images, the three-dimensional structure of the second three-dimensional object included in the first three-dimensional object can be specified. Further, even when the slice image includes distortion, the three-dimensional structure can be accurately specified.

  According to the second aspect of the present invention, the regions of the second three-dimensional object are aligned in the continuous deformed images, and the same second three-dimensional object image is specified. Thereby, the same 2nd solid object image | photographed in the some slice image can be pinpointed based on the position of the area | region of a 2nd solid object.

  According to the third aspect of the present invention, one second three-dimensional object calculated based on the interval between the cross-sectional positions at which the slice images are taken and the statistical value of the size of the second three-dimensional object is included. The image of the same second three-dimensional object is specified using the number of slice images. Thereby, even when a plurality of second solid objects are continuously present in the same cross-sectional area, it is possible to determine whether or not they are the same solid objects in consideration of the size of the second solid object.

  According to invention of Claim 4, the image of the same 2nd solid object is pinpointed using the change of the continuous area of the area | region of the 2nd solid object contained in each deformation | transformation image. Thereby, even when a plurality of second solid objects are continuously present in the same cross-sectional area, it is possible to determine whether or not they are the same solid object in consideration of the shape of the second solid object.

  According to the fifth aspect of the invention, the number of the second three-dimensional objects included in the first three-dimensional object is calculated using the same second three-dimensional object image specified in the continuous deformed images. Thereby, since the quantity of a 2nd solid object can be grasped | ascertained, the internal condition of a 1st solid object can be grasped | ascertained. For example, when a tissue specimen is used as the first three-dimensional object and a pathological cell is used as the second three-dimensional object, the degree of progression of the disease state can be evaluated based on the number of pathological cells.

  According to the sixth aspect of the present invention, the control means acquires the contour designated for the range specifying the second three-dimensional object in one slice image, and the contour is based on the deformation information of each slice image. And deform. Thus, by designating one contour, the contour can be specified also in other slice images, so that the evaluation target range for specifying the second three-dimensional object can be set efficiently.

  According to the seventh aspect of the present invention, a low resolution image is created for a slice image. Then, the regions of the second three-dimensional object included in the slice image are aligned using the continuous low resolution images, and the alignment of the continuous slice images is performed using the alignment result. Thereby, after performing rough alignment with the low-resolution image, detailed alignment with the high-resolution image is performed, so that high-speed and high-precision alignment can be realized.

  According to the present invention, it is possible to provide an image processing system, an image processing method, and an image processing program for specifying a small solid object region included in a three-dimensional object in a slice image obtained by photographing a cross section of the three-dimensional object. Can do.

The system schematic of one Embodiment of this invention. It is explanatory drawing of the data recorded on each data storage part, (a) is a slice image database, (b) is a deformation | transformation information database, (c) is a tissue outline database, (d) is a cell database in a slice, ( e) Explanatory drawing of the data recorded on the grouping database. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. It is explanatory drawing of the deformation | transformation of the image of other embodiment of this invention, Comprising: (a) is an initial image, (b) is the image after whole deformation | transformation, (c) is explanatory drawing of the image after dividing and deform | transforming . Explanatory drawing of the process sequence of other embodiment of this invention.

  Hereinafter, an embodiment embodying the present invention will be described with reference to FIGS. In the present embodiment, an image for counting the number of inclusions (second three-dimensional objects) that are small three-dimensional objects included in an evaluation object (first three-dimensional object) that is a large three-dimensional object by image processing. The processing apparatus will be described.

  Here, it is assumed that the number of pathological cells (second solid object) included in the specimen tissue (first solid object) is counted. For this purpose, the sliced tissue is sliced at equal intervals (slice intervals), and a slice image obtained by enlarging the cross section is prepared. This pathological cell has an ellipsoidal shape. For this reason, each pathological cell is represented by an elliptical shape on the slice image by staining the pathological cell in the cross section of the specimen tissue.

For this image processing, as shown in FIG. 1, an image processing apparatus 20 connected to an input unit 10 and an output unit 15 is used.
The input unit 10 includes input means for inputting various instructions such as a keyboard and a pointing device. The output unit 15 includes an output unit such as a display for outputting information processing results.

  The image processing apparatus 20 is a computer terminal that calculates the number of cells of a pathological tissue included in a plurality of slice images obtained by photographing a cross section of a specimen tissue. The image processing apparatus 20 includes a control unit 21, a slice image database 22, a deformation information database 23, a tissue contour database 24, an intra-slice cell database 25, and a grouping database 26.

  The control unit 21 functions as a control unit including a CPU, a RAM, a ROM, and the like, and will be described later (processing including a management stage, an image adjustment stage, a contour adjustment stage, a two-dimensional evaluation stage, a three-dimensional evaluation stage, and the like). To do. By executing an image processing program for this purpose, the control unit 21 functions as a management unit 211, an image adjustment unit 212, a contour adjustment unit 213, a two-dimensional evaluation unit 214, a three-dimensional evaluation unit 215, and the like.

  The management unit 211 acquires a slice image to be processed, and executes processing for controlling each unit to be described later. The management unit 211 includes a memory that temporarily stores a slice interval and a magnification when a slice image is created.

  The image adjustment unit 212 executes a process of aligning the pathological cells included in the slice images so as to ensure the continuity of the pathological cells included in the plurality of slice images. The image adjusting unit 212 generates a low resolution layer image when performing alignment. For this reason, the image adjustment means 212 holds data relating to the maximum number of layers (NLmax) including the same size layers. Further, the image adjusting unit 212 holds data relating to a reference value for determining the magnitude of the pattern distribution error on the image.

