WO2012128036A1 - Bone density measurement device - Google Patents

Abstract

In the present invention, when a user designates specific pixels in a test subject image using a cursor, correction assistance information for the designated pixels (target pixels) is displayed. The correction assistance information includes information such as the following: information indicating automatic identification results for the target pixels; local bone density at the target pixels and average bone density for a region that includes the target pixels; local attenuation at the target pixels and average attenuation within that region at time of low-energy radiation emission; and local attenuation at the target pixels and average attenuation in that region at time of high-energy X-ray emission. In a case where soft-tissue pixels are designated, in the test subject image, a local soft-tissue evaluation value and an average soft-tissue evaluation value are displayed instead of the local bone density and the average bone density.

Description

Bone density measuring device

The present invention relates to a bone density measuring apparatus (bone density measurement apparatus), and more particularly to a bone density measuring apparatus that displays a bone density image by irradiating a subject with X-rays.

The bone density measuring device is a device for diagnosing bone diseases such as osteoporosis in the medical field. Such a bone density measuring apparatus irradiates a subject with X-rays, detects X-rays transmitted through the subject, and forms a bone density image of the subject based on detection data obtained thereby. It is. Specifically, high energy X-rays and low energy X-rays are alternately irradiated according to a DEXA (dual-energy x-ray absorptiometry) method. Japanese Unexamined Patent Application Publication No. 2009-1000094 (Reference 1) and Japanese Unexamined Patent Application Publication No. 6-261894 (Reference 2) describe a calculation method according to the DEXA method. Measurement sites are the lumbar spine, forearm and the like. The bone density measuring device is also called a bone mineral content measuring device (bone (mineral content measurement apparatus).

In bone density measurement of the lumbar spine (lumbar spine), based on the pixel value (bone density value) of each pixel constituting the bone density image, for each pixel, the bone pixel (pixel corresponding to the bone) It is automatically identified whether it is a soft tissue pixel (a pixel corresponding to only soft tissue). At that time, a region of interest (ROI) is individually set for each vertebra constituting the lumbar vertebra. Within each region of interest, an average value of bone density (average bone density) of the bone pixel is calculated.

Conventionally, there is a case where an automatic identification result between a bone pixel and a soft tissue pixel is not always correct. For example, when a compression fracture site exists, the average bone density may not be calculated correctly. It is desirable to exclude the pixels constituting the compression fracture site from the average calculation target. In the case of a low bone mass, although it is originally a bone pixel, it may be recognized as a soft tissue pixel. In that case, it is desirable to change the type of the pixel subject to misidentification to a bone pixel. When blood vessels and lymph nodes are calcified, they may be recognized as bone. In that case, it is necessary to change the type of the pixel to be misidentified to a soft tissue pixel and exclude the pixel from the bone region. If other vertebrae enter the region of interest due to bone deformation or abnormal bone growth, or if there is an unnecessary part, exclude that part from the average bone density calculation target. There is a need to. Furthermore, when a metal is embedded, it is necessary to exclude the metal portion from the target of the average bone density calculation. In view of the above, conventionally, before calculating the average bone density, addition and deletion of bone pixels are manually performed on the bone density image by visual observation of the user.

However, considerable skill is required to add and delete bone pixels only by visual judgment of the bone density image. Or, according to such work, judgment tends to vary. Therefore, it is desired to support the work of adding and deleting bone pixels.

An object of the present invention is to assist a user who observes an image showing a bone density (bone mineral content) distribution when the user performs sorting of bone pixels. Alternatively, an object of the present invention is to enable quick and accurate determination of bone pixel sorting.

Desirably, the bone density measuring apparatus generates a subject image in which a two-dimensional distribution of bone mineral is reflected based on detection data obtained by irradiating the subject with X-rays. An identification processing unit that applies an identification process for identifying a bone pixel and a soft tissue pixel based on a pixel value of the pixel, and a pixel processing unit, A target pixel designating unit for designating a target pixel by a user, a correction support unit for providing correction support information to the user when the target pixel is designated, and a correction for the target pixel by the user A correction execution unit that corrects the result of the identification process when an instruction is given, and the correction support information includes a set representing the result of the identification process for the pixel of interest And type information, the local evaluation value calculated for the target pixel according to the result of the identification process, and the identification mean evaluation value in which the pixel of interest is computed for belonging region according to the results of processing includes.

