JP2001087290A - Filling artificial bone designing system and manufacture of filling artificial bone using the system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a filling artificial bone designing system capable of easily and accurately designing an artificial bone even to a defective part unapplicable by a reflection copy, and a manufacturing method of a filling artificial bone using the system. SOLUTION: Interpolative reference lines 149a, 149b reflected with a surface shape of a healthy bone part are set on the basis of bone part external shape line information 125, and a curve control point 152 for forming an interpolative curve is set along the interpolative reference lines 149a, 149b. Next, a defective part-interpolating curve 151 for curvedly interpolating a defective part is formed according to a predetermined curve deciding algorithm by using the preset curve control point 152. Three-dimensional shape data on an artificial bone outside surface for filling the defective part is formed on the basis of the defective part interpolating curve 151.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to X-ray CT, NMR
The present invention relates to a system for designing a bone prosthesis for filling a bone defect such as a skull based on a human tomographic image of a human body and the like, and a method for manufacturing the bone prosthesis using the same.

[0002]

2. Description of the Related Art In the field of orthopedic surgery or plastic surgery, for example, in order to treat a patient having a bone defect due to a traffic accident or other accident, the defect is filled with artificial bone. As a related art related thereto, for example, Japanese Patent Application Laid-Open No. Hei 7-284501 discloses an insertion part (for example, a defective part) of an internal fixation member such as an artificial head using an X-ray CT or MRI tomographic image (tomographic image). Is acquired as three-dimensional data, and while displaying a three-dimensional image of the insertion site using the three-dimensional data, the image of the inner fixed member is moved on the screen, and before the operation, the insertion site of the member is inserted. There has been disclosed a technique capable of simulating the suitability for the data.

On the other hand, Japanese Patent Publication No. 6-2137 discloses X
Using a tomographic image of X-ray CT or MRI, the insertion site of the internal fixation member such as an artificial bone head is acquired as three-dimensional data, and the three-dimensional data is output to a cutting device, thereby replicating the insertion site (model). An apparatus for making is disclosed.

[0004]

For example, when a defect occurs in a skull, as shown in FIG. 32, a human skull SKL is used.
Has a substantially symmetrical shape when viewed from the front (face side). Then, as shown in FIG. 32 (a), when the mirror reference plane MSP is set at the center of the face, if the defect 400 exists only on one side with respect to the mirror reference plane MSP, such as the temporal region, this defect is lost. As a method of manufacturing a supplementary artificial bone that fills the part 400, a technique called mirror copy (hereinafter, also referred to as “mirror copy”) has been used. This is as described in claim 5 of Japanese Patent Publication No. 6-2137.

[0005] However, there is no guarantee that bone loss due to injury or the like always occurs in the form shown in FIG. Actually, as shown in FIG. 7B or FIG. 7C, it is not rare that the defective portion 400 occurs on both sides of the mirrored reference plane MSP such as the forehead and the top of the head. Needless to say about the case, the mirror copy method is completely powerless. This is because there is no data to be referred to because the loss portion that spreads to one side of the mirror reference plane MSP also spreads to the opposite side expected as a copy source. Also, no matter how contrasting, the actual skull has subtle deviations from symmetry, and mechanical mirror copying may not always be the best method.

An object of the present invention is to provide a prosthetic artificial bone design system capable of easily and accurately designing an artificial bone even for a defective portion to which mirror copying cannot be applied, and a prosthesis using the same. And a method for producing an artificial bone for use.

[0007]

Means for Solving the Problems and Actions / Effects In order to solve the above problems, a replacement artificial bone design system according to the present invention provides a replacement artificial bone for filling a bone defect based on a human tomographic image. In the system to be designed, in each of the tomographic images photographed at a plurality of different tomographic positions, a bone region extracting means for extracting a bone image region, and a final bone region extracted based on the extracted bone region. A bone outline information generating unit that generates bone outline information that is outline information of an area to be determined as a bone, and a bone outline defect information based on the bone outline information for each tomographic position. And three-dimensional shape data generating means for generating three-dimensional shape data, the three-dimensional shape data generating means based on the bone outline information,
Interpolation reference line generating means for setting an interpolation reference line reflecting the surface shape of the healthy part of the bone, and curve control point setting means for setting a curve control point for generating an interpolation curve along the interpolation reference line; Using a set curve control point, according to a predetermined curve determination algorithm, comprising a missing part interpolation curve generating means for generating a missing part interpolation curve for curve interpolation of the missing part, based on the missing part interpolation curve And generating three-dimensional shape data of the outer surface of the artificial bone for filling the defect.

Further, the method for producing a prosthetic artificial bone according to the present invention provides a method for preparing a material to be processed by referring to the three-dimensional shape data created by the above-described artificial prosthetic bone design system. A step of processing into an artificial bone shape for use.

The bone part is, for example, a skull. In this case, the tomographic image can be a sliced image of the head obtained by setting the axial direction so as to penetrate the skull in the vertical direction. However, the present invention is not limited to this, and can be applied to any bone where a defect may occur. Further, the material to be processed may be an unfired molded material of ceramics which is an artificial bone material. In this case, the artificial bone for supplementation can be obtained by firing the material processed into the shape of the artificial bone for supplementation. Note that the created three-dimensional shape data can be used in a form in which the dimensions are enlarged (or reduced) at a predetermined ratio in order to take into account, for example, shrinkage due to firing. Further, in this specification, the pixel density refers to information in a broad sense that is indicated by any one of the density and color of a pixel and a combination of the two. Treated as "different".

According to the present invention, an interpolation reference line reflecting the surface shape of the healthy part on the side of the healthy part is set based on the bone outline information based on the tomographic image, and further along the interpolation reference line. Using the set curve control points, according to a predetermined curve determination algorithm, generate a missing portion interpolation curve for curve interpolation of the missing portion, and based on the missing portion interpolation curve, an artificial bone outer surface for filling the missing portion. 3D shape data was generated. The feature of this method is that, based on the shape information of the healthy part around the defective part, an interpolation curve matching the shape is generated in the defective part, so that the three-dimensional shape data of the artificial bone outer surface that compensates for the defective part can be obtained. The point is that it is created as estimated data from the shape. Therefore, an artificial bone can be easily and accurately designed and manufactured even for a defective portion where mirror copying cannot be applied. In addition, when there is a large shift in left-right symmetry with respect to a defective portion to which mirror copying can be applied, it is better to estimate the shape of the defective portion from the healthy portion around the defective portion as described above, rather than feeling uncomfortable In some cases, artificial bone for replacement can be produced.

In the following, the artificial bone design system for replacement is
Further, a configuration that can be added will be described. (Configuration 1) In each of the tomographic images photographed at a plurality of mutually different tomographic positions, a bone region extracting means for extracting a bone image region, and finally, based on the extracted bone image region, Bone outline information generating means for generating bone outline information that is outline information of an area to be determined as a bone, and a tertiary bone defect part based on the bone outline information for each tomographic position. Three-dimensional shape data generating means for generating original shape data, wherein the bone region extracting means changes a threshold density level for extracting a bone region with respect to a pixel density constituting a tomographic image according to a tomographic position. Set to a unique value.

When a tomographic image of a bone is taken by X-ray CT or the like, an image is overwhelmingly often formed based on accumulated tomographic information from a human body slice region having a predetermined width. For example, in the case of a skull, the shape of the skull decreases in diameter toward the top of the head. The boundary between the bone region and the non-bone region in the above becomes unclear. When judging a region having a pixel density equal to or higher than the determined pixel density to be a bone portion, the larger the blur spread, the smaller the bone region on the image becomes than the actual bone region. This may lead to a problem of misrecognition of.

In view of the above, in the above configuration, the threshold density level for extracting the bone region with respect to the pixel density forming the tomographic image is set to a unique value that differs depending on the tomographic position, in consideration of the influence of the blur spread. Was set to. This makes it possible to more accurately determine the outline of the bone portion, and furthermore, it is possible to further improve the suitability of the finally obtained artificial prosthesis for the defective portion.

(Structure 2) In each of the tomographic images photographed at a plurality of mutually different tomographic positions, a bone part candidate area extracting means for extracting a pixel area having a predetermined density level as a bone part candidate area, A region selecting means for selecting a bone region to be finally used as a bone region (hereinafter referred to as a defined bone region) among the bone candidate regions, and finally a bone region based on the determined bone region. Bone outline information generating means for generating bone outline information which is outline information of a region to be determined, and three-dimensional shape data of a bone defect based on bone outline information for each tomographic position. And three-dimensional shape data generating means for generating

For example, taking a skull as an example, the shape generally has a spheroidal or oval shape, but the internal structure is not so simple. Some parts are formed as
Since a large opening such as an eye socket is also formed, naturally, if a tomogram is taken, it is naturally possible that an integral bone part appears separately. Further, as a measurement problem, a small area having the same density level as the bone may appear in an area other than the bone due to the influence of noise or the like.

According to the above configuration, of the bone candidate regions appearing in the tomographic image, portions, such as the nose, which are considered to have a favorable effect on accurate shape interpolation of the missing portion shape,
Select the final bone area to be finally used as the bone area, and conversely,
Regions that are apparently due to noise or the like can be excluded from the definite bone region. This makes it possible to accurately determine the three-dimensional shape data of the bone defect.

(Structure 3) In each of the tomographic images photographed at a plurality of mutually different tomographic positions, a bone region extracting means for extracting a bone image region, and based on the extracted bone region, A bone outline information generating means for generating bone outline information which is outline information of an area to be finally determined as a bone, and a bone outline in which a defective part is involved, Based on the shape information, based on the bone outline information and the bone outline information, the defect estimated outline information generating means for generating the estimated loss outline information, which is the estimated outline information of the defect, based on the shape information. A bone outline display means for displaying the bone outline together with the estimated outline of the defective part;
Display control means for differentiating the display states of the outlines and the estimated outlines displayed on the bone outline display means so that the outlines and the estimated outlines can be distinguished from each other; And a three-dimensional shape data generating means for generating three-dimensional shape data of the bone defect based on the outline information.

According to the above configuration, the estimated outline information of the defective part is created based on the outline information of the bone generated based on the tomographic image, and the outline of the bone and the estimated outline can be identified. As described above, the display states of these outlines are displayed in different forms. According to this, at the stage before creating the final three-dimensional shape data of the defective part, the estimated outline information of the defective part based on the bone outline information is created, and furthermore, it is used for shading and color of the pixel. Are displayed in different display states. As a result, it is possible to intuitively grasp the rough three-dimensional shape of the defective part, and it is possible to reduce the inconvenience of mistaking the curved posture in space due to illusion, etc., so that the final three-dimensional shape data is designed・ When creating, you can work efficiently without confusion with a clear completed image.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the embodiments shown in the drawings. In this embodiment, an example is described in which a skull is to be filled with a defect, and three-dimensional shape data for designing an artificial bone for filling is created based on a tomographic image by X-ray CT. However, the means for photographing tomographic images is not limited to X-ray CT.
RI, positron CT, emission CT, ultrasonic tomography, or the like may be employed.

