Measurement image processing system and method using dual-energy x-ray absorptiometry

Abstract

The present invention discloses a measurement image processing method for calculating a motion degree value of a measurement target from a measurement image by using dual-energy X-ray absorptiometry (DXA) of an energy-switching type. The measurement image processing method using DXA of the present invention comprises the steps of: obtaining a binarized image by collecting a low-energy image; obtaining a binarized image by collecting a high-energy image; calculating a difference image of a measurement target region at two energies; calculating a ratio of a difference region relative to a size of the measurement target region; distinguishing a measurement target by combining the binarized image of the measurement target region at the low energy and the binarized image of the measurement target region at the high energy; correcting a numerical value according to characteristics of the measurement target; and outputting a motion degree value.

Description

Measurement image processing system and method using dual-energy X-ray absorptiometry

[0001] The present invention relates to a measurement image processing system and method for calculating a motion degree value of a measurement target from a measurement image using dual-energy X-ray absorptiometry (DXA) of an energy switching type.

[0002] Equipment utilizing energy-switching dual-energy X-ray absorptiometry (DXA) is used in all fields that examine the components of a target object. In particular, equipment utilizing energy-switching dual-energy X-ray absorptiometry (DXA) is utilized in animal testing, essential for new drug development, and in the medical field, where it measures the human body to diagnose diseases such as osteoporosis and sarcopenia.

[0003] Dual-energy X-ray absorptiometry (DXA) uses two X-rays of different energies to illuminate the target and uses the two X-ray images obtained.

[0004] Meanwhile, there are two methods for collecting X-ray images with different energies: one is to separate the X-ray energy received at one time from an image collection device such as a detector into two, and the other is to use an energy switching type method in which X-rays with different energies are irradiated twice from an X-ray irradiation device (HFG, High Frequency Generator).

[0005] Most commercial image acquisition devices (detectors) used in fields that require fast measurement times, such as animal testing using mice or rats, or medical fields that measure the human body to diagnose diseases such as osteoporosis or sarcopenia, do not have the function to distinguish the energy of X-rays. Therefore, equipment that uses dual-energy X-ray absorptiometry (DXA) using commercial image acquisition devices generally uses an energy switching method. When using this energy switching method, two different energies of X-rays are irradiated to the measurement target to sequentially acquire images. At this time, if there is movement of the measurement target during the image measurement process, the analysis accuracy of the measured image is significantly reduced.

[0006] Dual-energy X-ray absorptiometry (DXA) analyzes components by utilizing the difference in attenuation of two energies when passing through the same target. However, if the target's location changes in the images for each energy, the assumption that certain pixels represent the same target cannot be met. Failure to do so can lead to a reduction in the reliability of equipment using dual-energy X-ray absorptiometry (DXA).

[0007] Therefore, in order to trust the analysis results of equipment using dual-energy X-ray absorptiometry (DXA), it is necessary to provide the equipment user with the object moving value, which is information on how much the object moved during the measurement.

[0008] The present invention has been proposed to solve the above-mentioned problem, and the purpose of the present invention is to provide a measurement image processing system and method using a dual-energy X-ray absorptiometry (DXA) device that calculates a motion degree value of a measurement target from a measurement image measured using dual-energy X-ray absorptiometry.

[0009] In addition, the present invention provides a measurement image processing system and method using a dual-energy X-ray absorptiometry (DXA) device, which allows a user of the device to objectively determine the reliability of analysis results using a measurement image by checking the degree of movement of a measurement target.

[0010] In order to achieve the above object of the present invention, the present invention includes an X-ray irradiation unit that sequentially irradiates low-energy and high-energy X-rays, an image processing unit that calculates a motion degree value of a measurement object using a low-energy image and a high-energy image of a measurement object obtained by X-rays irradiated by the X-ray irradiation unit, and a display unit that displays the motion degree value of the measurement object calculated by the image processing unit.

[0011] The image processing unit includes a binary image generation unit that collects the low energy image and the high energy image and generates a binary image of a measurement target area of ​​low energy and high energy, a difference image calculation unit that calculates a difference image of a measurement target area with respect to the low energy binary image and the high energy binary image generated by the binary image generation unit, a difference area ratio calculation unit that divides the size of the difference in the measurement target area of ​​the low energy image and the high energy image calculated by the difference image calculation unit of the measurement target area by the total size of the measurement target to calculate a difference area ratio value, a measurement target distinction processing unit that loads data stored in a characteristic table storage database of X-ray measurement images by group of measurement targets and compares the low energy measurement target area binary image and the high energy measurement target area binary image to perform a process of distinguishing the measurement targets, a numerical correction unit that corrects the difference area ratio value calculated by the difference area ratio calculation unit to a numerical value according to the characteristics of the measurement target to calculate a motion degree value, and a motion degree value output unit that outputs the motion degree value calculated by the numerical correction unit to a display. Provides a system.

