JP4565796B2 - Diagnostic imaging equipment - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image diagnostic apparatus for measuring body fat based on tomographic images related to biological information, and more particularly to an image diagnostic apparatus capable of measuring body fat by automatically separating subcutaneous fat and visceral fat.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, diagnostic imaging apparatuses that measure body fat percentage based on tomographic images related to biological information captured by a CT apparatus, an MRI apparatus, or the like are being put into practical use. 1 and 2 are diagrams for explaining a conventional method for measuring body fat based on a tomographic image.
Conventionally, the abdominal wall muscle layer and body surface are manually traced using a tomographic image using a pointing device such as a mouse, and the areas of subcutaneous fat and visceral fat are calculated. That is, as shown in FIG. 1 (a), in order to determine the area of visceral fat in a tomographic image, the spinal column standing muscle 11, lateral abdominal muscles 12, 13, The abdominal wall muscle layer such as the rectus muscle 14 is surrounded by a white line 10. Then, the CT value of the portion surrounded by the white line 10 is extracted in the range from the maximum value −50 to the minimum value −150. The extracted part turns white as shown in FIG. The area V of the white portion 15 corresponds to visceral fat. Next, as shown in FIG. 2A, the entire abdomen is surrounded by a white line 20 using a pointing device. Then, the CT value of the portion surrounded by the white line 20 is extracted within the range from the maximum value −50 to the minimum value −150. The extracted portion turns white as shown in FIG. The area W of the white portion 25 corresponds to the whole fat.
Accordingly, the area S of the subcutaneous fat is obtained by subtracting the area V of the white portion 15 indicating visceral fat from the area W of the white portion 25 indicating total fat. Based on the total fat area W, visceral fat area V, and subcutaneous fat area S thus determined, the ratio of each fat is determined. That is, visceral fat: total fat is V: W, subcutaneous fat: total fat is S: W, and visceral fat: subcutaneous fat is V: S.
[0003]
As described above, conventionally, an operator has to manually operate the pointing device to trace the white line 10 and the white line 20. Therefore, when the operator to be traced is different, there is a problem that the measured value varies depending on the operator, and the measured value varies depending on the skill level of the operator. In order to solve such problems, an image diagnostic apparatus configured to automatically measure body fat has been proposed as disclosed in Japanese Patent Application Laid-Open No. 2000-93424. In this diagnostic imaging apparatus, the operator automatically sets the area of the subcutaneous fat area and the area of the visceral fat area by setting the minimum CT value and the maximum CT value of the fat portion and specifying one point of the subcutaneous fat area. Extract and measure body fat based on it.
[0004]
[Problems to be solved by the invention]
The diagnostic imaging apparatus described in the above Japanese Patent Application Laid-Open No. 2000-93424 does not require the skill of manual tracing, but the operator designates a point in the subcutaneous fat region, and the abdominal wall muscles When the subcutaneous fat region and visceral fat region are continuous, such as in a case of atrophy, it is necessary to manually perform a process in which an operator manually draws a separation line in the gap between the muscle layers. It was. Therefore, even if it is automated, there is still a problem that manual operation by an operator is still necessary, and measurement values vary depending on the operator.
[0005]
Therefore, the present inventor filed as Japanese Patent Application No. 2001-16937 an image diagnostic apparatus that can automatically measure body fat based on tomographic images without the operator having to operate the image diagnostic apparatus one by one. Yes. In this application, the subcutaneous fat region and the visceral fat region are automatically separated to calculate values such as the body fat percentage. At this time, a preset value preset by the user is used as a threshold range for extracting the fat region. In addition, the fat region automatically extracted from the CT image by setting the threshold range includes unnecessary minute fat regions around the abdominal muscles and spinal cord, which can be removed manually using a mouse or the like, Or, it was handled either by allowing it as an error and processing it as it is.
[0006]
The present invention has been made in view of the above points, and can determine the threshold range of the fat region of the patient based on the CT value of the fat region that is different for each patient, and improve the body fat measurement accuracy. An object is to provide a diagnostic apparatus.