The contour adjusting unit 213 executes processing for specifying an image range (contour) in which pathological cells are counted in the slice image.
The two-dimensional evaluation unit 214 executes a process of counting the number of pathological cell images taken in each slice image. This two-dimensional evaluation means 214 holds data relating to a short diameter candidate, a long diameter candidate, and a tilt candidate used for fitting using an elliptical shape. Further, the two-dimensional evaluation unit 214 includes a memory that temporarily stores a black image obtained by binarizing a deformed image obtained by deforming a slice image. In this black image, a pathological cell is specified by painting a region to which an elliptical shape is applied in white.

  The three-dimensional evaluation means 215 determines the same or different specific cells in consideration of the three-dimensional structure of the pathological cells, and executes a process of counting the number of pathological cells included in the specimen tissue. The three-dimensional evaluation means 215 holds data relating to a reference distance for determining whether or not the images of the same pathological cells are present in a plurality of slice images.

  Further, the three-dimensional evaluation unit 215 holds a maximum value calculation function for calculating the maximum predicted cell length as a statistical value for specifying the size of a pathological cell photographed at the magnification at which the slice image is photographed. Yes. The statistical value may be a statistical value representative of the size of the pathological cell, and is not limited to the maximum cell length. For example, an average value or the like can be used.

  The slice image database 22 functions as slice image storage means. As shown in FIG. 2A, slice image management data 220 is recorded in the slice image database 22 for an image obtained by photographing each cross section obtained by slicing a specimen tissue. The slice image management data 220 is recorded when a slice image is designated by the user. The slice image management data 220 includes data related to an image number, a slice image, and a use determination flag.

  Data relating to an identifier for specifying each slice image is recorded in the image number data area. As this image number, a sequential number in the order of the cross-sectional positions is set in order from the initial value (here, “0”).

In the slice image data area, data relating to an image obtained by photographing the tissue structure at the cross-sectional position specified by the image number is recorded.
In the use determination flag data area, data relating to an identifier for determining whether or not the slice image can be used is recorded. Specifically, when there is a large deviation from other slice images and alignment cannot be performed, a non-use flag indicating that this slice image is not used is recorded.

  In the deformation information database 23, as shown in FIG. 2B, deformation information management data 230 for deforming the slice image is recorded in order to ensure the continuity of the photographed cells. This deformation information management data 230 is recorded when the alignment processing of each slice image is executed. The deformation information management data 230 includes data relating to image numbers and deformation information.

Data relating to an identifier for specifying each slice image is recorded in the image number data area.
Data relating to a function for deforming each slice image is recorded in the deformation information data area. In this data area, a matrix for performing this projective deformation is recorded.

  As shown in FIG. 2C, the tissue contour database 24 records tissue contour management data 240 for specifying a range in which pathological cells are counted in the slice image. The tissue contour management data 240 is recorded when the contour alignment processing of the slice image is executed. The tissue contour management data 240 is configured to include data relating to image numbers and contour positions.

Data relating to an identifier for specifying each slice image is recorded in the image number data area.
In the contour position data area, data relating to coordinates for specifying a range in which pathological cells are counted in the slice image is recorded.

  In the intra-slice cell database 25, as shown in FIG. 2D, intra-slice cell management data 250 for specifying a pathological cell included in each slice image is recorded. This intra-slice cell management data 250 is recorded when the two-dimensional evaluation process is executed. The intra-slice cell management data 250 includes data relating to an image number, a cell image, and a binarized image. In addition, the cell management data 250 in the slice is associated with individual cell data 251 configured to include data related to the center position coordinate, the minor axis, the major axis, and the inclination.

Data relating to an identifier for specifying each slice image is recorded in the image number data area.
In the cell image data area, data relating to a deformed image obtained by deforming the slice image by alignment is recorded.
In the binarized image data area, data related to the binarized image of the deformed image is recorded.

  Data relating to the coordinates of the center position of the pathological cell specified in the binarized image is recorded in the center position coordinate data area.

In each data area of the short diameter and the long diameter, data relating to the short diameter and the long diameter when the pathological cell is expressed in an elliptical shape is recorded.
In the tilt data area, data relating to the tilt with respect to a predetermined axis (for example, the horizontal axis) of the binarized image in the elliptical shape is recorded.

  In the grouping database 26, as shown in FIG. 2E, grouping management data 260 for specifying a group of pathological cells photographed with a plurality of slice images in consideration of the three-dimensional structure is recorded. Is done. This grouping management data 260 is recorded when the three-dimensional evaluation process is executed. The grouping management data 260 includes data relating to a group code, an image number, and center position coordinates.

In the group code data area, data relating to an identifier for specifying a group of pathological cells is recorded.
Data relating to an identifier for specifying each slice image is recorded in the image number data area.

  In the center position coordinate data area, data relating to an identifier for specifying each pathological cell in the slice image is recorded. In the present embodiment, each pathological cell is identified using the coordinates of the center position.

Next, in the image processing apparatus 20 configured as described above, a processing procedure in the case of evaluating a small pathological cell contained in a specimen tissue will be described with reference to FIGS.
(Overview of the whole)
First, the overall outline will be described with reference to FIG. When evaluating small pathological cells contained in a sample tissue, the sample tissue is sliced at a predetermined interval (slice interval). Then, a pathological cell is stained in each cross section, and a slice image file photographed at a predetermined magnification is created. This slice image file is stored in a storage unit (hard disk or the like) of the image processing apparatus 20 or a storage medium. In this case, a plurality of slice images can be included in one file, or a plurality of slice image files in which file names are set in the sliced order can be used. Then, the image processing program of the image processing apparatus 20 is activated.