[0012] The database storing the characteristics table of the X-ray measurement images by group of the above measurement objects is

[0013] It is desirable to group measurable objects using equipment using dual-energy X-ray absorptiometry and store data including the size of the measured object in advance, the standard form model, the pixel value distribution, and the error correction formula for the difference area ratio value for each group.

[0014] A binarization image generation unit that collects the above low-energy image and high-energy image and generates a binarization image of the measurement target area

[0015] It is preferable to include a low-energy image collection unit that collects a low-energy image of a measurement target obtained by low energy irradiated from the X-ray irradiation unit, a high-energy image collection unit that collects a high-energy image of a measurement target obtained by high energy irradiated from the X-ray irradiation unit, a noise removal filter unit that filters noise in images acquired from the low-energy image collection unit and the high-energy image collection unit, and a probability image generation unit that generates a probability image for calculating a measurement target area of ​​the low-energy image and the high-energy image.

[0016] In addition, the present invention provides a measurement image processing method using dual-energy X-ray absorptiometry, including the steps of collecting a low-energy image to obtain a binarized image, collecting a high-energy image after the step of obtaining the low-energy binarized image, calculating a difference image of a measurement target area for two energies after the step of obtaining the high-energy binarized image, calculating a ratio of the difference area to the size of the measurement target area after the step of calculating the difference image of the measurement target area, distinguishing a measurement target by synthesizing the low-energy measurement target area binarized image and the high-energy measurement target area binarized image after the step of calculating the ratio of the difference area to the size of the measurement target area, correcting a value according to a characteristic of the measurement target after the step of distinguishing the measurement target, and outputting a motion degree value after the step of correcting a value according to the characteristic of the measurement target.

[0017] The step of distinguishing the measurement target that synthesizes the above two energies is

[0018] It is desirable to load data from the characteristic table of the X-ray measurement image by group of measurement objects and compare the loaded data with the low-energy binarized image and the high-energy binarized image to distinguish the type of measurement object.

[0019] The step of collecting the above low energy image and obtaining a binarized image is

[0020] It is preferable to include a step of filtering out noise, and a step of calculating a probability image for calculating a measurement target area after the step of filtering out noise.

[0021] The step of collecting the above high energy image and obtaining a binarized image is

[0022] It is preferable to include a step of filtering out noise, and a step of calculating a probability image for calculating a measurement target area after the step of filtering out noise.

[0023] The present invention specifies a region of a measurement target from two X-ray images having different energies that are sequentially measured, compares the regions of the specified measurement targets to quantify the degree of difference, corrects the value according to the characteristics of the measurement target to produce a motion degree value having an objective meaning, and outputs this motion degree value on a display. Therefore, a user of equipment using dual-energy X-ray absorptiometry can objectively determine the reliability of the analysis results by checking information about the movement of the measurement target during the measurement process in each measurement image.

[0024] FIG. 1 is a schematic diagram showing the main configuration of a measurement image processing system using dual-energy X-ray absorptiometry to explain an embodiment of the present invention.

[0025] FIG. 2 is a schematic diagram showing the main configuration of an image processing unit of a measurement image processing system to explain an embodiment of the present invention.

[0026] FIG. 3 is a diagram showing the main configuration of a binary image generation unit among the image processing units of a measurement image processing system to explain an embodiment of the present invention.

[0027] FIG. 4 is a flowchart illustrating a measurement image processing method using dual-energy X-ray absorptiometry to explain an embodiment of the present invention.

[0028] FIG. 5 is a flowchart illustrating a step of collecting a low energy image to obtain a binarized image to explain an embodiment of the present invention.

[0029] FIG. 6 is a flowchart illustrating a step of collecting a high-energy image to obtain a binarized image in order to explain an embodiment of the present invention.

[0030] Figure 7 is an image comparing a low-energy probability image (left) and a high-energy probability image (right) to explain an embodiment of the present invention.

[0031] FIG. 8 is an image comparing a low-energy binarized image (left) and a high-energy binarized image (right) to explain an embodiment of the present invention.

[0032] FIG. 9 is an example of a difference image of a measurement target area for two energies to explain an embodiment of the present invention.

[0033] FIG. 10 is another example of a difference image of a measurement target area for two energies to illustrate an embodiment of the present invention.

[0034] Figure 11 is an example of size distribution statistics for each measurement target to explain an embodiment of the present invention.

[0035] Figure 12 is an example of a standard form model of an animal to explain an embodiment of the present invention.

[0036] Figure 13 is an example of a standard form model of a human body to explain an embodiment of the present invention.