[0007]
Another object of the present invention is to provide an image diagnostic apparatus capable of improving body fat measurement accuracy in consideration of unnecessary minute fat regions around the abdominal muscles and spinal cord.
[0008]
[Means for Solving the Problems]
The present invention An image diagnostic apparatus according to the present invention includes a body surface region of interest extraction unit that extracts a body surface region of interest surrounding a whole living body based on a contour of the body surface in a tomographic image related to biological information, and a whole fat region in the body surface region of interest A whole fat region extracting means for extracting the contour of the abdominal wall muscle layer inside the subcutaneous fat layer in the tomographic image Based on the pixel value of the tomographic image Track and surround the internal organs Visceral Visceral region of interest extracting means for extracting a heart region, visceral fat region extracting means for extracting a visceral fat region in the visceral region of interest, and an average value of pixel values of the fat region from the body surface region of interest and the visceral region of interest And an average value and standard deviation calculating means for calculating the standard deviation, a fat threshold range calculating means for calculating an image-specific fat threshold range from the fat pixel average value and the standard deviation, and the fat threshold range calculating means. A fat region re-extracting unit that extracts the whole fat region and the visceral fat region again based on the image-specific fat threshold range and the body surface region of interest and the visceral region of interest, and the fat region re-extracting unit And a body fat calculating means for calculating body fat based on the whole fat region and the visceral fat region. A tomographic image relating to biological information is taken by a CT apparatus, an MRI apparatus, or the like. In this tomographic image, the CT value of the fat region differs for each patient. Therefore, in the present invention, the average value of the pixel value of the fat region and the standard deviation thereof are calculated in advance from the body surface region of interest and the visceral region of interest, and the threshold value range of the patient's fat region is calculated from the average value and standard deviation. I decided to decide. As a result, patient-specific body fat measurement accuracy can be improved.
[0009]
In addition, the present invention An image diagnostic apparatus according to the present invention includes a body surface region of interest extraction unit that extracts a body surface region of interest surrounding a whole living body based on a contour of the body surface in a tomographic image related to biological information, and a whole fat region in the body surface region of interest A whole fat region extracting means for extracting the contour of the abdominal wall muscle layer inside the subcutaneous fat layer in the tomographic image Based on the pixel value of the tomographic image Visceral region of interest extraction means for extracting a visceral region of interest surrounding the viscera, visceral fat region extracting means for extracting a visceral fat region in the visceral region of interest, and the whole extracted by the whole fat region extracting means Based on the unnecessary fat area deleting means for deleting unnecessary fat areas from the fat area, the whole fat area from which unnecessary areas have been deleted by the unnecessary fat area deleting means, and the visceral fat area extracted by the visceral fat area extracting means Body fat calculating means for calculating body fat. In the previous application, a small fat area scattered around the abdominal muscles, intestinal tract, and spine was recognized as an error in the visceral fat area, and the calculation was made to include it, but in this invention, the error in measuring body fat is reduced. In order to do this, parts that should not be included in these calculations, that is, unnecessary fat regions, were removed. Thereby, the body fat measurement accuracy can be improved.
[0010]
In addition, the present invention An image diagnostic apparatus according to the present invention includes a body surface region of interest extraction unit that extracts a body surface region of interest surrounding a whole living body based on a contour of the body surface in a tomographic image related to biological information, and a whole fat region in the body surface region of interest A whole fat region extracting means for extracting the contour of the abdominal wall muscle layer inside the subcutaneous fat layer in the tomographic image Based on the pixel value of the tomographic image Track and surround the internal organs Visceral Visceral region of interest extracting means for extracting a heart region, visceral fat region extracting means for extracting a visceral fat region in the visceral region of interest, and an average value of pixel values of the fat region from the body surface region of interest and the visceral region of interest And an average value and standard deviation calculating means for calculating the standard deviation, a fat threshold range calculating means for calculating an image-specific fat threshold range from the fat pixel average value and the standard deviation, and the fat threshold range calculating means. A fat region re-extracting means for extracting the whole fat region and the visceral fat region again based on the fat threshold range specific to the image and the body surface region of interest and the visceral region of interest, and the whole extracted by the fat region re-extracting unit Unnecessary fat area deleting means for deleting unnecessary fat areas from the fat area, and the whole fat area and internal organs from which unnecessary areas have been deleted by the unnecessary fat area deleting means Is obtained by a body fat calculation means for calculating the body fat based on 肪領 zone.
