JP5149930B2 - X-ray CT apparatus and image processing method - Google Patents

Description

  The present invention relates to an X-ray CT apparatus and an image processing method, and more particularly to a CT image processing technique for extracting a predetermined tissue.

  It is known that there are two types of adipose tissues (adipocytes) in the body of an organism (animal and human body). One is white adipose tissue and the other is brown adipose tissue. White fat has the function of storing energy in the form of neutral fat and providing energy to the whole body as needed. Brown fat has a highly developed body heat production function. The increase and activation of brown fat results in a decrease in white fat causing obesity. Therefore, in recent years, brown fat has attracted attention in the medical field. Incidentally, a large amount of white fat is present in the abdomen, buttocks, thigh, back, upper arm, visceral periphery, and the like. On the other hand, brown fat is present only in specific parts such as the back part of the neck and the vicinity of the scapula in the back according to current research.

  Conventionally, as described in Patent Document 1, an apparatus for discriminating subcutaneous fat and visceral fat based on a CT image has been proposed. However, an apparatus and method for automatically identifying and extracting brown fat has not yet been provided.

JP 2003-339694 A

  In order to quantify the brown fat on the CT image, it is conceivable to manually designate and identify the brown fat while visually observing the CT image. However, according to the manual work, the objectivity or reliability of the quantification result of brown fat decreases. In particular, when brown fat cannot be clearly distinguished from other tissues on a CT image, the accuracy of identification is greatly reduced. In addition, manual labor generates a great amount of labor, and rapid processing cannot be performed. In particular, the above problem becomes more serious when the volume of brown fat is obtained based on a plurality of CT images.

  An object of the present invention is to enable automatic identification of brown fat based on CT images.

  Another object of the present invention is to make it possible to accurately identify brown fat based on CT images.

(1) An X-ray CT apparatus according to the present invention includes an X-ray generator that irradiates an object with an X-ray beam, an X-ray detector that detects an X-ray beam transmitted through the object, and the object A rotating mechanism that rotates the X-ray beam relative to the specimen, an image forming unit that forms a CT image based on an output signal of the X-ray detector, and image processing on the CT image, the subject And an image processing unit that distinguishes brown fat from other tissues.

  In the above configuration, when an X-ray beam is irradiated by the X-ray generator, the X-ray beam is detected by the X-ray detector after passing through the subject. The X-ray beam is rotationally scanned relative to the subject. In this case, the X beam may be rotated or the subject may be rotated. In order to form a plurality of CT images, the subject is moved and scanned with respect to the X-ray beam. In this case, the subject may be moved, or the X-ray beam may be moved. The image forming unit forms a CT image based on the output signal of the X-ray detector. The image processing unit executes image processing for extracting brown fat from the CT image. A quantitative calculation may be performed on the extracted brown fat, or the extracted brown fat may be displayed as an image. The subject is an animal or a human body.

  The X-ray CT apparatus according to the present invention has a brown fat measurement function that is not mounted on a conventional X-ray CT apparatus. By executing this function, brown fat can be automatically extracted by analysis of CT images, so that various problems that occur in manual extraction can be solved. That is, according to the present invention, the extraction can be performed with high accuracy, and the extraction process is automatically performed according to an objective standard, so that the reproducibility is good, and therefore the reliability of the measurement result is high. Can be increased. Note that brown fat may be automatically extracted in a two-dimensional or three-dimensional region of interest designated by the user.

(2) An image processing method according to the present invention is an image processing method for processing a CT image obtained by X-ray CT measurement on a subject, and a CT included in each pixel from a set of pixels constituting the CT image. A first extraction step for extracting a candidate pixel group that may belong to brown fat based on the value, and applying a false recognition pixel exclusion process to the candidate pixel group, thereby extracting the brown fat pixel group A second extraction step.

  According to the above configuration, it is possible to identify a brown fat pixel group by identifying a candidate pixel group on the CT image on the basis of the CT value and excluding a false recognition pixel included in the candidate pixel group. For example, a pixel having a CT value similar to the CT value of a brown fat pixel may appear at the boundary between white fat and muscle, a thin muscle layer, and the like. If processing for excluding such false recognition pixels is performed, the brown fat identification accuracy and the reliability of quantitative calculation results can be improved.

  As described above, according to the present invention, brown fat can be automatically identified based on a CT image. According to the present invention, brown fat can be accurately identified based on a CT image.

It is a perspective view which shows an example of the measurement part in the X-ray CT apparatus concerning this invention. 1 is a block diagram showing a preferred embodiment of an X-ray CT apparatus according to the present invention. It is a figure which shows an example of CT image. It is a flowchart for demonstrating a brown fat extraction process. It is a figure which shows CT image as a model. It is a figure which shows the result of a candidate pixel group extraction process. It is an enlarged view which shows the candidate pixel row | line | column on a boundary. It is a figure for demonstrating the filter used by a coexistence determination process. It is a figure for demonstrating the specific example of a coexistence determination process. It is a figure which shows the result of having performed the boundary pixel exclusion process. It is a figure for demonstrating a contraction process and an expansion process. It is a figure which shows the result of having performed the misidentification pixel exclusion process. It is a figure which shows the candidate pixel group extraction process result in the case of high spatial resolution. It is a figure which shows the candidate pixel group extraction process result in the case of low spatial resolution. It is a figure for demonstrating the process which excludes the circumference | surroundings of a high CT value pixel. It is a figure for demonstrating the process which excludes the circumference | surroundings of a low CT value pixel. It is a flowchart for demonstrating the evaluation value calculation method. It is a figure for demonstrating how to obtain | require the coefficient in a bone density calculation formula.

  Hereinafter, preferred embodiments of the present invention will be described.

(1) Outline of X-ray CT Apparatus According to Embodiment As described later in detail, the X-ray CT apparatus of the present embodiment includes an X-ray generator, an X-ray detector, a rotation mechanism, a scanning mechanism, an image forming unit, An image processing unit and a calculation unit; The image forming unit includes a first extraction unit (first extraction function) and a second extraction unit (second extraction function). A 1st extraction part extracts a candidate pixel group from the pixel set which comprises CT image based on CT value which each pixel has. Each candidate pixel is a pixel that may belong to brown fat. A 2nd extraction part applies a misidentification pixel exclusion process with respect to a candidate pixel group, and extracts a brown fat pixel group by this. The extracted brown fat pixel group constitutes a brown fat image. When the candidate pixel group does not include or hardly includes a misidentified pixel, it is possible to provide only the first extraction unit and regard the candidate pixel group as a brown fat pixel group as it is. However, since the candidate pixel group usually includes various misidentified pixels, various misidentified pixel exclusion processes as described later are applied. The calculation unit calculates a brown fat amount based on the identified brown fat, and calculates an evaluation value related to the abundance ratio of the brown fat in the subject based on the brown fat amount. The amount of brown fat (and other tissue amounts described later) corresponds to the number of pixels, area, volume, or weight.