[0037] Figure 14 is an example of an error correction formula for the difference area ratio value of an animal to explain an embodiment of the present invention.

[0038] Figure 15 is an example of an error correction formula for the difference area ratio value of a human body to explain an embodiment of the present invention.

[0039] Fig. 16 is an example of a motion degree value when the movement of a measurement target is measured in a state where the movement is most restricted to explain an embodiment of the present invention.

[0040] Fig. 17 is an example of a motion degree value when the movement of a measurement target is measured in a state in which the movement is appropriately restricted to explain an embodiment of the present invention.

[0041] Fig. 18 is an example of a motion degree value when the movement of the measurement target is measured in a large state to explain an embodiment of the present invention.

[0042] Fig. 19 is a photograph showing an example of outputting an animal's movement degree value on a display to explain an embodiment of the present invention.

[0043] Fig. 20 is a photograph showing an example of outputting the movement degree value of the femur of a human body on a display to explain an embodiment of the present invention.

[0044] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in various other forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description have been omitted for clarity of description, and the same reference numerals will be assigned to identical or similar components throughout the specification.

[0045] FIG. 1 is a drawing for explaining an embodiment of the present invention, showing a measurement image processing system using dual-energy X-ray absorptiometry.

[0046] The measurement image processing system using the dual-energy X-ray absorptiometry of the embodiment of the present invention includes an X-ray irradiation unit (100), an image processing unit (200), a display unit (300), and a database (400) storing a characteristic table of X-ray measurement images for each group of measurement objects.

[0047] The X-ray irradiation unit (100) can sequentially irradiate low-energy X-rays or high-energy X-rays. The image processing unit (200) can calculate the motion degree value of the measurement object using the low-energy image and the high-energy image of the measurement object obtained by the X-ray irradiated by the X-ray irradiation unit (100). The display unit (300) can display the motion degree value of the measurement object calculated by the image processing unit (200).

[0048] Meanwhile, the database (400) storing the characteristics table of the X-ray measurement image by group of measurement objects is a database that groups the types of objects that can be measured by equipment using dual-energy X-ray absorptiometry and stores data including the size of the measurement object measured in advance for each group, a standard type model, pixel value distribution, and an error correction formula for the difference area ratio value. The group of measurement objects can be, for example, mice, rats, rabbits, etc. in animals, and the spine, femur, forearm, calcaneus, etc. in humans, which are major examination sites. The data stored in the characteristic table storage database (400) of the X-ray measurement image by group of measurement objects include size distribution statistics by measurement object (conceptual diagram by target group, see FIG. 11), standard shape models (images of rats and mice among the small animal group (see FIG. 12), and images of the spine, femur, forearm, and calcaneus sequentially among the human group (see FIG. 13)), and data of correction coefficients corresponding to the area of ​​a specific object and ratio values ​​of difference areas before and after correction (conceptual diagrams in the animal and human groups, see FIGS. 14 and 15).

[0049] The database (400) storing the characteristic table of the X-ray measurement image by group of measurement objects stores data of correction coefficients corresponding to the area of ​​a specific object (e.g., mice, rats, rabbits, etc. in animals, and spine, femur, forearm, calcaneus, etc. in humans), as illustrated in FIGS. 14 and 15. Meanwhile, the corrected difference area ratio value can be obtained by multiplying the above-described correction coefficient by the difference area ratio value before correction.

[0050] The image processing unit (200) described above is a computer program module that calculates a motion degree value of a measurement target using a low-energy image and a high-energy image of the measurement target and outputs this motion degree value to a display unit (300).

[0051] The image processing unit (200), as illustrated in FIG. 2, includes a binarized image generation unit (201), a difference image calculation unit (241) of a measurement target area, a difference area ratio calculation unit (251), a measurement target distinction processing unit (261), a numerical correction unit (271), and a movement degree value output unit (281).

[0052] The binary image generation unit (201), as illustrated in FIG. 3, includes a low energy image collection unit (211), a high energy image collection unit (213), a noise removal filter unit (215), and a probability image generation unit (217).

[0053] The low-energy image collection unit (201) is a computer program module that collects low-energy images irradiated by the X-ray irradiation unit (100). The high-energy image collection unit (213) is a computer program module that similarly collects high-energy images irradiated by the X-ray irradiation unit (100). The noise removal filter unit (215) is a computer program module that performs an image processing process to prevent calculation errors before estimating the measurement target area. The probability image generation unit (217) is a computer program module that performs measurement target area calculation, and assigns a value close to 0 if the probability of the measurement target (object) existing is low based on pixel information in the image, and a value close to 1 if the probability of the measurement target (object) existing is high. That is, 0 here means that the probability of the object existing is 0%, and 1 means that the probability of the object existing is 100%.