This is a combination of claims 1 and 2, which has both effects, and can significantly improve body fat measurement accuracy.
[0011]
In addition, the present invention The diagnostic imaging device ,in front The body fat calculating means calculates at least one of the values of visceral fat: total fat, subcutaneous fat: total fat, and visceral fat: subcutaneous fat. The body fat calculating means not only calculates the whole fat region, the visceral fat region, and the subcutaneous fat region, but also calculates each fat ratio based on these values.
[0012]
In addition, the present invention The diagnostic imaging device ,in front Display means for displaying a value indicating the body fat calculated by the body fat calculating means in the tomographic image is provided. Since each value indicating the body fat calculated by the body fat calculating means is displayed to the operator by the display means, the operator can easily recognize each fat ratio only by looking at it.
[0013]
In addition, the present invention The diagnostic imaging device ,in front The display means displays the total fat, visceral fat and subcutaneous fat on the tomographic image so as to be identifiable. The display means can easily recognize whether the visceral fat or the subcutaneous fat has been appropriately extracted by displaying the value of each fat ratio and displaying the subcutaneous fat or the visceral fat in an identifiable manner.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Preferred embodiments of an image diagnostic apparatus according to the present invention will be described below with reference to the accompanying drawings. FIG. 3 is a block diagram showing a hardware configuration of the entire diagnostic imaging apparatus to which the present invention is applied. This image diagnostic apparatus measures body fat based on a plurality of tomographic images collected for a target region of a subject with a medical image modality such as an X-ray CT apparatus, an MRI apparatus, and an ultrasonic apparatus.
[0015]
The diagnostic imaging apparatus includes a central processing unit (CPU) 30 that controls the operation of each component, a main memory 31 in which a control program for the diagnostic imaging apparatus is stored, and a plurality of tomographic image data and operational programs for each patient. A stored magnetic disk 32, a display memory 33 for temporarily storing image data for display, a CRT display 34 as a display device for displaying an image based on the image data from the display memory 33, and software on the screen A mouse 35 and a mouse controller 36 for operating the switches, a keyboard 37 having keys and switches for setting various parameters, a speaker 38, and a common bus 39 for connecting the above components. In this embodiment, only the magnetic disk 32 is connected as a storage device other than the main memory 31, but other than this, an FDD, a hard disk drive, a CD-ROM drive, a magneto-optical disk (MO) drive, a ZIP drive, A PD drive, a DVD drive, or the like may be connected. Furthermore, it can be connected to various communication networks such as a LAN (local area network), the Internet, and a telephone line via a communication interface (not shown) so that image data can be exchanged with other computers and databases. It may be.
[0016]
Hereinafter, an operation example of the diagnostic imaging apparatus of FIG. 3 will be described with reference to the drawings. FIG. 4 is a diagram illustrating a main flow executed by the diagnostic imaging apparatus of FIG. The CPU 30 in FIG. 3 operates according to this main flow. 7 and 8 are diagrams showing an example of a display screen showing how tomographic images are displayed on the display by this main flow, and correspond to FIG. Hereinafter, details of the main flow will be described in the order of steps.
[0017]
[Step S40]
First, since the patient ID input screen is displayed on the display 34 of the diagnostic imaging apparatus, the operator inputs the patient ID number. Then, a tomographic image as shown in FIG. 7A corresponding to the ID number of the patient to be diagnosed is read out from the tomographic images taken in advance by the medical image modality from the magnetic disk 32 and displayed on the display 34. Is done. Looking at the displayed tomographic image, the operator sets a threshold range of CT values for fat pixel extraction. Since the CT value of fat is usually in the range of −150 to −50, this range may be set as a default value and the threshold setting by the operator may be omitted. Further, by specifying a location corresponding to fat in the tomographic image, ± tens of percent of the CT value may be set as the threshold range.
