JP2001307065A - Image processor and image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor capable of easily grasping dislocation and stereoscopic positional relation of two kinds of three-dimensional images for an optional cross section. SOLUTION: The image processor is provided with a first image input part to input first image data to indicate a first three-dimensional image, a second image input part to input second image data to indicate a second three- dimensional image, a data generating part to generate synthesized voxel data in which the first image data and the second image data are periodically and alternately appear on each of three-dimensionally arranged orthogonal coordinate axes by synthesizing the first image data with the second image data and a display image output part to generate and output a two-dimensional image data on the optional specific cross section from the synthesized voxel data. Since the synthesized voxel data in which the first image and the second image are stereoscopically and alternately combined in a single three-dimensional data arrangement is obtained, the dislocation and the positional relation of the two kinds of three-dimensional images are easily grasped.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for processing a three-dimensional image, and more particularly, to two three-dimensional images.
The present invention relates to an image processing apparatus capable of synthesizing a three-dimensional image, generating three-dimensional image data capable of three-dimensionally grasping a positional relationship between the two, and outputting an arbitrary cross-sectional image.

[0002]

2. Description of the Related Art It is often desirable to know the three-dimensional positional relationship between two types of three-dimensional images. For example, X-ray C
T, MRI, SPECT, and the like are medical image diagnostic apparatuses that photograph a human body or the like as a continuous cross-sectional image, and can obtain three-dimensional image information relating to the form or function information of the human body using these. Since these medical image diagnostic devices detect physical information differently, if it is possible to compare image information obtained from different devices, it becomes easy to grasp the anatomical structure of the lesion and to obtain a comprehensive image. Diagnosis can be performed efficiently.

Hitherto, many proposals have been made on a method of aligning images of different formats (modalities) obtained from different medical image diagnostic apparatuses. For example, there are a method of performing positioning using an artificial marker or a frame attached at the time of photographing, a method of performing positioning using anatomical position information of a human body, and the like.

On the other hand, with respect to images of two types of modalities that have been aligned in this way, there are few reports on an image display method suitable for evaluating the alignment accuracy and grasping the positional relationship between the images. As an example of such a report,
"Robert J. Riker et al.:"Human Factors Problems in Opt
imization and Evaluation of RT Image Coregistratio
n ", XIIth, ICCR, pp. 27-30, 1997", it is proposed to display a checkerboard image in which cross-sectional images of CT images and MR images are arranged in a grid pattern.

[0005]

However, since this method compares cross-sectional images, blood vessels of a human body,
It is insufficient when grasping anatomical position information on an object having a complicated three-dimensional structure such as a nerve or a rib. Also, it is difficult to grasp the three-dimensional positional relationship between a plurality of objects such as an organ and a tumor by comparing the cross-sectional images.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image processing apparatus capable of outputting a display image which can easily grasp a positional shift and a three-dimensional positional relationship between two types of three-dimensional images for an arbitrary cross section. And

[0007]

In order to achieve the above object, an image processing apparatus according to the present invention comprises a first image input unit for inputting first image data representing a first three-dimensional image, 3
A second image input unit for inputting second image data representing a three-dimensional image, combining the first image data and the second image data, and combining the first image data and the second image data on each orthogonal coordinate axis of a three-dimensional array. A data generating unit that generates synthesized voxel data in which the second image data alternately appears; and a display image output unit that generates and outputs two-dimensional image data in an arbitrary designated section from the synthesized voxel data. It is characterized by having. According to the image processing device of the present invention,
Since it is possible to generate combined voxel data in which the first image and the second image are combined alternately three-dimensionally in a single three-dimensional data array and output an image at an arbitrary cross section, two types of three-dimensional data are output. It is possible to easily grasp the positional displacement of the image and the three-dimensional positional relationship.

Further, it is preferable that the display image output unit generates display image data from the combined voxel data by volume rendering.

Further, it is preferable that the data generator performs a contrast enhancement process on at least one of the first image data and the second image data. Thereby, even if one of the input images is an image having a low contrast, the contrast of the final composite image is easily adjusted.

[0010] In addition, it is preferable that the data generation unit performs an intensity distribution normalization process for matching the average value and the variance value of the intensity distribution between the first image data and the second image data. Accordingly, even when the intensity distribution of the first image and the intensity distribution of the second image are significantly different, it is possible to prevent a problem that one image has insufficient gradation in the final synthesized image.