[0054] Figure 7 is a diagram illustrating a low-energy probability image (left) and a high-energy probability image (right). The probability image generation unit (217) generates the low-energy probability image and the high-energy probability image using the following equation.

[0055] Probability = (pixel value - pixel mean) / (f1 X pixel standard deviation + f2 X pixel standard deviation)

[0056] Here, pixel value = pixel value of a specific part (X-ray count value),

[0057] Pixel Mean = Average value of all pixels in the image,

[0058] Pixel standard deviation = standard deviation value of all pixels in the image,

[0059] f1, f2 = calculation constants

[0060] am.

[0061] The binarization image generation unit (201) statistically determines pixels where the measurement target (animal or human body) is determined to be located in the low-energy probability image and the high-energy probability image generated by the probability image generation unit (217), and performs a process of distinguishing the pixels as the maximum value and distinguishing the pixels where the measurement target (animal or human body) is determined not to be located as the 0 value. Through this process, the binarization image generation unit (201) generates a low-energy binarization image of the measurement target area (left image of FIG. 8) and a high-energy binarization image of the measurement target area (right image of FIG. 8).

[0062] The binary image generation unit (201) can obtain the number of pixels of the measurement target by the following formula.

[0063] Number of pixels in the measurement target = nPixelsObject,

[0064] Number of valid pixels in low-energy binarized image = nPixelsObject_Low,

[0065] Number of valid pixels in the high-energy binarized image = nPixelsObject_High,

[0066] nPixelsObject = average(nPixelsObject_Low, nPixelsObject_High)

[0067] The difference image calculation unit (241) of the measurement target area is a computer program module that calculates the difference between the binary images of the measurement target area calculated for the low-energy binarized image and the high-energy binarized image generated by the binarized image generation unit (201). Fig. 9 is an example of a difference image obtained by the difference image calculation unit (241) of the measurement target area, and Fig. 10 is another example of a difference image obtained by the difference image calculation unit (241) of the measurement target area. The difference image calculation unit (241) of the measurement target area can obtain the number of valid pixels (nPixelsDiff) in the difference image. The number of valid pixels in the difference image calculated by the difference image calculation unit (241) of the measurement target area can be expressed as nPixelsDiff. For reference, with reference to Fig. 9, the number of white pixels is the number of valid pixels (nPixelsDiff).

[0068] The difference area ratio calculation unit (251) is a computer program module that calculates a difference area ratio value by dividing the difference size (number of pixels) of the measurement target area of ​​a low-energy binarized image and a high-energy binarized image by the total size (number of pixels) of the measurement target. The difference area ratio calculation unit (251) can display the difference area ratio value as ratioDiff. That is, the difference area ratio value (ratioDiff) can be displayed as (nPixelsDiff / nPixelsObject).

[0069] Meanwhile, the main values ​​obtained in Examples 1 and 2 shown in the process calculated in the difference image calculation unit (241) of the measurement target area can be expressed as follows.

[0070] Example 1

[0071] *nPixelsDiff_Example1 = 479

[0072] nPixelsObject_Example1 = 17,282

[0073] ratioDiff_example1 = 0.0277

[0074] Example 2

[0075] nPixelsDiff_Example2 = 1,328

[0076] nPixelsObject_Example2 = 19,251

[0077] ratioDiff_example2 = 0.0690

[0078] The measurement target classification processing unit (261) is a computer program module that loads data stored in a characteristic table storage database (400) of X-ray measurement images by group of measurement targets, and compares the loaded data with a binarized image of a low-energy measurement target area and a binarized image of a high-energy measurement target area to determine which group the measurement target belongs to. The measurement target classification processing unit (261) can use a template matching technique for shape, an equivalence evaluation technique using a pixel value distribution, etc. to determine which group the current measurement target belongs to.

[0079] For example,

[0080] The measurement target classification processing unit (261)

[0081] The highest matching value = maxMatching,

[0082] maxMatching = can be expressed as a template matching technique (standard shape model, size distribution statistics by measurement object, measurement image),

[0083] maxMatching_Example 1 = 0.9 (mouse), 0.2 (rat), if less than 0.1 (remaining),

[0084] The maxMatching_Example1 value with the mouse model is 0.9, which is larger than other models, so the measured image can be judged as a mouse.

[0085] The maximum matching value (maxMatching) indicates how similar the currently measured image is to each model corresponding to an animal such as a mouse, rat, or rabbit, or a human body part such as the spine, femur, forearm, or calcaneus. In Example 1, the maximum matching value (maxMatching_mouse) with the mouse model of 0.9 means that there is a 90% probability that the measured object is a mouse. Similarly, the maximum matching value (maxMatching_rat) with the rat model of 0.2 means that there is a 20% probability that the measured object is a rat. The same applies to the remaining models.