[0018]
[Step S41]
When the setting of the threshold range in step S40 is completed, a body surface contour extraction process for setting a region of interest that surrounds the entire body surface in the tomographic image is then executed. FIG. 5 is a diagram showing details of the body surface contour extraction processing. Details of this body surface contour extraction process will be described below in the order of steps.
[Step S50]
An edge of the original image is detected to create a contour-enhanced image. Edge detection is performed by applying a difference filter such as a Laplacian filter to the original image. The edge portion is converted to have a high value by Laplacian filter processing.
[Step S51]
Threshold processing is performed on the contour-enhanced image to create a binarized image as shown in FIG.
Here, an optimum value is set in advance so that the outline remains clearly.
[Step S52]
A starting point 71 for tracing the contour line is searched by raster scanning from the pixel in the upper left corner of the binarized image created in step S51 as indicated by the arrow line in FIG. 7B. That is, scanning is sequentially performed in the horizontal direction from the upper left corner of the tomographic image, and the first level “1” pixel encountered is set as the start point 71.
[Step S53]
Contour tracking is started based on the starting point 71 searched in step S53, and is continued until it goes around the body surface. The contour line thus formed becomes the body surface region of interest 72 surrounding the whole. In addition to this, it is also possible to scan in the horizontal direction and the vertical direction from each of the upper, lower, left and right ends of the tomographic image toward the center of the image and use the first level “1” pixel as the outline.
[0019]
[Step S42]
Next, a visceral contour extraction process for setting a region of interest in the abdominal wall muscle layer surrounding the viscera existing inside the subcutaneous fat layer is executed. FIG. 6 is a diagram showing details of the internal organ contour extraction processing. Hereinafter, details of the internal organ contour extraction processing will be described in the order of steps.
[Step S60]
For the pixels in the body surface region of interest 72 extracted in step S41, pixels within the fat threshold range are extracted. The value set in step S40 is used as the fat extraction range.
[Step S61]
The fat pixels extracted in step S60 are removed from the original image and binarized. That is, the outside of the entire body surface region of interest 72 is cleared to zero from the original image, the fat pixels are cleared to zero inside the body surface region of interest 72, and the other pixels corresponding to the abdominal wall muscle layer are replaced with level “1”. Thus, a binarized image as shown in FIG. In the binarized image of FIG. 8A, the body surface region of interest 72 is indicated by a dotted line for the sake of understanding, but this line does not actually exist. The binarized image created in this way is an image in which the abdominal wall muscle layer is exposed to the outside. By tracking the contour of the binarized image, the contour of the abdominal wall muscle layer including the entire internal organs can be extracted. However, as described in the section of the prior art, the abdominal wall muscle layer does not necessarily surround the viscera continuously, and there are many gaps in several places. Accordingly, it is not possible to extract the outline of the abdominal wall muscle layer including the entire internal organs simply by tracing the outline of the abdominal wall muscle layer in which such a gap exists. Therefore, in this embodiment, contour correction processing is performed to fill the gap between the abdominal wall muscle layers by the following steps S62 to S68 and extract the contour line of the abdominal wall muscle layers including the entire internal organs. .
[0020]
[Step S62]
Raster scanning is performed from the pixel in the upper left corner of the binarized image created in step S61 as indicated by the arrow line in FIG. That is, scanning is sequentially performed in the horizontal direction from the upper left corner of the tomographic image, and the first level “1” pixel encountered is set as the start point 81. When the search for the starting point is completed, the contour line tracking process in steps S63 to S68 is executed. FIG. 9 is a diagram for explaining the outline tracking process, and is an enlarged view of the vicinity of the start point 81 in FIG.
[Step S63]
Draw a small circle so that it touches the outside of the outline on the point of interest. That is, as shown in FIG. 9, a small circle 91 that touches the outside of the contour line is drawn at the point of interest (start point 81). The small circle 91 is a circle having the start point 81 as a contact point.
[Step S64]
It is searched whether there is a pixel of level “1” on the circumference of the small circle drawn in the previous step S63.
