JPH09248292A - Diagnostic device and method for osteoporosis - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce time required for image analysis and to improve the accuracy of image analysis, by indicating the direction of bone trabecula, analyzing the bone trabecula image, and quantifying the result of the analysis. SOLUTION: When activation of a diagnostic program to measure a bone trabecula structure is inputted from a keyboard 11, a CPU 1 gives an instruction to a scanner 11 to take in image information of the bone part from a set X-ray film and stores it in a work area of a VRAM 5. Next, the image information is processed by filtering to eliminate the noise component in the image information. Following to that, as an operator instructs that he/she directly designates an area, the CPU 1 displays a map for area designated. By inputting an area number on the map, the operator instructs the bone trabecula direction to be the object of measurement to the CPU 1. The CPU 1 calculates a power spectrum value of the area shown by the instructed area number and further calculates a specific value of the bone trabecula structure in the instructed direction using the power spectrum value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for diagnosing osteoporosis and a method for diagnosing trabecular bone, which image-analyzes an X-ray image of a bone to quantify the trabecular bone structure, and particularly to obtain a trabecular direction The present invention relates to a technique that makes it possible to automate the quantification of the states of other trabecular structures.

[0002]

2. Description of the Related Art Conventionally, the most common method for discovering osteoporosis is for a doctor to see the result of X-ray photography of the bone part and judge whether the sign of osteoporosis appears in the result of photography. . However, the visual judgment requires many years of experience, and different doctors may have different diagnostic contents. Bone strength (fracture risk) is known to be related to bone mineral content and trabecular running, and the bone mineral content of bone and the condition of trabecular structure of bone (eg trabecular bone A device for diagnosing osteoporosis using a quantified distribution) has been proposed.

Japanese Patent Application No. 2-114 for measuring bone mineral content
A bone image analysis method of No. 611 has been proposed. This method is a method in which the density of the bone mineral reference and the lightness of the X-ray image of the bone part are associated with each other, and the bone mineral content is calculated from the density of the actual X-ray image.

As a method for quantifying the degree of trabecular bone, Japanese Patent Laid-Open No. 6-133223 proposes a bone part image analysis method. In this method, a spectrum analysis is performed on an X-ray image, a vertical and horizontal dense trabecular image (an image that appears as a thin line) is detected from the analysis result, and the degree of the denseness is determined based on the trabecular structure. It is used as an index of the degree. In addition, the bone mineral content of the lumbar spine is used to diagnose osteoporosis.
It is also measured by method A.

[0005]

With such a proposal, the amount of bone mineral and the trabecular structure can be quantified, and osteoporosis can be diagnosed more accurately than by visual inspection.

However, the above proposal still has room for improvement in the following points. First, the trabecular bone structure differs depending on the bone site, and as described in the above proposal, the image analysis target is a two-dimensional plane image. It takes time to automatically recognize. In other words, when you try to examine the trabecular structure of the bone at a specific site,
For example, it is necessary to detect thin lines having inclinations of 1 degree, 2 degrees ... 360 degrees, and compare these lines with each other to detect the line having the clearest contour. The more X-ray results to be investigated, the longer the processing time. In addition, since the methods for measuring the amount of bone mineral and the trabecular bone structure are different, it takes more time to measure both. Also, the accuracy of quantification was not good.

Secondly, since the trabecular bone structure differs depending on the bone site as described above, the doctor or the like in charge of diagnosis gives information on the trabecular direction to the diagnostic device.
It was necessary to specify the trabecular direction. For this reason, it is necessary to input the image data of the X-ray image to the diagnostic device while making the direction thereof accurate, and the device operation is troublesome.

Therefore, in view of the above points, the present invention is
With improved measurement accuracy and processing time, more specifically, it automatically acquires the trabecular direction and quantifies the state of other trabecular structures, and provides osteoporosis for the convenience of osteoporosis diagnosis. An object of the present invention is to provide a diagnosing apparatus for osteosis and an osteoporosis diagnosing method.

[0009]

In order to achieve such an object, the invention of claim 1 quantifies the state of the trabecular structure by image analysis of an image obtained from an X-ray imaging result of a bone part. In the osteoporosis diagnosing device, indicating means for instructing the direction of the trabecular bone to be measured, image processing means for image-analyzing the trabecular bone image in the indicated direction, and quantifying the result of the image analysis. And an information processing means.

The invention of claim 2 is further characterized in that the instructing means directly instruct a direction. According to a third aspect of the present invention, in addition to the first aspect of the invention, the instructing means has table information in which a portion of a bone portion and the direction suitable for the portion are associated with each other, and a form of the portion of the bone portion is formed. Is received from the outside, the part of the bone part is converted to the direction according to the table information, and the image processing means is instructed.

According to a fourth aspect of the present invention, in addition to the first aspect of the invention, the image processing means executes a spectrum analysis as an image analysis, and the instructing means partially instruct a power spectrum distribution region. Characterized by indicating a direction.

The invention of claim 5 is characterized in that, in addition to the invention of claim 1, the feature amount of the trabecular bone structure is calculated using the value of the power spectrum for the designated power spectrum distribution region. To do.

According to a sixth aspect of the invention, in addition to the first aspect of the invention, a correspondence relation is defined between the feature quantity quantified by the information processing means and the bone mineral content, and the correspondence relation is used. The information processing means acquires the amount of bone mineral.

According to a seventh aspect of the present invention, in addition to the first aspect of the invention, a correspondence relation is defined between the characteristic amount quantified by the information processing means and the diagnostic content, and the correspondence relation is used. It is characterized in that the information processing means further comprises display means for acquiring the diagnosis content and displaying the diagnosis content.

According to an eighth aspect of the present invention, in an osteoporosis diagnostic method for quantifying the state of trabecular bone structure by image analysis of an image obtained from an X-ray imaging result of a bone part by an image analyzer, the bone to be measured Instructing the direction of the beam to the image analysis device in the form of a power spectrum distribution region, and calculating the feature amount of the trabecular bone structure using the value of the power spectrum for the specified distribution region in the image processing device. Characterize.

Further, according to the present invention, the inventions of the following aspects are also provided. That is, in order to achieve the above-mentioned object, the osteoporosis diagnostic device according to claim 9, when an X-ray image of a bone portion is given, image-processes the X-ray image to determine the state of the trabecular bone structure. An osteoporosis diagnosing device for quantifying, comprising image processing means for performing image processing for quantifying the state of trabecular bone structure based on the X-ray image, and a quantifying result with reference to the quantifying result. And an output unit that performs an output process of information including at least.