  In this case, the control unit 21 of the image processing apparatus 20 executes slice image recording processing (step S1-1). Specifically, the management unit 211 of the control unit 21 outputs to the output unit 15 an image designation screen for designating a slice image obtained by photographing the sample tissue to be evaluated. Here, the slice image to be evaluated recorded in the storage unit (hard disk or the like) of the image processing apparatus 20 or the storage medium is designated using the image designation screen. In this case, the management unit 211 acquires the designated slice image and assigns an image number based on the order included in the file or the order of the file names. Then, the management unit 211 generates slice image management data 220 in which the slice image associated with the image number is recorded, and registers it in the slice image database 22.

  Next, the control unit 21 of the image processing apparatus 20 executes input processing of a slice interval and a magnification (Step S1-2). Specifically, the management unit 211 of the control unit 21 outputs an image attribute setting screen to the output unit 15. In this image attribute setting screen, an input field for a slice interval when slicing a specimen tissue and a magnification when a slice image is created is provided. When the completion of input to each input field is detected, the management unit 211 temporarily stores the slice interval and the magnification in the memory.

  Next, the control unit 21 of the image processing apparatus 20 executes slice image alignment processing (step S1-3). Specifically, the image adjusting means 212 of the control unit 21 uses the slice images recorded in the slice image database 22 to perform image deformation for alignment of each image so as to ensure the continuity of the cell structure. To do. The deformation information calculated by this process is recorded in the deformation information database 23. Details will be described later with reference to FIGS.

  Next, the control unit 21 of the image processing apparatus 20 executes contour alignment processing (step S1-4). Specifically, the contour adjusting unit 213 of the control unit 21 acquires the contour specified by the user in one slice image. Next, the contour adjusting unit 213 uses the deformation information recorded in the deformation information database 23 for the specified contour, identifies the contour in all other slice images, and records it in the tissue contour database 24. . Details will be described later with reference to FIG.

  Next, the control unit 21 of the image processing apparatus 20 executes a two-dimensional evaluation process (step S1-5). Specifically, the two-dimensional evaluation unit 214 of the control unit 21 specifies pathological cells included in the contour recorded in the tissue contour database 24 using the slice image deformed by the deformation information. Then, the two-dimensional evaluation unit 214 records the specified pathological cell in the in-slice cell database 25. Details will be described later with reference to FIG.

  Next, the control unit 21 of the image processing apparatus 20 executes a three-dimensional evaluation process (step S1-6). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 calculates the number of cells included in the specimen tissue in consideration of the three-dimensional structure of the pathological cells recorded in the intra-slice cell database 25. Details will be described with reference to FIG.

  Next, the control part 21 of the image processing apparatus 20 performs a cell number output process (step S1-7). Specifically, the management unit 211 of the control unit 21 displays the number of cells calculated by the three-dimensional evaluation unit 215 on the output unit 15.

(Slice image alignment processing)
Next, the slice image alignment process will be described with reference to FIG. Here, in order to ensure the continuity of the cell structure, the positional relationship between successive slice images is adjusted.

  First, the control unit 21 of the image processing apparatus 20 executes a low-resolution layer image creation process for the first image (step S2-1). Specifically, the image adjusting unit 212 of the control unit 21 sets the layer number [L] to the maximum number of layers (NLmax) for the slice image with the smallest image number (here, 0th) in the slice image database 22. The low-resolution layer image Img [0] [L] with the resolution lowered is sequentially generated at a predetermined magnification until the value reaches. Here, [0] and [L] are variables indicating the image number and the layer number, respectively. In this embodiment, the difference in resolution between layers is ½, and the minimum resolution is [2 ^ (− NLmax + 1)].

Then, the following processing is repeated for the image number.
Here, when the process for the I-th image is performed, the control unit 21 of the image processing apparatus 20 executes a low-resolution layer image creation process (step S2-2). Specifically, the image adjusting unit 212 of the control unit 21 sequentially sets the resolution of the slice number of the image number [I] in the slice image database 22 at a predetermined magnification up to the maximum number of layers (NLmax). A low-resolution layer image Img [I] [L] is generated.

For the layer number [L], the following processing is repeated in order from the image with the lowest resolution.
Here, the control unit 21 of the image processing apparatus 20 executes alignment processing for each layer image (step S2-3). Specifically, the image adjustment unit 212 of the control unit 21 performs processing on the previous slice image in which the non-use flag is not recorded in the slice image management data 220 with respect to the processing target image Img [I] [L]. An image (effective image Img [I-x] [L]) is used. Here, “I-x” means an image number traced back to an effective image in which no unused flag is recorded. The image adjusting unit 212 compares the image Img [I] [L] to be processed and the effective image Img [Ix] [L], and determines the deformation matrix Matmin that minimizes the pattern distribution error on the image. A deformed image Imgmin is obtained. Details will be described later with reference to FIG.

  Next, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether or not the error is equal to or less than a reference value (step S2-4). Specifically, the image adjusting unit 212 of the control unit 21 compares the error calculated in the previous step with a reference value.

  When the error is not equal to or less than the reference value (in the case of “NO” in step S2-4), the control unit 21 of the image processing device 20 executes the discard processing of the slice image (step S2-5). Specifically, the image adjustment unit 212 of the control unit 21 records a non-use flag in the slice image management data 220. Then, the alignment for this slice image is completed.