  The image forming unit, the image processing unit, and the arithmetic unit are each configured as dedicated hardware or realized as a software function. The subject to be measured is an animal or a human body. Examples of animals include small animals such as rats, mice, and hamsters. In addition, dogs, cats, pigs and the like can be measured.

  Preferably, the first extraction unit determines, for each pixel constituting the pixel set, whether or not the CT value falls within a predetermined range, and specifies a pixel having the CT value within the predetermined range as a candidate pixel. . That is, since each tissue has a CT value within a certain range (however, the certain range is determined depending on the configuration and operating conditions of each device), it is possible to identify each tissue based on the CT value. The upper limit of the predetermined range is defined as a CT value between the standard CT value of muscle and the standard CT value of brown fat, and the lower limit of the predetermined range is a CT value between the standard CT value of white fat and the standard CT value of brown fat. It is determined as a value. However, since such a candidate pixel group generally includes pixels that are not brown fat pixels (misidentified pixels) as described above, a second extraction unit is provided to exclude them.

  Desirably, the misidentification pixel exclusion process in the second extraction unit includes a coexistence determination process. In the coexistence determination process, when there is a pixel having a CT value higher than the first threshold and a pixel having a CT value lower than the second threshold around the target candidate pixel, the target Candidate pixels are excluded as misidentified pixels. According to the inventor's experimental research, for example, a pixel having a CT value that appears to be similar to that of brown fat may occur between muscle and white fat. This is because the CT value of muscle is higher than the CT value of brown fat and the CT value of white fat is lower than the CT value of brown fat, so an intermediate CT value may be calculated at the boundary between the two. . On the other hand, if the above-described coexistence determination process is applied, misidentified pixels can be excluded using the characteristics of the boundary structure.

  Desirably, the misidentification pixel exclusion process in the second extraction unit includes an adjacent process. In the adjacent processing, a non-brown fat pixel that satisfies a predetermined condition in the CT image is specified, and when there is an adjacent candidate pixel adjacent to the non-brown fat pixel, the adjacent candidate pixel is excluded as a false recognition pixel. Desirably, the predetermined condition is a condition in which a pixel having a CT value higher than a third threshold is set as the non-brown fat pixel, or a pixel having a CT value lower than the fourth threshold is set as the non-brown fat pixel. This is a condition for pixels. According to this configuration, when a candidate pixel exists adjacent to the tissue that is not brown fat (that is, a non-brown fat pixel) as a reference, it is specified as a false recognition pixel. For example, in a tissue boundary or a thin muscle layer, a pixel having a CT value similar to that of brown fat may be generated. However, according to the above configuration, such a misidentified pixel can be excluded.

  Desirably, the false recognition pixel exclusion process in the second extraction unit includes a contraction / expansion process. In the contraction / expansion process, after the contraction process is performed on the candidate pixel group, the expansion process is performed on the candidate pixel group after the contraction process. Although the thin misidentification layer can be erased by the shrinking process, the area around the true brown fat area is also removed, but if the expansion process is performed thereafter, only the true brown fat area can be restored to the original state. Therefore, the accuracy of quantitative calculation can be increased. When the contraction process is executed N times (where N ≧ 1), the expansion process is also executed N times. In any case, this contraction / expansion process is effective when the misidentified tissue region is thinner or smaller than the true brown fat region. Further, according to this contraction / expansion process, it is possible to remove falsely recognized pixels that exist in isolation.

  Preferably, a voltage switching unit that switches the drive voltage of the X-ray generator is provided. In the normal measurement mode, a high voltage is selected as the drive voltage for the X-ray generator, while in the brown fat measurement mode, a low voltage is selected as the drive voltage for the X-ray generator. Preferably, a rotation speed switching unit that switches the rotation speed of the X-ray beam is provided. In the normal measurement mode, a high speed is selected as the rotation speed of the X-ray beam, while in the brown fat measurement mode, a low speed is selected as the rotation speed of the X-ray beam.

  The X-ray CT apparatus of the present embodiment has a function of calculating an evaluation value based on the amount of brown fat obtained by quantitative calculation of brown fat. That is, the image processing unit identifies a plurality of tissues including brown fat in the subject based on the CT image, and the calculation unit includes a plurality of tissues including the brown fat amount based on the image processing result of the image processing unit. An amount is calculated, and an evaluation value is calculated based on a plurality of tissue amounts. According to this configuration, an evaluation value reflecting not only the amount of brown fat but also the amount of tissue other than brown fat can be obtained. Therefore, it is possible to evaluate the subject comprehensively or objectively. According to this configuration, for example, it is possible to objectively compare measurement results between subjects of different sizes (physiques), and to comprehensively evaluate the degree of obesity or the difficulty of becoming obese. Become. Further, the constitution can be evaluated from the viewpoint of the abundance of brown fat.

  Desirably, an evaluation value is a brown fat rate calculated | required by the calculation which divides the amount of brown fat by the sum total of several tissue quantity. Desirably, the evaluation value is a brown fat ratio obtained by calculation by dividing the amount of brown fat by the weight of the subject. Desirably, the plurality of tissue amounts include a white fat amount in addition to a brown fat amount, and the evaluation value is obtained by calculating the brown fat amount by a value obtained by adding the brown fat amount and the white fat amount. The content of brown fat in the fat. Desirably, the plurality of tissue amounts include muscle mass and white fat mass in addition to the brown fat mass, and the evaluation value is calculated by dividing the value obtained by multiplying the brown fat mass and the muscle mass by the white fat mass. This is a required anti-obesity parameter. Desirably, the plurality of tissue amounts include muscle mass and visceral fat mass in addition to the brown fat mass, and the evaluation value is calculated by dividing the value obtained by multiplying the brown fat mass and the muscle mass by the visceral fat mass. This is a required anti-obesity parameter.

(2) Details of X-ray CT Apparatus According to Embodiment FIG. 1 shows an example of the measurement unit 10 in the X-ray CT apparatus. The X-ray CT apparatus of the embodiment is an apparatus for performing CT measurement of mice such as mice, rats, guinea pigs, hamsters and the like used in animal experiments. A tissue separated from those small animals may be a measurement target. As shown later in FIG. 2, the X-ray CT apparatus includes a measurement unit 10 and a calculation control unit.

