JPH07120438B2 - 3D image processor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application] The present invention processes three-dimensional image data of a voxel structure of an object, for example, a human head (hereinafter referred to as “voxel data”). The present invention relates to a three-dimensional image processing device.

(Prior Art) Conventionally, maxillofacial surgery, which is the main purpose of cosmetics, has been performed in surgery. In this operation, as shown in FIG. 17, the circumference of the orbit B1, B2 is cut with a bone saw A1 as shown by a broken line, and the nasal vertical bone B3 is cut with a periosteal dissector A2 as shown in FIG.
Then, the orbital parts B1 and B2 are moved to the center of the face as shown in FIG. The voids C1 and C2 generated by the movement of the orbits B1 and B2 are filled with the one obtained by cutting a part of the hip bone or the like.

In such maxillofacial surgery, the positioning of the cutting site is extremely important in order to properly move and fit the orbital part. This is because it is necessary to move the orbital portion to a predetermined area and fit it into the area appropriately by cutting with the minimum necessary amount.

Therefore, in order to support the maxillofacial surgery, it is earnestly desired to provide an image processing apparatus capable of simulating the surgery.

According to the conventional image processing apparatus, pseudo three-dimensional image display can be performed based on the head slice image group obtained by, for example, the X-ray CT apparatus. This pseudo three-dimensional image display is
From voxel data obtained by appropriately interpolating data between slices of a slice image group, a distance image on a desired viewing plane (also called a depth image, a front image or a perspective image, which corresponds to the distance, and the distance is long. It has a density value that becomes darker as it becomes) and is displayed on the CRT display. The cutting site can be designated by ROI (region of interest) setting, and the partial cutting of the image can be performed by partially clearing the voxel data by applying the ray tracing method in creating the range image.
It is considered that the orbital part can be moved by a partial image moving process (coordinate conversion process).

(Problems to be Solved by the Invention) However, in the conventional device, even if the ROI can be set in the cutting portion, it is determined whether or not the cutting position is appropriate, that is, whether the cutting position is specified properly. It is difficult to confirm sex.

Also, when cutting a three-dimensional image, it is performed by partially clearing the voxel data by applying the ray tracing method, but conventionally, ray tracing is performed in the normal direction from each point on the preset visual plane. However, there is a drawback that the cutting direction is different from that in the actual surgery. Therefore, in the conventional device, it is impossible to perform an accurate simulation such as a maxillofacial surgery.

Therefore, the present invention eliminates the above-mentioned drawbacks, and an object thereof is to make it possible to easily confirm the adequacy of the designation of the cutting position, and to perform the same cutting process as in the case of actual surgery. The purpose is to

[Configuration of the Invention] (Means for Solving the Problems) The present invention provides, in a three-dimensional image processing apparatus, input means for setting visual planes at different positions outside the three-dimensional image, and Distance image creating means for creating a distance image for each viewing plane by measuring the distance from the viewing plane to the surface of the three-dimensional image, and display means for displaying a plurality of created distance images on the same screen are displayed. ROI setting processing means that enables ROI setting on the three-dimensional image by ROI drawing on the range image.

Further, a plane memory for storing the distance image and an overlay memory for storing the ROI depiction information are provided, and a display control means for synthesizing the storage contents of both memories and displaying them on the display means is provided.

Here, as the display control means, one having a function of displaying the ROI drawing information in the overlay memory on the display means in a color different from that of the distance image may be applied.

In addition to the above configuration, the image cutting processing device is provided which performs cutting processing from each point of ROI in the three-dimensional image toward the center of the three-dimensional image.

Here, in the cutting processing device of the cutting processing means,
It is advisable to determine the cutting direction from the gray value of the distance image at each ROI point.

(Operation) When the input unit sets the view planes at different positions outside the three-dimensional image, the range image creation unit creates a range image for each of the view planes. This distance image is displayed on the same screen by the display means. The operator looks at this display image
ROI drawing can be done. The ROI setting information is used to set the ROI for the three-dimensional image. This ROI setting processing is executed by the ROI setting processing means. As described above, in the present invention, the ROI setting can be performed for a plurality of viewing planes by using the distance images, so that the validity of the cutting position designation can be easily confirmed. Here, by providing the plane memory for storing the range image and the overlay memory for storing the ROI depiction information, the ROI can be corrected without the loss of the range image. Further, by assigning a color different from the range image to the ROI depiction information in the overlay memory, ROI display with excellent visibility can be performed.