First, prior to describing the configuration of the apparatus, the principle of known X-ray CT will be briefly described. FIG.
FIG. 3 is a principle diagram showing an example of an X-ray CT. An X-ray beam narrowed down from an X-ray tube through a collimator is emitted to a human body (head in this case) as a subject. The X-ray beam is slightly absorbed by the living tissue in the middle and attenuated, and then most of the remaining beam is transmitted. The transmitted X-rays are measured by a detector.

Assume that the intensity of the radiated X-ray is I0, the intensity after passing through the subject is I, and the absorption coefficient of the X-ray absorbed by the subject, that is, the tissue inside the human body is μ (x,
If y), it is known that the relationship of Equation 1 is approximately established.

[0022]

(Equation 1)

Here, the integral indicates a line integral of the transmitted X-ray beam along the direction L. In this equation, the logarithm of both sides is given by Equation 2.

[0024]

(Equation 2)

That is, the logarithm of the ratio between the incident intensity and the transmitted intensity of the X-ray is substantially equal to the integral value of the absorption coefficient along the X-ray beam. Therefore, while keeping the direction of the X-ray beam constant,
When the source and the detector are moved synchronously (translational movement), the projection data of this angle is obtained by the operation by the translational movement. Then, set the angle to a predetermined interval (for example, 1 °).
By repeating the same operation while rotating each time, projection data (projection integral value) from all directions can be obtained. All of these data are stored in a computer, and the X-ray absorption coefficient μ (x, x, x) at each point is calculated from the projection data by a known image reconstruction method such as a successive approximation method, a Fourier transform method, or a filtered back projection method. y) is calculated, and a tomographic image is obtained.

In ordinary clinical diagnosis, a relative absorption coefficient called a CT value is introduced for convenience. For example, water is taken as 0, air is taken as 500, and bone is taken as 500.
Show the value by dividing equally. In the present invention, it is particularly necessary to extract a bone image. However, basically, the extraction processing is performed depending on the distance from the CT value of another tissue.
Then, in the tomographic image, as shown in FIG. 14B, the CT value is divided by an appropriate gradation threshold, and the C
By dividing the pixel outputs of different display colors (or densities) into the T value section by using the display color index, the pixel areas belonging to the same CT value section are imaged in the form of being displayed in the same color. is there. In addition, regarding image display, a window function is often introduced, in which a gradation is provided only within a specific absorption coefficient width and the other is expressed in black (or white). For example, it is also possible to obtain an image with good contrast for the bone part).

FIG. 2 shows a prosthetic artificial bone design system according to one embodiment of the present invention (hereinafter simply referred to as a design system).
FIG. 1 is a block diagram showing an electrical configuration of a machining system 40 for producing a prosthetic bone for supplementation using the same (hereinafter, simply referred to as a machining system) 1. The core of the design system 40 is the computer 50, the I / O port 2,
It includes a CPU 3, a ROM 4, a RAM 5, and a storage device 6 composed of a hard disk device and the like connected thereto. The I / O port 2 has a data reading device 8 for reading a data recording medium on which digital tomographic image data is recorded, such as a floppy disk, a magneto-optical disk, or an optical disk (CD-ROM). An image scanner 7 for reading a copy output, a display control unit 10, and input units such as a display device 11, a keyboard 12, and a mouse 13 connected thereto are connected.

The storage device 6 stores a control program 200 for realizing the basic functions of the design system 40 and the processing system 1. The control program 200
Are a bone part data extraction program 201, a bone part outline three-dimensional synthesis program 202, a three-dimensional shape data creation program 203, a cutting (machining) optimum position determination program 20
4. Includes a cutting (machining) program 205 and a simulation program 206. Further, a design data storage unit 207 is formed in the storage device 6, and as shown in FIG. 5, the design data storage unit 207 is associated with operation target person specifying data for specifying the operation target person (patient). The design data of the prosthesis for filling the defect is stored. And CPU
3 is each program 201 to 2 stored in the storage device 6
06 is performed using the program work area 5a of the RAM 5, thereby performing basic operation control processing of the design system 40 and the processing system 1, that is, control for realizing a function realizing unit required for each process. . Also,
The design data necessary for creating the three-dimensional shape data is stored in the RAM 5
And stored in the data storage area 5b for program processing.

FIG. 8 is an example of an X-ray photograph showing the head of an operation subject (patient). (At intervals of about 5 mm) indicate tomographic positions SC where tomographic imaging is performed. As shown in FIG. 6, the design data includes a slice No. specifying each of these tomographic positions (hereinafter, also referred to as a slice). In the form corresponding to the above, the multi-tone original image data of the tomographic image obtained by tomography and the bone extraction binary data obtained by binarizing this by extracting only the bone by the processing described later Including. In addition, the bone outline data generated using the binarized bone extraction binarized data, the bone outline three-dimensional synthesized data three-dimensionally synthesized according to the tomographic position array, and later used by using this The three-dimensional shape data of the artificial bone for replacement designed and created by the processing described below is also included.

As the image data of the human body tomographic image, for example, as shown in FIG.
For example, an image signal directly recorded as a digital image signal based on a photographing signal acquired by a tomographic image photographing apparatus such as MR) can be used. For example, in the case of a CT scanner,
The original signal data reflecting the X-ray absorptance taken through the CT scanner is taken into the CT device analysis computer via the input signal cable, where the known processing for image reconstruction described above is performed. Thus, digital grayscale image data (tomographic data) for outputting a tomographic image is created. This tomographic data may be transferred directly to the computer 50 of the design system 40 via a data output cable, but if this is not possible, the tomographic data may be transferred using a communication network or the data recording described above. The information may be temporarily stored in the medium 9 (FIG. 2) and read through the data reading device 8.

The advantages of the above method are that the image is extremely clear because the digital gray image data is directly used, and the image is hardly affected by noise and the like. It is not affected by errors at all. However, there are the following disadvantages. First, the format of tomographic data often differs depending on the type of tomographic apparatus, and data in a special format may not be used in some cases. Further, when data is erased at the hospital where the tomography was performed, or when digital image data cannot be output depending on the device model, it is basically impossible to cope.

Therefore, as a method for compensating for this drawback,
As shown in FIG. 3A, an image hard copy 80 output as a photographing film of a monitor image or an image printout in a tomographic image photographing apparatus such as X-ray CT or NMR is used as image data of a human body tomographic image. An example in which image data converted by the scanner 7 is used can be exemplified. With this method, if data is erased,
Alternatively, even in the case of a CT device or the like that cannot output data, an image hard copy 8 such as a photographic film or an image printout
If 0 remains, digital grayscale image data can be obtained by reading the image.

Returning to FIG. 2, a processing apparatus 15 is connected to the design system 40 via the data interface 14, and forms the processing system 1 together with the design system 40. This processing apparatus is not particularly limited as long as it is suitable for cutting a ceramic material to be processed. In this embodiment, for example, the processing apparatus is constituted by an NC processing machine using a vertical milling machine. FIG. 4 is a block diagram schematically showing an example of the hardware configuration. That is, the ceramic work material W is fixed on the XY table 60 that controls the work feed (XY directions) in a horizontal plane by a known screw shaft mechanism, and on the other hand, the tool (flying tool) TL
Is rotated by a motor 64, and can be moved in a vertical direction (Z direction) by a motor 63 via a screw shaft mechanism.

Motors 61 and 62 that control the X-Y table 60 in the X-axis direction and the Y-axis direction, and the Z
In the motor 63 for controlling the axial feed and the motor 64 for rotating the tool, the rotation angle position and the rotation speed are all detected by the pulse generator PG which rotates in conjunction with the rotation axis of each motor, and the detection information is fed back. The rotation is controlled by a servo controller receiving the rotation. Each servo controller includes a control computer 55 (an I / O port 56, a CPU 57 connected thereto, a ROM 58,
(Including the RAM 59). The control computer 55 is connected to the design system 40 via the data interface 14, receives a processing control signal (for example, a signal of processing path data) output based on the created three-dimensional shape data, and controls each servo. A drive command signal for the motors 61 to 64 is sent to the controller.

Hereinafter, the flow of the process of creating three-dimensional shape data will be described in order. First, processing for generating bone outline information based on a tomographic image is executed by the bone data extraction program 201 in FIG. Here, the program 201 plays a role of causing the computer 50 mainly composed of the CPU 3 to function as the following means. Bone region extracting means: Bone image region 85 is extracted from each of tomographic images 90 taken at a plurality of different tomographic positions. Bone outline information generation means: Based on the extracted bone image region 85, bone outline information 103 which is outline information of a region to be finally determined as a bone is generated.
Further, the threshold density level KS for extracting the bone region with respect to the pixel density forming the tomographic image 90 is set to a unique value different depending on the tomographic position SC.

FIG. 9 shows an example of a tomographic image 90 by X-ray CT. FIG.
(B) is the fault of the 14th layer (position several layers higher than the upper edge of the orbit), and (c) is the fault of the 24th layer (close to the top of the head). 9 is a diagram showing a multi-tone original image of a position. In each case, the upper side of the image is the face side (front side), the sliced outline of the head appears as a slightly dark gray area, and the image area estimated as the skull (bone estimated image area; The density level of the pixel is set so that the ring-shaped portion) appears as white as possible. Referring to the schematic diagram of FIG. 12, the estimated bone portion image area 100 can be visually distinguished from the background at a predetermined luminance K0 or more corresponding to the approximate position of the focused bone portion. Area with high contrast. A gradation due to the blur described later is formed on the outer edge portion, and the boundary with the background portion may be unclear. This boundary plays a very important role in determining the bone outline 103. There is no. On the other hand, the reference density level area 102 is located inside the bone part estimation image area 100 and has a predetermined pixel density level (for example,
This is a region having a reference density level that is a pixel density level at which the luminance is equal to or higher than Ki, and is a region that can be almost definitely determined to belong to a bone part (however, even if black and white are inverted in image expression, bone Identification of the parts is likewise possible without problems.
In this case, a density level lower than a certain luminance Kj may be set as a reference density level.

9 (a) to 9 (c), it can be seen that the width of the estimated bone part image area (corresponding to the thickness of the skull) is larger at the upper layer side. is there. This is not a difference in bone thickness,
It is considered that image blur is affected by the following factors. That is, as shown in FIG.
In many X-ray CT apparatuses, a tomographic image has a certain thickness as shown in FIGS. 11A and 11C because an X-ray beam used for tomography cannot be narrowed down to a certain diameter or less. Plate-shaped slice region SC, for example, a slice region S sandwiched between a certain tomographic position and an adjacent tomographic position
An image is formed based on accumulated tomographic information (a kind of integral information over the thickness direction of the human body slice region SC) from the human body slice region SC having a predetermined width, such as C1 and SC2. In this case, particularly in a portion above the upper edge of the orbital opening of the skull (for example, above the forehead), in the lower slice region SC1, the wall of the skull SKL is separated from the direction of the head axis. Of the skull SK in the upper slice area SC2.
The wall of L has a large inward inclination.