According to the above configuration, when a pixel of interest (that is, coordinates on the image) is designated on the subject image by the user, correction support information for the pixel of interest is displayed. Therefore, based on the content of the correction support information, it is possible to accurately determine whether or not the target pixel needs to be corrected. For example, the contents of the correction work include a type change from a bone pixel to a soft tissue pixel, a type change from a soft tissue pixel to a bone pixel, and exclusion from a calculation target. The correction support information preferably includes a type identification processing result for the target pixel. In that case, it is possible to match the result of visual judgment with the result of automatic judgment. At that time, the local evaluation value calculated for the pixel of interest and the average evaluation value calculated for the entire region to which it belongs (for example, the bone region or soft tissue region in the region of interest) are displayed together. It is possible to objectively determine whether a local value for a pixel is out of or substantially the same as the surrounding average value. In some cases, the difference between the local evaluation value and the average evaluation value may be displayed instead of or together with the local evaluation value and the average evaluation value. Usually, when the local value of the pixel of interest protrudes from the surroundings and is large or small, tissue abnormalities, measurement errors, calculation errors, etc. can be considered. According to the correction support information, it is possible to determine whether correction is necessary based on these possibilities. For example, the evaluation value for the bone part is the bone density. For example, the evaluation value for soft tissue is _RL / _RH described later. It corresponds to the ratio of two attenuation rates (attenuation amounts) for two types of X-ray energy. The evaluation value may simply be a pixel value. The evaluation value may be an intermediate coefficient value generated in the calculation process. In any case, it is desirable to provide an evaluation value that can evaluate or judge the relationship between the pixel of interest and the surroundings (or whether correction is necessary). Note that the target pixel is an abstract concept, and may be a single pixel when viewed physically or a set of a plurality of pixels when viewed physically. In the latter case, the local evaluation value is a local average value, a local intermediate value, a local median value, or the like. The evaluation unit and the correction unit may be different. You may comprise so that the content of correction assistance information can be customized by the user. For example, if the organization type is known, the organization type information can be excluded from the correction support information.

Desirably, when a bone pixel is identified by the identification process, a local bone density is displayed as the local evaluation value, and an average bone density is displayed as the average evaluation value. For example, the local bone density corresponds to a pixel value in the subject image, and is an element concept that is considered in comparison with the average bone density of the entire region. Although it does not represent the actual accurate bone density, it is useful information for comparison.

Desirably, when the pixel is identified as a soft tissue pixel by the identification process, a local soft tissue evaluation value is displayed as the local evaluation value, and an average soft tissue evaluation value is displayed as the average evaluation value. Since the concept of bone density does not exist for soft tissue, an evaluation value indicating the property of soft tissue is displayed instead. Although each element constituting the correction support information is basically a numerical value, a graph or the like that assists intuitive recognition may be displayed instead of or together with it.

Preferably, the correction support information further calculates a local attenuation amount during low-energy X-ray irradiation and a local attenuation amount during high-energy X-ray irradiation calculated for the target pixel, and a region to which the target pixel belongs. The average attenuation amount at the time of low-energy X-ray irradiation and the average attenuation amount at the time of high-energy X-ray irradiation are included. According to this configuration, it is possible to comprehensively determine whether or not correction is necessary by displaying intermediate numerical values used in the evaluation value calculation process. When an unnatural value is generated in only one of the energies, it is possible to recognize the possibility that a problem occurs in the measurement with the energy.

Preferably, the local soft tissue evaluation value corresponds to a ratio of local attenuation at the time of low energy X-ray irradiation and local attenuation at the time of high energy X-ray irradiation, and the average soft tissue evaluation value is at the time of low energy X-ray irradiation. It corresponds to the ratio of the average attenuation amount and the average attenuation amount at the time of high energy X-ray irradiation.

Desirably, the correction execution unit executes at least one of a type change from a bone pixel to a soft tissue pixel, a type change from a soft tissue pixel to a bone pixel, and an exclusion from a calculation target.

Desirably, the subject includes a plurality of vertebrae, a plurality of regions of interest are set for the plurality of vertebrae, and each region of interest is identified as a bone region and a soft tissue region. Of course, you may apply the said image processing to site | parts other than a lumbar vertebra. The boundary detection may be performed so as to bisect the region of interest, and the bone region and the soft tissue region may be specified, respectively, or only the pixel type may be specified in units of pixels.