Further, the osteoporosis diagnosing device according to claim 10 is, in addition to the device according to claim 9, provided with a storage means for registering a correspondence relationship between the quantified value and the diagnosis content of osteoporosis. The output means is an osteoporosis diagnostic apparatus characterized by performing a process of outputting diagnostic content corresponding to the obtained quantification result with reference to registered content of the storage means.

Furthermore, the osteoporosis diagnosing device according to claim 11 is the device according to any one of claims 9 and 10, wherein the image processing means quantifies the direction of the trabecular bone in a state of trabecular bone structure. In order to realize the above, a two-dimensional frequency plane for showing the power spectrum distribution for the X-ray radiographed image is represented by polar coordinates in which the radius is the frequency and the rotation angle from the predetermined reference axis is the existence direction of the power spectrum distribution. , The polar coordinate plane is divided into a plurality of regions defined by different rotation angles and the same frequency range, and the process of calculating the sum of the power spectra for each region, and the calculation result is the peak The osteoporosis diagnosing device is characterized by performing a process in which a direction perpendicular to a direction indicated by a rotation angle corresponding to a region indicating is the direction of the trabecular bone.

Furthermore, the osteoporosis diagnosing device according to claim 12 is the device according to any one of claims 9 and 10, wherein the image processing means determines the strength of the trabecular bone in a trabecular structure. In order to quantify, the two-dimensional frequency plane for showing the power spectrum distribution for the X-ray radiographed image, the radius is the frequency, and the rotation angle from the predetermined reference axis is the existing direction of the power spectrum distribution. Expressed in polar coordinates, the polar coordinate plane is divided into a plurality of regions defined by different rotation angles and the same frequency range,
The process of calculating the sum of the power spectra for each region, the peak value of the calculation result and the minimum value existing on both sides of the peak value are obtained, and the average value of the two minimum values is divided by the peak value. The osteoporosis diagnostic device is characterized by performing processing with the strength of the beam.

According to a thirteenth aspect of the present invention, in the apparatus for diagnosing osteoporosis according to any one of the ninth and tenth aspects, the image processing means quantitatively determines the width of the trabecular bone in a state of trabecular bone structure. In order to realize the above, the value of the power spectrum in the direction perpendicular to the direction in which the trabecular bone exists in the X-ray radiographic image is represented by the polar coordinate function F (f, θ) (where f is frequency and θ is power spectrum distribution). The rotation angle corresponding to the existing direction of), and in a predetermined frequency range f1 to f2,
Y = (Σf · F (f, θ)) / ΣF (f, θ) (however,
Σ means to take the sum total from f1 to f2)
And the obtained Y (average frequency) is used as an index of the width of the trabecular bone.

Further, the osteoporosis diagnosing device according to claim 14 is the device according to any one of claims 9 and 10, wherein the image processing means quantifies the width of the trabecular bone in a state of trabecular bone structure. In order to realize the above, the value of the power spectrum in the direction perpendicular to the direction in which the trabecular bone exists in the radiographic image,
Polar coordinate function F (f, θ) (where f is frequency and θ is
Rotation angle corresponding to the existence direction of the power spectrum distribution)
In the specified frequency range f3 to f4 included in the predetermined frequency range f5 to f6.
= Σf · F (f, θ) / ΣF (f, θ) (where the first Σ is the sum of f3 to f4, and the second Σ is f5 to f)
Means to sum up to 6: f5 ≦ f3 <f4
The osteoporosis diagnostic device is characterized in that ≦ f6) is obtained, and the obtained Z (specified range frequency) is used as the width of the trabecular bone.

According to the invention of claim 15, the following osteoporosis diagnostic method is also proposed. That is, it is a method for diagnosing osteoporosis in which, when an X-ray image of a bone is given, the X-ray image is image-processed to quantify the state of the trabecular bone structure. In order to quantify the direction of the beam, a two-dimensional frequency plane for showing the power spectrum distribution with respect to the X-ray radiographed image, a radius as a frequency, and a rotation angle from a predetermined reference axis as a power spectrum distribution Expressed in polar coordinates as the existence direction, the polar coordinate plane is divided into a plurality of regions defined by different rotation angles and the same frequency range, and the sum of power spectra is calculated for each region, and the calculation is performed. The direction perpendicular to the direction indicated by the angle information corresponding to the area where the result shows a peak is the direction of the trabecular bone, and using the obtained power spectrum distribution in the trabecular direction and the vertical direction, Quantify the condition It is an osteoporosis diagnostic method.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. First, the first embodiment will be described. This embodiment corresponds to the invention described in claims 1 to 8, and by operating the pointing means, it is possible to specify the direction of the trabecular bone image that is the target of image analysis, and regarding the direction in which it is meaningless to analyze. The main feature is that it is not necessary to perform the analysis processing of.

FIG. 1 shows an example of system configuration of an osteoporosis diagnostic apparatus to which the present invention is applied. In FIG. 1, a generally well-known personal computer can be used as the image processing apparatus 10. In this embodiment, a circuit according to the present invention will be described. The CPU 1 controls the entire device according to the system program stored in the ROM 2,
The bone image is analyzed according to the image analysis program in the diagnostic program stored in the hard disk drive (HDD) 4, the trabecular bone structure is quantified, and the bone mineral content is acquired. Furthermore, it is possible to diagnose osteoporosis from the acquired data by the above-mentioned diagnostic program. H
The DD 4 further stores a power spectrum region information table 4A and a diagnostic message table 4B according to the present invention.

In the present embodiment, a region to be subjected to spectrum analysis is determined for each bone part, and information indicating the bone part,
Correspondence is made with information indicating a region to be subjected to spectrum analysis. This correspondence is registered in the power spectrum region information table 4A. In addition, the diagnostic message of osteoporosis is stored in the various diagnostic message table 4A in the form of a character code corresponding to the feature amount of the trabecular bone structure obtained as a result of the image analysis.

The RAM 3 temporarily stores information processing related data executed by the CPU 1. The HDD 4 can save and store the image information of the bone part captured by the scanner 14 or the like in addition to the programs and table information. V
The RAM 5 stores the image information displayed on the display 13 and the image information of the image analysis target. The display image information is read by a controller (not shown) and displayed on the display screen of the display 13. Input / output interfaces (I / O) 6, 7 and 8 are keyboard 1 respectively
1 (including mouse 12), display 13 and scanner 14 are connected to transfer information.

The keyboard 11 inputs an operation instruction to the CPU 1 and information (described later) about image analysis according to the present invention. This information input is also realized by operating the mouse 12 to position the display information on the display screen. The display 13 displays guidance information related to information input, input information from the keyboard 11, the amount of bone mineral obtained as a processing result of the CPU 1, the characteristic amount of trabecular bone structure, and the like. The scanner 14 reads the X-ray image film, and outputs multi-valued image information indicating light and shade in the form of an electric signal to the image processing apparatus 10 side. When the X-ray imaging apparatus is connected to the I / O 8, the X-ray imaging apparatus itself outputs the imaging result to the image processing apparatus 10.