Furthermore, in the image cutting processing means, the cutting processing is performed from each point of the ROI toward the center of the three-dimensional image, so that the cutting is the same as in the case of actual surgery, which is extremely difficult to perform in simulation. It is convenient for. Here, such a cutting direction can be easily determined based on the gray value of the distance image at each point of the ROI.

(Examples) Hereinafter, the present invention will be specifically described with reference to Examples.

FIG. 1 (a) shows an embodiment of the present invention.

1 is the first voxel memory (512 × 512 × 512),
The first voxel memory 1 stores three-dimensional image data (voxel data) having a voxel structure formed based on a plurality of head slice images acquired by, for example, an X-ray CT scanner. The voxel data referred to here is assumed to be formed based on the head slice image that has been binarized using the CT value 700 as a threshold value, for example.

Reference numeral 2 denotes an input processing means. The input processing means 2 executes recognition processing of input information input by the keyboard 14a by the trackball 14b. The recognition processing result of the input processing means 2 is the first distance image creating means 3, the ROI setting processing means 5,
The second distance image creating means 9 and the fitting processing means 13 are incorporated. The keyboard 14a and the trackball 14b are used to input various processing conditions to the apparatus of this embodiment.
Done by For example, the keyboard 14a and the trackball 14b can be used to set the visual plane in creating a range image, set the ROI (meaning the region of interest), and input the moving direction and the moving amount in the process of fitting the cutout part such as the orbital part. it can. In this sense, the keyboard 14a and the trackball 14b are collectively referred to as the input means 14.

Here, in the apparatus of this embodiment, the input planes 14 can set the visual planes at different positions outside the three-dimensional image. For example, when the three-dimensional image in the first voxel memory 1 is a skull image 15 as shown in FIG. 2, the first visual plane is in front of the skull image 15.
16 is set, the second visual plane 17 is set to the side of the skull image 15, and the third visual plane 18 is set above the skull image 15.
The first, second and third viewing planes 16, 17, 18 are orthogonal to each other.

Then, the first range image creating means 3 measures the distances from the first, second, and third viewing planes 16, 17, 18 to the surface of the skull image 15 (for example, by the ray tracing method) to obtain each viewing plane. It has a function to create a range image for each. The first, second and third viewing planes 16,17,
Distance images at 18 are shown in FIGS. 3, 4, and 5, respectively.

Further, in FIG. 1A, reference numeral 4 is a first plane memory, and the first plane memory 4 stores the distance image created by the first distance image creating means 3. 512 × 5 matrix size for each range image
If the number is 12, the first plane memory 4 has a size of 1024 × 1024, and the distance images shown in FIGS. 3 to 5 are combined and stored as shown in FIG. Here, each pixel value of the distance image is represented by 8 bits. Therefore, this first
The plain memory 4 is 1024 × 1024 × as shown in FIG.
It has an 8-bit structure. The contents stored in the first plane memory 4 are displayed on the display means, for example, the CRT display 12, under the control of the display control means 11 arranged in the subsequent stage.

The ROI setting processing means 5 takes in ROI input information from the input means 14 via the input processing means 2 and performs ROI setting processing. The processing in the ROI setting processing means 5 specifies the area for setting the ROI on the three-dimensional image by drawing the ROI on the distance image and the area for calculating the void area in the processing by the fitting processing means 13 described later. There are two types of cases. The ROI setting information in the former case is transmitted to the image cutting processing means 7, and the ROI setting information in the latter case is transmitted to the fitting processing means 13. The ROI depiction information is written in the overlay memory 6, and the display control means
Displayed on CRT display 12 under control of 11.

Here, the overlay memory 6 and the first plane memory 4
The relationship with is explained. The overlay memory 6 is 1024 × 1024 × 1 bit as shown in FIG. 8, and coincides with the first plane memory 4 in terms of address. The ROI drawing information stored in the overlay memory 6 is displayed by the display control means 11
Under the control of, the superimposed image is displayed on the distance image in the first plane memory 4. When the distance image is displayed in black and white light and shade (the light and shade level corresponds to the distance from the viewing plane), the ROI depiction information in the overlay memory 6 is displayed in color. For example, red is assigned to the ROI drawing information in the overlay memory 6, and the CRT display 12
In ROI, only the ROI description information is displayed in red. The color processing of the ROI drawing information is based on RGB conversion in the display control means 11.