[0086] If two or more models have the same maximum matching value (maxMatching), the animal model is selected over the human model. This is because the characteristics of the human model are generally clearer than those of the animal model, so if the same matching value is obtained, the probability of the animal model being selected is greater. In other words, if a human body part is measured, the probability that it will be interpreted like an animal model is generally low. If two or more animal models have the same maximum matching value (maxMatching), the animal is selected in the order of mouse, rat, and rabbit. For example, if the values ​​of maxMatching_mouse, maxMatching_rat, and maxMatching_rabbit are 0.6, 0.6, and 0.1, respectively, the animal is determined to be a mouse. If two or more human models have the same maximum matching value (maxMatching), the calcaneus, forearm, femur, and spine are selected in the order of calcaneus, forearm, femur, and spine. For example, if the values ​​of maxMatching_calcaneus, maxMatching_forearm, and maxMatching_femur are 0.6, 0.6, and 0.1, respectively, the measurement target is determined to be the calcaneus.

[0087] The numerical correction unit (271) is a computer program module that calculates a movement degree value by correcting the difference area ratio value calculated by the difference area ratio calculation unit (251) to a numerical value according to the characteristics of the measurement target. In other words, the numerical correction unit (271) adjusts the numerical value for objectification of the difference area ratio value regardless of the characteristics of each group (error correction).

[0088] The numerical correction unit (271) is an example

[0089] If we apply the error correction formula for the difference area ratio value to the mouse model,

[0090] ratioDiff_example1 = 0.0277

[0091] ratioDiff_example1_correction = error correction formula (ratioDiff_example1) = error correction formula (0.0277) = 0.0351

[0092] You can get the movement degree value.

[0093] The error correction formula of the numerical correction unit (271) is a formula that multiplies the difference area ratio value by obtaining a correction coefficient when the type of the measurement target and the measured area information are input (see FIGS. 14 and 15).

[0094] That is, the corrected difference area ratio value = correction coefficient X difference area ratio value.

[0095] In the example, the correction coefficient of the error correction formula is 1.267, and is a value obtained by inputting the type of measurement target and measured area information (see FIGS. 14 and 15).

[0096] The movement degree value is an estimated value that has a value in the range of 0 to 1, and the closer it is to 0, the more it means there is no movement.

[0097] Figures 16 to 18 show the same trend as the graph above in terms of the degree of movement and the degree of dispersion of the measured data. Taking the test of measuring fat weight as an example, in Figures 16 to 18, the x-axis represents the fat weight without error (true value), and the y-axis represents the fat weight measured by the DXA equipment (measured value).

[0098] For example, a movement degree value of 0.03 or less indicates that the measurement was made in a state of extreme restriction of movement. This typically applies to animals that have been euthanized and are in a state of immobility, while for humans, it typically refers to subjects that are well-secured to an assistive device (see Figure 16). However, when measuring the human spine, movement degrees below 0.03 are rare due to abdominal movement caused by respiration.

[0099] If the movement degree value is greater than 0.03 and less than 0.06, it means that the movement was measured in a state where the movement was moderately restricted. In other words, in the case of normal animals, it can be considered that this applies when measuring an animal under respiratory anesthesia, and in the case of humans, it can be considered that this applies when measuring without using an assistive device or when there is slight movement even when fixed to an assistive device (see Figure 17).

[0100] If the movement degree value is greater than 0.06, it means that the measurement was made with significant movement. This can be considered to be the case when measuring an animal that is not well anesthetized, or in the case of a human subject that has difficulty controlling movement (see Figure 18).

[0101] The motion degree value output unit (281) is a computer program module that outputs the motion degree value calculated by the numerical correction unit (271) to a display. The motion degree value output unit (281) outputs and displays the motion degree value to the display unit (300), as shown in FIG. 19 (indicated by A) and FIG. 20 (indicated by B).

[0102] Users can review the output motion intensity values ​​and use them to select data from the measurement data that meet the required precision of the test. For example, to analyze only data with a similar precision to that obtained from euthanized animals or humans who are completely motionless due to being securely fixed to an assistive device, measurement data with a motion intensity value greater than 0.03 should be excluded as outliers. The criteria for determining the motion intensity value that is considered an outlier can be determined autonomously by the user, taking into account factors such as the required precision and convenience of the test.

[0103] The method of processing a measurement image using the measurement image processing system using the dual-energy X-ray absorptiometry of the present invention, which is performed in this manner, is described in detail with reference to FIGS. 4 to 6 as follows.