Preferably, the correction support means generates a histogram indicating the number of pixels for each pixel value based on the subject image, and a marker for generating a marker indicating the pixel value of the pixel of interest on the histogram And a generation unit. According to this configuration, it is possible to determine whether or not the pixel of interest is to be corrected in consideration of where the pixel of interest is located on the histogram, so that more accurate determination can be made.

Conventionally, the user has made corrections such as pixel exclusion only from visual observation of a monochrome grayscale image. That is, since the correction work based on sensory judgment is performed, there is a problem that the judgment result varies greatly by a user (doctor or the like) or the work load is heavy. On the other hand, according to the above configuration, information that supports the determination of whether correction is necessary, in particular, information that makes it easy to compare the attention point and the entire background is displayed. It can be made objective, and the burden of correction can be greatly reduced. The correction support information may be displayed in a pop-up so that the correspondence relationship with the pixel of interest can be understood in the vicinity of the pixel of interest. When a pixel is specified by the pointing device, the correction support information may be displayed using it as a trigger, and correction execution or correction postponement may be identified according to the content of the next instruction. When the pixel type is corrected, the correction content may be reflected in the data that is the basis of the average bone density calculation at that time, or when all corrections are completed or the user's explicit The contents of each correction may be reflected in the basic data at the stage of the instruction.

According to the above configuration, exclusion of compression fractures from bone regions, correction of soft tissue misidentification due to low bone mass, correction of bone region misperception due to tissue calcification, exclusion of bone deformed sites such as osteophyte, bone spur In addition, it is possible to exclude foreign parts such as metal parts and unnecessary parts that have entered the region of interest due to scoliosis.

It is a block diagram which shows the whole structure of the bone density measuring apparatus which concerns on this invention. It is a figure for demonstrating the function of the data calculating part shown in FIG. It is a figure which shows the 1st example of a display. It is a figure which shows the 2nd example of a display. It is a figure which shows the 3rd example of a display. It is a figure for demonstrating the pixel correction method according to a condition.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described with reference to the drawings.
First, the principle of bone density measurement (DEXA method) will be described. For two types of X-rays that pass through the human body, the total attenuation is defined as follows.

I _L = I _0L · EXP (-μ _BL X _B ) ・ EXP (-μ _SL X _S ) (1)
I _H = I _0H · EXP (-μ _BH X _B ) ・ EXP (-μ _SH X _S ) (2)

Here, for the subscripts, “L” indicates low energy, “H” indicates high energy, B indicates bone, and S indicates soft tissue. I _L and I _H both indicate transmitted X-ray intensity (low energy X-ray transmission intensity, high energy X-ray transmission intensity), and I _0L and I _0H both indicate incident X-ray intensity (low energy X-ray intensity). (Incident intensity, incident intensity of high energy X-rays). μ _SL , μ _SH , μ _BL, and μ _BH each indicate a linear absorption coefficient (cm ⁻¹ ). X _B and X _S each indicate a thickness (cm).

When taking the natural logarithm of both sides for the above formulas (1) and (2), the following two formulas are derived.

ln (I _0L / I _L ) = μ _BL X _B + μ _SL X _S (3)
ln (I _0H / I _H ) = μ _BH X _B + μ _SH X _S (4)

If X _B is solved using the above equation (2), the following is obtained.
X _B = C · (R _L −α · R _H ) (5)
Here, each coefficient is defined as follows.

R _L = ln (I _0L / I _L ) (6)

R _H = ln (I _0H / I _H ) (7)

α = μ _SL / μ _SH (8)

C = 1 / (μ _BL −α · μ _BH ) (9)

In the above equation (5), the left side is 0 in the area of only soft tissue. Therefore, the following is derived.

α = R _L / R _H (10)

Among the above R _L / R _H , the molecule R _L is ln (I _0L / I _L ), which corresponds to the attenuation factor (attenuation amount) of low energy X-rays. Of the above R _L / R _H , the denominator R _H is ln (I _0H / I _H ), which corresponds to the attenuation factor (attenuation amount) of high energy X-rays. Therefore, R _L / R _H corresponds to a ratio of two attenuation rates (attenuation amounts) for two types of energy with respect to soft tissue. Unlike the linear absorption coefficient ratio (μ _SL / μ _SH ) defined by Equation (7), it is obtained as an actual measurement value. It can be referred to as a soft tissue evaluation value, or can simply be represented by “R”.