Still further, the data generator performs a contrast enhancement process on at least one of the first image data and the second image data, and
An intensity distribution normalization process for matching the average value and the variance value of the intensity distribution between the image data and the second image data may be performed.

Further, the data generation unit converts the intensity distribution of the first image data and the second image data into a new intensity distribution that does not overlap with each other, so that the intensity distribution range of the first image data is It is preferable to separate the intensity distribution range from the second image data. This allows
After combining the images, the first image and the second image can be characterized differently with respect to coloring, opacity, and the like.

The two-dimensional data extracted from the first image data by the data generation unit may include an unmasked area having the same size as the two-dimensional data and data of 0, and a data a (where a is a constant and a (1) multiplying the first mask data in which the mask areas which are (0) alternately exist in a lattice pattern, and multiply the two-dimensional data extracted from the second image data by the first mask data.
The mask data of the mask data is multiplied by the second mask data obtained by exchanging the unmasked area, and the two-dimensional data extracted from the first image data and the two-dimensional data extracted from the second image data are combined. Generating composite two-dimensional data in which first image data and the second image data alternately exist in a grid pattern, and generating the composite two-dimensional data is performed by periodically arranging the first image data and the second image data. It is preferable to generate the synthesized voxel data by repeating so as to be alternately performed.

Further, when the data generating section generates the synthesized voxel data, the data generating section repeats generation of the synthesized two-dimensional data such that the arrangement of the first image data and the second image data is switched. It is preferable to change based on the resolution of the image data or the second image data on each orthogonal coordinate axis. This makes it possible to generate isotropic combined voxel data even when the resolution of the input image data changes or when the input image data has different resolutions in the cross-sectional direction and the slice axis direction. .

Further, the image processing method according to the present invention comprises n (n
Is a natural number), the first image data and the second image data, which are three-dimensional data composed of a set of two-dimensional data, are combined, and the first image data and the second image data are combined on each orthogonal coordinate axis of a three-dimensional array. An image processing method for generating synthetic voxel data in which image data and periodic alternately appear,
(A) a sub-step of extracting the i-th (i is a natural number, 1 ≦ i ≦ n) two-dimensional data from each of the first image data and the second image data; and (b) extracting the two-dimensional data from the first image data. The i-th two-dimensional data is multiplied by first mask data having the same size as the two-dimensional data and having unmasked areas where data is 0 and mask areas where data is a are alternately arranged in a grid pattern. And (c) adding the i-th two-dimensional data extracted from the second image data to
A sub-step of multiplying a second mask data obtained by exchanging a mask area and an unmask area of the first mask data; (d) an i-th two-dimensional data extracted from the first image data; From the i-th two-dimensional data extracted from the sub-steps to generate combined two-dimensional data. Starting from i = 1, (A) sub-steps (a) to (d) While increasing by
repeating k times (k is a natural number, k <n);
(B) The first and the second used in the step (A)
Is the step of repeating sub-steps (a) to (d) k times while increasing i by 1 using the first mask data and the second mask data obtained by exchanging the unmasked area and the masked area. And (C) Steps (A) and (B) are repeated m times (m is a natural number, n / k ≦ m <
n) a step of repeatedly generating and obtaining composite voxel data by stacking and combining the obtained composite two-dimensional data. As a result, it is possible to obtain combined voxel data in which the first image and the second image are combined alternately in a three-dimensional manner in a single three-dimensional data array. The displacement and the positional relationship can be easily grasped.

[0016]

Embodiments of the present invention will be described below with reference to the drawings. The image processing apparatus of the present invention is preferably used for image processing of three-dimensional image data obtained by a medical image diagnostic apparatus such as X-ray CT, MRI, and SPECT. Therefore, in this specification, a case will be mainly described as an example in which image processing for simultaneously displaying a CT image and an MR image having different modalities is performed. Note that the three-dimensional image data is obtained by arranging a plurality of two-dimensional image data in which intensity information is assigned to pixels having position information in the vertical direction and the horizontal direction in the depth direction. It is represented as voxel data in which intensity information is assigned to voxels having direction position information.

Embodiment 1 FIG. 1 is a block diagram illustrating an image processing apparatus according to Embodiment 1 of the present invention. FIG.
, 10 is a first image input unit, 12 is a second image input unit, 22 is a data generation unit that generates synthesized voxel data from the input first image and second image, and 24 is an arbitrary cross section of the synthesized voxel data. A display image output unit that outputs two-dimensional image data. The data generation unit 22 includes contrast enhancement units 14 and 16 that perform a contrast enhancement process on the first image and the second image, and an intensity distribution normalization unit that normalizes the intensity distribution of the first image and the second image. 18 and a combination processing unit 20 for combining the first image and the second image.