[0104] First, a low-energy image of a measurement target obtained by low energy irradiation from an X-ray irradiation unit is collected to obtain a binarized image (S1). The step (S1) of collecting a low-energy image to obtain a binarized image further includes a step (S11) of filtering noise and a step (S13) of calculating a probability image, as illustrated in FIG. 5.

[0105] The noise filtering step (S11) is a process of collecting a low-energy image and filtering the noise of the image to prevent calculation errors before estimating the measurement target area. Then, the probability image calculating step (S13) assigns a value close to 0 if the probability of the measurement target (object) existing is low based on the pixel information in the low-energy image, and a value close to 1 if the probability of the measurement target (object) existing is high. In other words, 0 here means that the probability of the object existing is 0%, and 1 means that the probability of the object existing is 100%.

[0106] The step of calculating the probability image (S13) is a step of obtaining a low-energy probability image (left image of Fig. 7).

[0107] The step (S1) of collecting low-energy images and obtaining a binarized image statistically determines the values ​​of all pixels in the low-energy probability image to determine a reference value (threshold) at which the measurement target (object) is determined to exist. The algorithm for determining this reference value uses the widely known Otsu Algorithm. Afterwards, pixels above the reference value are set to the maximum value, and pixels below the reference value are divided into 0 to obtain a low-energy binarized image of the measurement target area (left image of Fig. 8).

[0108] And, a high-energy image of a measurement target obtained by high energy irradiation from an X-ray irradiation unit is collected to obtain a binarized image (S3). The step of collecting a high-energy image to obtain a binarized image (S3) further includes a step of filtering noise (S31) and a step of calculating a probability image (S33), as illustrated in Fig. 6.

[0109] The noise filtering step (S31) is a process of collecting high-energy images and filtering out noise in the images to prevent calculation errors before estimating the measurement target area. In addition, the probability image calculating step (S13) assigns a value close to 0 if the probability of the measurement target (object) existing is low based on the pixel information in the high-energy image, and a value close to 1 if the probability of the measurement target (object) existing is high. In other words, 0 here means that the probability of the object existing is 0%, and 1 means that the probability of the object existing is 100%.

[0110] The step of calculating the probability image (S13) is a step of obtaining a high-energy probability image (right image of Fig. 7).

[0111] Step (S3) of collecting high-energy images and obtaining a binarized image statistically determines the values ​​of all pixels in the high-energy probability image to determine a threshold value (threshold) at which the measurement target (object) is determined to exist. The algorithm for determining this threshold value uses the widely known Otsu Algorithm. Afterwards, pixels above the threshold value are set to the maximum value, and pixels below the threshold value are divided into 0 to obtain a high-energy binarized image of the measurement target area (left image of Fig. 8).

[0112] In the step (S1) of collecting a low-energy image to obtain a binarized image and the step (S3) of collecting a high-energy image to obtain a binarized image, the number of pixels of the measurement target can be obtained by the following formula.

[0113] Number of pixels in the measurement target = nPixelsObject

[0114] Number of valid pixels in the low-energy binarized image = nPixelsObject_Low

[0115] Number of valid pixels in the high-energy binarized image = nPixelsObject_High

[0116] nPixelsObject = average(nPixelsObject_Low, nPixelsObject_High)

[0117] After the step (S3) of obtaining a binarized image of high energy, a step (S5) of calculating a difference image of a measurement target area for two energies is performed. The step (S5) of calculating a difference image of a measurement target area calculates the difference between the binarized images of the measurement target area calculated for the low-energy binarized image and the high-energy binarized image, respectively. Fig. 9 is an example of a difference image obtained by the difference image calculation unit (241) of the measurement target area, and Fig. 10 is another example of a difference image obtained by the difference image calculation unit (241) of the measurement target area. That is, the step (S5) of calculating a difference image of a measurement target area can obtain the number of valid pixels (nPixelsDiff) in the difference image by the difference image calculation unit (241) of the measurement target area. The number of valid pixels in the difference image calculated by the difference image calculation unit (241) of the measurement target area can be expressed as nPixelsDiff.

[0118] After the step (S5) of calculating the difference image of the target area to be measured, the ratio of the difference area to the size of the target area to be measured is calculated (S7). The step (S7) of calculating the ratio of the difference area to the size of the target area to be measured divides the difference sizes of the target area to be measured between the low-energy binarized image and the high-energy binarized image by the total size of the target area to calculate the difference area ratio value. The step (S7) of calculating the ratio of the difference area to the size of the target area to be measured can be expressed as ratioDiff. That is, the difference area ratio value (ratioDiff) can be expressed as (nPixelsDiff / nPixelsObject).

[0119] Meanwhile, in the step (S7) of calculating the ratio of the difference area to the size of the measurement target area, the main values ​​obtained in Examples 1 and 2, which show examples of calculating the difference image of the measurement target area, can be expressed as follows.