The image analysis processing of the bone portion executed by such a system configuration will be described. When an instruction to activate the diagnostic program is input from the keyboard 11, the CPU 1 causes the HDD to
The diagnostic program is loaded onto the RAM 3 from 4, and the execution is started. The main processing procedure of this diagnostic program is shown in FIG. The CPU 1 first displays the menu screen on the display 13 by writing the menu screen information in the VRAM 5 (step S10). In this embodiment, a menu for measuring the characteristic amount of the trabecular bone structure, a menu for additionally measuring the amount of bone mineral, and a menu for diagnosing osteoporosis are prepared. The operator designates a necessary menu via the keyboard 11. The processing of the CPU 1 for each menu will be described below. The measuring process of the characteristic amount of the trabecular bone structure is executed in common for each menu. A) Measurement of trabecular bone structure (step S30) FIG. 3 shows the detailed procedure of this menu. The CPU 1 gives an instruction to the scanner 14 to take in the image information of the bone part from the set X-ray film and store it in the work area of the VRAM 5. Next, the CPU 1 filters this image information to remove noise components in the image information (noise signals mixed in during image reading, images such as dust in the X-ray film).

Subsequently, the CPU 1 displays a menu screen and accepts an instruction from the operator (steps S100 to S100).
110). On this menu screen, a guide message is displayed for selecting whether the operator directly designates a region to be subjected to image analysis, which will be described later, or the operator designates a bone portion of the X-ray film. When the operator directly gives an instruction to specify an area, the CPU 1 displays a map for area specification as shown in FIG. 5, for example. This map shows the distribution region of the power spectrum, and the regions 1 to 5 in the horizontal direction in the figure show the frequency component regions of the trabecular image in the vertical direction in the image of the bone part. In addition, the regions 11 to 11 in the vertical direction in the drawing
Reference numeral 15 indicates the frequency component region of the trabecular image in the lateral direction in the image of the bone part. The operator inputs the area number to instruct the CPU 1 of the direction of the trabecular bone to be measured (step S120 → S125).

The CPU 1 calculates the value of the power spectrum for the region indicated by the designated region number, and using this power spectrum value, in the above-mentioned example of instruction, the feature amount of the trabecular structure in the direction designated by the following equation. Find X.

[0031]

[Equation 1] Here, i and j are numbers of the designated area, and indicate that two types of directions are designated. Further, Ai is a weighting coefficient of the area number i, which is predetermined for each area. When a plurality of directions are designated, the power spectrum value of the area for each direction is accumulated (step S150).

On the other hand, when the operator specifies the bone part, the CPU 1 retrieves the region number (see FIG. 5) corresponding to the instructed bone part from the power spectrum region information table 4A of the HDD 4 and formula The feature amount X of the trabecular bone structure is calculated using 1 (steps S130 → S160). The calculation result is temporarily stored in the RAM 3. B) Measurement of Bone Mineral Content Formula 1 is also used to obtain the bone mineral content, but in this case, the power spectrum region is designated as necessary, and the weighting factors Ai, j. . . The value of is simply switched to that for the amount of bone mineral (step S45 in FIG. 2). Therefore, it will be understood that the bone mineral content can be easily calculated even if the value of the power spectrum is obtained. For reference, a bone portion whose bone density (BMD value) is known in advance is subjected to image analysis, and the correlation between the feature amount of the trabecular bone structure and the BMD value obtained from the result is shown in FIG. . If the correlation of the bone part of each part is acquired, the coefficient A of the above equation 1 can be obtained from this correlation.
i can be determined. C) Diagnosis of osteoporosis Since the feature amount X of the trabecular bone structure has been calculated in the processing procedure (step S30) so far, the CPU 1 refers to the diagnostic message table 4B and corresponds to the feature amount of the trabecular bone structure. After extracting the message and converting it to an image, VRA
Write to M5.

At the same time, the CPU 1 transfers the image captured by the scanner 14 and the guide information for visual diagnosis from this image from the HDD 4 to the display area of the VRAM 5 to display it on the display 13 as shown in FIG. Steps S200 to S220 of FIG. 4).

By the way, in FIG. 7, 101 is an image captured from the scanner 14, 102 is an explanation for diagnostic criteria, 102A is a sample image of trabecular bone corresponding to a feature amount of trabecular bone structure called Itami index, and 102B is The Itami index as a feature amount, and 102C are explanatory texts showing the state of the trabecular structure corresponding to the Itami index. Reference numeral 103 is diagnostic content determined from the feature amount represented by the value of the power spectrum in the present embodiment.

According to the present invention, the following modes can be implemented in addition to this embodiment. 1) In order to specify the direction of the trabecular bone image to be image-analyzed, in addition to using the power spectrum distribution region as shown in FIG. 5, the angle at which the direction of the trabecular bone image intersects the reference direction is numerically indicated. You can also do it. 2) Needless to say, the feature amount and bone mineral content of the trabecular bone structure obtained from the result of the image analysis may be converted into various index values, for example, the Itami index and other indexes and displayed. 3) The display form of the feature amount obtained in the present embodiment may be displayed by using a numerical value or a graph, or may be displayed in the form of a pattern figure.

The above is the description of the first embodiment of the present invention. Next, a second embodiment according to the present invention will be described. The present embodiment corresponds to the inventions described in claims 9 and thereafter, and automatically acquires the trabecular direction and quantifies the states of other trabecular structures without operating the indicating means to diagnose osteoporosis. It is characterized in that it realizes a device for convenience of use.

The hardware configuration of the apparatus according to this embodiment may be the same as the system configuration shown in FIG. However, software such as the diagnostic program in the HDD 4 and the image analysis program in the diagnostic program
The processing according to this embodiment may be modified so that it can be executed.

Further, in this embodiment, as the state of the trabecular structure, the direction of the trabecular bone, the strength of the trabecular bone and the width of the trabecular bone are obtained,
In order to quantify these characteristic quantities for diagnosis, each of the trabecular direction, the strength of the trabecular bone, the width of the trabecular bone, and the diagnostic messages such as the progression level of osteoporosis are stored in the diagnostic message table 4B. And are registered in association with each other.

FIG. 17 shows an example of the diagnostic message. As shown in FIG. 17, the Itami index is predetermined and registered as the diagnosis content for each of the direction of the trabecular bone, the strength of the trabecular bone, and the width of the trabecular bone. Of course, the registration example of the diagnostic message is not limited to that shown in FIG.