The image cutting processing means 7 takes in the ROI setting information by the ROI setting processing means 5 and, with respect to the three-dimensional image data, this ROI setting information.
The cutting process is applied to the three-dimensional area specified by. In the cutting processing by the image cutting processing means 7, ray tracing processing is performed toward each center of the three-dimensional image (skull image 15) by each point of ROI in the three-dimensional image, and voxel data is sequentially cleared to zero. It becomes possible by going. The cutting direction is determined based on the gray value of the distance image at each ROI point. The result of this processing is written in the second voxel memory 8 arranged in the subsequent stage.

Reference numeral 9 is a second distance image creating means, and this second distance image creating means 9 creates a distance image on a desired visual plane from the three-dimensional image data after the cutting processing. The visual plane in creating this range image is set by the input means 14. The distance image thus created is displayed on the second plane memory 10 via the display control means 11 and the fitting processing means.
It is designed to be incorporated into 13.

The fitting processing means 13 moves the three-dimensional area (cutout portion) cut out by the image cutting processing means 7 to a predetermined three-dimensional area and performs fitting processing on the area.
The processing result is transmitted to the display control means 11.

Next, the detailed configuration of the fitting processing means 13 will be described with reference to FIG.

As shown in FIG. 7B, this fitting processing means 13
y-direction moving means 13a, z-direction moving means 13b, overlapping portion detecting means 13
c, overlapping portion area measuring means 13d, void detecting means 13e, void area measuring means 13f, z direction deviation detecting means 13g, z value measuring means 13h
Have.

The x- and y-direction moving means 13a executes a moving process of the cutout portion in accordance with the cutout portion movement information received from the input processing means 2. The movement of the cutout portion is performed by performing coordinate conversion of the cutout portion based on the coordinate information input by operating the input means 14, for example, the trackball 14b. Also this
In the x, y-direction moving means 13a, the cutout portion can be automatically fitted. This automatic fitting moves the cutout vertically and horizontally (that is, in the y direction and the x direction) around the reference point designated by the input means 14 to minimize the void area when the cutout is fitted in the desired area. It is performed by ending the movement of the cutout portion in the state.

Further, the z-direction moving means 13b moves the cutout portion in the z-direction (here, the z-direction is a direction orthogonal to the view plane in the second distance image creating means 9).
The movement of the cutout portion in the z direction appears as a change in the gray value on the range image.

The overlapping portion detecting means 13c detects an image overlapping portion accompanying movement of the cutout portion, and the detection result is sent to the display control means 11 and the overlapping portion area measuring means 13d. The overlapping portion is displayed in a display mode that can be distinguished from the surroundings under the control of the display control means 11. Specifically, the overlapping portion is blinked or the display brightness of the overlapping portion is changed.

The overlapping portion area measuring means 13d measures the area of the overlapping portion based on the detection result of the overlapping portion detecting means 13c.

The void detecting means 13e is based on the ROI setting processing means 5.
The void portion generated in the ROI is detected, and the void area measuring means 13f measures the area of this void.

Further, the z-direction deviation detecting means 13g is for detecting deviation of the cutout portion in the Z direction after the movement of the cutout portion, and the z value measuring means 13h is for measuring the deviation value in the z direction. .

The measurement results of the overlapping portion area measuring means 13d, the void area measuring means 13f, and the z value measuring means 13h are numerically displayed on the CRT display 12 under the control of the display control means 11.

Next, the operation of the embodiment apparatus configured as described above will be described with reference to the flowchart of FIG.

First, the visual plane is set via the input means 14 (step
S1). This view plane is set to the front, side, and upper side of the skull image 15 as shown by 16, 17, and 18 in FIG.

When the viewing planes 16, 17, and 18 are set, the first distance image creating means 3 creates a distance image for each viewing plane (step S2, see FIGS. 3 to 5). The layout relationship of each distance image written in the plane memory 4 of 1 (step S3) and displayed on the CRT display 12 under the control of the display control means 11 (step S4) is similar to that shown in FIG. In the four divided display areas, the distance image on the first viewing plane 16 is displayed at the lower left (this is referred to as "first distance image 16a").
Is displayed), a distance image on the second viewing plane 17 (hereinafter referred to as “second distance image 17b”) is displayed, and a distance image on the third viewing plane 18 is displayed on the upper left. (This is referred to as "third distance image 18a") is displayed.

Here, the operator is the first, second, and third distance images 16a, 17a, 18
The cursor 19 is set to any of a and the ROI 20 is drawn by moving the cursor 19 on the distance image. This ROI
The depiction of 20 is performed by operating the mouse 14b to move the cursor 19 (step S5). In FIG. 6, a cursor 19 is set on the first distance image 16a, and ROI drawing is performed centering on the first distance image 16a.