FIG. 2 is a flowchart showing the operation of the image processing apparatus shown in FIG. In the present embodiment, n (n is a natural number) slice images (= n is a natural number)
In the case of a set of two-dimensional cross-sectional images, for example, CT
The operation will be described by taking an example of an image or an MRI image as an example.

First, in step S201, the first slice image data (i = 1) of the first image and the second image is input from the first image input unit 10 and the second image input unit 12. The first image and the second image are input from an X-ray CT apparatus, an MRI apparatus, and another information storage medium (not shown in FIG. 1) via an electric communication line. The slice image data may be sequentially extracted from a memory or a hard disk provided in the first and second image input units.

Next, in step S202, contrast enhancement processing of the input slice image data is performed. The contrast enhancement processing can be performed by, for example, performing a linear conversion so that the difference between the maximum value and the minimum value of the intensity in the slice image data becomes the entire gradation width, or smoothing the histogram of the intensity distribution. If the contrast of the input image data is sufficient, the contrast enhancement processing may be omitted.

Next, in step S203, an intensity distribution normalization process is performed on the slice image data of the first image and the second image so as to convert the average value and the variance of the intensity distribution to predetermined values. The normalization processing can be performed using, for example, Equation 1. Instead of performing the normalization process on both the first image and the second image, a normalization process may be performed such that the average value and the variance value of one intensity distribution match the other. If the intensity distribution of the first image and the intensity distribution of the second image are substantially the same, the normalization of the intensity distribution may be omitted.

(Equation 1)

Next, in step S204, the slice image data of the first image and the second image include an unmasked area having the same size as the slice image data and data of 0 and data a (where a is a constant and a ≠ 0). ) Is multiplied by mask data in which mask areas alternately exist in a grid pattern. In this specification, "multiplying" image data means performing integration processing of intensity data between pixels. Further, “masking” data means that only data in a masked area is left. FIGS. 3A and 3B are schematic diagrams showing an example of the mask data. FIG. 3A shows first mask data 30 to be multiplied with the first image, in which unmasked areas 30a and masked areas 30b are alternately arranged in a grid pattern. By multiplying the slice image data of the first image by the first mask data 30, the unmask area 3 in the slice image is obtained.
The area corresponding to 0a has the pixel intensity of 0, and the area corresponding to the mask area 30b has the pixel intensity a times as large as the original.
Therefore, it is possible to obtain slice image data in which the original image is masked in a lattice shape.

On the other hand, FIG. 3B shows the second mask data 32 to be multiplied with the second image. However, the arrangement of the unmasked area and the masked area in the first mask data 30 are interchanged. Therefore, by multiplying the slice image data of the second image by the second mask data, it is possible to obtain a slice image masked in a lattice shape opposite to that of the first image.

Next, in step S205, the slice image data of the first image and the second image subjected to the mask processing are synthesized. The synthesis of the image data can be performed by, for example, an addition process between pixels. Thereby, the first
Synthetic two-dimensional data in which image data and second image data are alternately arranged in a grid pattern can be generated.

The processing from step 201 to step 205 is repeated until it is determined in step S206 that the slice image data is the last slice. This repetition process is performed so that the arrangement of the first image data and the image data in the combined two-dimensional data is periodically switched. For example, for the first mask data to be multiplied with the first image, as shown in FIG. 4A, the first k times (k is a natural number, k <n) have the same data arrangement, and the next k times are unmasked. The data arrangement is such that the area and the mask area are exchanged, and this is alternately repeated. As for the second mask data to be multiplied with the second image, as shown in FIG. 4B, the relationship between the unmasked area and the masked area is always inverted with respect to the first mask data.

Next, in step S207, synthesized voxel data is generated by stacking the synthesized two-dimensional data thus obtained. As schematically shown in FIG. 5, the obtained synthesized voxel data is obtained by combining the first image data 50a and the second image data 50b on each of x, y, and z axes, which are orthogonal coordinate axes of a three-dimensional array. They appear periodically and alternately.