Now, a series of processes of this embodiment will be described. FIG. 8 is a flowchart showing a processing procedure executed by the diagnostic program. The diagnostic program includes at least software that executes display processing on the display 13, a program that controls the operation of each component according to a given instruction, and the like.

When the operator gives a command to activate the device through the keyboard 11, the CPU 1 loads the diagnostic program into the area of the RAM 3 and operates according to the loaded diagnostic program.

First, in step S12, the CPU 1
Displays a menu screen on the display 13. Specifically, the CPU 1 stores the information for displaying the menu screen in the area of the VRAM 5, so that the osteoporosis diagnosis menu is displayed on the display screen of the display 13.

In this embodiment, the state of the trabecular structure is shown.
"Menu for quantifying trabecular structure", which is a menu for quantifying and displaying the direction, strength, and width of trabecular bone, and a menu for displaying and outputting the progression level of osteoporosis based on the quantification result There are two menus, “a menu for diagnosing osteoporosis”, and these two types of menus are displayed on the display 13.

Next, in step S22, the operator designates a desired menu via the keyboard 11. At this time, the CPU 1 stores the information of the instructed menu in the work area of the RAM 3, for example.

The calculation process of the feature amount of the trabecular bone structure in the next step S32 is a process in which the CPU 1 quantifies the direction, strength and width of the trabecular bone, which shows the state of the trabecular bone structure,
This is a process that is executed regardless of which menu is selected.
Since this processing is a main part of the present invention, it will be described in detail later.

Next, in step S42, the CPU1
Determines whether a diagnosis of osteoporosis is necessary. Specifically, the CPU 1 performs the following processing. That is, CP
The U1 refers to the menu information stored in the RAM 3, and if the instructed menu is the “menu for quantifying the trabecular bone structure”, the U1 proceeds to step S52 and determines the trabecular bone obtained in step S32. Direction of trabecular bone, which is a feature of the structure,
The quantification result of the intensity and the width is displayed and output on the display 13 and the process ends (end).

On the other hand, if the instructed menu is "a menu for diagnosing osteoporosis", the process proceeds to step S43, and based on the measured data, that is, the quantification result,
The progression level of osteoporosis is diagnosed, the process further proceeds to step S52, the diagnostic result is displayed and output on the display 13, and the process ends (end).

The process in step S43 will be described in detail as shown in FIG. 4 as in the first embodiment. That is, since the feature amount of the trabecular bone structure has already been obtained in step S32, the CPU 1 refers to the diagnostic message table 4B, extracts the corresponding diagnostic message from the HDD 4 (step S200), and extracts it from the image information. To the VRAM (step S210)
It stores in 5 and displays on the display 13 (step S220). At this time, the CPU 1 may be configured to display the image information captured by the scanner 14 and the guide information used for osteoporosis diagnosis on the display 13.

Similar to the first embodiment, FIG. 7 shows a state of the display screen when such a display is performed. Reference numeral 101 denotes image information (bone image) captured by the scanner 14. ), 102 is guide information used for osteoporosis diagnosis, and the guide information 102 is five samples (102A) of trabecular bone images corresponding to the feature amount of trabecular bone information called Itami index. , Itami index (102B) corresponding to each sample, and an explanation message (102C) of the state of the trabecular structure corresponding to each Itami index. And 103 is a diagnostic message (diagnosis name) obtained in step S42.

In this way, by displaying the diagnostic message and the guide information used for the diagnostic on the same screen, the convenience of the diagnostic can be achieved. As described above, by performing the series of processes shown in FIG. 8, the direction, strength, and width of the trabecular bone, which indicates the state of the trabecular bone structure, are quantified and osteoporosis is diagnosed.

Now, the process of step S32 of FIG. 8, which is a main part of the present invention, will be described with reference to the drawings.
FIG. 9 is a flowchart for explaining the process of step S32 of FIG. 8 in more detail.

In step S102, the operator sets the X-ray film on the scanner 14 and the keyboard 11
When a start-up instruction for the scanner 14 is given via the scanner, the scanner 14 acquires image information of the bone part from the set X-ray film, stores it in the work area of the RAM 3, and stores it in the VRAM 5 to store the image of the bone part. The image information is displayed on the display 13.

It is to be noted that the CPU 1 is configured to filter the acquired image information in order to remove noise components such as unnecessary signals mixed in during image reading and dust on the X-ray film. It is preferable.

Further, in step S102, a guide screen for inputting parameters required for the subsequent analysis processing is displayed. The parameters input here will be described later.

Next, step S112 performed by the CPU 1,
The acquisition processing of the direction of the trabecular bone, which is the feature amount of the trabecular bone structure, performed at 122, 132, and 142 will be described. D-1) "Acquisition of trabecular direction" First, in step S112, the operator designates an area to be analyzed in the acquired image information. For example, when image information of 1200 pixels in the vertical direction and 1200 pixels in the horizontal direction is displayed on the display 13, a maximum of 1200 pixels in the vertical direction is displayed.
A quadrilateral region having a predetermined number of pixels is designated so that the number of pixels is 1200 pixels in the horizontal direction. Such designation is performed by inputting the number of vertical and horizontal pixels and the position coordinate of the upper left point of the rectangular area with the keyboard 11, or operating the mouse 12 so that the rectangular area is formed in the display image. Then, a diagnostic program may be created so that the quadrangle area having the upper left point and the lower right point as vertices can be designated as the analysis target area by clicking the positions of the upper left point and the lower right point. .

After the area to be analyzed is designated, the two-dimensional Fourier transform is performed on the area to be analyzed to obtain the power spectrum. In this embodiment, the two-dimensional plane for showing the power spectrum distribution is obtained. Consider dividing into a plurality of regions defined by the rotation angle corresponding to the existing direction of the power spectrum distribution and the frequency of the power spectrum distribution. This will be described with reference to FIG. It should be noted that obtaining the power spectrum by performing the two-dimensional Fourier transform is a publicly-used technique, and since no special device has been devised in the present invention, description thereof will be omitted. Also,
The CPU 1 stores the value of the power spectrum of each pixel in the RAM 3
To be stored.

FIG. 10 shows a polar coordinate plane which is a two-dimensional plane for showing the power spectrum distribution.
It is composed of areas 1 to N. Although the coordinate axis showing the magnitude of the power spectrum is not shown, if it is shown, the coordinate axis showing the magnitude of the power spectrum is set in the direction perpendicular to the paper surface.

FIG. 10 shows a power spectrum distribution 2
The dimensional frequency plane is expressed in a polar coordinate system, so that the point O corresponds to the origin, the radial direction corresponds to the frequency of the power spectrum distribution, and the rotation direction corresponds to the angle information corresponding to the existence direction of the power spectrum distribution. Since the power spectrum distribution of the two-dimensional image has a property of being point-symmetric with respect to the point O, as shown in FIG.
It is sufficient to consider the spectral distribution from 180 to 180 degrees.