  The measurement unit 10 includes a main body 16 including a gantry 18. An opening is formed in the upper surface 16A of the main body 16, and an arm 26 projects upward from the opening. The arm 26 forms a part of a slide mechanism to be described later. The arm 26 is connected to the container 24, and slides (moves and scans) the container 24 in the direction of the rotation center axis.

  On the other hand, a measurement unit (X-ray generator, X-ray detector), which will be described later, is housed in the gantry 18 and rotates around the rotation center axis. A hollow portion 18A is formed in the central portion of the gantry 18 in the direction of the rotation center axis. The hollow portion 18A is a non-penetrating type, but may be a penetrating type.

  In the present embodiment, the container 24 is a capsule for storing a specimen (a small animal or a tissue extracted therefrom), and the container 24 has a substantially cylindrical shape in the present embodiment. The container 24 is arranged in a state where the container center axis coincides with the rotation center axis. Specifically, the base end portion of the container 24 is detachably attached to the upper end portion of the arm 26 described above. In this case, examples of the attachment / detachment mechanism include various engagement mechanisms or screwing mechanisms. As described above, the container 24 has a hollow cylindrical shape, and one or a plurality of small animals are disposed in the container 24 in the present embodiment. With such a configuration, the hair of a small animal can be prevented from coming into direct contact with the gantry 18. In addition, it is possible to prevent the problem that small animal excrement or detached hair is released to the outside. Furthermore, since it becomes possible to restrain a small animal in the container 24 with a fixture, problems such as image blurring when a CT image is reconstructed can be prevented. It is desirable to prepare a plurality of types of containers having different sizes and shapes and selectively use the containers.

  After the container 24 is attached to the arm 26, the arm 26 is driven forward along the direction of the rotation center axis, whereby the container 24 is inserted into the cavity 18A of the gantry 18. At this time, the container 24 is positioned so that the X-ray beam is set at the measurement position in the specimen. Further, such a measurement position is changed continuously or stepwise. As a result, a large number of CT sections spatially aligned at a predetermined pitch are formed.

  An operation panel 20 is provided on the upper surface 16A of the main body 16, and the operation panel 20 includes a plurality of switches, indicators, and the like. Using this operation panel 20, the user can operate the operation of the apparatus at the measurement site. A plurality of casters 22 are provided below the main body 16.

  In this embodiment, the rotational speed of the measurement unit in the gantry 18 can be varied stepwise or continuously. Further, the drive voltage of the X-ray generator in the gantry 18 can be varied stepwise or continuously. As will be described later, in the normal measurement mode, a high rotation speed (or normal rotation speed) and a high voltage (or normal voltage) are selected. On the other hand, in the brown fat measurement mode, a low rotation speed is selected and a low voltage is selected. An arithmetic control unit to be described later has a rotation speed switching function and a voltage switching function.

  FIG. 2 is a block diagram showing the configuration of the X-ray CT apparatus according to the present embodiment. In the measurement unit 10, an X-ray generator 52 is provided on one side and the X-ray detector 60 is provided on the other side with the rotation center axis O therebetween. A collimator 54 is provided on the irradiation side of the X-ray generator 52. The X-ray generator 52 generates a divergent or fan-shaped (ie, fan beam-shaped) X-ray beam 56 as shown. On the other hand, the X-ray detector 60 is configured as a plurality of (for example, 100) X-ray sensors arranged in a line, and an X-ray receiving opening is set according to the opening angle of the X-ray beam 56. Incidentally, the arrangement of the plurality of X-ray sensors may be linear or arcuate. In the present embodiment, a highly sensitive X-ray sensor is used. In FIG. 2, the voltage source connected to the X-ray generator 52 and the signal processing circuit connected to the X-ray detector 60 are not shown.

  In FIG. 2, reference numeral 58 indicates an effective field of view. This is a circular area in which a CT image can be formed when the X-ray beam 56 is rotated and scanned. Incidentally, the effective visual field 58 is determined according to the positional relationship between the rotation center axis, the X-ray generator 52, and the X-ray detector 60. In this embodiment, since the displacement mechanism 62 is provided, it is possible to change the positional relationship between them and mechanically vary the magnification of the CT image.

  That is, the X-ray generator 52 and the X-ray detector 60 are connected to the displacement mechanism 62, and the displacement mechanism 62 maintains the distance between the X-ray generator 52 and the X-ray detector 60. They (that is, the measurement unit) are displaced in the beam axis direction of the X-ray beam 56. In this case, the rotation center axis O is not changed, that is, the magnification can be changed by moving the measurement unit without moving the container described above. The displacement mechanism 62 includes a motor 62A for generating a displacement force.

  The gantry rotation mechanism 66 is a mechanism that rotates the entire base including the displacement mechanism mounted thereon by rotating the rotation base. Since the measurement unit is mounted on the displacement mechanism 62, the measurement unit positioned at a desired position by the displacement mechanism 62 is rotationally driven while maintaining the position. The gantry rotating mechanism 66 has a motor 66A for generating the driving force.

  The slide mechanism 68 is a moving mechanism that slides the arm 26 shown in FIG. 1, and the driving force is generated by the motor 68A. The operation panel 20 is provided on the upper surface of the main body as described above. The operation panel 20 may be connected to a local controller (not shown) provided on the measurement unit 10 side so that the local controller and the arithmetic control unit 12 communicate with each other.

  Incidentally, although various mechanisms 62, 66, 68, etc. are shown in FIG. 2, it is desirable to provide a sensor to detect the position or position change by these mechanisms. And it is desirable for the arithmetic control part 12 to perform feedback control based on the output signal of those sensors. The magnification change by the displacement mechanism 62 may be performed by user input. For example, the subject size or the container size is automatically detected, and the magnification is automatically set based on the detected data. Also good. Furthermore, when the container type or the like is registered in advance, the magnification may be set using the registered information. Further, in the example shown in FIG. 2, the slide mechanism 68 has the motor 68A as a drive source, but the slide force may be generated artificially.

  Next, the arithmetic control unit 12 will be described. Connected to the processor 30 are a display 32, a storage device 34, a keyboard 36, a mouse 38, a printer 40, and the like. In addition, a communication unit 42 for communicating with an external device via a network is connected.