In this case, as shown in FIG. 11C, the wall of the skull SKL forms the bone part estimated image area 100. As described above, the slice area SC
Because the image is formed by the integral information in the thickness direction of
The projection area of the skull SKL is increased by the amount of the inclination, which leads to image blur. In particular,
As shown in FIG. 11D, in the case of the slice region SC2 in which the degree of inclination of the skull SKL is large, the portion CP of the skull SKL located over the entire thickness of the slice region SC2 is located near the center of the skull SKL in the width direction. It is a limited form. This portion is a portion that becomes the reference density level region 102 in which the luminance of the pixel is equal to or higher than Ki. On the other hand, since both side portions CQ occupy only a part in the thickness direction, the luminance is K
Although it appears as a portion less than i, the upper limit luminance K0 of the background portion
It is still higher and the bone estimation image region 100
Have to be incorporated into As a result, the width W ″ of the estimated bone part image area 100 is equal to the width W of the actual skull SKL.
It is much larger than it is (that is, blurred). On the other hand, since the width W ′ of the reference density level region 102 is smaller than the width W of the skull SKL, it cannot be adopted as the correct width W. On the other hand, in the slice region SC1 in which the degree of inclination of the skull SKL is large, the above tendency is not so remarkable.

Therefore, as shown in FIG.
In the portion above the upper edge of the orbital opening of L, the reference level region 102 should be finally determined so that the influence of the blurring of the estimated bone region 100 due to the inclination of the skull wall is reduced. If the positional relationship with the bone outline 103 is determined, the determined position of the bone outline 103 can be adjusted according to the magnitude of image blur due to the tomographic position,
The bone outline 103 can be determined more accurately. As shown in FIG. 12, the bone part estimation image area 100
In the thickness direction of the skull, the reference density level region 102 appears at the center of the skull, and the pixel density level increases outward (decreases when the inverted density setting is performed), and blurs and spreads outward. Then, the position of the bone outline 103 can be determined as a pixel position having a threshold density level KS separated by a predetermined value from the reference density level Ki in the direction of blurring.

When the blur spread is small, the position of the bone outline 103 is set so that the gap between the boundaries of the reference level area 102 is small, and when it is large, the gap is set to be large. . In order to determine the position more accurately, reference is made to the reference level region shape of the lower-layer tomographic position where the blur extension of the bone-portion estimated image region 100 is relatively small, and the bone outer shape of the upper-layer tomographic position is referred to. Line 1
It is effective to determine 03. For example, at the tomographic position on the upper layer side where the blur spread is large, the bone width is estimated from the width of the lower reference level region, the density level of the pixel corresponding to the end point position of the bone width is read, and this is used. It is also possible to determine the threshold density level KS.

The setting of the threshold density level KS may be effective, for example, even for a lower-layer tomographic image having a small blur spread. For example, as shown in FIG. 11B, when a tomographic image of X-ray CT is obtained by filming a monitor output, the skull is positioned because the monitor screen and the film surface are separated to some extent. Image spreading is more likely to occur at the outer part of the image, and image blurring due to this may inevitably occur even at a lower tomographic position. Therefore, if the boundary position of the bone outline 103 is corrected by the threshold density level KS in consideration of this, a more accurate outline can be obtained.

Next, in order to make the skull wall less susceptible to the effects of blurring, the threshold density level KS is
It is desirable to set the value such that the distance from the reference density level Ki increases as the tomographic position is located above the skull in the axial direction. As described above, the reference density level region width becomes narrower as the tomographic position is located in the upper layer. In particular, at the tomographic position (slice) near the top of the head, FIG.
As can easily be inferred from the case where the inclination of the skull wall is further increased, the region satisfying the reference density level Ki or more, that is, the reference density level region 102 itself may be lost. However, in reality, the bones always appear on the cross-section, and the cross-sectional width should be increased by the inclination. Therefore, by setting the threshold density level KS as a value apart from the reference density level Ki, it is possible to determine the bone outline more accurately.

Hereinafter, a specific method of setting the threshold density level KS will be described. For example, the display device 11 of FIG.
It functions as tomographic image display means for displaying a tomographic image 90 at each tomographic position. The threshold density level KS can be manually input while referring to the tomographic image 90 at each tomographic position. For example, as shown in FIG. 12, the bone outline information generating means realized by the program 201 includes a threshold density level input means 99 for variably or steplessly setting the threshold density level KS, and a variable A bone image region display control means for variably outputting the extracted bone image region 85 to the tomographic image display means (display device) 11 in accordance with the input threshold density level KS so as to be distinguishable from other image regions; To
Functions can be added.

As shown in FIG. 12A, the display device 11
In the tomographic image displayed on the screen, a bone portion estimation image region 100 having a pixel density level equal to or higher than the above-described luminance K0 appears. A threshold density level input bar 99 as threshold density level input means is formed on the screen. This is so that the set threshold density level KS can be changed steplessly continuously (or intermittently by a finite number of steps) by sliding the pointer P in the bar length direction by mouse dragging or the like. Has become. Then, on the screen of the display device 11, the set threshold density level KS is used as a binarization threshold, and pixels having a higher luminance than the remaining pixels are output in a state that can be identified with respect to the remaining pixels (for example, identification of red or the like). Color) and the bone region 8
Displayed as 5. The row of pixels that gives the outer edge position of the bone region 85 forms the bone outline 103. That is, the bone outline information is given as point group data including a set of outer edge points of the bone region 85 formed by the above-described binarization.

When the setting of the threshold density level KS is changed by operating the threshold density level input bar 99, FIG.
By (b), the binarization process is performed again with the changed threshold density level KS, and the display state of the bone region 85 is changed on the screen in real time. The operator operates the threshold density level input bar 99 and observes a change in the display state of the bone region 85, and considers, for example, the width of the bone region 85 determined in another slice and the position of the slice. Then, the set threshold density level KS which is considered to be optimal is selected. Thereby, the final point group data of the bone region 85 and the bone outline 103 is determined and stored.

On the other hand, the bone outline information generating means described above
As shown in FIG. 13, at least one of the storage value 105 of the individual threshold density level KS for each tomographic position and the arithmetic program 105 ′ for calculating the threshold density level value for each tomographic position is determined by the threshold. A threshold density level generation source storage means stored as a density level generation source; and a threshold density level value determination means for determining a threshold density level value corresponding to a tomographic position of interest based on the threshold density level generation source. It can also be configured as one. In this case, the pixel position corresponding to the determined threshold density level value is defined by the bone outline 10
3 position. As shown in FIG. 13C, the threshold density level value KS for each slice (tomographic position)
(= KS1, KS2 ‥‥) can be stored in the memory 105, and the threshold density level value KS of the slice of interest can be read and determined as appropriate. Further, in order to calculate the threshold density level value for each tomographic position, the relationship between the tomographic position (slice No.) and the threshold density level value KS is represented by a function expression 10.
It is also possible to store in the form of 5 'and use this to determine the threshold density level value KS.

Even if the value of KS is not directly stored, it is sufficient that the position of the bone outline 103 is finally determined. There are also ways to use the information. That is, as shown in FIG. 13B, the first layer (slice N
o. Although there is not much difference in the bone width between the first layer and the fifth layer (slice No. 5), the reference density level region widths W1 and W5 include W1 for the above-described reason.
> W5. For example, by multiplying the reference density level area W by a coefficient α, it can be converted to a bone width. If the value of α is obtained for each tomographic position, this is calculated as a threshold density level value. It can be used as information reflecting KS. In this case, the value of α for each slice (tomographic position) may be stored in the memory 104 and read out and used as appropriate.

As described above, the bone region 85 can be automatically extracted by the binarization process by setting the threshold density level KS. However, not all of the regions extracted in this way can be used as the bone region 85. For example, a portion that does not originally belong to the skull or a high-luminance region caused by the influence of noise or the like must be excluded. On the contrary, there is a case where it is better to dare to incorporate even a part clearly not belonging to the bone part into the skull part in order to create accurate three-dimensional shape data. A function convenient for such processing is also realized by the bone part data extraction program 201. That is, here, the computer 50 mainly constituted by the CPU 3 functions as the following units. Bone region extracting means: In each of the tomographic images 90 taken at a plurality of different tomographic positions, a pixel region having a predetermined density level is extracted as a bone candidate region 108. Region selecting means: one of the bone candidate regions 108 which is finally used as a bone region (determined bone region) 93, 1
Select 10. Bone outline information generation means: fixed bone regions 93 and 110
, The bone outline information 103, which is the outline information of the area to be finally determined as the bone, is generated.

When the bone is a skull, the bone candidate area 1
08 can be extracted as follows. That is, as shown in FIG. 14A, the multi-tone original image data is converted into binary bitmap image data at a certain threshold density level (here, the aforementioned threshold density level KS). Then, as shown in FIG. 16A, the bitmap data is scanned in a predetermined direction (for example, the x direction),
Depending on whether or not a break of “1” bits occurs for a certain number (for example, 3 bits) or more, each bit is labeled with a labeling code (this embodiment) while determining whether it is the same bone region or another bone region. In the example, it is represented by a number such as 1, 2 ‥‥)
Will be applied. In the second and subsequent scan rows, when the detection state of the “0” bit is changed to the detection of the “1” bit, the labeling state of, for example, eight bits surrounding the “1” bit is determined, and the already recognized state is determined. If a bit labeling code is detected, the same labeling code is applied.
If nothing is detected, a new labeling code is applied. Then, sets of bits with different labeling codes are identified as different bone regions. FIG. 15 shows an example in which different bone regions are output in different colors by such labeling processing (the left side is a multi-tone original image, and the right side is an image in which the bone region is color-coded and output by labeling).

Now, as shown in FIG. 17, the region selecting means adopts at least the largest candidate region 93 having the largest area among the extracted bone candidate regions 108 as the determined bone region. Become. On the other hand, as shown in FIG. 16 (b), a region having a predetermined allowable lower limit area S0 or less is not adopted at least as a definite bone region. As a result, all the small bone regions appearing due to noise or the like are excluded, and the influence of the influence is eliminated.

Next, in the case of a prosthetic bone for filling a defective part of the skull, the main problem is only the outer surface shape that appears in the appearance of the head. If the above-described bone outline information is generated using only the outlines that are facing, the data amount of the bone outline information can be reduced as a whole.

In this case, by using a part of the outline of the defined bone region, an estimated outline 112 to be incorporated as a part of the bone outline is interpolated and formed in a region where the defined bone region does not exist. It is preferable to add a function (estimated contour line interpolation means). Specifically, the function is to interpolate and form an estimated outline in such a manner that the opening that the skull should originally have besides the defect, such as the orbit, mouth, and nostrils, is closed. (Hereinafter, this is referred to as “hole filling processing”). With such a filling process, it is possible to avoid such a problem that the opening is mistakenly confused with a missing portion in the subsequent three-dimensional shape data creating process, thereby increasing the creating efficiency. In addition, when performing shape interpolation, which is a process of restoring the bone shape of the defective part from the bone outline data of the upper and lower slices, interpolation accuracy may not be ensured in the presence of an opening in some cases. In some cases, closing in advance can contribute to an improvement in interpolation accuracy.