In step S122, the number of areas N
The operator inputs two parameters (also referred to as “the number of divisions in the angular direction”) and a frequency designation range (designation range in FIG. 10) that is a predetermined frequency range designated from the lowest frequency to the highest frequency. To do.

In step S132, first, in step S132
A two-dimensional Fourier transform is performed on the two-dimensional image data in the analysis symmetric region designated by 112 to obtain a power spectrum. The CPU 1 stores the power spectrum obtained for the subsequent processing in the work area of the RAM 3.

The distribution of the obtained power spectrum is represented by a coordinate system composed of an X axis showing the frequency in the vertical direction, a Y axis showing the frequency in the horizontal direction, and a Z axis showing the magnitude of the power spectrum. Although it is easy to understand considering that, in the present embodiment, the distribution of the power spectrum is
Trabecular structure, paying attention to the fact that the radial direction corresponds to the frequency of the spectrum distribution, and the rotation direction corresponds to the existence direction of the spectrum distribution, and can be expressed by polar coordinates corresponding to the rotation angle from a predetermined reference axis. To obtain the feature amount of.

If the image analysis program is prepared so that the CPU 1 uses the work area of the RAM 3 to convert the power spectrum distribution from the Cartesian coordinate system to the polar coordinate system using a known mathematical formula. Good.

Then, in step S142, the parameters specified in step S122 are used to define a region into which the two-dimensional plane for indicating the power spectrum distribution is divided.

For example, "N = 10", frequency designation range f
If the parameters 1 to f2 are input,
It will be divided into 10 areas. Further, the sum of the sizes of the power spectra existing in each area is obtained by setting the area specified by the frequency specified ranges f1 to f2 in each area as an “area”. Specifically, CPU
1, the data stored in the RAM 3 can be used to obtain the size of the power spectrum of each pixel by Fourier transform. Therefore, by summing the sizes of the power spectra of the pixels forming each region, The sum of the magnitudes of the power spectra in each area is obtained. The process of obtaining the total sum of the magnitudes of the power spectra is performed for each of the 10 regions according to the above parameters.

The direction perpendicular to the rotation angle corresponding to the region where the sum total shows a peak is the direction of the trabecular bone. This is because the Fourier transform has a mathematical meaning of obtaining the magnitude of the power spectrum of the image existing in the direction perpendicular to the analysis direction (direction in which the Fourier transform is performed). This situation will be described with reference to FIG.
This graph shows horizontal lines (horizontal in the real space) and vertical lines (vertical in the real space) that are perpendicular to each other as an example of the parts where the trabecular bone exists in the vertical direction (real space) and the horizontal direction (real space). Existing in the direction) is the analysis target of the lumbar spine.

FIG. 11 is a graph in which the horizontal axis represents angle information and the vertical axis represents intensity (sum of power spectra in each region). The angle information on the horizontal axis increases from 0 degrees, which corresponds to the angles with respect to region 1, region 2, ..., N in FIG.

Therefore, the peak in the 90 degree direction is due to the transverse muscle (transverse trabecular bone) and is 180 degrees (0 degree).
The directional peaks are due to the longitudinal streaks (vertical trabeculae). The vertical streak peak appears near 0 degree because of the point symmetry of the power spectrum obtained by the two-dimensional Fourier transform as described above.

Further, depending on the part where the bone part exists, the trabecular direction may not be the longitudinal direction or the lateral direction but may be the oblique direction. In this case as well, the peak value of the total sum in the power spectrum region is obtained. The direction of the trabecular bone can be obtained by obtaining the direction perpendicular to the direction of.

With the above processing, the direction of the trabecular bone can be automatically acquired. The CPU 1 stores the information on the trabecular direction in the work area of the RAM 3 for use in diagnosis or the like, or stores it in the VRAM 5 and displays it on the display 13.

Next, the process of acquiring the strength information of the trabecular bone, which is the characteristic amount of the trabecular structure, which is executed by the CPU 1 in step S152 will be described. D-2) “Acquisition of Trabecular Strength Information” The quantification of trabecular strength will be described with reference to FIG. 11.

The strength of the trabecular bone is a value related to the number and width of the trabecular bone. The larger the number and the wider the width, the stronger the strength. As described above, in FIG. 11, the horizontal axis represents angle information and the vertical axis represents intensity (sum of power spectra in each region).
Is a graph in which the lumbar spine, which is composed of transverse and vertical muscles perpendicular to each other, is the analysis target.

The intensity (sum of power spectra) shown on the vertical axis of the graph changes while having a peak value and a minimum value with respect to a change in angle. For generalization, let us say that the peak value is "MAX", one minimum value existing on both sides of the peak point is "MIN1", and the other minimum value is "MIN2".
This is represented by the following equation. That is, MAX = peak value MIN1 = one local minimum value on both sides of the peak point MIN2 = other local minimum value on both sides of the peak point.

Then, the strength of the trabecular bone is quantified by the following general formula. MIN = (MIN1 + MIN2) / 2 Trabecular strength = MAX / MIN In this way, the peak value of the sum in the power spectrum region and the minimum values existing on both sides of the total value are obtained, and the peak value is divided into two minimum values. The value obtained by dividing the average value is used as the strength of the trabecular bone.

In FIG. 11, the longitudinal trabeculae of the lumbar spine,
There are two peaks for the trabecular bone in the lateral direction,
These are referred to as “MaxV” and “MaxH”, respectively. Further, one minimum value existing on both sides of the peak point of the trabecular bone in the vertical and horizontal directions and the other minimum value existing on both sides of the peak point of the trabecular bone in the vertical and horizontal directions are respectively “Min1” and “Min2”. And These can be expressed by the following equations. That is, MaxV = peak value of trabecular bone in the longitudinal direction MaxH = peak value of trabecular bone in the lateral direction Min1 = vertical, one local minimum value existing on both sides of the trabecular bone in the lateral direction Min2 = vertical trabecular value of the trabecular bone in the lateral direction It is the minimum value of the other existing on both sides.

Therefore, the strength of the trabecular bone in the lateral direction shown in FIG. 11 is as follows. Min = (Min1 + Min2) / 2 Transverse trabecular strength = MaxH / Min In general, the larger the power spectrum value, the more ordered the trabecular bones of the corresponding frequency are, so Max is set. It is sufficient to quantify it as the strength of the trabecular bone.

However, although MAX may be used as a parameter representing the strength of the trabecular bone and used as strength information, this value is a value proportional to the overall image brightness at the time of X-ray photography, In the present embodiment, in order to eliminate this effect, the strength of the trabecular bone is quantified based on the value of MIN.