  The processor 30 includes a CPU and various programs. FIG. 2 shows typical functions thereof, and the processor 30 includes an operation control unit 44, an image forming unit 46, an image processing unit 70, a calculation unit 72, and the like. You may make it provide a scout image formation part etc. as needed.

  The operation control unit 44 controls the overall operation of the measurement unit 10. The operation control unit 44 has a function of switching the rotation speed, a function of switching the driving voltage of the X-ray generator 52, and the like. The rotation speed and driving voltage may be designated by the user. Alternatively, when the measurement mode is selected by the user, the rotation speed and the drive voltage may be automatically set according to the selected measurement mode.

  The X-ray CT apparatus according to the present embodiment includes a normal measurement mode and a brown fat measurement mode. In the normal measurement mode, a high rotation speed is selected and a high voltage is selected. In the brown fat measurement mode, a low rotation speed is selected and a low voltage is selected. According to the inventor's experimental research, it has been found that brown fat can be well distinguished on a CT image by using the high-sensitivity X-ray sensor and changing the operating conditions as described above.

  When a small animal such as a mouse is to be measured, in the normal measurement mode, for example, 10 rotations per minute is selected as the rotation speed, and 70 kV is selected as the drive voltage, for example. On the other hand, in the brown fat measurement mode, for example, one rotation per minute is selected as the rotation speed. In this case, the drive voltage can be set within a range of 20-100 kV, but is preferably set within a range of 30-70 kV, and particularly preferably within a range of 30-50 kV. The apparatus used for the experiment of the present inventor is an X-ray CT apparatus (product name: LaTheat (registered trademark), model: LCT-100) manufactured by Aroka Corporation. Each numerical condition may vary depending on the type of subject, the configuration of the X-ray CT apparatus, and the like.

  The image forming unit 46 performs an operation for forming a CT image based on a lot of data obtained by rotational scanning of the X-ray beam. Various known methods can be used for the CT image reconstruction calculation.

  The image processing unit 70 has a function of identifying each tissue (bone, muscle, white fat, brown fat, etc.) based on the CT value of each pixel on the CT image. In particular, the first extraction process and the second extraction process are applied to the CT image, thereby having a new function of extracting only the brown fat group (brown fat image). The extracted brown fat is imaged or quantitatively calculated. The method for extracting brown fat will be described in detail later.

  The calculation unit 72 performs a calculation for quantitative evaluation of each tissue based on the image processing result. In particular, it has a function of calculating a brown fat amount (area, volume, weight, etc.) for brown fat, and further has a function of calculating a predetermined evaluation value based on the brown fat amount. The evaluation value calculation method will be described in detail later.

  The display 32 displays CT tomographic images, measurement results, and the like. The display unit 32 may superimpose and display a brown fat image using the scout image as a background image. In that case, the background image may be a black and white image and the brown fat image may be a colored image. It is also possible to construct such a superimposed image as a three-dimensional image.

FIG. 3 shows an example of a CT image formed by the X-ray CT apparatus of the present embodiment. This CT image is a tomographic image of a mouse. WAT indicates white fat and BAT indicates brown fat. Under the optimum operating conditions described above, brown fat appears relatively clearly on the CT image. Under the operating conditions, the average (standard) CT value and identification range of each tissue are as follows.

  As described above, the CT value of brown fat is higher than that of white fat and lower than that of muscle. However, if the operating conditions of the device are not optimized, the discrimination between brown fat and white fat becomes very difficult. That is, the X-ray absorption amount of both is almost the same value. Therefore, it is desirable to appropriately change the operating conditions of the apparatus so as to utilize the X-ray characteristics and find the optimum conditions for brown fat measurement (especially the CT value between white fat and brown fat). It is desirable to find conditions that can widen the difference). In the state where the operating conditions are optimized, as can be understood from the above table, it is possible to specify the tissue to which each pixel belongs based on the CT value of each pixel. However, as will be described later, for example, pixels (misidentified pixels) having intermediate CT values occur at the boundary between white fat and muscle. In order to extract brown fat with high accuracy, it is desirable to exclude such misidentified pixels.

(3) Brown Fat Extraction Next, a brown fat extraction method from a CT image will be described with reference to FIGS. First, FIG. 4 conceptually shows the flow of extraction processing.

  The CT image shown in S101 is formed by executing the reconstruction calculation (image forming calculation) shown in S100 based on the data output from the X-ray detector. FIG. 5 described later shows a CT image represented as a model.

  S102 represents the entire brown fat extraction process. This extraction process S102 is roughly divided into a candidate pixel group extraction process S103 corresponding to the first process, and a misidentified pixel exclusion process S104 as the second extraction process. In candidate pixel group extraction processing S103, candidate pixel groups that may correspond to brown fat are extracted from the CT image, as will be described later with reference to FIG. In this case, a pixel having a CT value within a certain range is set as a candidate pixel in a pixel set constituting the CT image. In the misidentification pixel exclusion process S104, a misidentification pixel other than the brown fat pixel included in the candidate pixel group is specified, and a process of excluding it (a process of making a pixel that is not a candidate pixel) is executed. As specific methods for the misidentification pixel exclusion process S104, several techniques can be given. In FIG. 4, the boundary pixel exclusion process S105 described later with reference to FIGS. 8 and 9 is used, and FIG. 11 is used later. Contraction / expansion process S108 described later, process S106 for excluding the periphery of the high CT value pixel, which will be described later with reference to FIG. It is shown.

  In the erroneous pixel exclusion process S104 shown in FIG. 4, S108 is executed after S105 is executed. Instead, S108 may be executed after S106 is executed. Alternatively, instead of them, only S107 may be executed.

  In any case, as shown in S109, the brown fat pixel group can be extracted as a result of executing the candidate pixel group extraction process S103 and the misidentified pixel exclusion process S104. This brown fat pixel group constitutes a brown fat image. This will be described with reference to FIG. When the brown fat image is obtained, display processing for the brown fat image is performed or processing for quantitatively calculating the brown fat is executed in S110. Then, a brown fat image is displayed on the display screen of the display and / or a numerical value representing the brown fat amount is displayed. The X-ray CT apparatus according to this embodiment further includes an evaluation value calculation function, which will be described later with reference to FIG.