A specific part of the bone candidate regions other than the maximum candidate region is selected as a selection candidate region.
By determining the confirmed bone region by incorporating the selected candidate region into the maximum candidate region, the accuracy of the missing part interpolation can be further improved. A specific example will be described below. That is, FIG. 17 is a bone extraction image at a tomographic position passing through the orbit and the nose shown in FIG. 9A, and the nasal bone part 110 located adjacent to the orbit 111 as an opening is selected as a selection candidate area. You have selected. The determined bone region is divided into the nasal region 110 and the nasal region 110 with the orbit 111 interposed therebetween.
Of the skull on the outer corner of the eye (opposite corner of the eye)
An interpolation curve 112 that blocks the eye socket 111 is generated using the outline of the eye 11a. As shown in FIG. 18, in the filling processing, the interpolation curve 112 is obtained by setting the interpolation curve 112 on the outline (the bone outline 103 in FIG. 12) of the nasal region 110 and the outer corner 111 a.
The operation is performed by hitting the control point 115 by operating the mouse 13 and recursively generating the interpolation curve 112 by using a known free curve generation tool (such as a spline curve). By incorporating the nasal bone region 110 into the defined bone region, the longitudinal dimension of the skull contour reflects a more accurate dimension, and consequently, a defect using the skull contour (bone outline). Can be further improved.

As a more accurate method, FIG.
As shown in (5), the estimated contour interpolation means can be realized as a function having the following means. Search circle setting means: With respect to the outline of the nose bone region 110 and the outline of the outer corner of the eye 111a, a search circle 117 having a predetermined radius that minimizes the geometric displacement from the outline is set. Interpolation reference part extraction means: For the search circle 117,
An outline part (matching part) whose distance is less than a predetermined reference value is extracted as an interpolation reference part. Curve control point setting means: sets a plurality of curve control points 115 for forming an interpolation curve on the extracted interpolation reference portion. Interpolated curve generation means: Generates an interpolated curve using these curve control points.

The present inventor has empirically found that the cross-sectional profile of the cross section of the head crossing the orbital portion can be approximated to a substantially circular arc including the nose bone portion in many individuals. Then, as described above, this is approximated by the search circle 117, and only the matching portion is used as the interpolation reference portion, thereby making it possible to increase the accuracy of the interpolation curve particularly filling the orbital portion. Can be suitably used as a part of the reference shape when generating the three-dimensional shape data of the defective portion.

The center O of the search circle 117 can be determined as the geometric center of gravity of the smallest circumscribed figure of a fixed shape with respect to the outline of the determined skull region. As shown in FIG. 20, the shape of the circumscribed figure, as shown in FIG. 20, is simplest in terms of processing and the accuracy of determining the center is high, so that it can be suitably used in the present invention. However, the circumscribed figure is not limited to a rectangle, and another figure such as an ellipse may be adopted. Also, without using a circumscribed figure, for example, a biological tissue that is located near the center of the head and has an absorption coefficient similar to that of a bone,
For example, the pineal gland may be referred to as the center position of the search circle.

An example of the processing flow of the bone data extraction program 201 described in detail above is shown in the flowchart of FIG. In steps S1 to S4, data obtained directly from an X-ray CT apparatus or the like is used as original image data (digital grayscale image data), or data once hard-copyed on a film or the like is captured by an image scanner to obtain data. It is the part that decides. And
In S7, the original image data of each slice is read, and in S8
Then, the bone outline 103 is extracted by setting the threshold density level KS. Then, in step S9, the bone candidate region 108 is extracted by labeling, and the determined bone region is further determined. In a slice such as the orbit that requires filling, the above filling processing is performed (S10). In this way, the final shape of the bone outline of the slice is determined and stored in the design data storage unit 207 (FIG. 2) as bone extraction binary data. Then, the above processing for each slice is sequentially performed,
When the processing for all slices is completed, the processing ends (S6, S1).
3, S14).

Thus, the process of generating the bone outline information is completed, and the process then proceeds to the bone outline three-dimensional synthesis processing. this is,
This is performed by the bone outline three-dimensional synthesizing program 202 shown in FIG. 2, and the flow of the processing is shown in FIG. S30 ~
In S31, the bone outline data (the point group data arranged at a predetermined interval as described above) of each slice is read out, and if there is a missing portion, the process is interrupted by the missing portion. The end point coordinates of the bone outline are determined (S33), and the data is stored.

In the following processing, the bone outline three-dimensional synthesizing program 202 causes the computer 50 mainly composed of the CPU 3 to function as the following units together with the bone outline data extraction program 201 described above. Bone region extracting means: Bone region 85 is extracted from each of tomographic images SC captured at a plurality of different tomographic positions. Bone outline information generating means: extracted bone region 8
5, bone outline information which is outline information of an area to be finally determined as a bone is generated. Means for generating missing outline estimated contour line information: missing part 400 (FIG. 2)
With respect to the bone outline 125 involving 3) and the like, based on the shape information of the bone outline 125, the estimated missing outline information that is the estimated outline information of the defective portion 400 is generated. Bone outline display means: The bone outline 125 is displayed together with the estimated outline 129 of the defect based on the bone outline information and the estimated defect outline information. Display control means: Display states of the bone outline 125 and the estimated outline 129 displayed on the bone outline display are made different from each other so that they can be identified.

More specifically, the bone outline 125 of each tomographic position and the estimated outline 129 of the defect are synthesized on the same screen based on the bone outline information and the estimated defect outline information. In addition to the display, the bone outline 125 and the estimated outline 129 of the defective part can be displayed in different colors and / or densities. However, a display mode in which identification is possible by a method other than color coding, such as shading or a solid line / broken line, is also possible.

Returning to FIG. 21, a specific processing flow will be further described. First, in S36, point group data of the bone outline of each slice is read, and three-dimensionally synthesized with reference to the tomographic position. 121 and 12 in FIG.
Reference numerals 3 and 124 denote composite display examples (display device 1 in FIG. 2).
1 is displayed on the screen by switching windows). Here, in the skull region, contour point group data as bone outline information is generated using only the outline facing the human outer surface side, and a plane projection 121 when the skull is viewed from the top of the head. , Front projection 123 when the skull is viewed from the face side
And a side projection 124 when the skull is viewed from the ear side.
The bone outline is displayed by point cloud data by the type of projection. By looking at this, the operator can roughly grasp the three-dimensional shape of the skull and the schematic shape of the defect 400, and can clearly image the expected completion form of the artificial bone for filling the defect 400. it can.

FIG. 23 is an enlarged view of the planar projection 121, and the outline of the bone at each tomographic position appears in an annual ring shape (or contour line shape). The most central part is the outline of the bone near the top of the head, which has a closed shape.However, the outline of the lower bone is interrupted on the right side of the drawing, and the end point is formed Understand. This is because a skull defect is formed here. The composite image of the plane projection 121 is as shown in FIG. Here, the plane projection plane is the XY plane, and the projection of the bone outline 125 (the end point is 126) of each slice is adjusted using specific coordinates (X0, Y0) as an alignment reference point. Note that reference numeral 126 is an end point.

In each of the projections thus synthesized, the estimated outline 129 of the missing portion is rather hard to see when displayed in a planar projection. Therefore, as shown in FIG. At least one of them (here, both of them). Since both projections use a projection plane that is orthogonal to the tomographic plane,
The bone outline 125 at each tomographic position appears as a parallel horizontal line. Of these, in the front projection 123, the defective portion is projected so as to be located at the outer edge of the projected image. And
The outer edge position of the estimated deficient outline 129 is formed based on the shape of the projected outer edge of the bone outline 125 of the healthy part in which the deficient part is not involved. Specifically, it is formed by mirror-inverting the projected outer edge of the bone outline 125 of the healthy part with respect to the center line (mirror plane) MO. In addition, since the end point 126 appearing on the front side (information of the end point of the bone outline 125 in which the defective part is involved) is also the end point of the estimated outline of the defective part, this also forms the estimated outline of the defective part. You can see that you are. On the other hand, side projection 124
Then, since the end points 126 and 126 of the bone outline 125 interrupted by the defect appear, the estimated outline 129 of the defect can be easily obtained by connecting these points. In this example,
The bone outline 125 is, for example, blue, and the estimated defect outline 129.
Are displayed, for example, in red (FIG. 21: S38, S
39).

By referring to the projection as described above,
In addition to being able to intuitively grasp the rough three-dimensional shape of the defective portion and reducing the possibility of mistaking a curved posture in the space due to an illusion or the like, there is an advantage that the operator can immerse himself in the work without confusion. . This is because, as shown in FIG. 26, the three-dimensional positional relationship of the bone outline 125 appears as a ring whose center is shifted in the vertical direction just like the latitude line of the earth, as shown in FIG. For example, as a result of erroneously recognizing the front-rear relationship between the near side and the rear side of the ring on the screen, whether the skull is viewed from above (A direction) as shown in FIG. To the lower side (B
It is often an illusion that you are looking up from direction. However, by referring to the projection as shown in FIG. 25 in advance and printing the approximate shape of the defective part and the positional relationship with the skull as an image, it is possible to effectively avoid such an illusion. You can do it.

As shown in FIG. 27, at least one of front projection and side projection (here, front projection 1
In 23), the composite projection position of each bone outline is determined so that smoothing correction is performed on the arrangement of the outer edge positions 130 of the bone outline projection 125 corresponding to each tomographic position on the side where no defect exists. An adjustment function (bone outline projection position adjustment means) can also be added. Since the three-dimensional combined data of the bone outlines described above is used as a basis for later generation of the three-dimensional shape data of the defective part, the positioning accuracy when synthesizing each bone outline 125 is determined by the three-dimensional data to be created. This has a direct effect on the accuracy of shape data creation. According to the above function, when synthesizing the bone outline 125,
When some of the bone outlines 125 are extremely deviated from the arrangement tendency of the other bone outlines 125 due to a sudden factor, this can be resolved by smoothing correction, and thus the three-dimensional shape The accuracy of data creation can be further improved.

The above-mentioned problem is caused by a positioning mark image (hereinafter also simply referred to as a mark image) 1 together with a human body tomographic image 90 as an image hard copy as shown in FIG.
This is likely to occur when the photographing film 80 on which the 32a is formed is used, and the bone outline of each slice is synthesized based on the mark image 132a. In the example of FIG. 29, an L-shaped mark image (target mark) 132a and a dimensional scale 1
31a are formed on the film 80. By fixing the position of the patient at the time of imaging, the mark image 132a and the scale 131a are such that the relative positional relationship with the patient, that is, the human tomographic image 90 is essentially the same for any slice. It is formed in. After the image is captured by the image scanner, the coordinates of the image in the Z-axis direction (slice arrangement direction) are specified from the slice number, while the X-axis direction and Y
The coordinates in the axial direction are determined by using a specific position on the mark image 132a, in this case, as shown in FIG. 30B, the outer vertex of the intersection of the two straight lines forming the L-shape as a positioning reference point. Is defined as the origin (X0, Y0), for example.