Here, even if the cursor 19 is moved on the first distance image 16a as shown in FIG. 6, the second,
The ROI is drawn on the third range images 17a and 18a. this is,
As shown in FIG. 7, the first range image (first viewing plane 16
Corresponding to the cursor set point 21 on (), the address (x, y, z) on the three-dimensional image in the normal direction is obtained,
Further, on the second and third distance images (corresponding to the second and third viewing planes 17 and 18), the address (x, y, z) on the above three-dimensional image
This is possible by calculating the addresses of points 22 and 23 corresponding to. That is, when the point 21 is specified by the cursor 19 on the first range image, the corresponding points 22, 23 are specified on the second and third range images. The locus of movement of this point is ROI20. Further, in this ROI setting process, the ROI on the three-dimensional image (cranial image 15) is also obtained at the same time. The above ROI setting processing is performed by the ROI setting processing means 5, and the ROI drawing information is displayed on the CRT display 12 via the overlay memory 6 (steps S6, S7, S8). Here, the ROI depiction information is the display control means 11
Are displayed in color (for example, red) under the control of. 1st, 2nd,
The third range image is a grayscale display in black and white, and therefore the color ROI 20 superimposed on this range image has excellent visibility.

In the above ROI setting, the operator determines whether or not the ROI is appropriate, that is, whether or not the cutting position in the maxillofacial surgery is more appropriate than the ROI 20 set on the first, second, and third distance images 16a, 17a, and 18a. It is determined (step S9). According to the first, second, and third distance images 16a, 17a, 18a and the ROI 20 superimposed on this, the cutting site is displayed from three directions, and the validity of the cutting site designation (ROI setting) can be easily confirmed. it can.

If it is determined in step S9 that the ROI is not appropriate (NO), the ROI is partially erased (S11), and then the process returns to step S5 to correct the ROI. Since the ROI descriptive information is written in the overlay memory 6 (see FIG. 8), ROI correction can be performed without loss of the range image information in the first plane memory 4.

If the ROI is determined to be appropriate (YES) in the determination in step S9, the device of this embodiment is made to recognize that fact via the keyboard 14a. Then, the image cutting processing means 17 starts the cutting processing of the three-dimensional image (S1
0).

In this cutting process, a ray tracing process is performed from each point of ROI on the three-dimensional image (which has already been obtained at the stage of the setting process (S6) by the ROI setting processing means 5) toward the center of the three-dimensional image. It becomes possible by executing it and sequentially clearing voxel data to zero.

Here, the principle of determining the cutting direction in this cutting process will be described in detail.

For example, when the sphere 25 shown in FIG. 9 is cut from its surface toward the center, a contact surface 26 contacting the sphere 25 at one point is obtained,
Cutting is made in the direction of the normal line 27 from the contact point on the contact surface 26. Contact surface 26
The upper contact point corresponds to one point of the ROI set on the sphere 25. The contact surface 26 is obtained as follows.

First, assuming that Zi, j is one point (designated point) of ROI in the coordinate system of FIG. 9, a point Zi, j−1 adjacent to this Zi, j in the y direction is
, And Zi, j + 1, and as shown in FIG. 10, Zi, j and Zi, j
A line 28 connecting −1 and a line 29 connecting Zi, j and Zi, j + 1 are obtained, and a line 30 passing through Zi, j and having an average inclination with respect to the lines 28,29 is obtained (this Line 30 is referred to as the "first tangent".

Next, the points Zi−1, j and Zi that are adjacent to Zi, j in the x direction are
+1 and j are obtained, and as shown in FIG. 11, a line 31 connecting Zi, j and Zi−1, j and a line 32 connecting Zi, j and Zi + 1, j are obtained. A line 33 that passes through and has an average slope with respect to the lines 31 and 32 is obtained (this line 33 is referred to as a "second tangent line").

The surface including the first and second tangent lines 30 and 33 becomes the tangent surface 26 in FIG.

The cutting direction can also be determined as follows.

That is, a line 34 connecting points Zi, j-1 and Zi, j + 1 as shown in FIG. 12 is obtained, and points Zi-1, j and Zi as shown in FIG.
The line 35 connecting + 1, j is found, and the designated point Zi, j is a normal line from the plane including the lines 34,35 (the plane coincides with the tangent plane 26 in FIG. 9). Find a line to pass through. This line is
It is equal to the normal line 27 in FIG. 9, and thus the cutting direction can also be determined in this way.