[0120] Example 1

[0121] nPixelsDiff_Example1 = 479

[0122] nPixelsObject_Example1 = 17,282

[0123] ratioDiff_example1 = 0.0277

[0124] Example 2

[0125] nPixelsDiff_Example2 = 1,328

[0126] nPixelsObject_Example2 = 19,251

[0127] ratioDiff_example2 = 0.0690

[0128] After calculating the ratio of the difference area to the size of the measurement target area (S7), the measurement target area binarization image of low energy and the measurement target area binarization image of high energy are combined to distinguish the measurement target area (S9). That is, the step (S9) of distinguishing the measurement target area by combining the low energy measurement target area binarization image and the high energy measurement target area binarization image distinguishes the measurement target area by combining the two energies.

[0129] The step (S9) of distinguishing the measurement target by synthesizing the low-energy measurement target area binarized image and the high-energy measurement target area binarized image is to load data stored in the characteristic table storage database (400) of the X-ray measurement image for each group of the measurement target, and to compare the low-energy measurement target area binarized image and the high-energy measurement target area binarized image with the loaded data to determine which group the measurement target belongs to. The step (S9) of distinguishing the measurement target by synthesizing the low-energy measurement target area binarized image and the high-energy measurement target area binarized image is to determine which group the current measurement target belongs to, such as a template matching technique for shape or an equivalence evaluation technique using pixel value distribution.

[0130] For example,

[0131] The measurement target classification processing unit (261)

[0132] The highest matching value = maxMatching,

[0133] maxMatching = can be expressed as a template matching technique (standard shape model, size distribution statistics by measurement object, measurement image),

[0134] maxMatching_Example 1 = 0.9 (mouse), 0.2 (rat), if less than 0.1 (remaining),

[0135] The maxMatching_Example1 value with the mouse model is 0.9, which is larger than other models, so the measured image can be judged as a mouse.

[0136] After the step of distinguishing the measurement target (S9), the numerical value is corrected according to the characteristics of the measurement target (S11).

[0137] The step (S11) of correcting the numerical value according to the characteristics of the measurement target corrects the difference area ratio value calculated by the difference area ratio calculation unit (251) to a numerical value according to the characteristics of the measurement target to produce a movement degree value. In other words, the step (S11) of correcting the numerical value according to the characteristics of the measurement target adjusts the numerical value for objectification regardless of the characteristics of each group (error correction) for the difference area ratio value. In the step (S11) of correcting the numerical value according to the characteristics of the measurement target, if the error correction formula for the difference area ratio value for the mouse model is applied as an example,

[0138] ratioDiff_example1 = 0.0277

[0139] ratioDiff_example1_correction = error correction formula (ratioDiff_example1) = error correction formula (0.0277) = 0.0351

[0140] You can get the movement degree value.

[0141] The movement degree value is an estimated value that has a value in the range of 0 to 1, and the closer it is to 0, the more it means there is no movement.

[0142] Figures 16 to 18 show the movement degree value and the degree of dispersion of the measured data following the same trend as the graph above. For example, a movement degree value of 0.03 or less indicates that the measurement was made with the most restricted movement. This typically corresponds to measurements made in animals that have been euthanized and are motionless, and in humans, it corresponds to measurements made in subjects that can maintain a stable state while fixed to an assistive device (see Figure 16). However, when measuring the spine, movement degrees below 0.03 are rare due to abdominal movement caused by respiration. A movement degree value greater than 0.03 and less than 0.06 indicates that the measurement was made with moderately restricted movement. This typically corresponds to measurements made in animals under respiratory anesthesia, and in humans, it corresponds to measurements made without the use of an assistive device or with slight movement while fixed to an assistive device (see Figure 17). A movement degree value greater than 0.06 indicates that the measurement was made with significant movement. In this case, it can be seen that it applies to measuring animals that are not well anesthetized, and in the case of humans, it can be seen to apply to subjects that have difficulty controlling their movements (see Figure 18).

[0143] After the step (S11) of correcting the numerical value according to the characteristics of the measurement target, the movement degree value is output (S13). The step (S13) of outputting the movement degree value outputs the movement degree value calculated in the step (S11) of correcting the numerical value according to the characteristics of the measurement target to the display unit (300). The step (S13) of outputting the movement degree value outputs and displays the movement degree value on the display unit (300) (see A of FIG. 19 and B of FIG. 20).

[0144] The user can select data among the measurement data that meet the required precision of the test by referring to this motion degree value. For example, if only data with a similar precision to that when measuring a euthanized animal or a human body that is well fixed to an assistive device and has no movement at all are to be used for analysis, measurement data with a motion degree value greater than 0.03 should be excluded as outliers. The standard for the motion degree value that determines it as an outlier can be autonomously determined by the user considering the required precision or convenience of the test. That is, the user of the equipment using dual-energy X-ray absorptiometry can objectively determine the reliability of the analysis results by checking information about the movement of the measurement subject in each measurement image during the measurement process.