  The specific contents of each process shown in FIG. 4 will be described in detail below. FIG. 5 shows a CT image 74 represented as a model. Image processing described below is executed for each CT image. In FIG. 5, the X axis and the Y axis are defined as the coordinate system of the CT image. Reference numeral 76 indicates white fat. Reference numeral 80 indicates a muscle including skin. 82 and 84 both represent muscles. Here, the muscle indicated by reference numeral 84 is a thin muscle layer. Reference numeral 78 indicates brown fat to be extracted. Reference numeral 92 indicates a boundary between the white fat 76 and the muscle 82, and reference numeral 94 indicates a boundary between the muscle 80 and the white fat 76.

  FIG. 6 shows a result of executing the candidate pixel group extraction process S103 shown in FIG. That is, reference numeral 90 represents an image in which only candidate pixel groups are identified. In the present embodiment, in accordance with Table 1 above, focusing on the CT value of each pixel constituting the CT image, the pixel having the CT value within a predetermined range is specified as a candidate pixel. Specifically, pixels satisfying the following conditions are used as candidate pixels.

      CTmin <CTxy <CTmax (1)

  In the above equation (1), CTxy represents the CT value (HU) of the pixel located at the xy coordinate on the CT image. CTmin represents the lower limit of the predetermined range, for example, the value is -170. CTmax represents the upper limit of the predetermined range, and for example, its value is -20. That is, in the CT image formed under the above-described optimum operating conditions, it is possible to identify brown fat by referring to the CT value of each pixel. However, in practice, the pixel group identified may include pixels belonging to a tissue other than brown fat, and thus the above-described misidentification pixel exclusion process is applied. Specifically, as shown in FIG. 6, brown fat is extracted as indicated by reference numeral 78A by candidate pixel group extraction processing based on the CT value. At the same time, a tissue or a region other than brown fat is extracted. Are extracted as candidate pixels. In the example shown in FIG. 6, at the boundary 92A between the muscle 82 and the white fat 76, although it is not brown fat, it exhibits a CT value similar to that of brown fat. It will be extracted as a candidate pixel. Further, an intermediate CT value also appears at the boundary 94A between the white fat 76 and the muscle 80, and each pixel on the boundary 94A is extracted as a candidate pixel. Furthermore, each pixel on the thin muscle 84A is extracted as a candidate pixel. Therefore, it is required to exclude such a misidentified pixel group.

  In addition, since the part where brown fat exists is generally only the specific part of the rear part of the neck and the back, if quantitative measurement of brown fat is performed, CT measurement is performed only on the specific part of the subject. Good. Of course, CT measurement may be performed on the whole body.

  The intermediate CT value occurs because, as shown in Table 1 above, white fat has a lower CT value than brown fat, while muscle has a higher CT value than brown fat. This is because a CT value equivalent to brown fat is observed at the boundary between fat and muscle. Therefore, a process for excluding erroneous recognition pixels will be described in detail below.

  First, the boundary pixel excluding process S105 shown in FIG. 4 will be described. In FIG. 7, a boundary 92A between the muscle 82 and the white fat 76 is shown as a partially enlarged view. The muscle 82 is composed of a plurality of muscle pixels 96, and the white fat is composed of a plurality of white fat pixels 98. A plurality of pixels 100 having the above-described intermediate CT values are arranged on the boundary 92A.

  In the boundary pixel exclusion process, a filtering process is applied to each candidate pixel. In the filtering process, as shown in FIG. 8, a predetermined number (for example, eight) of pixels R1-R8 existing around the candidate pixel Q of interest is referred to, and the first of the pixels R1-R8 includes the first pixel R1-R8. When at least one pixel having a CT value higher than a threshold (for example, CTmax) is included and at least one pixel having a CT value lower than a predetermined second threshold (for example, CTmin) is included, that is, high CT When the value pixel and the low CT value pixel coexist, the target candidate pixel is specified as a misidentified pixel, and is excluded from the candidate pixel group.

  That is, because muscle and white fat are adjacent to each other at the specific boundary as described above, setting a filter on the boundary means that it is set across two regions. High CT value pixels and low CT pixels are captured simultaneously. Therefore, a misidentified pixel is specified using such a characteristic of the boundary structure.

  FIG. 9 shows some application examples of the filtering, that is, the coexistence determination process. In the case of (A), all the pixels 104 existing around the target candidate pixel Q are candidate pixels. In this case, the target candidate pixel Q is not excluded. In the case of (B), there is one white fat pixel 98 and one muscle pixel 96 around the attention candidate pixel Q. In this case, since the coexistence determination condition is satisfied, the attention candidate pixel Q is Excluded. In the case of (C), since there are some white fat pixels 98 and one muscle pixel 96 in addition to the candidate pixel 104 around the target candidate pixel Q, the above coexistence determination condition is satisfied, and the result As a result, the target candidate pixel is excluded. Further, in the case of (D), there are a plurality of candidate pixels 104 and a plurality of muscle pixels 96 around the attention candidate pixel Q, but no high CT value pixels exist. Candidate pixel Q is stored as it is.

  In the above processing example, eight pixels around the candidate pixel of interest are referred to, but only four pixels adjacent in the vertical and horizontal directions around the candidate pixel of interest may be referred to, and the distance from the candidate pixel of interest is A range within 2 pixels (that is, 15 pixels) may be referred to. It is also possible to use a three-dimensional filter.

  FIG. 10 shows the result of executing the above boundary pixel exclusion process. That is, in the image 106, as is clear from the comparison with the image 90 shown in FIG. 6, the candidate pixel columns existing on the boundaries 92A and 94A are removed (see 92B and 94B). However, even if the boundary pixel exclusion process as described above is applied, as shown in FIG. 10, the thin muscle layer 84A may not be effectively removed depending on the spatial resolution of the image. Therefore, as shown in FIG. 4, in order to exclude such candidate pixels belonging to the thin muscle layer 84A, a contraction / expansion process is executed.

  FIG. 11 illustrates the contents of the contraction / expansion process. (A) shows an image (a part) before the contraction / expansion process is performed. Reference numeral 110 represents brown fat, and a plurality of candidate pixels 104 are specified covering the whole. Reference numeral 108 represents a thin muscle layer, which should be filled with muscle pixels, but it is thin due to the presence of a white fat layer around the thin muscle layer 108. The muscle layer 108 is provided with a plurality of candidate pixels 104. Reference numeral 98 denotes a white fat pixel.