Depending on the part where the bone part exists, the direction of the trabeculae may not be the vertical direction or the horizontal direction but may be the oblique direction. Even in this case, the peak value of the sum in the region of the power spectrum is The strength of the trabecular bone can be measured by obtaining the minimum values existing on both sides of the minimum value and dividing the average value of the two minimum values by the peak value to obtain the strength of the trabecular bone.

FIG. 12 shows an example of measurement results of an actual trabecular bone image of the lumbar spine. As described above, since the trabecular bone exists in the lumbar spine in the longitudinal direction and the transverse direction, when the strength of the trabecular bone is high, as shown in FIG. 12A, 0 degrees (180 degrees) and 90 degrees are provided.
A strong peak occurs near this point.

On the other hand, in the measurement results shown in FIG. 12 (b), no peaks are seen near 0 degree (180 degrees) and 90 degrees, indicating that the strength of the trabecular bone is low. .

By the above processing, the strength information of the trabecular bone can be quantified. The CPU 1 stores the information on the trabecular strength in the work area of the RAM 3 or the VRAM 5 and displays it on the display 13 for use in diagnosis or the like.

Next, the acquisition process of the trabecular bone width information, which is the feature amount of the trabecular bone structure, performed by the CPU 1 in step S162 will be described. D-3) "Acquisition of Trabecular Width Information" First, the power spectrum distribution in the direction perpendicular to the trabecular direction acquired by the above-described processing is acquired.

For this purpose, the CPU 1 executes the step S1.
RAM of the power spectrum of the analysis target area obtained in 32
The image analysis program may be created so as to acquire the power spectrum distribution in the direction perpendicular to the trabecular direction acquired in step S142.

FIG. 13 shows an example of the power spectrum distribution thus obtained. FIG. 13 is a graph in which the horizontal axis represents frequency (hereinafter, frequency means spatial frequency) and the vertical axis represents power spectrum.
m) ”, there is a unimodal distribution with the maximum power spectrum.

Now, Y and Z defined by the following equations are referred to as "average frequency" and "specified range frequency", respectively, and these are values for quantifying the width of the trabecular bone. It should be noted that although the "specified range frequency" is multiplied by 100 to calculate the 100-percentage, this is for the convenience of the later description, and it is not always necessary to multiply by 100 to obtain the "specified range frequency ratio". Absent.

The reason why these values serve as an index showing the width of the trabecular bone will be described later.

[0086]

[Equation 2] However, F (f, θ) is a function expressing the power spectrum value in polar coordinates, and f is a frequency. Further, θ is a rotation angle, and corresponds to a direction in which the spectrum distribution of the trabecular structure information in the acquired trabecular direction appears, for example, a direction perpendicular to the acquired trabecular direction. In addition, f1 and f2 are
It is a frequency range designated in advance to obtain Y.

[0087]

(Equation 3) However, F (f, θ) is a function expressing the power spectrum value in polar coordinates, and f is a frequency. Further, θ is a rotation angle, and is a direction in which a spectrum distribution of the trabecular structure information in the acquired trabecular direction appears, for example, a direction perpendicular to the acquired trabecular direction. Note that f5 to f6 and f3 to f
Reference numeral 4 is a frequency range designated in advance to obtain Z, and the range of f3 to f4 is included in the range of f5 to f6 and has a relationship of “f5 ≦ f3 <f4 ≦ f6”.

The average frequency Y and the designated range frequency Z calculated in this way are indices for quantifying the width of the trabecular bone.
The reason will be described below. FIG. 14 is a graph in which an average value of normalized values of power spectra of a total of 45 lumbar spine image groups from series 1 to series 3 is plotted. First, this will be described.

Sequences 1, 2, and 3 are respectively BMD (Bone M
15 lumbar vertebra images with small ineral density, BM
These are three groups of 15 lumbar spine images with medium D values and 15 lumbar spine images with large BMD values.

The BMD indicates the amount of minerals per unit volume, is a value related to the width of the bone, and is not the data indicating the physical thickness of the trabecular bone visible from the outside.
First, in each series (group), the power spectrum value is normalized at frequencies 1 to 14 (book / analysis target image size) in the direction in which the longitudinal trabecular bone information appears.

In the normalization here, the sum of the power spectrum values at each frequency 1 to 14 in each of the 15 images forming the group is set to 100, and at each frequency 1 to 14 at each image. It is performed by dividing the power spectrum value by the sum of the power spectrum and multiplying by 100.

The average value obtained by averaging the normalized values in each group is plotted for each frequency in the graph shown in FIG. In FIG. 14, series 1
Are plotted with circle symbols, Series 2 with square symbols, and Series 3 with triangular symbols.

As can be seen from FIG. 14, frequencies 7 to 1
In No. 4, the larger the BMD value, the larger the power spectrum value. This is because there are more trabecular bones with a frequency of about 7 to 14, which are generally thicker, as the BMD value increases.

Therefore, it can be said that the magnitude of the value of BMD has a correlation with the number of thick trabeculae (the magnitude of the width of each trabecular bone). Therefore, a large value corresponding to the magnitude of the value of BMD is obtained. The above Y and Z were adopted in order to quantify the width of the trabecular bone by introducing some kind of index.

FIG. 15 is an explanatory diagram showing an example of the relationship between the BMD and the average frequency Y. Horizontal axis is BMD, vertical axis is f1 =
The value of the average frequency X calculated as 1, f2 = 14 is shown.

As a result of numerical analysis of this graph, the correlation coefficient becomes about 0.60, and when the value of BMD becomes large, the average frequency Y
Was confirmed to be large. In addition, in FIG.
An explanatory view showing an example of relation between D and designated range frequency rate Z is shown.

The horizontal axis is BMD, and the vertical axis is f3 = 7, f4 = 1.
The value of the specified range frequency ratio Z calculated as 4, f5 = 1 and f6 = 14 is shown. As a result of numerical analysis of this graph, it was confirmed that the correlation coefficient was about 0.62, and that the specified range frequency ratio Z increased as the value of BMD increased. In addition,
Needless to say, the values of f2 and f6 are not limited to the above-mentioned values, and may be increased to perform calculation in a high frequency region to improve measurement accuracy.

As described above, the average frequency Y and the designated range frequency ratio Z have large values corresponding to the magnitude of the BMD value, and serve as an index for quantifying the width of the trabecular bone. That is,
The width of the trabecular bone can be quantified by obtaining the average frequency and the specified range frequency.