[Step S65]
It is determined whether or not there is a pixel of level “1” on the circumference of the small circle. If it is determined that the pixel exists (yes), the process proceeds to the next step S66, where it is determined that it does not exist (no). If yes, jump to Step S67.
[0021]
[Step S66]
Since it is determined in the previous step S65 that the pixel of level “1” exists on the circumference of the small circle, the pixel closest to the point of interest (the pixel at the nearest point) among the pixels of level “1” and the pixel of interest A plurality of pixels corresponding to the arc of a small circle between the points is replaced with a pixel of level “1”. In FIG. 9, the small circle 91 to the small circle 92 have no level “1” pixel in the small circle, but the small circle 93 has a level “1” at the closest point 83 relative to the point of interest 82. Pixel exists. Therefore, in such a case, a plurality of pixels corresponding to the arc 93A of the small circle 93 connecting the point of interest 82 and the closest point 83 are replaced with the level “1”. Note that a pixel corresponding to a straight line connecting the point of interest 82 and the closest point 83 may be replaced with a pixel of level “1”.
[Step S67]
The point of interest is moved to the next pixel. If the result of determination in step S65 is that there is no level “1” pixel on the circumference of the small circle, the pixel on the next contour line adjacent to the point of interest is set as the next point of interest. When a plurality of pixels between the closest point and the point of interest are replaced with level “1” in step S66, the pixel of the closest point is set as the pixel of the next point of interest. That is, in the case of the small circle 93 in FIG. 9, the next point of interest after the point of interest 82 is the nearest point 83.
[0022]
[Step S68]
It is determined whether or not the next pixel is the same as the start point. If it is determined that the next pixel is not the start point (no), the process returns to step S63. If it is determined that the next pixel is the start point (yes), the internal organs are included. The contour extraction process ends, and the process proceeds to the next step S43. That is, it is determined whether or not the processing of step S63 to step S67 has made a round of the abdominal wall muscle layer and returned to the starting point. In this way, the small circle is rotated around the outline of the abdominal wall muscle layer once, and the outline tracking is performed while filling the gap of the abdominal wall muscle layer. Can be created. Such a contour extraction method is called a rolling ball method.
[0023]
Note that when contour tracking is performed by this rolling ball method, there is a problem that accurate contour tracking cannot be performed with a minute recess as shown in FIG. That is, as shown in FIG. 11A, in the case of the small circle 95, the point of interest 85 and the nearest point 86 exist on the circumference. In this case, the contour line is a contour line 111 as shown in FIG. However, the exact outline of this part is the dotted outline 112.
Therefore, in this embodiment, in order to correct such an inaccurate contour line, the visceral contour line extracted by the rolling ball method, that is, the periphery of the visceral region of interest 80 is made once more counterclockwise, and FIG. As shown in (b), a place where a branch point such as the outline 111 and the dotted outline 112 exists is searched. Then, at the branching point, follow the branch line to the left (dotted line outline 112) and the branch line to the right (contour line 111), and determine whether both branch lines merge at the nearest point 86 again. Investigate. In the case of joining, it is determined that the portion is a minute recess, and correction processing is performed with the left branch line (dotted outline 112) as a formal outline as shown in FIG. If they do not merge, it is determined that the gap is as shown in FIG. 9, and the right branch line remains. In addition, if the upper limit of the distance to be traced after branching is determined in advance and the merging is not performed within a certain distance, the tracking is suspended to stop the search from being wasted and the branch is regarded as not merging. .
[0024]
[Step S43]
The fat pixels in the body surface contour extracted in step S41, that is, in the body surface region of interest 72 in FIG. 8A are extracted. In this fat pixel extraction process, it is checked whether each pixel of the original image is in the body surface region of interest 72 and the pixel value is within the threshold range. count. At the same time, the value of the number of images is stored in a fat image buffer to be displayed superimposed on the original image, and the color of the pixel corresponding to the fat pixel is replaced with a predetermined color (for example, white).