As shown in FIG. 29, the mark image 132a is usually formed on the outer edge of the film 80 so as not to interfere with the tomographic image 90. The problem actually occurs here. That is, as shown in FIG. 11B, when taking a photograph of a tomographic image displayed on the monitor of the CT apparatus, as described above, the image is more likely to be blurred out of the outer portion of the film, as shown in FIG. As shown in ()), the landmark image 132a also slightly changes its shape, such as being affected by blurring, or the L-shape becoming thicker or thinner due to variations in monitor brightness and exposure conditions. Sometimes. As a result, as shown in FIG. 30B, an overlay error may occur at the position of the alignment reference point. Another factor is that the patient moves when photographing, or as shown in FIG. 30C, when the film 80 is rotated and set in the scanner during capture by an image scanner, the landmark image 132a
In some cases, a rotational position shift may occur.

As shown in FIG. 30A, the extent to which the above-described positional deviation has occurred can be determined by comparing the human body tomographic image 90 with the mark image 132a for positioning each tomographic position captured by the image scanner. Fig. 2
By displaying the superimposed image 122 as shown in FIG. 8, it is possible to relatively easily confirm (overlap image image display means). In FIG. 22, this superimposed image 122
Are displayed together with the planar projection 121, the front projection 123, and the side projection 124 described above. In FIG. 28, it can be seen that the mark image 132a and the scale 131, which are displaced from each other, are superimposed and considerably thick. In this case, processing is performed so as to select whether or not to perform the projection position adjustment processing (smoothing correction) by the bone outline projection position adjustment means according to the state of the positioning mark image 132a appearing in the superimposed image image 122. (Adjustment selecting means). The selection can be made by inputting a command from the keyboard, clicking a soft button on the screen, or the like (FIG. 21: S40 to S44).

FIG. 27 shows an example of the smoothing correction processing. In (a), the outer edge positions 130 of the plurality of bone outlines 125 are arranged in a curved shape protruding outward reflecting the outer shape of the skull. However, one of them 125x protrudes outward, deviating from the tendency. Therefore, as shown in (a) to (e), a set of three moving averaging processes is performed from one side (here, the upper side) along the arrangement of the bone outlines 125. That is, the central bone outline 125 at the center of the three bone outlines 125 is located on a straight line connecting the two outer edges 130 on both sides. To move. This processing is repeated while shifting the positions one by one on the arrangement of the bone outlines 125 as shown in (c) → (d) → (e). As can be seen from the state (e), the abnormal protruding state of the bone outlines 125 is eliminated by smoothing, and the arrangement tendency of the outer edge position 130 of each bone outline 125 is also substantially preserved. Understand.

The bone outline three-dimensional composite data (bone outline information) created as described above is stored and saved as a part of the design data as shown in FIG.

When the three-dimensional synthesizing process of the bone outline is completed, the process proceeds to a three-dimensional shape data creating process using the outline point group data. This is a computer 50 mainly composed of the CPU 3.
Function as a three-dimensional shape data generating means. In this process, the already created bone outline three-dimensional combined data (FIG. 6) of each slice is read out, and is displayed on the display device 11 (FIG. 2) using the three-dimensional coordinates of XYZ.
To be displayed. As a display form, for example, a front projection (shown in a bird's-eye view from above) as shown in FIG. 35 and the like, a side projection as shown in FIG. 36, and a plane as shown in FIG. Various types of switching, such as a projection type, can be performed as needed. In the front overhead view projection of FIG. 35 and the like, the bone outline (bone outline) 12 which appears in an annular shape
5 are displayed in different display colors so as to avoid the illusion of the front-rear relationship between the front side portion and the rear side portion. Here, the first color indicating the front side is yellow, and the second color indicating the rear side is red. However, a mode in which identification is possible by a method other than color coding, such as shading or a solid line / broken line, is also possible. In the side projection shown in FIG. 36 and the like, the projection of each bone outline 125 is color-coded in a form corresponding to the front projection.

The three-dimensional shape data is created by a different method depending on the position of the skull where the defect is formed. This can be roughly divided into a pattern 1: a reflection reference plane MS which is appropriately set so that the skull is divided into two parts at the approximate center as shown in FIG.
With respect to P, when the defect 400 is present only on one side of the skull; When the part 400 includes the top position;
And different creation algorithms are employed.

First, in the case of the pattern 1, the lost portion is basically restored by mirror copy (mirror copy).
In this case, the three-dimensional shape data creation program 203
The computer 50 functions as the following units together with the programs 201 and 202 described above. Skull region extraction means: Extracts a skull image region from each of the tomographic images taken at a plurality of different tomographic positions in the head axis direction. Bone outline information generating means: Based on the extracted skull candidate region, generates bone outline information 125 as outline data of an area to be finally determined as a skull as point group data at predetermined intervals. I do. Three-dimensional shape data generating means: generates three-dimensional shape data of a skull defect based on the bone outline information 125 for each tomographic position. The three-dimensional shape data generating means includes the following means. Reflection reference plane setting means: Sets a reflection reference plane MSP for dividing the skull into left and right portions at the center of the face. Compensation part outline data generation means: missing part 400 existing on one side with respect to the set mirroring reference plane MSP, or missing part 40 existing on both sides in an asymmetrical left and right shape
0 to at least partially compensate for bone outline 1
Deficient portion 40 to be filled out of the point cloud data forming
By mirror-copying the data corresponding to 0 with respect to the mirroring reference plane MSP, the copied point group data is generated as supplementary part outline data. Mirror copy correction means: The copy destination outline of the point group data is copied so that the supplementary part outline 125 ′ generated by the mirror copy conforms to the contour of a healthy part existing around the defective part 400 to be compensated. Correct the position. Then, based on the supplementary part outline data based on the corrected point cloud data,
Generate three-dimensional shape data of the outer surface of the artificial bone for filling the defect.

The "mirror copy" is a process in which a point in a three-dimensional space is used as a copy source and a point is generated at a mirror-symmetric (plane-symmetric) position with respect to a predetermined mirror reference plane. Say. If the copy source is a bone outline formed of a group of points, a new bone outline is generated at a position where the bone outline is mirror-inverted with respect to the reflection reference plane.
However, since the position of the copy destination of the point cloud data is corrected so as to match the external shape of the healthy part existing around the defect to be compensated, it is not necessarily mathematically exact mirror reflection inversion. Needless to say, it should not be.

As shown in FIG. 33, when the outer shape of the head has a shape with a somewhat poor symmetry with respect to the reflection reference plane MSP, the shape data of the healthy side defect corresponding portion 401 is represented by M
Even if the SP is mirror-copied as it is, the copied shape data does not always conform to the shape of the defective part 400 or the healthy part around it. Therefore, as shown in FIG. 34A, as a method unique to the present invention, the position of the copy destination of the copy point group sequence 125 'generated by the mirror copy is corrected, so that the defective portion 4 to be compensated is corrected.
This is to achieve more accurate adaptation to the external shape of the healthy part existing around 00 (function of mirror copy correction means).

As described above, since the information of the bone outline 125 is defined as point cloud data, the mirror copy is also performed as a copy of the point cloud data cut out therefrom. Then, the position correction of the copy destination of the point cloud data can be performed by any of parallel movement, rotational movement, and a combination thereof as shown in FIG. This is a kind of linear transformation for each coordinate of the copy source point group, and the basic shape of the bone outline 125 is also preserved.

For correction, for example, as shown in FIG. 34, a reference outline 125 s is set on the outer surface of a healthy part existing around the defect 400 to be compensated, and the mirror-copied point group 125 ′ The operation can be performed so that the sum of the geometric displacements with the reference outline 125s is minimized. In this example, as shown in FIG. 34 (a), the copy source point group sequence of the mirror copy is used to compensate for the missing portion from the bone outline 125 located on the opposite side of the missing portion 400 with respect to the mirroring reference plane. It is used in such a form that it is extracted by a predetermined length more than the required minimum length. Specifically, the copy source point group sequence is represented by the bone outline 1 by the copy target area 140 having a larger area than the defective part 400.
As shown in FIG. 34 (b), the copy area 140 'is superimposed on the defective area 400 so as to include the peripheral area of the defective area 400 as well. Then, the copy point group sequence 125 ′ generated by the mirror copy is
A portion located around the missing portion of the bone outline 125 to which the copy source point group sequence belongs is defined as a reference outline 125 s and the column end 1
25k is copied so as to have a positional relationship opposite thereto.

The position of the copy destination of the copy point group row 125 'is corrected so that the total geometric displacement between the point group at the end of the row and the reference outline is minimized. For example, as shown in FIG.
5, the points forming the column end 125k and the reference outline 1
25s (for example, the bone outline 125 is X-Y
When parallel to the plane, the distance d in the X direction or the Y direction) or the square of the distance d (d ² ) is regarded as the geometric displacement, and as shown in FIG. Is calculated as the sum of displacements for all bone outlines 125 in the copy area 140 ′ ((e)
Represents an AA cross section of (c), where 125sp is each reference outline, 125′p is a copy point group row, and d1 to d6.
Indicates the geometric displacement between them at each bone outline 125). Then, the correction is performed by performing any one of a parallel movement, a rotational movement, and a combination thereof on the point group included in the copy area 140 ′ such that the value of the displacement sum Σ is minimized.

According to the above-mentioned method, an extra portion of the bone outline 125 located on the opposite side of the defect reference portion 400 with respect to the mirror reference plane is extracted by a predetermined length from the minimum length necessary for filling the defect portion. The position of the mirror copy is corrected so that the column end 125′k, which is an extra part, is adapted to the reference outline 125s based on the bone outline 125. That is, since the position of the mirror copy is corrected using only the information of the bone outline 125 precisely determined from the tomographic image without the interpolation curve or the like, the normal state of the mirror copy point group sequence after the correction is corrected. It is possible to further improve the accuracy of fitting to the part.

The length of the end 125 ′ k of the copy point group row 125 ′ that overlaps with the reference outline 125 s is the length corresponding to the actual size from the viewpoint of increasing the accuracy of matching with the reference outline 125 s. 5 mm or more, desirably 10
mm or more is preferably secured. Further, although the calculation time of the displacement sum 長 く becomes longer, the copy target area 140 may extend over the entire outer surface of the head on the side where the missing portion does not participate with respect to the mirroring reference plane MSP (FIG. 33). The bone outline 125 generally has a locally small radius of curvature at a corner from the forehead to the temporal region, and the copy target area 140 has a set position and size to include the corner. Can be determined. For example, at a position off the corner, the bone outline 125 has a shape close to an arc, and when the point group in the copy target area 140 is mirror-copied, the bone outline 125 is displaced in the circumferential direction. However, there is a difficulty that the influence is hardly reflected on the displacement sum し て も even if the occurrence of occurs. However, when the corner portion is included, the displacement in the circumferential direction causes a rapid increase in the sum of displacement Σ, so that the circumferential displacement can be effectively identified and prevented.

When the target of the mirror copy is the point group data of the bone outline at a plurality of tomographic positions (slices), FIG.
As shown in FIG. 7, the point group (that is, the point group in the area 140) extending over the plurality of tomographic positions is grouped, and the correction is made to the integral point group so that the relative positional relationship between the points is preserved. Can be done against By doing so, it is possible to copy in a form in which the shape of the loss corresponding portion 401 is substantially preserved. On the other hand, it is also possible to perform position correction individually for each bone outline without performing such grouping. In this case, since the calculation must be performed for each bone outline (each slice), the calculation load on the computer 50 increases somewhat, but the compatibility with the shape of the healthy part is prioritized. Fine correction is possible.