The CPU 1 stores the trabecular width information in the work area of the RAM 3 for use in diagnosis or the like, or stores it in the VRAM 5 and displays it on the display 13. In this embodiment, the storage means is an HDD.
4. The output means includes at least step S52, and the image processing means includes steps S132 and 142,
The method according to claim 15 corresponds to steps 152, 162, and the steps S132, 142, 152, 1 shown in FIG.
Corresponding to the step including 62.

As described above, according to the present invention, the apparatus and method for automatically quantifying the direction, strength and width of the trabecular bone, which shows the state of the trabecular bone structure, and for diagnosing osteoporosis. Can be realized.

[0101]

As described above, according to the first aspect of the invention, it is meaningless to analyze by specifying the direction of the trabecular bone image to be image-analyzed through the indicating means. Since it is not necessary to perform the analysis processing for, the image analysis time can be shortened.

According to the second aspect of the present invention, since the direction can be directly designated by operating the pointing means, it is possible to analyze the bone part images of different parts and improve the image analysis accuracy.

According to the third aspect of the present invention, a person who is unfamiliar with the operation can perform the diagnosis by designating the part of the bone part. According to the invention described in claim 4, by using the power spectrum distribution area for designating the direction, not only the direction but also the characteristics of the frequency component (shading of the image or the size of the line) can be targeted for image analysis. .

According to the invention of claim 5, the density of the trabecular image in a specific direction is indicated by the value of the power spectrum, and the degree of osteoporosis can be judged by comparing the values. Become. According to the invention described in claim 6, since the amount of bone mineral is easily derived from the feature amount of the trabecular structure, the acquisition time of these two feature amounts is shortened.

According to the invention of claim 7, the feature amount of the trabecular bone structure represents the degree of osteoporosis. Therefore, by associating it with the diagnostic contents, the diagnostic contents are derived from the features of the trabecular bone structure. This makes it possible to diagnose osteoporosis using an information processing device.

According to the invention described in claim 8, it is possible to perform image analysis rich in variations while shortening the image analysis time. Further, according to the inventions of claims 9 and thereafter, there are the following effects.

According to the ninth aspect of the present invention, since the state of the trabecular bone structure is automatically quantified by the image processing means, it is possible to realize an osteoporosis diagnosing device provided for the convenience of diagnosis. Further, according to the invention of claim 10, the quantified value of the trabecular bone structure represents the degree of osteoporosis, so that the osteoporosis can be diagnosed by associating it with the diagnosis content.

According to the eleventh aspect of the present invention, the image processing means sets the two-dimensional frequency plane for showing the power spectrum distribution for the X-ray image to
Also, the rotation angle from a predetermined reference axis is expressed in polar coordinates with the existence direction of the power spectrum distribution, and the polar coordinate plane is divided into a plurality of regions defined by different rotation angles and the same frequency range. , A process of calculating the sum of power spectra for each region, and a process of setting the direction perpendicular to the direction indicated by the angle information corresponding to the region where the calculation result shows a peak to the direction of the trabecular bone, It is possible to realize an osteoporosis diagnostic device that can automatically acquire the direction of the beam.

Further, according to the invention of claim 12,
The image processing means uses a two-dimensional frequency plane for indicating the power spectrum distribution for the X-ray radiographed image in polar coordinates with the radius vector as the frequency and the rotation angle from the predetermined reference axis as the existence direction of the power spectrum distribution. Expressing, dividing the polar coordinate plane into a plurality of regions defined by different rotation angles and the same frequency range, calculating the sum of power spectra for each region, and the peak value of the calculation result And the minimum values existing on both sides of the trabecular bone are calculated, and the average value of the two minimum values is divided by the peak value to obtain the strength of the trabecular bone, thereby automatically acquiring the strength information of the trabecular bone. It is possible to realize a possible osteoporosis diagnostic device.

Further, according to the invention described in any one of claims 13 and 14, the image processing means sets the value of the power spectrum in the direction perpendicular to the direction in which the trabecular bone exists in the X-ray radiographed image to polar coordinates. By expressing the function F (f, θ) and using it to calculate the average frequency Y and the specified range frequency Z, it is possible to realize an osteoporosis diagnostic device that can automatically acquire the width information of the trabecular bone.

Further, according to the invention of claim 15, X
A two-dimensional frequency plane for showing a power spectrum distribution for a line image is represented by polar coordinates with a radius being a frequency and a rotation angle from a predetermined reference axis being a power spectrum distribution existence direction, and the polar coordinate plane Is divided into a plurality of regions defined by different rotation angles and the same specified frequency range, the sum of power spectra is calculated for each region, and the calculation result corresponds to the region showing a peak. The direction perpendicular to the direction indicated by the angle information is taken as the trabecular direction, and the direction of the trabecular bone can be automatically acquired. Furthermore, by using the power spectrum distribution of the trabecular direction and the direction perpendicular thereto, other trabecular bone An osteoporosis diagnostic method that quantifies the structural status is realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing a system configuration of an embodiment according to the present invention.

FIG. 2 is a flowchart illustrating a processing procedure executed by a CPU 1;

FIG. 3 is a flowchart showing a processing procedure executed by a CPU 1.

FIG. 4 is a flowchart showing a processing procedure executed by the CPU 1.

FIG. 5 is an explanatory diagram showing an example of a display screen used for designating a direction.

FIG. 6 is an explanatory diagram showing a correlation between image feature amounts and BMD values.

FIG. 7 is an explanatory diagram illustrating a display example on a display.

FIG. 8 is a flowchart showing a processing procedure executed by the CPU 1 in the second embodiment.

FIG. 9 is a flowchart showing a processing procedure for quantifying a trabecular bone structure.

FIG. 10 is an explanatory diagram for explaining a calculation process using a power spectrum.

FIG. 11 is an explanatory diagram showing an example of angle dependence of a power spectrum distribution.

FIG. 12 is an explanatory diagram showing an example of angle dependence of a power spectrum distribution obtained from an X-ray image.

FIG. 13 is an explanatory diagram showing an example of a power spectrum distribution in a direction perpendicular to a trabecular direction.

FIG. 14 is an explanatory diagram showing an example of angle dependence of a power spectrum distribution in a direction in which trabecular bone information appears in a trabecular bone image obtained from an actual X-ray image.

FIG. 15 is an explanatory diagram showing the relationship between BMD and average frequency.

FIG. 16 is an explanatory diagram showing a relationship between a BMD and a designated range frequency ratio.

FIG. 17 is an explanatory diagram of an example of a diagnostic message table.