On the other hand, when the bone thickness X _B defined by the above equation (5) is multiplied by the bone physical density ρ _B and integrated in the bone region, the bone mineral content BMC (bone mineral content) is as follows. ).

BMC ＝ ∫∫ρ _B・ X _B dxdy… (11)

Further, by dividing BMC by the area S of the bone region as described below, the bone density (planar bone density) BMD (bone mineral density) is finally calculated.

BMD = BMC / S ... (12)

Actually, in order to calculate an accurate bone density due to a beam hardening phenomenon and other influences, corrections are applied to individual coefficients and final calculation results. Bone density image corresponds to a representation of the distribution of the pixel values determined from the above X _B. On the bone density image, the pixel value is usually the lowest value in the soft tissue part. Even so, the soft tissue evaluation value (R) can be calculated from the above equation (8). On the bone density image, for example, a plurality of regions of interest are automatically or manually set for a plurality of vertebrae (blocks), and a bone region is automatically recognized in each region of interest. Is calculated. Then, the bone density for each vertebra is calculated according to the above equations (9) and (10). It is basically “average bone density (= average bone evaluation value)” over the entire bone region. In contrast, each pixel value in the bone region on the bone density image can be referred to as “local bone density (= local bone evaluation value)”. However, it is a value that does not reflect the difference in structure in the thickness direction, and should be understood as a guide. Even so, it can be said that it is a local value that can be compared with the background average value. For soft tissue, an “average soft tissue evaluation value” over the entire region and a “local soft tissue evaluation value” in pixel units can be defined.

FIG. 1 shows a preferred embodiment of a bone density measuring apparatus according to the present invention, and FIG. 1 is a block diagram showing the overall configuration thereof. The bone density measuring apparatus shown in FIG. 1 is an apparatus that is installed in a medical institution and performs bone density measurement on human bones, particularly lumbar vertebrae.

The bone density measuring apparatus is roughly divided into a measuring unit 10 and a calculation unit 12. First, the measurement unit 10 will be described. A subject 16 as a human body is placed on the bed 14. In the present embodiment, X-ray irradiation is performed on a region including the lumbar spine. The bed 14 may be a radiographic table or the like. An X-ray generator 18 is provided below the bed 14. The X-ray generator 18 is an apparatus that alternately generates low energy X-rays and high energy X-rays. In their generation, voltage switching, filter switching, and the like are applied.

Reference numeral 19 denotes an X-ray beam. In the present embodiment, a fan-beam having a divergent shape is formed. Reference numeral 20 denotes an X-ray detector, which is constituted by a plurality of X-ray sensors arranged in a straight line corresponding to the fan beam. The X-ray generator 18 and the X-ray detector 20 constitute a movable body, and the movable body is connected to the scanning mechanism 22. The movable body is scanned in the body axis direction (the direction in which the spine extends) by the scanning mechanism 22. Detection data acquired from a two-dimensional region can be obtained by performing the above-described mechanical scanning while alternately repeating low-energy X-ray irradiation and high-energy X-ray irradiation. Specifically, two-dimensional detection data corresponding to low energy X-rays and two-dimensional detection data corresponding to high energy X-rays are obtained. They are output to the data calculation unit 24.

Next, the arithmetic unit 12 will be described. As described above, the detection data is input to the data calculation unit 24. The function will be described in detail later with reference to FIG. The data calculation unit 24 is a module for calculating the planar bone density, that is, the average bone density in the bone region according to the above-described calculation formula. In the present embodiment, the bone density (average bone density) is calculated for each of the plurality of vertebrae, and various other information is calculated. The data calculation unit 24 generates a subject image representing a two-dimensional distribution of bone mineral in the subject, that is, a bone density image. The image is a black and white grayscale image, and each pixel value represents bone density. However, it is the local bone density referred to to determine the above average bone density. Prior to such calculation, the data calculation unit 24 individually sets a region of interest for each vertebra as described later, and identifies a bone region and a soft tissue region within each region of interest. Yes. In the present embodiment, the identification process is automatically executed based on the pixel value of each pixel.

The display processing unit 26 is a module that configures an image to be displayed on the display unit 30. The display example will be described later with reference to FIGS. On the display unit 30, the subject image is displayed as a black and white image, and correction support information described in detail below is displayed as necessary. Information such as a histogram is also displayed.