Next, in step S208, an arbitrary section of the composite voxel data is designated according to the input from the user or the like, and two-dimensional display image data of the designated section is output in step S209. In order to generate two-dimensional display image data for an arbitrary cross section of the synthesized voxel data, it is preferable to use a volume rendering technique. Volume rendering raises a ray from a point on the plane on which the three-dimensional image data is projected toward the three-dimensional image data, calculates the total amount of light reflected from voxels passing through this ray, and calculates This is a method for setting the image intensity. The amount of light reflected from each voxel is
This is performed based on the parameter of opacity set for each voxel (0 for transparent, 1 for opaque). That is, the amount of attenuated light incident on each voxel is calculated from the opacity of the voxel through which the light ray has passed, and the intensity of light and the opacity of the voxel are multiplied by the amount of light to calculate the amount of light reflected from each voxel. . Therefore, the designated cross section can be displayed by setting the intensity to 0 for voxels in front of the specified cross section and setting the opacity to 1 for all remaining voxels.

FIGS. 6 to 8 show examples of display images output through the above-described processing, with the first image being an MRI image and the second image being a CT image. The MRI used here
The image and the CT image are obtained by photographing the human head.
In the RI image, soft tissue of the cerebrum hardly reflected in the CT image is depicted, and in the CT image, bone not reflected in the MRI image is depicted. FIG. 6 shows the original MRI image and C
It is an image in a cross section parallel to the slice image plane of the T image. FIG. 7 shows an image in another section parallel to FIG. FIG. 8 shows an image in a cross section orthogonal to the original slice image plane. Although not shown in the figure, a cross-sectional image in which CT images and MRI images are alternately arranged can be output not only on a plane parallel or orthogonal to the slice image plane but also on an oblique plane. As described above, since an image in which MRI images and CT images are alternately arranged for an arbitrary cross section of the observation target can be obtained, evaluation of the alignment accuracy of both images and three-dimensional structure of the tissue appearing in both images can be performed. It is extremely easy to grasp the positional relationship.

The combination of images to be input may be images having different modalities as described in the present embodiment, or may be the same modality, such as an imaging date or an imaging protocol. For example, if images of the affected part at different imaging dates are input, it is possible to three-dimensionally grasp the state of the affected part over time.

Embodiment 2 FIG. From the viewpoint of the visibility of the display image, when the cross section parallel to the slice image plane is displayed (FIGS. 6 and 7) as in the examples shown in FIGS. 6 to 8, the cross section parallel to the slice axis of the image is displayed. It is preferable that the size of the lattice unit when () is displayed (FIG. 8) matches. Therefore,
In the present embodiment, a description will be given of image processing for obtaining combined voxel data in which the size of a lattice unit is isotropic even when the resolution or the like of input image data changes.

The image processing in the present embodiment is the same as that in the first embodiment except that the repetition period using the same mask data is changed so as to satisfy a predetermined relationship. In the present embodiment, in order to match the size of the lattice unit of the cross section parallel to the slice image plane with the cross section parallel to the slice axis, the number g _z of repeating the same two-dimensional data of the same lattice pattern in the slice axis direction Is changed in accordance with Expression 2 according to the input image data.
Note that g _z is the number of repetitions k using the same data arrangement for mask data to be multiplied on the original slice image, as described with reference to FIG.
Becomes equal to

(Equation 2)

Here, x _reso0 is the resolution (= length per unit pixel) of the original image on the slice image plane.
_scale is the magnification of the slice image plane, g _x is the grid unit size represented by the number of pixels on the slice image plane, z _reso
Is the resolution in the slice axis direction (= slice image interval).

Here, the "magnification ratio of the slice image plane" refers to a conversion magnification of the image size between the original image on the slice image plane and the display image. The reason why the enlargement ratio of the slice image plane is included in Expression 2 is that the lattice unit size is isotropic even when the original image is enlarged or reduced and displayed. By providing the data generation unit 22 with means for changing the number of repetitions g _z (= k) so as to satisfy such a relationship, when the resolution of the input image data in the slice axis direction changes, or in the slice image plane, And the resolution in the slice axis direction is different,
A display image having an isotropic lattice unit size can be output.

Embodiment 3 After the first image and the second image are combined to generate combined voxel data, if the first image and the second image can be separately colored and feature values such as opacity can be set, the first image and the second image can be combined. Images can be easily compared. Therefore, in the present embodiment, when generating the composite voxel data, the intensity distribution of the first image data and the second image data are converted so as not to overlap each other, so that the intensity distribution range of the first image data and the second The intensity distribution range of the image data is separated.