  After N contraction processes are applied to such an image, N expansion processes are applied. N is an integer greater than or equal to 1, and is set by the user or automatically. In the contraction process, each candidate pixel is set as a candidate pixel of interest, and the surrounding eight pixels are referred to for each pixel of interest, and if the eight pixels include at least one white fat pixel, the candidate of interest The pixel is determined to be a misidentified pixel and is excluded from the candidate pixel group. In this case, for example, the target candidate pixel is replaced with a white fat pixel. FIG. 11B shows the result of applying one contraction process to the image shown in FIG. As shown in the figure, the thin muscle layer 108 existing in the image shown in (A) has been erased, and accordingly, the periphery of the brown fat 110 is also deleted and the size thereof is reduced. (See 110A).

  In the dilation processing, each white fat pixel is set as a target white fat pixel on the image after the contraction processing, and the surrounding eight pixels are referenced for each target white fat pixel, and candidate pixels are included therein. In this case, a process of replacing (restoring) the target white fat pixel with a candidate pixel is executed. The processing result is shown in FIG. As is clear from the comparison with the image shown in (B), the brown fat indicated by reference numeral 110B has been restored to its original size. Of course, after the contraction and expansion process, it is generally difficult to completely restore the original shape of brown fat, but the original shape can be almost restored.

  Incidentally, when the thickness of the thin muscle layer to be excluded is 2 pixels or less, it seems that it is sufficient to perform only one contraction process and one expansion process. When the thickness is larger than two pixels, it is desirable to perform a plurality of contraction processes and the same number of expansion processes.

  As a result of performing the above-described contraction and expansion processing, an image 112 after brown fat extraction processing as shown in FIG. 12 can be obtained. Only the candidate pixel group on the brown fat indicated by reference numeral 78A is left. The candidate pixel group is regarded as a brown fat pixel group. Therefore, the area can be obtained by counting the number of pixels constituting the brown fat pixel group for each CT image, and if the same processing is applied to a plurality of CT images, the total area (or the total number of pixels) is obtained. ) To determine the volume of brown fat. It is also possible to calculate the weight of brown fat by multiplying its volume by a constant factor.

  The above-described contraction / expansion process is particularly effective when the spatial resolution of the image is not sufficient, as will be described below. 13 and 14 show a state in which a thin muscle layer 116 exists in the white fat 114 on the CT image. 13 and 14, (A) shows an image before the candidate pixel group extraction process is performed, and (B) shows an image after the candidate pixel group extraction process is performed.

  As shown in FIG. 13, in the case of low spatial resolution, b ′, c ′, e ′, f ′ shown in (B) among twelve pixels from a to l illustrated in (A). , G ′, i ′, and j ′ are recognized as candidate pixels. That is, a pixel that should be recognized as a muscle pixel is recognized as a candidate pixel under the influence of surrounding white fat. Therefore, in the case of such a low spatial resolution, it is difficult to exclude those erroneously recognized pixels only by the boundary pixel exclusion process shown in FIG. 9, and the above-described contraction / expansion process is additionally applied. Is desirable. On the other hand, as shown in FIG. 14, in the case of high spatial resolution, as a result of performing the candidate pixel group extraction process among the 12 pixels from a to l shown in (A), (B) The six pixels b ′, d ′, e ′, h ′, i ′, and l ′ shown in FIG. Since they generally constitute two lines (two candidate pixel rows) and the thickness of each line is very thin, a plurality of candidate pixels constituting each line can be obtained by applying the boundary pixel exclusion process described above. Can be effectively excluded as a misidentified pixel. Therefore, in such a high spatial resolution, it is not always necessary to apply the contraction / expansion processing (see S111 in FIG. 4).

  Next, the processing contents of S106 and S107 shown in FIG. 4 will be described with reference to FIGS. Those S106 and S107 are applied instead of S105 described above. Of course, these processes may be applied in multiple ways. Basically, S106 and S107 are selectively executed. In S106, if a pixel having a CT value equal to or greater than a predetermined third threshold (for example, CTmax) is specified on the CT image, and a candidate pixel exists among a plurality of pixels around the pixel, the candidate Pixels are excluded. Here, the surrounding 8 pixels may be referred to, or the surrounding 15 pixels may be referred to. Alternatively, other numbers of pixels may be referred to. The pixel to be excluded is excluded from the candidate pixel group. In that case, it may be replaced with a tissue pixel other than brown fat as necessary.

  FIG. 15 shows a specific example of the process S106. As shown in (A), a plurality of candidate pixels a to o exist on the boundary 118 between the muscle 117 having a high CT value and the white fat 120. When the surrounding pixel exclusion process is applied to such an image, a result as shown in (B) is obtained. That is, all the candidate pixels a to o are excluded from the candidate pixel group, that is, they are replaced with pixels other than the candidate pixels (in this case, muscle pixels). This is represented by a 'to o' in (B).

  In the above processing, attention is paid to a non-brown adipose tissue having a high CT value, which is clearly different from brown fat, and on the basis of that, candidate pixels around it are excluded. In addition, attention may be paid to a tissue having a low CT value (for example, white fat) that is clearly different from brown fat, and the surrounding candidate pixels may be excluded. This process corresponds to the process S107 for excluding the periphery of the low CT value pixel shown in FIG.

  The above S107 will be specifically described with reference to FIG. FIG. 16A shows the same image as the image shown in FIG. On the other hand, when the above-described processing S107 is applied, each pixel constituting each white fat (each white fat pixel) is specified, and for each white fat pixel, a predetermined number of pixels existing around it are referred to. If there is a candidate pixel among them, the candidate pixel is excluded. The processing result is shown in (B), and similarly to the above, candidate pixels a to o are replaced with pixels that are not candidate pixels. When such processing is performed, in order to specify a white fat pixel, a pixel having a CT value equal to or smaller than a predetermined fourth threshold (for example, CTmin) may be specified as a target pixel. In this case, the reference range around the target pixel can be arbitrarily set according to the situation as described above.

  When the process of S107 is performed, as a result, the same result as the contraction process is obtained at the same time, so that the process of S108 shown in FIG. 4 can be omitted. Although it is possible to remove candidate pixels and the like on the boundary also by the steps of S106 and S107, according to these processes, the periphery of brown fat may be deleted over a certain width. Therefore, when such a problem becomes conspicuous, it is better to execute the above steps S105 and S108.

  In any case, according to the brown fat extraction process as described above, the brown fat can be objectively extracted, so compared with the case where the brown fat region is manually extracted, Variations in analysis results can be prevented, and the reliability of measurement results can be increased. Further, since each of the above processes is based on a simple image calculation, a rapid process can be expected. That is, a brown fat image can be formed in a short time.