The control unit 28 controls the operation of each component shown in FIG. An input unit 32 is connected to the control unit 28. The user can give an operation command to the control unit 28 using the input unit 32. In addition, it is possible to designate a pixel on the subject image using the input unit 32 and give a modification instruction such as changing the type of the pixel or excluding it from the calculation target. The correction instruction given from the input unit 32 is sent to the data calculation unit 24 via the control unit 28, and the data calculation unit 24 executes correction of the pixel type or the like at a predetermined timing in accordance with the correction instruction. That is, for example, when a metal or the like is imaged and it has an influence on the average bone density, pixels corresponding to the metal are excluded from the data serving as the basis of the average bone density calculation. Further, when the calcified soft tissue is misidentified as a bone part, the pixel type is changed from the bone pixel to the soft tissue pixel for the pixel in the misidentified region. The necessity of the correction depends on the visual judgment of the user. In the present embodiment, since correction support information is displayed on the screen in addition to the subject image, the judgment by the user is accurate and quick.

FIG. 2 conceptually shows the function of the data calculation unit 24 shown in FIG. The data calculation unit 24 has a function of executing the above-described equations (1) to (12). The entity of the data calculation unit 24 is software. As shown in FIG. 2 as a plurality of blocks, the data calculation unit includes a region-of-interest setting unit 34, a pixel type determination unit 36, and a correction unit 50. In FIG. 2, each block represents a software function. The region-of-interest setting unit 34 is a module that executes automatic processing for individually setting a plurality of regions of interest for a plurality of vertebrae on the subject image. Of course, the region of interest may be set by the user. Within each region of interest, it is automatically determined pixel by pixel whether each pixel belonging to it is a bone pixel or a soft tissue pixel. The pixel type determination unit 36 performs this. That is, the pixel type determination unit 36 determines which type the pixel belongs to based on the pixel value of each pixel in the subject image. Of course, the type may be determined together with the pixel value or by referring to other information instead. The correction unit 50 is a module that executes a process of changing the type of each pixel or a process of excluding a specific pixel from the calculation target in accordance with an instruction from the user.

2, a large number of blocks are shown on the right side of the data calculation unit 24. They schematically show information output from the data calculation unit. “Type information” 38 is information indicating the type identified for each pixel. “Local bone density” 40A is the bone density obtained for each pixel or a value corresponding thereto. The “average bone density” 40B is a planar bone density obtained in the bone region, that is, an average bone density. Usually, the average bone density is used as an index representing the properties of each vertebra.

“Local R” indicated by reference numeral 42A is a local evaluation value in units of pixels obtained for soft tissue (see the above equation (10)), and “average R” indicated by reference numeral 42B is obtained for soft tissue. Average evaluation value within a predetermined area (see the above equation (10)). Here, the predetermined region is a soft tissue region other than the bone region in the region of interest. The “L local attenuation rate” 44A is an attenuation rate per pixel obtained by irradiation with low energy X-rays, and the “L average attenuation rate” 44B is within the bone region or soft tissue region at the time of low energy X-ray irradiation. Is the average attenuation rate. “H local attenuation rate” 46A is an attenuation rate in pixel units during high energy X-ray irradiation, and “H average attenuation rate” 46B is an average attenuation rate in the region during high energy X-ray irradiation. A “histogram” 48 is a histogram constructed by representing the number of pixels for each local bone density. As will be described later, such a histogram is displayed together with the subject image, and the position of the bone density (or R) for the pixel of interest is marked on the histogram.

3 to 5 show display examples in the bone density measuring apparatus according to the present embodiment.

In FIG. 3, a subject image 56 is displayed on the display screen 54. The subject image 56 is generally a bone density image, and is a black and white grayscale image. In the illustrated example, a plurality of vertebrae are shown. For these, as indicated by reference numeral 58, a plurality of regions of interest (sub-ROIs) are set. L1 to L4 indicate respective regions of interest. The setting of the plurality of regions of interest L1-L4 is automatically executed in the present embodiment, and the technique itself is known. Within each region of interest L1-L4, whether the pixel is a bone pixel or a soft tissue pixel is identified for each individual pixel by the above-described operation of the data calculation unit, and thereby region division, that is, pixel group division, is executed. Then, the average bone density for the region is calculated from the local bone density obtained for the bone pixel group. On the other hand, for the soft tissue region, the local R is calculated for each pixel, and the average R is calculated for the entire region. In addition to this, the various information described above is calculated.