  On the other hand, if the error is equal to or smaller than the reference value (“YES” in step S2-4), the control unit 21 of the image processing device 20 repeats the above processing until the processing for the equal layer number is completed. .

Next, the control unit 21 of the image processing apparatus 20 executes a process for outputting deformation information and a deformed image (step S2-6). Specifically, the image adjustment unit 212 of the control unit 21 generates the deformation information management data 230 and records it in the deformation information database 23.
The above processing is repeated until the processing for all image numbers is completed.

(Alignment processing of each layer image)
Next, the alignment process of each layer image will be described with reference to FIG. In the present embodiment, alignment is performed using projective deformation in which the coordinates of the four corners of the image are independently moved and deformed. Here, the lowest resolution images are compared, roughly aligned, and the deformation determined at a lower resolution is used to calculate a more closely matched deformation to increase the alignment accuracy. .

First, the control unit 21 of the image processing apparatus 20 executes a maximum movement width calculation process (step S3-1). Specifically, the image adjusting unit 212 of the control unit 21 determines the maximum movement width dm that can be moved according to the resolution here. Specifically, since the alignment is performed on the low-resolution image first, the image is moved within a range not exceeding the minimum movement amount used in the previous alignment. Therefore, the following setting is made according to the layer.
In the case of L = 0, dm = dmax * 2 ^ (-NLmax + 1)
When L> 0, dm = 1

Next, the control unit 21 of the image processing apparatus 20 executes initial value calculation processing for the four vertices at the four corners of the image (step S3-2). Specifically, the image adjusting unit 212 of the control unit 21 sets the following values as initial values for the coordinates P0 [L] [I] of the four vertices.
When L = 0, position before movement When L> 0, P0 [L] [I] = P0 [L-1] [I] * 2

  Next, the control unit 21 of the image processing apparatus 20 executes a process of setting the sum of the least square errors (step S3-3). Specifically, the image adjusting unit 212 of the control unit 21 sets the minimum square error sum SSDmin. Here, a large value is set as an initial value as the minimum square error sum SSDmin.

Then, the following processing is repeated while moving the vertexes (P [I]) at the four corners of the slice image.
Here, the control unit 21 of the image processing apparatus 20 executes deformation information creation processing (step S3-4). Specifically, the image adjustment unit 212 of the control unit 21 creates a projection deformation matrix.

  Next, the control unit 21 of the image processing apparatus 20 executes a process for creating a deformed image of the image (step S3-5). Specifically, the image adjustment unit 212 of the control unit 21 applies the created matrix to the slice image to create the deformed image Img2.

  Next, the control unit 21 of the image processing apparatus 20 executes a calculation process of the sum of square errors (step S3-6). Specifically, the image adjustment unit 212 of the control unit 21 calculates a square error sum SSD of the deformed image Img2 and the image Img [0] [L].

  Next, the control unit 21 of the image processing apparatus 20 executes a determination process as to whether or not the sum of the square errors is small (step S3-7). Specifically, the image adjustment unit 212 of the control unit 21 compares the calculated square error sum SSD with the minimum square error sum SSDmin.

When it is determined that the sum of the square errors is small (in the case of “YES” in step S3-7), the control unit 21 of the image processing apparatus 20 executes a data update process (step S3-8). Specifically, the image adjustment unit 212 of the control unit 21 updates the following temporarily stored data.
SSDmin = SSD
Pmin [I] = P [I]
Matmin = Mat
Imgmin = Img2
On the other hand, when it is determined that the total sum of the square errors is not small (in the case of “NO” in step S3-7), the control unit 21 of the image processing device 20 skips the data update process (step S3-8).
The above processing is repeated for the vertices at the four corners of the image.

Next, the control unit 21 of the image processing apparatus 20 executes update processing of vertex position information and images at the four corners (step S3-9). Specifically, the image adjustment unit 212 of the control unit 21 updates each parameter as follows.
Img [0] [L] = Img [1] [L]
P0 [L] [I] = Pmin [I]
As a result, the deformation matrix can be obtained by comparing the preceding and subsequent images.

(Outline alignment processing)
Next, the contour alignment process will be described with reference to FIG.
First, the control unit 21 of the image processing apparatus 20 executes a first tissue contour input process (step S4-1). Specifically, the contour adjusting unit 213 of the control unit 21 displays the first slice image (image number I = 0) recorded in the slice image database 22 on the output unit 15. The contour adjusting unit 213 displays a graphical user interface (GUI) for setting a contour in the slice image. Using this GUI, the user sets a tissue contour R [0] that is a range for evaluating pathological cells. In this case, the contour adjusting unit 213 generates tissue contour management data 240 based on the set tissue contour R [0] and stores the tissue contour management data 240 in the tissue contour database 24.

Then, the following processing is repeated for the image number.
Here, when the process for the I-th image is performed, the control unit 21 of the image processing apparatus 20 executes a contour calculation process in the image to be processed (step S4-2). Specifically, the contour adjusting unit 213 of the control unit 21 acquires the deformation information management data 230 of the image number I recorded in the deformation information database 23, and deforms the coordinates of the tissue contour R [I-1]. A tissue contour R [I] is calculated. The contour adjusting unit 213 temporarily stores the calculated tissue contour R [I] in the memory.

  Next, the control unit 21 of the image processing device 20 executes display processing of an image and a contour (step S4-3). Specifically, the contour adjusting unit 213 of the control unit 21 displays on the output unit 15 a contour confirmation image in which a contour image based on the calculated tissue contour R [I] is superimposed on the slice image with the image number I. In the contour confirmation screen, a confirmation completion button and a correction button are provided.