The image processing in the present embodiment is the same as that in the first embodiment except that a predetermined intensity distribution conversion is performed before the first image data and the second image data are combined.
FIG. 9 is a schematic diagram showing how the combined voxel data generation unit 22 of the image processing apparatus shown in FIG. 1 converts the intensity distribution of the first image data and the second image data.
This conversion can be performed, for example, after normalization by the intensity distribution normalization unit 18 and before synthesis by the synthesis processing unit 20. As shown in FIG. 9, the intensity gradation of the first image 402 on which contrast enhancement and normalization has been performed is compressed, shifted and converted into an intensity range of 0 to (N / 2-1). The intensity gradation of the second image 404 on which contrast enhancement and normalization have been performed is converted into an intensity range of (N / 2) to (N-1). After performing the conversion process in this way,
By synthesizing the first image and the second image, the first image and the second image are stored in the image data 406 after the synthesis in such a manner that they are separated in gradation.

Therefore, when the characteristic values such as coloring and opacity are set, the set values are changed depending on the gradation, so that the first image and the second image can be characterized independently. In addition, since the first image and the second image are displayed with different gradations when displayed on an arbitrary cross section of the synthesized voxel data, it is possible to easily determine which image each grid is from. Become. 7 and 8
Is an example in which the first image and the second image are differently colored according to the present embodiment.

[0037]

According to the image processing apparatus of the present invention, synthesized voxel data in which a first image and a second image are three-dimensionally combined in a single three-dimensional data array is obtained, and an arbitrary cross section is displayed. Since images can be output, it is possible to grasp the characteristics of the positional relationship between the two types of three-dimensional images over the entire three-dimensional object and three-dimensionally grasp the difference between the two images based on this data. Become.

Further, since the data display unit displays an arbitrary cross section by volume rendering, various characterizations on a display image are facilitated.

Further, the data generation unit performs a contrast enhancement process on at least one of the first image data and the second image data, so that even if one of the input images is an image having a low contrast, the final composite image is obtained. The contrast of the image can be easily adjusted.

In addition, the data generation unit performs an intensity distribution normalization process for matching the average value and the variance value of the intensity distribution between the first image data and the second image data, thereby obtaining the first image and the second image. Even if the intensity distributions of the images are significantly different, it is possible to prevent the problem that one image has insufficient gradation in the final synthesized image.

Still further, the data generating section performs a contrast enhancement process on at least one of the first image data and the second image data, and performs the first image data and the second image data.
By performing the intensity distribution normalization processing for matching the average value and the variance value of the intensity distribution with the image data, the image quality of the final composite image can be further easily adjusted.

Further, the data generation section separates the intensity distribution range of the first image data from the intensity distribution range of the second image data, so that the first image and the second image are colored after the images are combined. Can be differently characterized, and the difference between the first image and the second image can be more easily grasped.

Further, the data generation unit generates data using mask data in which mask areas and unmask areas alternately exist in a grid pattern, so that the first image and the second image are three-dimensionally combined. Voxel data can be generated by simple processing.

Further, by changing the cycle of repeating the generation of the combined two-dimensional data so that the arrangement of the first image data and the second image data is switched based on the resolution of the input image data on each orthogonal coordinate axis. Even when the resolution of the input image data changes or when the input image data has different resolutions in the cross-sectional direction and the slice axis direction, it is possible to generate isotropic combined voxel data.

According to the image processing method of the present invention, it is possible to obtain combined voxel data in which the first image and the second image are three-dimensionally combined in a single three-dimensional data array. Of the first image and the second image based on
The positional deviation and the positional relationship of the two-dimensional image can be easily grasped.

[Brief description of the drawings]

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

FIG. 2 is a flowchart illustrating an operation of the image processing apparatus according to the first embodiment of the present invention.

FIGS. 3A and 3B are schematic diagrams showing first and second mask data according to the first embodiment of the present invention.

FIGS. 4A and 4B are schematic diagrams showing an array of mask data according to the first embodiment of the present invention.

FIG. 5 is a schematic diagram showing combined voxel data according to Embodiment 1 of the present invention.

FIG. 6 is an image photograph showing an example of an image output by the image processing apparatus according to the present invention.

FIG. 7 is an image photograph showing another example of an image output by the image processing apparatus according to the present invention.

FIG. 8 is an image photograph showing still another example of an image output by the image processing apparatus according to the present invention.

FIG. 9 is a schematic diagram illustrating intensity distribution conversion processing according to Embodiment 3 of the present invention.

[Explanation of symbols]

10 1 image input unit, 12 second image input unit, 14 and 16 contrast enhancement unit, 18 intensity distribution normalization unit, 20 synthesis processing unit, 22 data generation unit, 24 display image output unit, 30 and 32 mask data, 50 Synthetic voxel data.