As shown in FIG. 3, a specific pixel, that is, a specific coordinate can be designated by a user by performing a click input after moving the cursor 60 using a pointing device. Then, as shown in FIG. 3, correction support information 62 is displayed. Specifically, the correction support information 62 is displayed as a pop-up. It has a form like a balloon. In the example illustrated in FIG. 3, a bone pixel is designated, and the correction support information 62 includes information 64 indicating that the pixel is a bone pixel as an automatic identification result, a local bone density obtained for the target pixel, and the target pixel. The average bone density 66 calculated for the bone region, the attenuation amount at the target pixel at the time of low energy X-ray irradiation, the average attenuation amount 68 at the region including the target pixel, and the local at the target pixel at the time of high energy X-ray irradiation The attenuation amount and the average attenuation amount 70 in the region including the target pixel are displayed.

Therefore, the user can confirm the result of automatic identification that has already been performed on the designated pixel of interest, that is, the determined pixel type, and then check the local bone density and the average bone density. From the comparison, it can be determined whether or not the pixel of interest has a protruding pixel value compared to the surrounding area. Further, in such an evaluation, it is possible to comprehensively determine whether correction is necessary by comparing the L local attenuation rate and the L average attenuation rate and comparing the H local attenuation rate and the H average attenuation rate. Incidentally, since the correction support information 62 displayed in a pop-up has a balloon-type form as described above, that is, it indicates the cursor 60, the correspondence between the correction support information 62 and the target pixel is displayed on the screen. It can be intuitively recognized above. For example, a plurality of target pixels may be designated and a plurality of correction support information may be displayed simultaneously.

4, soft tissue pixels are designated by the cursor 72 in the display example shown in FIG. In that case, information as shown in the figure is displayed as the correction support information 74. That is, the correction support information 74 includes information 64A indicating soft tissue as an automatic identification result of the pixel type, the local R calculated for the target pixel, the average R66A calculated for the region including the target pixel, and the L calculated for the target pixel. L average attenuation rate 68A calculated for the region including the local attenuation rate and the target pixel, H local attenuation rate calculated for the target pixel, and an H average attenuation rate 70A calculated for the region including the target pixel, etc. Yes. The user can comprehensively determine the necessity or method of correction using such information. In addition, since the determination can be made accurately and quickly, the burden on the user can be greatly reduced as compared with the conventional case.

In the example shown in FIG. 5, as indicated by reference numeral 102, a specific region of interest is selected by the user and is highlighted. A specific pixel of interest is designated by the cursor 100. A histogram 76 for a specific region of interest is displayed at a position adjacent to the subject image 56. In the histogram 76, the horizontal axis represents the bone density (local bone density), and the vertical axis represents the number. That is, histograms are displayed for a plurality of bone pixel groups that constitute bone regions within the designated region of interest. When a specific bone pixel is designated as a pixel of interest by the cursor 100, a marker 78 is displayed on the histogram 76, and it is possible to easily specify where the bone pixel is located on the histogram. In the display example as shown in FIG. 5, the correction support information for the target pixel designated by the cursor 100 is displayed in the column indicated by reference numeral 104. Of course, such a display example is only an example.

In Fig. 6, various correction methods are organized as a table. Preferably, the correction operation is performed by the user while comparing the local value with the average value. First, as a result of automatic identification, when a bone pixel is identified, as shown in (A1), if the local value, that is, the local bone density is less than the average value, that is, the average bone density (that is, local If the bone density is lower than the average bone density by more than a predetermined value), for example, there is a possibility that a bone pixel is misidentified due to calcification of soft tissue. In that case, the pixel of interest is calculated. Correction to exclude from the target or correction to change the type of the target pixel from the bone pixel to the soft tissue pixel is executed. As shown in (A2), if the local bone density and the average bone density are equal (that is, if the difference between the two is within a predetermined value), it is determined that no special correction is necessary. As shown in (A3), if the local bone density is higher than the average bone density (that is, if the local bone density is larger than the average bone density by a predetermined value), for example, the target pixel is in the metal region In other words, it is presumed that the target pixel is in the compression fracture region, so that the correction for excluding the target pixel from the calculation target is executed.