  Next, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether correction is necessary (step S4-4). Specifically, the contour adjusting unit 213 of the control unit 21 determines that correction is necessary when detecting the pressing of the correction button.

  When it is determined that correction is necessary (in the case of “YES” in step S4-4), the control unit 21 of the image processing apparatus 20 executes a tissue contour correction process (step S4-5). Specifically, the contour adjusting unit 213 of the control unit 21 adjusts the tissue contour R [I] by moving the coordinates of the contour image superimposed on the slice image using the GUI. The contour adjusting unit 213 temporarily stores the adjusted tissue contour R [I] in the memory. If it is determined that correction is not necessary (NO in step S4-4), the control unit 21 of the image processing apparatus 20 skips the process of step S4-5.

Next, the control unit 21 of the image processing apparatus 20 executes tissue contour output processing (step S4-6). Specifically, the contour adjusting unit 213 of the control unit 21 generates tissue contour management data 240 that records the tissue contour R [I] of the contour displayed on the contour confirmation screen, and records the tissue contour management data 240 in the tissue contour database 24.
The above processing is repeated until the processing for all image numbers is completed.

  Next, the control unit 21 of the image processing apparatus 20 performs a tissue contour smoothing process (step S4-7). Specifically, the contour adjusting unit 213 of the control unit 21 smoothes the contour coordinates by averaging the contour positions of the central slice image using a plurality of consecutive front and rear slice images. Then, the contour adjusting unit 213 updates the contour position data recorded in each tissue contour management data 240 using the coordinates of the smoothed contour.

(2D evaluation process)
Next, the two-dimensional evaluation process will be described with reference to FIG. Here, pathological cells included in each slice image are identified by fitting using an elliptical shape.

The following processing is repeated for the image number.
When processing the I-th image, first, the control unit 21 of the image processing device 20 binarizes the image and executes black region registration processing (step S5-1). Specifically, the two-dimensional evaluation unit 214 of the control unit 21 binarizes the deformed image recorded in the slice image database 22. Here, the method of Otsu, which is a method for determining an appropriate threshold when a set of values is classified into two classes, is used. Here, in order to minimize the variance within the two classes and maximize the variance between the classes, a threshold value that minimizes the ratio between the intra-class variance and the inter-class variance is obtained. Next, the two-dimensional evaluation unit 214 binarizes the deformed image using this threshold value. Then, the two-dimensional evaluation unit 214 records the black image generated by binarization in the in-slice cell management data 250.

  Next, the following processing is repeated in the black region included in the binarized image for the minor axis candidate, major axis candidate, and inclination candidate of the ellipse. In this case, the short diameter candidate and the long diameter candidate are used in order from the long candidate.

  Here, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether or not an ellipse enters the black area (step S5-2). Specifically, the two-dimensional evaluation unit 214 of the control unit 21 applies an ellipse composed of a short diameter candidate, a long diameter candidate, and a tilt candidate to each black region, and confirms whether or not the ellipse protrudes.

  When it is determined that an ellipse enters the black area (in the case of “YES” in step S5-2), the control unit 21 of the image processing device 20 executes a process of painting the fitted ellipse white (step S5-3). . Specifically, the two-dimensional evaluation unit 214 of the control unit 21 repaints the fitted elliptical area to white in the black area of the binarized image recorded in the black area database.

Next, the control unit 21 of the image processing apparatus 20 executes registration processing using the fitted ellipse as a cell (step S5-4). Specifically, the two-dimensional evaluation unit 214 of the control unit 21 generates individual cell data 251 in which an elliptical shape (short diameter, long diameter, inclination) is recorded in the intra-slice cell management data 250, and the intra-slice cell management data The data is registered in the in-slice cell database 25 in association with 250.
When it is determined that an ellipse does not enter the black area (“NO” in step S5-2), the control unit 21 of the image processing device 20 skips the processes in steps S5-3 to S5-4.
The above processing is repeated for every image number until the processing for the elliptic candidates of the inclination candidate, the major diameter candidate, and the minor diameter candidate is completed for all black regions.

(Three-dimensional evaluation process)
Next, the three-dimensional evaluation process will be described with reference to FIG. Here, the number of pathological cells specified in the two-dimensional image is counted in consideration of the three-dimensional structure of the pathological cells.

Again, the following processing is repeated for the image number.
When the process for the I-th image is performed, first, the control unit 21 of the image processing apparatus 20 executes a tissue (inside the contour) alignment process (step S6-1). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 uses a slice image in which the non-use flag is not recorded in the use determination flag data area in the slice image management data 220 recorded in the slice image database 22 (valid A preceding slice image) to which the previous image number is assigned is specified. Next, the three-dimensional evaluation unit 215 obtains the tissue contour R [I] of continuous effective slice images from the tissue contour database 24. Then, the three-dimensional evaluation unit 215 performs alignment of cells included in the range specified by the tissue contour R [I].

Next, the control unit 21 of the image processing apparatus 20 executes a registration process for cells in substantially the same position as a cell group (step S6-2). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 identifies the cell having the closest center position coordinate as the same cell group when the cell having the center position coordinate within the reference distance is detected in the preceding slice image. To do. In this case, the three-dimensional evaluation unit 215 generates grouping management data 260 in which the same group code as that of the closest cell is assigned to the image number and the center position coordinates, and records them in the grouping database 26. On the other hand, in the new slice image, when a cell whose center position coordinate is within the reference distance cannot be detected in the preceding slice image, grouping management data 260 in which a new group code is assigned to the image number and the center position coordinate. Is generated and recorded in the grouping database 26.
The above processing is repeated until the processing for all image numbers is completed.