[Step S44]
Fat pixels in the visceral contour extracted in step S42, that is, the visceral region of interest 80 in FIG. 8B are extracted. This fat pixel extraction process is performed in the same way as described above, by checking whether the original image is one pixel at a time within the internal region of interest 80 and the pixel value is within the threshold range. Count the number of pixels. At the same time, the value of the number of images is stored in a fat image buffer to be displayed superimposed on the original image, and the color of the pixel corresponding to the fat pixel is replaced with a predetermined color (for example, yellow).
[0025]
[Step S45]
Based on the total fat pixel number W extracted in step S43 and the visceral fat pixel number V extracted in step S44, the subcutaneous fat pixel number S is calculated, and based on the calculated ratio of these values. Visceral fat: Total fat value V / W, Subcutaneous fat: Total fat value S / W, Visceral fat: Subcutaneous fat value V / S, respectively.
[Step S46]
Visceral fat: total fat value V / W, subcutaneous fat: total fat value S / W, visceral fat: subcutaneous fat value V / S, which are the calculation results of step S45, are displayed on the screen, respectively. A tomographic image representing whole fat as in (a) or a tomographic image capable of distinguishing visceral fat and subcutaneous fat as shown in FIG. 10B is displayed. As a result, as shown in FIG. 10B, the visceral fat 15 is displayed in yellow in the visceral region of interest 80, the letters “V = 9545” as the number of pixels of the visceral fat, and “S = “17056” and “W = 26601” are displayed as the total fat pixel count. Further, as shown in the figure, values representing the respective fat ratios are also displayed. In FIG. 10, since the color of the image cannot be displayed, the yellow color corresponding to the visceral fat 15 is indicated by hatching, and the subcutaneous fat 100 is indicated by white.
[0026]
In the above-described embodiment, the case where the small circle used in the rolling ball method is set in advance so that the size of the small circle does not enter the gap of the abdominal wall muscle layer is described. You may make it select automatically. For example, a contour is extracted by the low link ball method using a circle having a relatively large radius R1 that does not enter the gap between the abdominal wall muscle layers, and an approximate area A1 of the visceral region of interest is obtained based on the contour. Next, a contour is extracted by the rolling ball method using a half-sized circle of radius R2, and the area A2 of the visceral region of interest based on the contour is obtained. If this area A2 is extremely small compared to the previously obtained area A1, the circle with the radius R2 is estimated to have entered the gap in the abdominal wall muscle layer, so that it is intermediate between the radius R1 and the radius R2. A contour is extracted using a circle with a radius R3, and an area A3 of the visceral region of interest based on the contour is obtained. If this area A3 is slightly smaller than the previously obtained area A1, the contour is extracted using a circle having a radius R4 between the radius R2 and the radius R3. By repeating the same processing a plurality of times, it becomes possible to specify a circle having the optimum size. Therefore, it is only necessary to perform contour extraction by the rolling ball method using this circle.
[0027]
Moreover, although the above-mentioned embodiment demonstrated the case where the outline of a visceral region of interest was extracted by the rolling ball method, you may use methods other than this. For example, a tomographic image on which the abdominal wall muscle layer is displayed is scanned in the horizontal and vertical directions from the top, bottom, left, and right ends toward the center of the image, and the first level “1” pixel that is encountered is outlined. It may be a line. In this case, since the outline of the gap between the abdominal wall muscle layers is interrupted, it is possible to extract the outline of the visceral region of interest by connecting the closest ends of the interrupted ends. In addition, when the outline of the inner abdominal wall muscle layer is extracted through the gap, the outline becomes extremely short, so the short one may be deleted. In this way, the outline of the visceral region of interest may be extracted.
[0028]
The above-described embodiment is the content of Japanese Patent Application No. 2001-16937 previously proposed by the present inventors. In general, since the CT value of the fat region varies from patient to patient, if the average CT value of the fat region is originally obtained in advance and the threshold range of the fat region of the patient is determined from the average value and the standard deviation, Increases body fat measurement accuracy. In the previous application, as shown in FIG. 12, the abdominal muscles surrounded by dotted lines in the visceral fat region, the intestinal tract, and minute fat regions scattered in white around the spine (unnecessary fat regions) are recognized as errors while being recognized as errors. The calculation is performed with this included. Therefore, in order to perform accurate body fat measurement, it is necessary to remove beforehand such an error (unnecessary fat region) so as not to be included in the calculation.