The cutting direction is determined as described above for all points of the ROI on the three-dimensional image, and the three-dimensional image is cut in that direction.
The area corresponding to the vertical bone B3 of the nasal bone (shown by 36 in FIG. 6) in FIG. 18 is entirely included in the cutting area by performing ROI filling processing or the like.

The three-dimensional image data after the cutting process is stored in the second voxel memory 8 (S12).

Next, the visual plane is set again via the input means 14 (S1
3). Then, the second range image creating means 9 creates a range image on the viewing plane (S14), and the range image is written in the second plane memory 10 (S15). The distance image created here is for simulating movement and fitting of the cutout portion (this corresponds to the orbital portions B1 and B2 in FIG. 19) and not for ROI setting. Similar to the first viewing plane 16, it is considered that only the range image of the viewing plane set in front of the skull image (after the cutting process) is sufficient, but if necessary, the first, second, third View plane 1
It is also possible to set view planes corresponding to 6, 17 and 18 and create a distance image in each view plane. To enable this, the second plane memory 10 having the same size as the first plane memory 4 is applied. The second lever
The distance image in the plane memory 10 is displayed on the CRT display 12 under the control of the display control means 11 (S16). 14th
The figure shows an enlarged main part of the display image in this case.

Next, fitting is performed by moving the cutout portions (orbital portions) B1 and B2 on this range image. This fitting processing is executed by the fitting processing means 13 (S17).

That is, as shown in FIG. 15, the cutout portions B1 and B2 are indicated by arrows by the moving means 13a in the x and y directions in the fitting processing means 13.
They are moved in the 38 and 39 directions, respectively, and are fitted into a predetermined area. The movement of the cutout portions 38, 39 is performed by coordinate conversion. The moving direction and the moving amount of the cutouts 38, 39 can be input via the input means 14. Also, if a reference point is specified via the input means 14, the cutout portion is centered on this reference point.
It is also possible to automatically move B1, B2 up, down, left and right, and end the movement in a state where the void area when the cutouts 38, 39 are fitted is minimized.

Further, the cut-out portions B1 and B2 are moved in the z-direction by the z-direction moving means 13b. This movement in the z direction changes the distance values of the cutouts B1 and B2 so that the distance value of the cutouts B1 and B2 and the distance value of the bone around the cutouts B1 and B2 become equal in the distance image. It is done by doing.

In the movement of the cutout portion, if there is a bone-overlapped portion (for example, a portion indicated by 40, 41, 42, 43 in FIG. 15), the overlapped portion is detected by the overlapping portion detection means 13c, and the CRT display At 11, the display is clearly distinguished from other parts by blinking or changing display brightness (for example, blanking). Such a display is a display control means.
Controlled by 11. Further, the area of the overlapped portion is measured by the overlapped portion area measuring means 13d.

Further, when the ROI is set as shown by the broken line 44 in FIG. 15, the void 45 in the ROI 44 is detected by the void detecting means 13e, and the area of this void is measured by the void area measuring means 13f. To be done. Here, the ROI 44 setting processing is performed by the ROI setting processing means 5, and the ROI drawing information display is performed via the overlay memory 6.

The deviation of the cutouts B1 and B2 in the z direction is detected by the z direction deviation detecting means 13g, and the deviation value is z value measuring means 13h.
Measured by.

Then, the measurement results of the respective measuring means 13d, 13f, 13h are numerically displayed on the CRT display 12 under the control of the display control means 11 (S18).

The above-described processing and display enable simulation of maxillofacial treatment.

As described above, in the apparatus of this embodiment, it is possible to determine whether or not the cutting position in the maxillofacial surgery is appropriate from the ROI 20 set in the first, second, and third distance images 16a, 17a, and 18a. Therefore, it is possible to easily confirm the adequacy of the cutting position. Further, since the range image and the ROI depiction information are stored in different memories, ROI can be corrected without loss of the range image, and the ROI depiction information is displayed in color. Therefore, ROI display with excellent visibility can be performed.

Further, in the processing of the image cutting processing means 7, since the cutting is performed from each point of the ROI toward the center of the three-dimensional image (skull image 15), the cutting is the same as in the case of actual maxillofacial surgery. , It is convenient for simulation.

Therefore, according to the apparatus of this embodiment, the simulation of the maxillofacial surgery can be performed accurately.