  Next, the following processing is repeated for the cell groups recorded in the grouping database 26.

  Here, the control unit 21 of the image processing apparatus 20 executes the smoothing process of the cell size in the group (step S6-3). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 acquires the short diameter and long diameter of the cells registered in the group from the individual cell data 251. Next, the three-dimensional evaluation unit 215 calculates the size using the minor axis and the major axis in the grouped cells. For this size, for example, an area (= π [minor axis] · [major axis]) is used. Then, the three-dimensional evaluation unit 215 calculates a moving average of the sizes for successive image numbers, and smoothes each size.

  Next, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether or not the cell size has reversed from a decrease to an increase (step S6-4). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 calculates a change tendency of the size when rearranged in the order of the image numbers. Here, the three-dimensional evaluation unit 215 determines whether or not the current change tendency has been reversed to the increase tendency when there is a decrease tendency in the preceding image.

When it is determined that the cell size is reversed from the decreasing tendency to the increasing tendency (in the case of “YES” in step S6-4), the control unit 21 of the image processing apparatus 20 executes the cell group dividing process at the corresponding location ( Step S6-5). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 divides and registers a new group in the slice image in which the change tendency of the group is reversed in the grouping database 26. Here, the three-dimensional evaluation unit 215 re-assigns a new group code to the image numbers after the image number of the slice image, and updates the group code of the grouping management data 260.
On the other hand, when it is determined that the cell size has not reversed from the decreasing tendency to the increasing tendency (in the case of “NO” in step S6-4), the control unit 21 of the image processing device 20 skips the process of step S6-5. To do.

  Next, the control unit 21 of the image processing apparatus 20 executes a calculation process of a continuous number upper limit value corresponding to the slice interval and the magnification (step S6-6). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 acquires the slice interval and the magnification recorded in the management unit 211. Then, the three-dimensional evaluation means 215 introduces the acquired magnification into the maximum value calculation function and calculates the maximum cell length. Next, the three-dimensional evaluation unit 215 uses the calculated maximum cell length and the slice interval to calculate the number of images (the maximum number of consecutive images) in which one maximum cell may be included in a continuous slice image. To do.

  Next, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether or not the slice number is greater than the continuous number upper limit value (step S6-7). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 calculates the number of slice images (slices) in which cells are photographed from the difference between the minimum value and the maximum value of the image numbers assigned with the same group code. calculate. Then, the three-dimensional evaluation unit 215 compares the number of slices with the continuous number upper limit value.

When the number of slices exceeds the continuous number upper limit value (in the case of “YES” in step S6-7), the control unit 21 of the image processing device 20 executes the cell group division process with the continuous number upper limit value (step S6-7). S6-8). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 registers a new group by dividing the slice image in the grouping database 26 that exceeds the upper limit value of the continuous number in this group. Here, the three-dimensional evaluation unit 215 re-assigns a new group code to the image numbers after the image number of the slice image, and updates the group code of the grouping management data 260. On the other hand, when the number of slices does not exceed the upper limit of consecutive numbers (in the case of “NO” in step S6-7), the control unit 21 of the image processing device 20 skips the process of step S6-8.
The above processing is repeated until the processing for all the cell groups recorded in the grouping database 26 is completed.

  And the control part 21 of the image processing apparatus 20 performs the calculation process of the number of cells in a tissue (step S6-9). Specifically, the three-dimensional evaluation unit 215 of the control unit 21 counts the number of records of the individual cell data 251 recorded in the grouping database 26. Then, the three-dimensional evaluation unit 215 supplies the counted number of records to the management unit 211 as the number of pathological cells.

According to the image processing system of the present embodiment, the following effects can be obtained.
(1) In this embodiment, the control unit 21 of the image processing apparatus 20 performs slice image recording processing (step S1-1), slice image positioning processing (step S1-3), and contour positioning processing (step S1). -4) is executed. Next, the control unit 21 executes a two-dimensional evaluation process (step S1-5) and a three-dimensional evaluation process (step S1-6). And the control part 21 performs a cell number output process (step S1-7). Thereby, the inclusion situation of the 2nd solid object in the 1st solid object can be grasped. For example, the degree of progression of a disease state can be evaluated based on the number of pathological cells.

  (2) In the present embodiment, the control unit 21 of the image processing device 20 executes slice image alignment processing (step S1-3). Here, the control unit 21 of the image processing apparatus 20 executes a low-resolution layer image creation process (step S2-2). And the control part 21 of the image processing apparatus 20 performs the alignment process of each layer image about all the layer numbers (step S2-3). As a result, by creating a plurality of images with reduced resolution and comparing them in order from the smallest image, it is possible to achieve high speed in rough alignment processing of the two images. Furthermore, high accuracy can be achieved by aligning the high resolution image using the alignment result of the low resolution image. Furthermore, it is also possible to generate a three-dimensional model of a cell unit by performing well-known volume rendering using a slice image aligned in a high resolution image and to display it three-dimensionally.

  (3) In the alignment processing of each layer image in the present embodiment, the control unit 21 of the image processing device 20 executes deformation information creation processing (step S3-4). Specifically, the image adjustment unit 212 of the control unit 21 creates a projection deformation matrix. And the control part 21 of the image processing apparatus 20 performs the production | generation process of the deformation | transformation image of an image (step S3-5). By using projective deformation, unlike affine deformation, arbitrary deformation is possible and accurate alignment can be performed.