[0029]
FIG. 13 shows the body in an image diagnostic apparatus that calculates the average and standard deviation of CT values for each patient, determines the threshold range of the fat region, further removes the unnecessary fat region, and accurately measures the body fat. It is a figure which shows the flowchart of a fat measurement process. FIG. 14 is a diagram showing to which region on the CT image the process of FIG. 13 is applied. The processing contents will be described below according to this flowchart.
[Step S300]
As in step S40 described above, the patient ID input screen is displayed on the display 34 of the diagnostic imaging apparatus, so the operator inputs the patient ID number. As a result, a tomographic image as shown in FIG. 13 corresponding to the ID number of the patient to be diagnosed is read from the magnetic disk 32 and displayed on the display 34. In this step, a fat pixel threshold range reading process for reading a threshold range for extracting fat pixels from the initial setting file together with the display of the tomographic image is executed. This threshold range is a temporary threshold range used in later steps S303 and S304. Therefore, the true threshold range is obtained in step S307 of the subsequent process. The temporary threshold range may be recorded in the initial setting file as an initial value from −150 to −50 so that the user can appropriately rewrite it.
[0030]
[Step S301]
Similar to step S40 described above, body surface contour extraction processing is performed to extract the body surface contour and extract the body surface region of interest 141.
[Step S302]
Similarly to the above-described step S41, a visceral contour extraction process for extracting a region of interest (internal organs of interest) 142 surrounding the internal organs is executed. The internal region of interest 142 is indicated by a white dotted line in FIG.
[Step S303]
Using the temporary threshold range set in step S300 from the body surface region of interest 141 obtained in step S301, the whole fat pixel extraction process is performed to extract the entire fat pixels. The extracted result is output as a binarized image (whole fat mask image) in which the fat pixel is “1” and the other pixels are “0”.
[Step S304]
Using the temporary threshold range set in step S300 from the visceral region of interest 142 obtained in step S302, visceral fat pixel extraction for extracting fat pixels (visceral fat mask image) 143 in the visceral region is executed. The extracted result is output as a binarized image (visceral fat mask image) in which fat pixels are “1” and other pixels are “0”.
[0031]
[Step S305]
Subcutaneous fat region (subcutaneous fat mask image) 144 is extracted by subtracting the visceral fat region (visceral fat mask image) extracted in step S304 from the whole fat region (whole fat mask image) extracted in step S303. A fat pixel extraction process is executed.
[Step S306]
The average of the subcutaneous fat region pixel values for obtaining the mean value (MN) and standard deviation (SD) of the pixel values of the subcutaneous fat region based on the original image corresponding to the subcutaneous fat region (subcutaneous fat mask image) extracted in step S305 Execute value and standard deviation calculation processing.
[Step S307]
A fat threshold range calculation process is performed to calculate a fat pixel threshold range based on the average value (MN) and standard deviation (SD) of the pixel values of the subcutaneous fat region obtained in step S306. This calculation process is executed according to the following equation.
[0032]
Fat threshold range = MN ± nSD (where n = 1, 2, 3)
n is a value that can be selected by the operator. The default is n = 2.
[0033]
[Step S308]
The same process as S303 is repeated using the fat threshold range obtained in step S307. That is, the whole fat pixel re-extraction process for re-extracting the fat pixels inside the body surface region of interest is executed. The re-extracted fat pixel is output as a re-extracted whole fat mask image.
[Step S309]
Using the fat threshold range obtained in step S307, the same processing as in step S304 is repeated. That is, visceral fat pixel re-extraction processing for re-extracting fat pixels inside the visceral region of interest is executed. The re-extracted fat pixel is output as a re-extracted visceral fat mask image.
[Step S310]
Subcutaneous fat pixel re-extraction processing is performed to extract a subcutaneous fat image by taking the difference between the re-extracted whole fat mask image extracted in step S308 and the re-extracted visceral fat mask image extracted in step S309. The re-extracted subcutaneous fat pixel is output as a re-extracted subcutaneous fat mask image.