[Explanation of symbols]

 1 CPU 2 ROM 3 RAM 4 HDD 5 VRAM 6, 7, 8 I / O 10 Image Processing Device 11 Keyboard 13 Display 14 Scanner

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Manabu Tanigawa 1 2-2 Samejima, Fuji City, Shizuoka Prefecture Asahi Kasei Corporation (72) Inventor Yoshio Fukunaga 193-1 Futako, Kurashiki City, Okayama Prefecture

Claims (15)

[Claims] 1. An osteoporosis diagnostic device for quantifying the condition of trabecular bone structure by image analysis of an image obtained from an X-ray image of a bone part, and an indicating means for indicating the direction of the trabecular bone to be measured. An osteoporosis diagnostic apparatus comprising: an image processing means for image-analyzing a trabecular bone image in the designated direction; and an information processing means for quantifying a result of the image analysis. 2. The osteoporosis diagnosing device according to claim 1, wherein the indicating means directly indicates a direction. 3. The instructing means has table information in which a site of a bone part and the direction suitable for the site are associated with each other,
An osteoporosis diagnostic apparatus, which receives an instruction from the outside in the form of a part of a bone part, converts the part of the bone part into the direction according to the table information, and gives an instruction to the image processing means. 4. The image processing unit executes spectral analysis as image analysis, and the instructing unit indicates the direction by partially instructing a power spectrum distribution region. Osteoporosis diagnostic device. 5. The osteoporosis diagnosing device according to claim 4, wherein the feature amount of the trabecular bone structure is calculated using the value of the power spectrum for the designated power spectrum distribution region. 6. A correspondence relation is defined between a feature quantity quantified by the information processing means and a bone mineral content, and the information processing means acquires the bone mineral content using the correspondence relationship. The osteoporosis diagnostic device according to claim 1, wherein the device is an osteoporosis diagnostic device. 7. A correspondence relation is defined between a feature quantity quantified by the information processing unit and a diagnosis content, and the information processing unit acquires the diagnosis content using the correspondence relation, and the diagnosis content is acquired. The osteoporosis diagnostic device according to claim 1, further comprising display means for displaying. 8. An osteoporosis diagnostic method for quantifying the condition of trabecular bone structure by image analysis of an image obtained from an X-ray image of a bone part by an image analyzer, in which the direction of the trabecular bone to be measured is set to power. An osteoporosis characterized by instructing the image analysis device in the form of a spectral distribution region, and calculating the characteristic amount of the trabecular bone structure using the power spectrum value for the instructed distribution region in the image processing device Method for diagnosing osteoporosis in a diagnosing device for osteoporosis. 9. An osteoporosis diagnostic device for quantifying the state of trabecular bone structure by image-processing an X-ray image of a bone portion, the X-ray image being obtained. Based on, the osteoporosis comprising an image processing means for performing image processing for quantifying the state of trabecular structure and an output means for referring to the quantification result and outputting information including at least the quantification result Diagnostic device. 10. The storage device according to claim 9, further comprising storage means for registering a correspondence relationship between the quantified value and the diagnosis content of osteoporosis, and the output means refers to the registration content of the storage means. hand,
An osteoporosis diagnostic device characterized by performing a process of outputting diagnostic content corresponding to the obtained quantification result. 11. The image processing unit according to claim 9, wherein the image processing unit indicates a power spectrum distribution for the X-ray image in order to quantify the direction of the trabecular bone, which is a state of trabecular bone structure. The two-dimensional frequency plane is expressed by polar coordinates in which the radius is the frequency and the rotation angle from the predetermined reference axis is the existence direction of the power spectrum distribution, and the polar coordinate plane is the same as the different rotation angles. Dividing into a plurality of regions defined by the frequency range, calculating the sum of the power spectrum for each region, and the direction perpendicular to the direction indicated by the rotation angle corresponding to the region where the calculation result shows a peak. An osteoporosis diagnostic device characterized by performing processing in the direction of the trabecular bone. 12. The image processing means according to claim 9, wherein the image processing means indicates a power spectrum distribution for the X-ray image in order to quantify the strength of trabecular bone in a state of trabecular bone structure. The two-dimensional frequency plane is expressed by polar coordinates in which the radius is the frequency and the rotation angle from the predetermined reference axis is the existence direction of the power spectrum distribution, and the polar coordinate plane is the same as the different rotation angles. A process of dividing into a plurality of regions defined by a frequency range, calculating the sum of power spectra for each region, and obtaining a peak value of the calculation result and a minimum value existing on both sides of the peak value, and obtaining two minimum values. An osteoporosis diagnosing device, characterized in that a value obtained by dividing the average value by the peak value is used as the strength of the trabecular bone. 13. The presence of trabecular bone in the radiographic image according to claim 9, wherein the image processing unit quantifies the width of the trabecular bone in a state of trabecular bone structure. The value of the power spectrum in the direction perpendicular to the direction is expressed by the polar coordinate function F (f, θ) (where f is the frequency and θ is the rotation angle corresponding to the existing direction of the power spectrum distribution),
In a predetermined frequency range f1 to f2, Y = (Σf · F (f, θ)) / ΣF (f, θ) (however,
Σ means to take the sum total from f1 to f2)
The osteoporosis diagnosing device characterized by performing the process of using the calculated Y (average frequency) as the width of the trabecular bone. 14. The presence of trabecular bone in the radiographic image according to claim 9, wherein the image processing unit quantifies the width of the trabecular bone in a state of trabecular bone structure. The value of the power spectrum in the direction perpendicular to the direction is expressed by the polar coordinate function F (f, θ) (where f is the frequency and θ is the rotation angle corresponding to the existing direction of the power spectrum distribution),
In the specified frequency range f3 to f4 included in the predetermined frequency range f5 to f6, Z = Σf · F (f, θ) / ΣF (f, θ) (However, the first Σ is from f3 The sum up to f4, the next Σ means to take the sum from f5 to f6: f5 ≦ f3 <f
An osteoporosis diagnostic device, characterized in that 4 ≦ f6) is obtained, and the obtained Z (specified range frequency) is used as the width of the trabecular bone. 15. Given a radiographic image of the bone,
A method for diagnosing an osteoporosis in which the condition of trabecular bone structure is quantified by image-processing the radiographic image, wherein the X-ray radiograph is used to quantify the direction of trabecular bone in the condition of trabecular bone structure The two-dimensional frequency plane for showing the power spectrum distribution for the image is represented by polar coordinates in which the radius is the frequency and the rotation angle from the predetermined reference axis is the existing direction of the power spectrum distribution, and the polar coordinate plane is Divide into multiple regions defined by different rotation angles and the same frequency range, calculate the sum of the power spectra for each region, and the direction indicated by the angle information corresponding to the region where the calculation result shows a peak. A method for diagnosing osteoporosis, in which the direction of the trabecular bone is defined as a direction perpendicular to the trabecular bone and the state of other trabecular bone structures is quantified using the obtained power spectrum distribution in the trabecular direction and the vertical direction.