  (4) In the contour alignment process of the present embodiment, the control unit 21 of the image processing apparatus 20 executes the first tissue contour input process (step S4-1) and the contour calculation process in the processing target image ( Step S4-2). Thereby, the number of pathological cells in a desired region can be counted in the slice image. Furthermore, the tissue outline specified first can be developed into another slice image.

  Further, the control unit 21 of the image processing apparatus 20 executes an image and contour display process (step S4-3) and a determination process (step S4-4) as to whether correction is necessary. When it is determined that correction is necessary (in the case of “YES” in step S4-4), the control unit 21 of the image processing apparatus 20 executes a tissue contour correction process (step S4-5). Thereby, when the development of the tissue contour with respect to another slice image is not appropriate, it can be corrected efficiently.

  Further, the control unit 21 of the image processing apparatus 20 performs a tissue contour smoothing process (step S4-7). Thereby, the blurring of the shape of the tissue contour registered in the tissue contour database 24 can be adjusted.

  (5) In the two-dimensional evaluation process of the present embodiment, the control unit 21 of the image processing apparatus 20 binarizes the image and executes a black area registration process (step S5-1). Then, the control unit 21 of the image processing apparatus 20 sets the short diameter candidate and the long diameter candidate in order from the long candidate for each black region included in the binarized image with respect to the short diameter candidate, the long diameter candidate, and the inclination candidate of the ellipse. Then, a determination process as to whether or not an ellipse enters the black area is executed (step S5-2). When it is determined that an ellipse is included in the black area (in the case of “YES” in step S5-2), the control unit 21 of the image processing device 20 applies the process of painting the applied ellipse white (step S5-3). The registration process (step S5-4) is executed with the ellipse as a cell. Thereby, each pathological cell contained in a slice image can be specified.

  (6) In the three-dimensional evaluation process of the present embodiment, the control unit 21 of the image processing apparatus 20 performs tissue (inside the contour) alignment process (step S6-1), and cells at substantially the same position are defined as cell groups. A registration process (step S6-2) is executed. When the pathological cell is large, it is included in a plurality of slice images. Even in this case, the images of pathological cells taken in a plurality of slice images can be collected.

  (7) In the three-dimensional evaluation process of the present embodiment, the control unit 21 of the image processing apparatus 20 determines whether the cell size within the group is smoothed (step S6-3), and whether the cell size is reversed from increase to decrease. The determination process (step S6-4) is executed. When it is determined that the cell size is reversed from the decreasing tendency to the increasing tendency (in the case of “YES” in step S6-4), the control unit 21 of the image processing apparatus 20 executes the cell group dividing process at the corresponding location. (Step S6-5). Thereby, even when a plurality of pathological cells are continuously photographed in the same region in a plurality of slice images, it is possible to determine whether the cells are the same based on the shape of the pathological cells.

  (8) In the three-dimensional evaluation process according to the present embodiment, the control unit 21 of the image processing apparatus 20 executes a process for calculating the continuous number upper limit value corresponding to the slice interval and the magnification (step S6-6). Then, the control unit 21 of the image processing apparatus 20 performs a determination process as to whether or not the slice number is greater than the continuous number upper limit value (step S6-7). When the number of slices exceeds the continuous number upper limit value (in the case of “YES” in step S6-7), the control unit 21 of the image processing device 20 executes the cell group division process with the continuous number upper limit value (step S6-7). S6-8). Thereby, even when a separate cell cannot be identified based on the last shape, it can be determined whether it is the same cell based on the statistical size of a pathological cell.

Moreover, you may change the said embodiment as follows.
In the above embodiment, it is assumed that the number of pathological cells (second solid object) included in the specimen tissue (first solid object) is counted. The application target of the present invention is not limited to this, and can be applied to the identification of the second three-dimensional object included in the first three-dimensional object.

  In the above embodiment, in the two-dimensional evaluation process, cells are specified by an ellipse having a short diameter and a long diameter. The shape to be applied is not limited to an ellipse. It is also possible to perform fitting by a circular shape. In this case, each process of steps S5-2 to S5-4 is repeated for the diameter candidate.

  In the above embodiment, in the alignment process for each layer image, the control unit 21 of the image processing apparatus 20 executes a deformation information creation process (step S3-4). Specifically, the image adjustment unit 212 of the control unit 21 creates a projection deformation matrix. The deformation method is not limited to this, and free deformation based on a ratio, affine deformation, or a combination thereof can be used. In the free deformation by the ratio, the deformation by the ratio is performed using four reference points. In affine deformation, deformation is performed by combining linear transformation (rotation, enlargement / reduction, shear) and parallel movement.

  Further, as a modification method, it is also possible to use a phase only correlation method that measures the similarity by taking the correlation of only the phase characteristics. This phase-only correlation method can obtain information on rotation, parallel movement, and enlargement / reduction, but cannot perform arbitrary deformation. Therefore, as shown in FIG. 9, an arbitrary quadrangular deformation is realized by dividing the slice image into four and comparing the divided images into four.

Here, it is assumed that the image A0 and the image B0 shown in FIG. The deformation information generation process (step S3-4) in this case will be described.
In FIG. 9A, since the image is biased, the control unit 21 of the image processing apparatus 20 performs overall alignment by using the phase only correlation method. As a result, as shown in FIG. 9B, an image B1 obtained by performing overall deformation on the image B0 is acquired. Further, the control unit 21 of the image processing apparatus 20 divides the image A0 and the image B1 into four as shown in FIG. Then, the control unit 21 of the image processing apparatus 20 acquires an image B2 obtained by transforming each divided image of the image A0 using the phase only correlation method for each divided image of the image B1.