  According to the brown fat extraction process as described above, as described above, only the brown fat pixel group is extracted. Therefore, if the number of pixels constituting the brown fat pixel group is counted, the processing is performed. It becomes possible to obtain the area of brown fat on the CT image. If the same processing is applied to a plurality of CT images, the volume or weight of the brown fat that exists spatially can be calculated as the sum of the areas.

  According to the above embodiment, the amount of brown fat that could not be obtained by a conventional apparatus can be measured, and thus there is an advantage that useful information can be provided in disease diagnosis and health management in the medical field. Various variations are conceivable for the candidate pixel group extraction process as the first extraction process and the misidentified pixel exclusion process as the second extraction process. In any case, a high-accuracy extraction result can be obtained by extracting candidate pixels and excluding false recognition pixels as necessary.

(4) Calculation of Evaluation Value Next, calculation of the evaluation value based on the amount of brown fat will be described with reference to FIGS. FIG. 17 conceptually shows the flow of the evaluation value calculation.

  S201 indicates a CT image of each frame. The process described below is executed for each frame CT image. S202 represents a tissue discrimination process. That is, in the tissue discrimination process S202, each tissue is identified on the CT image. Here, S203 indicates brown fat extraction processing, and the processing shown in FIG. 4 is a typical example. S204 indicates white fat extraction processing, and the processing shown in FIG. 4 can be given as a representative example. In other words, white fat can be extracted by the same process as brown fat.

  S205 indicates visceral fat extraction processing, and S206 indicates subcutaneous fat extraction processing. As such a processing method, for example, the method described in Patent Document 1 can be used. S207 indicates a muscle extraction process, and S208 indicates a bone extraction process. A known method can be applied to them, that is, each tissue can be identified by a discrimination process based on the CT value. If necessary, another extraction process may be incorporated into the tissue discrimination process S202.

  S209 indicates quantitative calculation processing. That is, when each tissue is extracted on the image by each extraction process S203 to S208, the tissue amount is calculated for each tissue in the quantitative calculation of S209. The amount of tissue corresponds to the number of pixels, area, volume, or weight. For example, the volume can be easily calculated by obtaining the sum of the area or the number of pixels calculated over a plurality of frames for each tissue.

S203A to S208A represent the tissue amount as a result of the quantitative calculation performed in S209. Incidentally, the weight of each tissue can be obtained by multiplying the volume of each tissue by a predetermined conversion coefficient (specific gravity). For example, the specific gravity of white fat is 0.92 (g / cm ³ ), and the specific gravity of muscle is 1.06 (g / cm ³ ). Thus, if specific gravity of each structure | tissue is known, weight conversion can be performed about each structure | tissue using those specific gravity.

  However, the bone mass can be obtained by the following calculation. In the following calculation formula, BMC indicates bone mass (weight of bone mineral), and its unit is gram.

In the above equation (2-2), BMDxyz represents the bone density of the bone pixel located at the coordinates (x, y) in the z-th frame. The unit is (g / cm ³ ). X represents the CT value (HU) for the bone pixel, a represents the slope as the bone density conversion coefficient, and b represents the intercept or offset value as the bone density conversion coefficient. That is, when the CT value is obtained for the bone pixel, the bone density for the pixel can be obtained from the above equation (2-2).

In the above equation (2-1), V represents the volume (cm ³ ) per pixel, r and s represent the number of pixels in the X direction and Y direction constituting the frame, t represents the number of frames. Therefore, if the equation (2-1) is executed, the weight of the whole bone can be obtained from the CT value of each pixel.

  Incidentally, the above-described coefficients a and b can be obtained in advance by performing tomography on a plurality of phantoms. Each of the plurality of phantoms has a known bone density and a different bone density.

  FIG. 18 shows the result of the phantom experiment. The horizontal axis is CT value, that is, X, and the vertical axis is BMD. By plotting the measurement values for each phantom on the coordinate system shown in FIG. 18, a straight line connecting the respective measurement points can be found, and a and b in the above formula can be easily specified.

  Returning to FIG. 17, in the evaluation value calculation S <b> 210, an evaluation value is obtained by executing the calculation described below based on the tissue amount of each tissue obtained as described above.

  In the present embodiment, “brown fat percentage” shown in S211 “anti-obesity parameter” shown in S212 and “brown fat content” shown in S213 are calculated as evaluation values. Below, each evaluation value is demonstrated.

  First, the brown fat percentage will be described. If the subject is large, generally the amount of brown fat it has increases. Therefore, an objective evaluation cannot be performed even if the amount of brown fat is directly compared for subjects having different sizes. Therefore, it is desirable to perform correction based on the physique. The evaluation value determined from such a viewpoint is the brown fat percentage. The brown fat percentage is obtained, for example, by the following equation (3).

      Brown fat mass / weight x 100 (%) (3)

  The body weight in the above can be determined from the sum of bone mass, brown fat mass, white fat mass and muscle mass. Here, each tissue amount may be a weight as described above, but may also correspond to a volume or an area. The body weight can also be directly measured using a scale or an electronic balance. In any case, it is possible to solve the problem due to physique by normalizing the amount of brown fat using some index representing the physique. It is also possible to define the brown fat percentage using a calculation formula other than the above.

  Next, the anti-obesity parameter will be described. From the viewpoint of anti-obesity, it is desirable that the amount of white fat is extravagant and low. On the other hand, the muscle mass is proportional to the basal metabolic rate, and it can be said that the larger the mass, the less likely it is to become obese. Moreover, since it can be said that the amount of brown fat is proportional to the amount of combustion of white fat, it can be said that it is desirable that the amount of brown fat is large. An anti-obesity parameter is obtained by reflecting those relationships in one calculation formula. The larger the value, the less likely it is to become obese, or it is recognized that it is not obese. The anti-obesity parameter is defined as follows, for example.

      Brown fat mass x muscle mass / white fat mass (4)

  On the other hand, from the viewpoint of obesity, an increase in visceral fat is said to be a particular problem. Therefore, the above formula may be modified to define the anti-obesity parameter by the following calculation formula.

      Brown fat mass x muscle mass / visceral fat mass (5)

  Next, the brown fat content in fat will be described. As described above, fat can be divided into white fat and brown fat, and the ratio of both can be used as an evaluation value. That is, the content defined by the following formula (6) can be used as the evaluation value.

      Brown fat mass / (white fat mass + brown fat mass) × 100 (%) (6)

  In the above, some evaluation values are illustrated, but in any case, when brown fat amount alone cannot make an objective assessment of constitution or health status, It is possible to provide clinically useful information by obtaining an evaluation value relating one or more tissue quantities. For example, it is possible to calculate a parameter such as brown fat amount / visceral fat amount as an evaluation value. Furthermore, the plurality of evaluation values described above may be calculated in parallel and displayed simultaneously as numerical values.