Next, pattern 2 or pattern 3 can be processed according to the flow shown in FIG. In this case, the three-dimensional shape data creation program 203 causes the computer 50 to function as the following units. Interpolation reference line generating means: The interpolation reference line 14 reflecting the surface shape of the healthy part of the bone based on the bone outline information 125.
9,163 is set. Curve control point setting means; interpolation reference line 149 (149a,
Curve control points 152 and 168 for generating an interpolation curve are set along 149b) and 163. Missing part interpolation curve generation means: set curve control point 15
2, 168 in accordance with a predetermined curve determination algorithm, for performing a curve interpolation of a missing portion.
1,164 is generated. The missing part interpolation curve 151,1
Based on 64, three-dimensional shape data of the outer surface of the artificial bone for filling the defect is generated.

The interpolation reference line is composed of a first reference line segment (FIG. 39: 149a, FIG. 46: 163a) having an end point on the inner peripheral side of the defective part in a healthy part adjacent to the defective part.
A second reference line segment that is located on the opposite side of the first reference line segment across the defective portion in the direction in which the first reference line segment extends toward the defective portion, and that has an end point on the inner peripheral side of the defective portion. (FIGS. 39: 149b, FIG. 46: 1)
63b) can be set. And
As shown in FIG. 40A, the missing part interpolation curve 151 is
The first and second reference line segments 149a, 149
Based on each curve control point on b, both segments 149a,
149b can be determined. With this configuration, it is possible to generate the three-dimensional shape data of the defective part having higher shape conformity to the healthy part from the information of the healthy part shape on both sides of the defective part.

The missing part interpolation curve can be generated by a so-called free curve tool. A plurality of control points to be set are divided into a predetermined number, and a curve passing through each set of control points is converted into a relatively simple mathematical formula, for example. It can be treated as a segment described by a second-order or higher polynomial. in this case,
Replace the missing part interpolation curve with the outline of the healthy part (reference line segment)
In order to smoothly connect the curve interpolation points to the curve control points, the missing portion interpolation curve is formed at each end point of the first and second reference line segments so that at least the first-order derivative substantially matches each of the reference line segments. It is desirable to determine recursively based on the coordinate information. As such an interpolation curve,
In particular, the curve displayed by the interior method, that is, the spline curve is intuitive and easy to handle, and can be suitably used in the present invention.

For example, in a cubic spline curve, FIG.
As shown in (b), a curve connecting each control point is represented by a cubic equation,
That, f is intended to approximate at (r) = k0 + k1 · r + k2 · r 2 + k3 · r 3 ‥‥ (1) ( a variable representing the position coordinates, taken here to be (r, f (r)) ) For example, if the first four points p0 to p3 to which a curve is to be applied are designated as control points, a simultaneous equation that determines the four coefficients k0 to k3 of the equation (1) is obtained. A cubic curve segment sg1 is determined. Next, three points p3 to p5 are selected and similarly applied to the equation (1). With this alone, only three equations are obtained, but the fourth equation is derived from the condition of tangent vector direction (ie, first derivative) matching at p3 for the already determined cubic curve segment sg1. By combining this with the above three equations, the coefficients k0 to k3 can be determined, and the cubic curve segment sg2 is obtained. In the same manner, by sequentially determining and connecting the cubic curve segment in the same manner while dividing into three points,
A cubic spline curve that smoothly connects an arbitrary number of control points can be obtained.

However, the above cubic spline curve has a disadvantage that the algorithm is simple, but, for example, moving or deleting even one control point on the curve affects the entire curve and increases the amount of calculation. The following B-spline curve can be used as a method for solving this disadvantage. In the B-spline curve, a point on the curve defined by the base spline function is set as a control point, and the curve shape can be more easily controlled by designating / changing the control point. This control point is also called a node, and the n-th order B-spline curve is defined by Expression 3 by treating the node conceptually as an element of a (n + k + 1) -dimensional contact vector.

[0087]

(Equation 3)

Here, Nk, r (r) is a node vector r
(= [R0, r1,..., Rm] (r0 ≦ r1 ≦ ‥ ≦ rm)). This is the control point Vk
, And is defined as in the following Expression 4.

[0089]

(Equation 4)

Therefore, the n-th order B-spline curve is represented by n-
It is a spline curve formed by connecting the primary curves (for example, if n = 4, it is close to a cubic spline curve). In this specification, the B-spline curve is
The concept includes not only a narrow B-spline curve with uniform node intervals but also a concept of a Nurbs curve with uneven node intervals.

Hereinafter, as a specific processing example, FIG.
A description will be given of a case where the defective portion 400 does not include the top of the head in FIG. In this case, the processing flow is from S52 to S53 in FIG. First, FIG. 35 shows a front projection, and it can be seen that end points 126 are generated in some of the lines crossing the forehead of the bone outline 125 due to breakage due to loss. By connecting these end points 126,
The planar shape of the defect can be inferred. FIG.
Reference numeral 6 denotes a side projection, in which the above-mentioned end point 126 appears in a left portion corresponding to the forehead. By connecting these end points 126, it is possible to infer the depth of the gouging of the defective portion.

First, in FIG. 35, all or some of three or more end points 126 of the interpolation reference line on the opening side of the missing portion are selected as the curve control points 126a, and as shown in FIG. Based on the curve control point 126a, the outer contour line 15 representing the outer peripheral shape of the artificial bone 401 for replacement.
0 is determined in a form corresponding to the opening shape 400 of the defective portion (function of the outer surface contour determining means: FIG. 31, S52 and S5).
3). At first glance, the use of all of the end points 126 seems to be advantageous because the shape of the defective opening can be accurately traced, but this is not always the case. In other words, as shown in FIG. 38, when all of the end points 126 are selected as control points in the case where the wavy shape or the unevenness of the defective opening shape is large, the determined outer surface contour 150 may partially enter the healthy part side. There is. In this case, it becomes impossible to fit the prosthetic artificial bone 401 into the defective portion. Therefore, the end points 126 are appropriately thinned out (that is, not adopted as control points) to prevent the outer contour 150 from entering the healthy part. (In FIG. 37, the end points with the reference numeral 126b are thinned out.) If the gap formed between the outer contour 150 and the inner edge of the defective opening becomes larger due to the thinning of the end points 126, the outer contour 15
For example, one or more new control points may be generated on the zero, and the shape of the outer contour 150 may be adjusted so that the gap formed is reduced by changing the position of the control point.

On the other hand, when the undulations and irregularities of the shape of the defective opening are extremely faithfully copied, as shown in FIG. . As is well known, the artificial bone for replacement 401 is formed by firing ceramics, so that there is a problem that cracks and cracks C are easily generated from such irregularities as starting points. In such a case, as shown in FIG.
Some end points at positions corresponding to the irregularities are not adopted so that the irregularities have a shape artificially adjusted.

Next, as shown in FIG. 39, the first reference line segment 149a and the second reference line segment 149b are separated from the bone outline 125 interrupted by the presence of the defect.
Both end portions facing the defective portion are used. And
According to the method described with reference to FIG. 40, the interpolation curve 151 is related to the missing portion using a part or all of the point cloud data of the first reference line segment 149a and the second reference line segment 149b (that is, the missing part). Are generated for each bone outline 125 having an end point 126 (FIG. 31: S55 to S5).
9). FIGS. 41 and 43 show the interpolation curve 151 formed in this way by plane projection or side projection. This is the estimated outline 1 of the defective part shown in FIG.
Unlike 29, since interpolation is generated based on the interpolation reference line shape of the healthy part, it reflects a more precise three-dimensional shape of the defective part.

A function for generating one or more new curve control points can be provided on the missing part interpolation curve 151 (curve control point generation means). For example, by looking at the planar projection in FIG. 41, it is possible to confirm the compatibility of the interpolation curve 151 with the healthy part, for example, the degree of smooth connection of the curves and the presence or absence of unnatural irregularities. If the shape is not satisfied, as shown in FIG. 42, the interpolation curve 151 is manually set by the pointer P or automatically generated by designating the interval / number.
A new curve control point 154 can be generated above. By selecting a desired one of the curve control points 154 with the pointer P and changing the position by a known mouse drag movement or the like (or manual input of coordinates), the shape of the missing part interpolation curve 151 can be freely adjusted. Yes (function of the missing part interpolation curve shape adjusting means).

Referring back to FIG. 31, in S68 and S69, three-dimensional shape data is generated based on the determined information on the interpolation curve 151 and the outer contour 150, and is stored as design data. In the present embodiment, the three-dimensional shape data is described by a so-called solid model. FIG. 50 conceptually illustrates this. In other words, the lines and vertices constituting the solid (in addition to the geometrically narrow vertices, 2
The common point of the two lines is also called a vertex in a broad sense), the relationship between the line and the vertex, and the correspondence between the surface and the line, and further, data defining the entity side for each surface is added. Here, as shown in FIG. 40, a process of generating curved surface data of a free-form surface in a portion surrounded by each of the curves by, for example, a sweep surface or a spline surface is performed (FIG. 31: S6).
8). Further, the interpolation curve 151 and the outer contour 150 are
Each end point or shared point position is defined by a coordinate value (the outer contour 150 may be regarded as a segment-like line divided by these points), and each line is associated with curved surface data. Then, for each surface, a method of designating one point of the entity side as shown in FIG. 50A, a method of defining a normal vector as shown in FIG. 50B, or a method of defining a surface as shown in FIG. The entity-side prescribed data is generated and added by any of the methods of designating the rotation direction of the surrounding ridge line, and the creation of the three-dimensional shape data is completed.

As shown in FIG. 32B, when the shape of the defective portion 400 is asymmetric with respect to the mirroring reference plane MSP, it may be possible to partially restore the mirroring copy. In this case, the process proceeds from S54 in FIG. 31 to S60, and the remaining side of the healthy part (which may be only one side or both sides) is mirror-copied to the other side. The correction of the copy destination position, which has already been described with reference to S34, is performed (S60, S61). As described above with reference to FIG. 32A, the processing in the case where the missing portion is completely compensated by the mirror copy has already been described. In this case, the process proceeds to S68.
On the other hand, when it is not completely compensated for as shown in FIG. 32B, the uncompensated portion is compensated for by the interpolation curve by the above-described processing from S55 onward in a state where the compensated portion by the mirror copy is incorporated into the healthy portion. It is.

Next, in the case of the pattern 3 which is a defective portion including the top of the head, the processing after S64 is performed. The basic idea of the processing here is as follows. That is,
As shown in FIGS. 44 to 47, healthy part three-dimensional surface data that defines the three-dimensional outer surface shape of the healthy part is created based on the information of the bone outline 125. Then, the healthy part outer surface object 16 stretched based on the healthy part three-dimensional surface data
An interpolation reference line 164 that intersects with the bone outline in a predetermined direction that crosses the defect 161 is set on 0.