On the other hand, as a result of automatic identification, if the pixel is identified as a soft tissue pixel, as shown in (B1), if the local R is less than the average R (that is, the local R exceeds the predetermined value than the average R) If it is small, a measurement error or the like is inferred. If necessary, correction for excluding the pixel from the calculation target is executed. As shown in (B2), if the local R is equal to the average R (that is, if the difference between the two is within a predetermined value), no special correction is performed. As shown in (B3), if the local R is larger than the average R (that is, if the local R is larger than the average R beyond the predetermined value), for example, in soft tissue due to low bone density, abnormal growth, etc. Since there is a possibility that there is a misperception that there is, correction for excluding the target pixel from the calculation target or correction for changing the type of the target pixel from the soft tissue pixel to the bone pixel is executed.

Of course, the correction method shown in FIG. 6 is an example, and may be determined by the user according to the situation. In the present embodiment, since the correction support information is displayed as described above, there is an advantage that the necessity of correction can be determined accurately and promptly as compared with a case where an intuitive determination is simply made on the image. By the way, in each region of interest, if the number of bone pixel groups and soft tissue pixel groups is too large or too small, some abnormality can be estimated, so there is a possibility of a region of interest setting error in particular. , Operations such as recalculation or resetting the region of interest are automatically executed.

Claims (10)

A subject image generation unit that generates a subject image in which a two-dimensional distribution of bone mineral is reflected based on detection data obtained by irradiating the subject with X-rays;
An identification processing unit that applies, to each pixel constituting the subject image, an identification process for identifying a bone pixel and a soft tissue pixel based on a pixel value of the pixel;
A target pixel designating unit for designating a target pixel by the user on the subject image;
A correction support unit that provides correction support information to the user when the pixel of interest is specified;
A correction execution unit that corrects the result of the identification process when a correction instruction is given by the user to the target pixel;
Including
The correction support information is
Organization type information representing the result of the identification processing for the pixel of interest;
A local evaluation value calculated for the pixel of interest according to the result of the identification process;
An average evaluation value calculated for the region to which the pixel of interest belongs according to the result of the identification process;
A bone density measuring apparatus comprising: The apparatus of claim 1.
When a bone pixel is identified by the identification process, a local bone density is displayed as the local evaluation value, and an average bone density is displayed as the average evaluation value. . The apparatus of claim 2.
Bone density characterized in that when the identification processing identifies a soft tissue pixel, a local soft tissue evaluation value is displayed as the local evaluation value, and an average soft tissue evaluation value is displayed as the average evaluation value measuring device. The apparatus of claim 1.
The correction support information further includes:
Local attenuation at the time of low energy X-ray irradiation and local attenuation at the time of high energy X-ray irradiation calculated for the pixel of interest;
The average attenuation at the time of low energy X-ray irradiation and the average attenuation at the time of high energy X-ray irradiation calculated for the region to which the pixel of interest belongs,
A bone density measuring device comprising: The apparatus of claim 3.
The local soft tissue evaluation value is a value indicating a ratio of local attenuation at the time of low energy X-ray irradiation and local attenuation at the time of high energy X-ray irradiation,
The average soft tissue evaluation value is a value indicating a ratio between an average attenuation amount during low energy X-ray irradiation and an average attenuation amount during high energy X-ray irradiation.
A bone density measuring device characterized by the above. The apparatus of claim 1.
The correction execution unit executes at least one of a type change from a bone pixel to a soft tissue pixel, a type change from a soft tissue pixel to a bone pixel, and an exclusion from a calculation target. apparatus. The apparatus of claim 2.
The subject includes a plurality of vertebrae;
A plurality of regions of interest are set for the plurality of vertebrae,
Each region of interest is identified as a bone region and a soft tissue region.
A bone density measuring device characterized by the above. The apparatus of claim 1.
The correction support unit
A histogram generation unit that generates a histogram indicating the number of pixels for each pixel value based on the subject image;
A marker generating unit that generates a marker indicating a pixel value of the target pixel on the histogram;
A bone density measuring apparatus comprising: In a method executed by a bone density measuring apparatus, and processing a subject image based on detection data obtained by irradiating a subject with X-rays,
Applying, to each pixel constituting the subject image, an identification process for identifying a bone pixel and a soft tissue pixel based on a pixel value of the pixel;
Recognizing the coordinates of the pixel of interest designated by the user on the subject image;
Providing correction support information to the user for the pixel of interest;
Including
The correction support information includes a local evaluation value calculated for the target pixel according to the result of the identification process, and an average evaluation value calculated for a region to which the target pixel belongs according to the result of the identification process. Feature method. The method of claim 9, wherein
The correction support information further includes tissue type information representing a result of the identification process for the pixel of interest.

Images