  As described above, according to the above method, the evaluation value related to the abundance ratio of brown fat can be calculated based on the amount of brown fat. Even if it is not always possible to accurately evaluate the subject with the amount of brown fat alone, the evaluation value is obtained by considering other information or reflecting other information while using the amount of brown fat as a basis. By doing so, it is possible to perform objective or comprehensive evaluation on the subject.

(5) Various Variations In the above embodiment, the X-ray CT apparatus for small animals has been described. However, the present invention can also be applied to an X-ray CT apparatus for human bodies. That is, brown fat in the human body can be measured by the same method as described above, and an evaluation value based on the amount of brown fat can be calculated by the same method as described above.

  In the above embodiment, a fan beam is used. However, a pencil beam or a cone beam may be used. Further, CT measurement may be performed by holding the container in an upright state and forming a beam in the horizontal direction. In that case, one of the container or the beam is rotationally scanned. In the above embodiment, the subject is moved and scanned, but the measurement unit, that is, the beam may be moved and scanned. The moving scanning range may be set so as to cover the whole body of the subject, or the moving scanning range may be set so as to cover only a specific part of the subject.

  In the above embodiment, the two-stage processing of the first extraction processing and the second extraction processing is applied to the CT image. However, when brown fat can be distinguished with high accuracy by the first extraction processing, 2 The extraction process can be omitted. In order to further improve the brown fat discrimination accuracy, various types of image processing can be applied after the two-stage processing or before the first extraction processing.

  The brown fat image is desirably displayed superimposed on a background image representing the entire tissue. In this case, the background image can be configured as a black and white luminance image, and the brown fat image can be configured as a colored image with predetermined coloring. Alternatively, when a CT image is displayed in color, a predetermined hue may be assigned to each tissue and expressed in a colored color so that each tissue can be clearly distinguished by the hue on the CT image.

  DESCRIPTION OF SYMBOLS 1 Measurement part, 12 Calculation control part, 70 Image processing part, 72 Calculation part, S102 Brown fat extraction process, S103 Candidate pixel group extraction process, S104 False recognition pixel exclusion process, S202 Tissue discrimination process, S209 Quantitative calculation, S210 Evaluation value calculation .

Claims (9)

An X-ray generator that irradiates the subject with an X-ray beam;
An X-ray detector for detecting an X-ray beam transmitted through the subject;
A rotation mechanism for rotating the X-ray beam relative to the subject;
An image forming unit that forms a CT image based on an output signal of the X-ray detector;
An image processing unit for identifying brown fat in the subject from other tissues by image processing on the CT image;
Only including,
The image processing unit
A means for extracting a candidate pixel group from a pixel set constituting the CT image based on a CT value possessed by each pixel, wherein the CT value is within a predetermined range for each pixel constituting the pixel set. A first extraction unit that determines whether or not the pixel falls within the predetermined range and that identifies a pixel having a CT value within the predetermined range as a candidate pixel;
Applying a false pixel exclusion process including an adjacent process to the candidate pixel group, thereby extracting a brown fat pixel group;
Including
The upper limit of the predetermined range is defined as a CT value between the standard CT value of muscle and the standard CT value of brown fat, and the lower limit of the predetermined range is between the standard CT value of white fat and the standard CT value of brown fat. Defined as the CT value,
In the adjacent processing, a non-brown fat pixel that satisfies a predetermined condition in the CT image is specified, and when there is an adjacent candidate pixel adjacent to the non-brown fat pixel, the adjacent candidate pixel is excluded as a misidentified pixel,
The predetermined condition is a condition in which a pixel having a CT value higher than the third threshold value for non-brown fat pixel determination is the non-brown fat pixel, or a CT value lower than the fourth threshold value for non-brown fat pixel determination The X-ray CT apparatus includes a condition that a pixel having a non-brown fat pixel is included . The apparatus of claim 1.
Including a voltage switching unit for switching the driving voltage of the X-ray generator;
In the normal measurement mode, a high voltage is selected as the drive voltage of the X-ray generator,
In the brown fat measurement mode, a low voltage lower than the high voltage is selected as the drive voltage of the X-ray generator.
An X-ray CT apparatus characterized by that. The apparatus of claim 2 .
X-ray CT apparatus characterized in that the low voltage is a voltage within a range of 30-70 kV. The apparatus of claim 1.
A rotation speed switching unit that switches a rotation speed of the X-ray beam;
In the normal measurement mode, high speed is selected as the rotation speed of the X-ray beam,
In the brown fat measurement mode, a low speed lower than the high speed is selected as the rotation speed of the X-ray beam.
An X-ray CT apparatus characterized by that. The apparatus of claim 1.
The X-ray CT apparatus, wherein the subject is an animal other than a human body. The apparatus of claim 1.
An X-ray CT apparatus, wherein the subject is a human body. In an image processing method for processing a CT image obtained by X-ray CT measurement on a subject,
Extracting a candidate pixel group that may belong to brown fat from the pixel set constituting the CT image based on the CT value of each pixel, and for each pixel constituting the pixel set A first extraction step of determining whether or not the CT value falls within a predetermined range, and specifying a pixel having the CT value within the predetermined range as a candidate pixel ;
Applying a false pixel exclusion process including an adjacent process to the candidate pixel group, thereby extracting a brown fat pixel group; and
Including
The upper limit of the predetermined range is set to a level for discriminating between muscle and brown fat, and the lower limit of the predetermined range is set to a level for discriminating between white fat and brown fat,
In the adjacent processing, a non-brown fat pixel that satisfies a predetermined condition in the CT image is specified, and when there is an adjacent candidate pixel adjacent to the non-brown fat pixel, the adjacent candidate pixel is excluded as a misidentified pixel,
The predetermined condition is a condition in which a pixel having a CT value higher than the third threshold value for non-brown fat pixel determination is the non-brown fat pixel, or a CT value lower than the fourth threshold value for non-brown fat pixel determination The image processing method characterized by including the conditions which make the pixel which has the said non-brown fat pixel . The method of claim 7 , wherein
The image processing method further includes a quantitative evaluation step of quantitatively evaluating brown fat based on the number of the extracted brown fat pixels. The method of claim 8 , wherein
An image processing method, wherein at least one of area, volume, and weight is calculated for brown fat by the quantitative evaluation.