The interpolation reference line can be specifically formed as follows. First, as shown in FIG. 44, a reference plane SP is set on the bottom side of the healthy part outer surface object 160, and FIG.
As shown in (5), a projection source reference line 162 is set on the reference plane SP, for example, in a radial pattern. Then, as shown in FIG. 46, the projection reference line 162 can be projected onto the healthy part outer surface object 160 to form the interpolation reference line 163.

For example, as shown in FIG. 45, if the projection source reference line 162 is set so as to cross the defective portion 161 (FIG. 44) in planar projection, the corresponding interpolation reference line is obtained as shown in FIG. The segment of the line 163 appears on both sides of the defect 161. These are the first baseline segment 16
3a and the second reference line segment 163b. Then, according to the same principle as in FIG. 40, an interpolation curve 164 can be generated as shown in FIG. In this case, the interpolation curve 164 appears radially in the shape of an umbrella. If they do not always intersect at one point, control points can be generated on each interpolation curve 164 by the above-described control point generation processing, and they can be connected to one point α by, for example, manual correction. Thereafter, as shown in FIG. 48, control points 165 for defining the surface are generated on the interpolation curve 164 at predetermined intervals, for example, and the surface segments 167 are generated in order from the lower side toward the intersection α, for example. .

As shown in FIG. 49, a parallel projection source reference line SP may be generated. In this case, the control point 168 generated on the projection source reference line SP
The interpolation curve 164 can be generated in exactly the same manner as in FIG.

By using the three-dimensional shape data created by the above method to control the operation of the machining device 15 by the cutting (machining) program 205, as shown in FIG. Molding material W
(For example, a powder molding material in which ceramic powder is bound by a binder) is directly cut to form an unfired cut body 401 ′, which is then fired to manufacture the artificial bone 401 for filling a defective portion. . As described above, the three-dimensional shape data includes the artificial bone 40 to be finally obtained.
1 defines the shape of the outer surface 401a. However, in consideration of shrinkage due to firing, the unfired cutting body 401 ′
Must be formed larger than the final artificial bone 401. In this case, data conversion for enlargement can be appropriately performed on the three-dimensional shape data according to the contraction rate.

Then, the cutting program 205 determines the outer surface 401a 'of the unfired cutting body 401' to be finally cut.
Is set as the cutting limit position, cutting path data of the tool T for the ceramic work material (work) W is created, and the cutting path data is used as the data interface 14.
To the processing device 15 of FIG. Processing equipment 15
In such a manner, the relative movement defined by the cutting path data occurs between the ceramic workpiece W and the tool TL,
The operation control of the motors 61 and 62 of the XY table for work feeding and the motor 63 for feeding the tool in the Z direction is performed.

FIG. 53 shows an example of the flow of processing by the cutting program. In S100, first, dimensions such as the length, width, and height of the ceramic workpiece W are input, and in S101, a tool type and a cutting medium type are selected.
In the storage device 6 of FIG. 2, the machining condition database shown in FIG. The tool cutting amount and the work feed speed based on the XY table are read out, and are automatically set in the processing device 15.

Referring back to FIG. 53, the three-dimensional shape data created and stored as the design data is read out in S102,
First, rough cutting processing is performed in S103 to S115. Here, the tool cutting direction (vertical direction) is the Z direction, the vertical work feed direction is the X direction, and the horizontal work feed direction is the Y direction. First, in S104, the tool is moved to the home position (X, Y, Z =
0), and feed unit ΔX determined according to the work feed speed
Is used to feed the workpiece in the X direction and perform cutting (S105-S
108). Then, it returns to the position immediately before entering the cutting limit position defined by the three-dimensional shape data, or returns to X = 0 when reaching the limit position Xmax (S108 → S1).
09), and is moved by the feed unit ΔY in the Y direction (S109).
〜S111 → S105), and the cutting in the X direction from S105 to S108 is repeated. When this process is repeated and the relative position between the tool and the workpiece reaches the limit position Ymax in the Y direction, the tool is raised by the tool cutting amount ΔZ and X, Y = 0.
(S114 → S105), S105-S111
Is repeated.

Referring to the screen display of the cutting simulation (to be described later) shown in FIG. 54, the work W is roughened in a substantially stair-like shape.
It should be noted that the material object indicating the work W is three-dimensionally displayed in gray tones, and the tool object indicating the tool appears as a white bar. The back side of the upper surface of the work is a rough cut surface, and the front side is a finished surface. On the other hand, in the case of cutting the outer surface shape whose central portion is convex, stepwise rough cutting cannot cut the other side of the convex portion in the X feed direction. Then, the origin is appropriately changed, and the processing from S104 is repeated again. As a result, rough cutting on the side opposite to the convex portion, which could not be cut by the previous rough cutting, is performed.

When the rough cutting is completed, the tool is appropriately changed,
The type of the tool and the type of the cutting medium (accordingly, the corresponding feed rate and cutting depth: see FIG. 7) are changed to those for finishing, and the process proceeds to finishing in S118. In the finishing process, by sending the tool so as to follow the final surface shape specified by the three-dimensional shape data, small irregularities on the rough cut surface are stroked away, and the surface finish is performed (see FIG. 54). See again).

As described above, the conventional cutting method has the following problems. Depending on the position of the missing part, when cutting uneven parts,
As shown in FIG. 57A, the tool interferes with a part of the material to be cut which should not be cut, which may cause a problem such as a decrease in cutting accuracy or an inability to cut. As shown in FIG. 52 (b), when the defect filling shape is vertical or oblique, the efficiency and the material yield are significantly reduced.

The first problem is that, based on the three-dimensional shape data of the defective portion, a processing volume parameter reflecting the processing volume of the ceramic material to be processed is calculated, and the value of the processing volume parameter is optimized. The problem can be solved by performing a predetermined rotation conversion on the three-dimensional shape data of the defective portion. This processing is controlled by the optimum cutting position determination program 204 shown in FIG. FIG. 51 shows an example of the processing flow.
Is shown in Rotational movement around the horizontal axis (X axis and Y axis) is mainly effective in optimizing the cutting volume. First, in S70 and S71, the three-dimensional shape data is subjected to rotation conversion in a predetermined rotation angle unit ΔθX, and the three-dimensional shape data after the rotation is spatially expanded. Dimension LX in the direction and the Z direction,
Find LY and LZ. For example, LX is defined as Lmax = Xm, where Xmax is the maximum value of the X coordinate values and Xmin is the minimum value of the coordinate points belonging to the surface, based on the surface data of the outer surface of the missing portion filling shape.
Calculated as ax-Xmin. LY and LZ are similarly LY = Y
It is calculated as max-Ymin, LZ = Zmax-Zmin. Then, in S72, the processing volume parameter V is calculated as V = LX × LY.
Calculate as × LZ. As shown in FIG. 52, this is the volume of the circumscribed rectangular parallelepiped on the outer surface 401a of the defect-repairing shape, and it is clear that the larger this is, the larger the processing volume, that is, the processing allowance.

In S73, the angular position around the X-axis at which V becomes minimum is found and fixed, and in this state, the same processing is performed for rotation around the Y-axis in S74 and below. As a result, the optimum rotation angle position of the three-dimensional shape data that minimizes the processing volume is determined. Then, in S80, the three-dimensional shape data is stored again in a state where the rotation conversion to the determined rotation position has been performed.

On the other hand, with regard to the above problem, it is necessary to avoid any stupid notice that processing is impossible after actually starting processing. Therefore, in the present embodiment, it is possible to execute a cutting simulation for the ceramic workpiece based on the three-dimensional shape data of the defective portion. This cutting (machining) simulation is performed by the simulation program 207 of FIG. 2, but the basic processing flow is exactly the same as that of FIG. Here, instead of the actual ceramic workpiece W, three-dimensional shape data of a material object by a solid model is used, and instead of a tool, three-dimensional shape data of a tool object by a solid model is also used. A cutting simulation by computer graphics is performed according to the tool cutting amount and the workpiece feed speed set based on the data shown in FIG. For example, a material material object,
If a spatial overlap with the tool object occurs, it is determined that the overlapped portion has been cut, and a process of deleting or invalidating the data portion corresponding to the cut portion from the material material object is performed. Thereby, as shown in FIG. 54 or 55, it is possible to display a simulation screen representing the processing progress. Incidentally, FIG.
(A) shows a state where the finishing process on the front side of the artificial bone has progressed halfway, and (b) shows a state where the finishing process has been completed. FIG. 55 (a) shows a state in which the finishing process on the back side of the artificial bone has progressed halfway, and FIG. 55 (b) shows a state in which the finishing process has been completed.

The tool cutting amount and the workpiece feed speed may be different depending on the type of the material to be processed. For example, in addition to the ceramic unfired molding material using the processing apparatus,
For example, a gypsum model can be cut. In this case, data shown in FIG. 7 is prepared for each type of material to be processed, such as a ceramic unfired molded material or gypsum, and data is appropriately selected for each material. Can also be used.

By the above-described cutting simulation, it is possible to easily know whether or not there is a possibility that the cutting cannot be performed due to the factors shown in FIG. 57 without actually performing the processing. For example, in the machining by the machining path set at the time of the simulation, it is monitored whether or not the tool enters inside beyond the machining limit position defined by the three-dimensional shape data in the already cut portion, and the tool enters inside. If so, it can be determined that cutting is impossible. In addition, rapid traverse processing, which is absolutely impossible with actual cutting, can be performed extremely easily by simulation processing, and the possibility of cutting can be quickly known. Also, referring to the feed speed of the tool or the work and the total path length from the start of cutting to the end of cutting as shown in S119 of FIG. 53, the time required for processing can be easily known.

If it is determined in the result of the cutting simulation that normal cutting cannot be performed, the three-dimensional shape data of the defective portion is added to the data so that normal cutting can be performed.
A predetermined movement conversion consisting of a parallel movement, a rotational movement, or a combination thereof can be performed. In addition, when the cutting optimal position determination processing is prioritized, the rotation angle position of the three-dimensional shape data that minimizes the machining allowance may not always be able to be cut. In order to solve the two problems of ensuring normal cutting and minimizing machining allowance, FIG.
The processing shown in FIG. First, in S120, a cutting simulation is performed by setting the three-dimensional shape data at a certain initial position, and thereafter, the three-dimensional shape data is gradually rotated (or translated; here, rotated to facilitate understanding). The cutting simulation is repeated. And
With reference to the result of the determination, the angle positions that can be cut and the angle positions that cannot be cut are registered and stored, for example, as part of the design data (S121 to S123).

When all the effective angle ranges have been examined, the cuttable angle range [θV] should have been stored. By performing the position determination process, it is possible to obtain a converted position of the three-dimensional shape data that allows normal cutting and minimizes the cutting allowance.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of X-ray CT.

FIG. 2 is a block diagram showing an example of an electrical configuration of a prosthetic artificial bone processing system using the prosthetic artificial bone design system of the present invention.

FIG. 3 is an explanatory view conceptually showing a method for acquiring image data of a human body tomography.

FIG. 4 is a block diagram illustrating an example of an electrical configuration of the processing apparatus.

FIG. 5 is a conceptual diagram showing an example of a storage form of design data.