Continued on the front page F term (reference) 4C096 AB50 AD12 AD14 DA11 DA30 DC11 DC27 DC33 DC36 DC37 5B057 AA07 BA03 BA07 BA24 CA13 CB12 CE08 CE09 CE11 CE20 5C076 AA02 AA11 AA40 BA06 CA02

Claims (9)

[Claims] An image processing apparatus for combining two three-dimensional image data and outputting display image data, wherein a first image data representing a first three-dimensional image is input.
An image input unit; and a second input unit that inputs second image data representing a second three-dimensional image.
An image input unit; combining the first image data and the second image data;
A data generation unit that generates composite voxel data in which the first image data and the second image data periodically and alternately appear on each orthogonal coordinate axis of a three-dimensional array; An image processing apparatus including a display image output unit that generates and outputs two-dimensional image data. 2. The image processing apparatus according to claim 1, wherein the display image output unit generates two-dimensional image data by volume rendering. 3. The image processing apparatus according to claim 1, wherein the data generator performs a contrast enhancement process on at least one of the first image data and the second image data. 4. The apparatus according to claim 1, wherein the data generation unit performs an intensity distribution normalization process for matching an average value and a variance value of an intensity distribution between the first image data and the second image data. The image processing apparatus according to any one of the preceding claims. 5. The data generation unit performs a contrast enhancement process on at least one of the first image data and the second image data, and further performs a contrast enhancement process on the intensity distribution between the first image data and the second image data. 2. The image processing apparatus according to claim 1, wherein an intensity distribution normalization process for matching the average value and the variance value is performed. 6. The image processing apparatus according to claim 1, wherein the data generation unit converts the intensity distributions of the first image data and the second image data into a new intensity distribution that does not overlap with each other. . 7. The two-dimensional data extracted from the first image data by the data generation unit,
Multiplying unmasked areas having the same size as the dimension data and having data of 0 and mask areas having data of a (a is a constant, a ≠ 0) by first mask data alternately in a grid pattern; The two-dimensional data extracted from the image data is multiplied by second mask data obtained by exchanging a mask area and an unmasked area of the first mask data. The two-dimensional data extracted from the first image data and the second image Combining the two-dimensional data extracted from the data,
Generating composite two-dimensional data in which the first image data and the second image data alternately exist in a grid pattern, and generating the composite two-dimensional data is performed by arranging the first image data and the second image data. The image processing apparatus according to claim 1, wherein the composite voxel data is generated by repeating the switching processing periodically. 8. The method according to claim 1, wherein the data generating unit generates the combined voxel data by combining the first image data with the second voxel data.
The image processing apparatus according to claim 7, wherein a cycle at which the arrangement of the image data is switched is changed based on a resolution of each of the first image data and the second image data on each orthogonal coordinate axis. 9. The first image data and the second image data, which are three-dimensional data including a set of n (n is a natural number) two-dimensional data.
An image processing method of combining image data to generate synthesized voxel data in which the first image data and the second image data periodically and alternately appear on each orthogonal coordinate axis of a three-dimensional array, a) i-th (i is a natural number, 1 ≦ i ≦) from each of the first image data and the second image data
(b) a sub-step of extracting two-dimensional data; and (b) an unmasked area having the same size as the two-dimensional data and having zero data is added to the i-th two-dimensional data extracted from the first image data. Substep of multiplying first mask data in which mask areas where a is a (a is a constant, a ≠ 0) are alternately arranged in a grid pattern; and (c) the i-th second mask data extracted from the second image data. In the dimensional data, the first
Sub-step of multiplying the second mask data obtained by exchanging the mask area and the unmask area of the mask data; and (d) the i-th second data extracted from the first image data.
A sub-step of combining the dimensional data with the i-th two-dimensional data extracted from the second image to generate combined two-dimensional data; starting from i = 1, (A) sub-steps (a) to (a) (D) is increased k times by 1 (k is a natural number,
k <n) a repeating step; and (B) the first mask data and the second mask data used in the step (A) are a first mask data and a second mask data obtained by exchanging an unmasked area and a masked area. Using,
Sub-steps (a) to (d) are repeated k times while incrementing i, and (C) steps (A) and (B) are repeated m times (m is a natural number, n / k ≦ m < n) an image processing method, comprising a step of repeatedly arranging the obtained combined two-dimensional data to generate combined voxel data.