Although one embodiment of the present invention has been described above, it is needless to say that the present invention is not limited to the above embodiment and various modifications can be made.

In the above-described embodiment, the distance images in the visual planes set in front, side, and above the skull image 15 are displayed on the same screen.
Although the ROI setting has been described, the distance between the front view plane of the head image 15 and the side or upper view plane, or the view planes set to the front, side, upper, and lower, respectively. You may make it display an image on the same screen and perform ROI setting.

Further, in the above embodiment, the voxel data formed based on the plurality of head slice images is the first voxel memory 1.
Although it has been described as being stored inside, voxel data based on a slice image other than the head may be stored.
That is, simulations other than maxillofacial surgery can be performed by using voxel data other than the head as the processing target.

[Effects of the Invention] Since the present invention is configured as described above, the following effects are achieved.

According to the three-dimensional image processing apparatus of the first aspect, the ROI setting can be performed by using the distance images of a plurality of viewing planes, so that the validity of the cutting position designation can be easily confirmed.

Further, according to the three-dimensional image processing device of the second aspect, a plane memory for storing the distance image and an overlay memory for storing the ROI depiction information are provided, and the storage contents of both memories are combined and displayed. Therefore, ROI can be corrected without loss of range image.

Then, according to the three-dimensional image processing device of claim 3,
Since the color different from that of the distance image is assigned to the ROI descriptive information, ROI display with excellent visibility is possible.

Further, according to the three-dimensional image processing device of claim 4, the RO
Since the cutting process is performed from each point of I toward the center of the three-dimensional image, the cutting is the same as in the case of actual surgery, which is very convenient for performing simulation.

Then, according to the three-dimensional image processing device of claim 5,
The cutting direction from each point of ROI to the center of the 3D image is
It can be easily determined based on the gray value of the range image.

[Brief description of drawings]

FIG. 1 (a) is a block diagram of the configuration of an apparatus according to an embodiment of the present invention, FIG. 1 (b) is a detailed configuration block diagram of the fitting processing means in FIG. 1 (a), and FIG. 2 is a three-dimensional image. FIG. 3 to FIG. 5 are explanatory views of the distance image on each viewing plane in FIG. 2, and FIG. 6 is a distance image on each viewing plane in FIG. 2 on the same screen. FIG. 7 is an explanatory diagram of the relationship between the address on the distance image and the address on the three-dimensional image, and FIG. 8 is the first plane memory and overlay memory in FIG. 1 (a). FIG. 9 to FIG. 13 are explanatory views of the principle of determining the cutting direction in the image cutting processing means of FIG. 1 (a), and FIGS. 14 and 15 are CRT displays of FIG. 1 (a). 16 is an explanatory view of the distance image display in FIG. 16, FIG. 16 is a flow chart for explaining the operation of the apparatus of this embodiment, and FIGS. FIG. 19 is an explanatory view of maxillofacial surgery. 3 ... First range image creating means (range image creating means),
4 ... First plane memory (plane memory), 5 ...
... ROI setting processing means (region of interest setting processing means), 6 ...
Overlay memory, 7 ... Image cutting processing means, 12 ... CR
T display (display means), 14 ... Input means.

Claims (5)

[Claims] 1. A three-dimensional image processing apparatus for processing three-dimensional image data having a voxel structure, input means for setting respective visual planes at different positions outside the three-dimensional image, and each visual plane. Distance image creating means for creating a distance image for each visual plane by measuring the distance from the surface to the three-dimensional image surface, display means for displaying a plurality of created distance images on the same screen, and the displayed distance image A region-of-interest setting processing unit capable of setting a region-of-interest to the three-dimensional image by drawing the region of interest described above. 2. A display control means is provided which comprises a plane memory for storing a distance image and an overlay memory for storing a region of interest delineation information, and which combines the stored contents of both memories and displays them on the display means. Item 3. The three-dimensional image processing device according to Item 1. 3. The three-dimensional image processing apparatus according to claim 2, wherein the display control means causes the display means to display the ROI descriptive information in the overlay memory in a color different from that of the distance image. 4. The three-dimensional image processing according to claim 1, 2 or 3, further comprising an image cutting processing device that performs a cutting process from each point of a region of interest in the three-dimensional image toward a central portion of the three-dimensional image. apparatus. 5. The three-dimensional image processing apparatus according to claim 4, wherein the image cutting processing means determines the cutting direction in the cutting processing based on the gray value of the distance image at each point in the region of interest.