FIG. 6 is a conceptual diagram showing an example of the contents of design data.

FIG. 7 is a conceptual diagram showing a configuration example of setting data of a tool cutting amount and a feed speed for a processing apparatus.

FIG. 8 is an X-ray photograph showing a setting example of a tomographic position where imaging by X-ray CT is performed.

FIG. 9 is a diagram showing some output examples of tomographic images by X-ray CT.

FIG. 10 is a flowchart illustrating an example of a flow of a bone part data extraction process.

FIG. 11 is a view for explaining a cause of image blurring in a bone region due to a tomographic position.

FIG. 12 is an explanatory diagram showing a method of manually inputting a bone outline.

FIG. 13 is an explanatory view showing a method of automatically generating a bone outline by selecting a threshold density level.

FIG. 14 is a diagram showing a principle of binarization of original image data and a setting example of a display color according to a CT value.

FIG. 15 is an explanatory diagram showing an example of extraction of original image data having shades of gray and bone part candidate regions.

FIG. 16 is a view for explaining the principle of region separation and extraction by labeling.

FIG. 17 is an explanatory diagram showing an example of processing for filling an opening of a skull with an interpolation curve.

FIG. 18 is an explanatory diagram showing a method of generating an interpolation curve.

FIG. 19 is an explanatory diagram showing a method for generating an interpolation curve with higher accuracy using a search circle.

FIG. 20 is an explanatory diagram showing a method for determining the center of a search circle.

FIG. 21 is a flowchart illustrating an example of the flow of a contour point grouping process.

FIG. 22 is a view showing an example of a screen displaying bone outlines in various projections.

FIG. 23 is an enlarged view showing the planar projection of FIG. 22;

FIG. 24 is an explanatory diagram showing a state in which bone outlines of a plurality of slices are combined with a plane projection.

FIG. 25 is a diagram schematically illustrating front projection and side projection of FIG. 22;

FIG. 26 is a view for explaining an example of an illusion caused by stereoscopic synthesis projection of a bone outline.

FIG. 27 is a diagram illustrating a position method of smoothing correction of a bone outline.

FIG. 28 is an enlarged view showing a superimposed image of a grayscale image in FIG. 22;

FIG. 29 is a diagram showing an example of an X-ray CT film on which a positioning mark image has been formed.

FIG. 30 is a view for explaining how a positioning mark image is blurred and a positioning error occurs.

FIG. 31 is a flowchart illustrating an example of the flow of a three-dimensional shape data creation process.

FIG. 32 is a view showing various forms of occurrence of defects in the skull.

FIG. 33 is a view for explaining a problem in a case where a defective portion is restored by mirror copying.

FIG. 34 is an explanatory diagram of a method of performing a mirror copy while correcting the position of a copy destination.

FIG. 35 is a diagram showing an example of a front projection display of a bone outline synthesizing image used when creating three-dimensional shape data of a defective portion generated on a forehead;

FIG. 36 is a diagram showing a side projection display example.

FIG. 37 is a diagram showing an example of forming an outer contour line of a defective portion.

FIG. 38 is a schematic view showing an example of the shape of an artificial bone for filling a defective portion having irregularities on the inner surface of the opening.

FIG. 39 is a diagram showing an example of occurrence of an interpolation reference line and a missing portion interpolation curve.

FIG. 40 is a view for explaining an interpolation curve and a method of generating surface data based on the interpolation curve.

FIG. 41 is a diagram showing an example of a planar projection display of a composite image of a bone outline used when creating three-dimensional shape data of a defective portion generated on a forehead;

FIG. 42 is an explanatory diagram showing how control points are generated on a missing part interpolation curve and shape correction is performed using the control points.

FIG. 43 is a view showing an example of a side projection display after forming a missing portion interpolation curve.

FIG. 44 is a perspective view schematically showing a healthy part outer surface object used when creating three-dimensional shape data of a defective part generated at the crown.

FIG. 45 is a diagram showing a display example of a healthy part outer surface object in which a projection reference line is set.

46 is a diagram showing a display example of a healthy part outer surface object in a state where an interpolation reference line is formed based on the projection source reference line in FIG. 45.

47 is a diagram showing a display example of a healthy part outer surface object in a state where a missing part interpolation curve is formed based on the interpolation reference line of FIG. 46.

FIG. 48 is an explanatory diagram showing how surface data is generated between the missing part interpolation curves in FIG. 47;

FIG. 49 is a three-view drawing of a healthy part external object showing an example in which a projection reference line is set in a parallel line shape.

FIG. 50 is a view for explaining the concept of three-dimensional shape data based on a solid model.

FIG. 51 is a flowchart showing an example of the flow of a cutting optimum position determination process.

FIG. 52 is a view for explaining rotation conversion of three-dimensional shape data for minimizing a cutting volume.

FIG. 53 is a flowchart showing an example of the flow of a processing process or a simulation process thereof.

FIG. 54 is a diagram showing an output example of a processing simulation.

FIG. 55 is a view showing another output example of the processing simulation.

FIG. 56 is a view for explaining a method for producing a prosthetic artificial bone by directly cutting a work material.

FIG. 57 is an explanatory diagram of a case where cutting becomes impossible due to a positional relationship between a work material and a tool.

FIG. 58 is a flowchart showing an example of the flow of a cutting optimum position determination process in consideration of the possibility of machining;

[Explanation of symbols]

 Reference Signs List 1 artificial bone processing system for supplementation 3 CPU 7 image scanner 11 display device (image display means) 15 processing device 40 artificial bone design system for supplementation 50 computer 80 film 90 tomographic image 93 maximum candidate area 99 pixel density monitor bar (pixel density display means 100) Estimated bone part image area 101 Curve control point 102 Reference density level area 103 Bone outline information 108 Skull part candidate area 117 Search circle 122 Superimposed image 125 Bone outline information 129 Estimated outline of defective part 132 Positioning mark Image 149 Interpolation reference line 149a First reference line segment 149b Second reference line segment 150 Outer surface contour 151,164 Missing part interpolation curve 152,168 Curve control point 160 Healthy part outer surface object 162 Projection reference line 200 Control program 400 defects SC tomographic position MSP mirroring the reference plane Kx threshold concentration level Ki reference density level

Claims (14)

[Claims] 1. A system for designing an artificial bone for filling a bone defect based on a tomographic image of a human body, wherein an image region of a bone is determined in each of tomographic images photographed at a plurality of different tomographic positions. Bone region extraction means for extracting, and bone outline information generation means for generating bone outline information, which is outline information of a region to be finally determined as a bone, based on the extracted bone region And three-dimensional shape data generating means for generating three-dimensional shape data of the missing part of the bone based on the bone outline information for each tomographic position, wherein the three-dimensional shape data generating means comprises: An interpolation reference line generating means for setting an interpolation reference line reflecting the surface shape of the healthy part of the bone based on the bone outline information, and a curve control point for generating an interpolation curve along the interpolation reference line. Curve control to set Point setting means, and using the set curve control points, according to a predetermined curve determination algorithm, comprising a missing part interpolation curve generating means for generating a missing part interpolation curve for curve interpolation of the missing part, A prosthetic artificial bone design system for generating three-dimensional shape data of an outer surface of an artificial bone that compensates for the defective part based on a defective part interpolation curve. 2. The method according to claim 1, wherein the interpolation reference line is defined as point group data arranged at a predetermined interval, and a part or all of the point group is used as the curve control points. Artificial bone design system for replenishment. 3. The interpolation reference line is directed to a first reference line segment having an end point on an inner peripheral side of the defective portion in a healthy portion adjacent to the defective portion, and a first reference line segment extending to the defective portion side of the first reference line segment. In the extension direction of, while located on the opposite side of the first reference line segment across the defect,
A second reference line segment generating an end point on the inner peripheral side of the missing portion, wherein the missing portion interpolation curve generation means controls the missing portion interpolation curve by controlling each curve on the first and second reference line segments. 3. The artificial bone design system for replacement according to claim 1, wherein the system is determined based on the points by connecting the two segments. 4. The missing portion interpolation curve generating means, wherein at each end point of the first and second reference line segments, the missing portion is adjusted so that at least the first-order derivative of each of the reference line segments substantially coincides. 4. The artificial bone design system for replacement according to claim 3, wherein the interpolation curve is recursively determined based on the coordinate information of the curve control points. 5. The first reference line segment and the second reference line segment are used at both ends of the bone outline which is interrupted by the presence of the defect, facing the defect. The artificial bone design system for replacement according to claim 3 or 4. 6. The artificial bone design system for replacement according to claim 5, wherein the bone outline is defined as point cloud data arranged at predetermined intervals. 7. The healthy part three-dimensional surface data that defines a three-dimensional outer surface shape of a healthy part based on the information of the bone outline, and the interpolation reference line generating means generates the healthy part three-dimensional surface data. The interpolation according to any one of claims 1 to 4, wherein an interpolation reference line that intersects with the bone outline in a predetermined direction crossing the defect is set on a healthy part outer surface object stretched based on the following. Artificial bone design system. 8. The interpolation reference line sets a reference plane in a three-dimensional coordinate space in which the healthy part outer surface object is arranged, and sets a projection source reference line on the reference plane in a radial or parallel line shape. 8. The artificial bone design system for replacement according to claim 7, wherein the projection source reference line is formed by projecting the reference line on the outer surface object of the healthy part. 9. The three-dimensional shape data generating means may include all or three of the end points of the interpolation reference line on the opening side of the defective portion.
A part of the above is selected as a curve control point, and based on the selected curve control point, an outer contour line representing the outer peripheral edge shape of the artificial bone for replacement is formed into a shape corresponding to the shape of the opening of the defect. Based on the determined outer surface contour data and the data of the missing portion interpolation curve, the curved surface data surrounded by those lines is determined based on the determined outer surface contour data and the data of the missing portion interpolation curve. The artificial bone design system for replacement according to any one of claims 1 to 8, wherein the system is generated as three-dimensional outer surface data. 10. A curve control point generating means for generating one or a plurality of new curve control points on the missing part interpolation curve, and manually changing the position of the generated curve control point to thereby perform the missing part interpolation. The artificial bone design system for supplementation according to any one of claims 1 to 9, further comprising means for adjusting a shape of a curve for adjusting the shape of a curve. 11. The method according to claim 1, wherein the bone part is a skull, and the tomographic image is a sliced image of a head obtained by setting the axial direction so as to penetrate the skull in a longitudinal direction. The artificial bone design system for replacement according to any one of the above. 12. The image data of the human body tomographic image, which is directly generated as digital image data based on a photographing signal acquired by a tomographic image photographing apparatus is used. 2. The artificial bone design system for replacement according to 1. 13. The tomographic image photographing apparatus according to claim 13, wherein image data obtained by converting an image hard copy output as a photographing film of a monitor image or an image printout with an image scanner is used as the image data of the human tomographic image. Item 13. An artificial bone design system for replacement according to any one of Items 1 to 12. 14. A prosthetic artificial bone that indicates a material to be processed by referring to the three-dimensional shape data created by the artificial prosthetic bone design system according to claim 1. A method for producing a prosthetic artificial bone, comprising a step of processing into a shape.