[0145] Although the preferred embodiments of the present invention have been described above, the present invention is not limited thereto, and various modifications can be made within the scope of the claims, the detailed description of the invention, and the attached drawings, and it is obvious that this also falls within the scope of the present invention.

Claims

1. An X-ray irradiation unit that sequentially irradiates low and high energy X-rays; An image processing unit that calculates the motion degree value of the measurement object by using the low energy image and high energy image of the measurement object obtained by the X-ray irradiated from the above X-ray irradiation unit, and A display section that displays the movement degree value of the measurement target produced by the image processing section above. Includes, The above image processing unit A binarization image generation unit that collects the low energy image and the high energy image and generates a binarization image of the measurement target area of ​​low energy and high energy, A difference image calculation unit for a measurement target area that calculates a difference image of a measurement target area for a low-energy binarized image and a high-energy binarized image generated by the above binarized image generation unit, A difference area ratio calculation unit that calculates a difference area ratio value by dividing the size of the difference between the low energy image and the high energy image calculated in the difference image calculation unit of the above measurement target area by the total size of the measurement target area, A measurement target classification processing unit that loads data stored in a database storing a characteristic table of X-ray measurement images by group of measurement targets and compares the binarized image of the low-energy measurement target area with the binarized image of the high-energy measurement target area to perform a process of distinguishing measurement targets. A numerical correction unit that calculates a movement degree value by correcting the difference area ratio value calculated in the difference area ratio calculation unit to a numerical value according to the characteristics of the measurement target, and A motion degree value output unit that outputs the motion degree value calculated by the above numerical correction unit to the display. A measurement image processing system using dual-energy X-ray absorptiometry including .

2. In claim 1, The database storing the characteristics table of the X-ray measurement images by group of the above measurement objects is A measurement image processing system using dual-energy X-ray absorptiometry, which groups measurable objects using equipment using dual-energy X-ray absorptiometry and stores data including the size of the measured object in advance, a standard shape model, a pixel value distribution, and an error correction formula for the difference area ratio value for each group.

3. In claim 1, A binarization image generation unit that collects the above low-energy image and high-energy image and generates a binarization image of the measurement target area A low energy image collection unit that collects a low energy image of a measurement target obtained by low energy irradiated from the above X-ray irradiation unit, A high-energy image collection unit that collects a high-energy image of a measurement target obtained by high energy irradiated from the above X-ray irradiation unit, A noise removal filter unit that filters out noise in images acquired from the low-energy image collection unit and the high-energy image collection unit, and A probability image generation unit that generates a probability image for calculating a measurement target area of ​​the above low-energy image and the above high-energy image. A measurement image processing system using dual-energy X-ray absorptiometry including .

4. Step of collecting low energy images to obtain binarized images. A step of collecting a high energy image to obtain a binarized image after the step of obtaining the low energy binarized image above, After the step of obtaining the binarized image of the above high energy, the step of calculating the difference image of the measurement target area for the two energies, After the step of calculating the difference image of the above measurement target area, the step of calculating the ratio of the difference area to the size of the measurement target area, After the step of calculating the ratio of the difference area to the size of the above measurement target area, the step of distinguishing the measurement target by synthesizing the low-energy measurement target area binarization image and the high-energy measurement target area binarization image, After the step of distinguishing the measurement target, a step of correcting the numerical value according to the characteristics of the measurement target, and A step of outputting a movement degree value after a step of correcting a numerical value according to the characteristics of the above measurement target. A method for processing measurement images using dual-energy X-ray absorptiometry including .

5. In claim 4, The step of distinguishing the measurement target by synthesizing the low-energy measurement target area binarization image and the high-energy measurement target area binarization image is A measurement image processing method using dual-energy X-ray absorptiometry, which loads data from a characteristic table of X-ray measurement images by group of measurement objects and compares the loaded data with a low-energy binarized image and a high-energy binarized image to distinguish the type of measurement object.

6. In claim 4, The step of collecting the above low energy image and obtaining a binarized image is Step to filter out noise, and A step of calculating a probability image for calculating the measurement target area after the step of filtering out the above noise. A method for processing measurement images using dual-energy X-ray absorptiometry including further.

7. In claim 4, The step of collecting the above high energy image and obtaining a binarized image is Step to filter out noise, and A step of calculating a probability image for calculating the measurement target area after the step of filtering out the above noise. A method for processing measurement images using dual-energy X-ray absorptiometry including further.

Images