  Next, how to obtain deformation information regarding an arbitrary quadrangle will be described with reference to FIG. Here, since the deformation of an arbitrary quadrangle may not be expressed by a matrix, “coordinates before and after the deformation for four points (representative points) in the image” is stored as the deformation information. Specifically, the center of gravity of the four squares after deformation is used as the representative point.

  First, the control unit 21 of the image processing device 20 executes the process of specifying the center of gravity in the image after the entire deformation (step S7-1). Specifically, the image adjusting unit 212 of the control unit 21 divides the image B0 into four, and specifies the coordinates of the center position (indicated by “+” in the drawing) of each divided region.

  Next, the control unit 21 of the image processing apparatus 20 performs a process of specifying the corresponding position of the center of gravity after the entire deformation in the original image (step S7-2). Specifically, the image adjusting unit 212 of the control unit 21 performs the inverse transformation of the overall deformation on the center position of each divided region after the entire deformation, thereby returning the position (the center of gravity in the original image). ).

  Next, the control unit 21 of the image processing apparatus 20 executes a process of specifying the corresponding position of the center of gravity after the entire deformation in the image after the division deformation (step S7-3). Specifically, the image adjusting unit 212 of the control unit 21 performs division transformation conversion on the center position of each divided area after the entire deformation, thereby obtaining a position in the image after the division transformation (in the image after the division transformation). Specify the coordinates of the center of gravity.

  Next, the control unit 21 of the image processing device 20 executes a calculation process of deformation information from the center of gravity of the original image to the center of gravity of the image after the split deformation (step S7-4). Specifically, the image adjusting unit 212 of the control unit 21 calculates deformation information for moving from the coordinates of the center of gravity in the original image to the coordinates of the center of gravity in the image after the split deformation.

  If the transformation of an arbitrary square can be expressed by a matrix, it can be obtained in the form of [deformation matrix] = [deformation matrix of the entire transformation] × [deformation matrix of the partial transformation]. In some cases, it cannot be expressed. Even in this case, the deformation can be expressed by using “coordinates before and after the deformation for four points in the image” as the deformation information.

DESCRIPTION OF SYMBOLS 10 ... Input part, 15 ... Output part, 20 ... Image processing apparatus, 21 ... Control part, 211 ... Management means, 212 ... Image adjustment means, 213 ... Contour adjustment means, 214 ... Two-dimensional evaluation means, 215 ... Three-dimensional evaluation Means 22 ... Slice image database, 23 ... Deformation information database, 24 ... Tissue contour database, 25 ... In-slice cell database, 26 ... Grouping database.

Claims (9)

Slice image storage means for registering slice images taken at a plurality of cross-sectional positions for the first three-dimensional object;
An image processing system including a control unit that identifies a region of a second three-dimensional object included in the first three-dimensional object,
The control means is
A modified image that is recorded in the slice image storage means, calculates a difference between slice images at successive cross-sectional positions, and performs alignment of a region of the second three-dimensional object included in the slice image based on the difference Means for calculating deformation information for generating
Means for specifying a region of the second three-dimensional object included in each deformation image;
An image processing system comprising: means for specifying an image of the same second three-dimensional object included in the continuous deformed image based on the specified region.   2. The image processing system according to claim 1, wherein in the continuous deformed image, the second solid object region is aligned and an image of the same second solid object is specified.   Using the same number of slice images including one second three-dimensional object calculated based on the interval between the cross-sectional positions at which the slice images were taken and the statistical value of the size of the second three-dimensional object. The image processing system according to claim 1, wherein an image of two three-dimensional objects is specified.   The image of the same 2nd solid object is specified using the change of the continuous area of the area | region of the 2nd solid object contained in each deformation | transformation image, The any one of Claims 1-3 characterized by the above-mentioned. The image processing system described.   The number of second three-dimensional objects included in the first three-dimensional object is calculated using images of the same second three-dimensional object specified in the continuous deformed images. The image processing system according to any one of the above. The control means is
Means for acquiring a contour designated for a range specifying the second three-dimensional object in one slice image;
The image processing system according to claim 1, further comprising: a unit that deforms the contour based on deformation information of each slice image. Create a low-resolution image for each slice image,
The continuous low-resolution image is used to align the region of the second three-dimensional object included in the slice image, and the alignment result is used to align the continuous slice image. The image processing system according to claim 1. Slice image storage means for registering slice images taken at a plurality of cross-sectional positions for the first three-dimensional object;
A method for performing image processing using an image processing system including a control unit that identifies a region of a second three-dimensional object included in the first three-dimensional object,
The control means is
A modified image that is recorded in the slice image storage means, calculates a difference between slice images at successive cross-sectional positions, and performs alignment of a region of the second three-dimensional object included in the slice image based on the difference Calculating deformation information for generating
Identifying a region of a second three-dimensional object included in each deformed image;
And a step of specifying an image of the same second three-dimensional object included in the continuous deformed image based on the specified region. Slice image storage means for registering slice images taken at a plurality of cross-sectional positions for the first three-dimensional object;
A program for executing image processing using an image processing system including a control unit that specifies a region of a second three-dimensional object included in the first three-dimensional object,
The control means;
A modified image that is recorded in the slice image storage means, calculates a difference between slice images at successive cross-sectional positions, and performs alignment of a region of the second three-dimensional object included in the slice image based on the difference Means for calculating deformation information for generating
Means for specifying the region of the second three-dimensional object included in each deformation image;
An image processing program that functions as means for specifying an image of the same second three-dimensional object included in the continuous deformed image based on the specified region.

Images