[0034]
[Step S311]
An unnecessary fat pixel removal process for deleting unnecessary regions from the re-extracted visceral fat mask image extracted in step S309 and the re-extracted subcutaneous fat mask image extracted in step S310 is executed. It may be possible to select whether or not the operator performs this process. FIG. 15 is a diagram showing details of the unnecessary fat pixel removal processing in step S311. In this process, the same process is performed on the visceral fat mask image and the subcutaneous fat mask image.
[Step S400]
A process of labeling each of the binarized visceral fat mask image and the subcutaneous fat mask image is executed. Here, labeling is a process of assigning a different number for each continuous area in a binarized image.
[Step S401]
An area calculation process for calculating an area (or the number of pixels) for each continuous region labeled in step S400 is executed.
[Step S402]
Unnecessary area deletion processing for deleting an area having an area smaller than a preset threshold from the labeling image is executed. Through this series of processing, minute fat regions scattered in white in each visceral fat mask image and subcutaneous fat mask image are deleted as unnecessary fat regions from each mask image. Note that as a post-processing, a labeled image may be returned to the original binary image.
[0035]
[Step S312]
A fat pixel area ratio calculation process for calculating the following parameters based on the re-extracted visceral fat mask image and the re-extracted subcutaneous fat mask image from which unnecessary fat regions have been deleted in step S311 is executed. The parameters to be calculated are visceral fat: total fat ratio (total fat = visceral fat + subcutaneous fat), subcutaneous fat: total fat ratio, and visceral fat: subcutaneous fat ratio.
[Step S313]
A calculation result display process is executed in which the parameter calculated in step S312 is displayed as a numerical value, or the visceral fat region and the subcutaneous fat region are filled with different colors and displayed over the original image.
[0036]
With the above processing, the threshold range of fat pixels that differ depending on the patient can be easily calculated, and unnecessary areas in each fat image can be automatically deleted without the operator having to delete them one by one. This improves the performance and improves the body fat measurement accuracy.
[0037]
【The invention's effect】
As described above, according to the diagnostic imaging apparatus of the present invention, it is possible to improve the body fat measurement accuracy by determining the threshold range of the fat region of the patient based on the CT value of the fat region that is different for each patient. effective. Moreover, according to the diagnostic imaging apparatus of the present invention, there is an effect that the body fat measurement accuracy can be improved in consideration of unnecessary minute fat regions around the abdominal muscles and the spinal cord.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining a conventional method for measuring body fat based on a tomographic image.
FIG. 2 is another diagram for explaining a conventional method of measuring body fat based on a tomographic image.
FIG. 3 is a block diagram showing a hardware configuration of the entire diagnostic imaging apparatus to which the present invention is applied.
4 is a diagram illustrating a main flow executed by the diagnostic imaging apparatus of FIG. 3;
FIG. 5 is a diagram showing details of body surface contour extraction processing in the main flow of FIG. 4;
FIG. 6 is a diagram showing details of a visceral contour extraction process in the main flow of FIG. 4;
7 is a diagram showing an example of a display screen showing how tomographic images are displayed on the display according to the main flow of FIG. 4;
FIG. 8 is another diagram showing an example of a display screen showing how tomographic images are displayed on the display by the main flow of FIG. 4;
9 is an enlarged view showing the vicinity of the start point in FIG. 8 in order to explain the contour line tracking process in FIG. 6;
FIG. 10 is a diagram showing a display example of how total fat, visceral fat and subcutaneous fat are displayed by the main flow of FIG.
FIG. 11 is a diagram for explaining a contour tracking correction process by a rolling ball method;
FIG. 12 is an example of a tomographic image showing a specific example of minute fat regions (unnecessary fat regions) scattered in white.
FIG. 13 shows a body in an image diagnostic apparatus in which an average and standard deviation of CT values is obtained for each patient, a threshold range of fat regions is determined, and unnecessary fat regions are also removed to accurately measure body fat. It is a figure which shows the flowchart of a fat measurement process.
FIG. 14 is a diagram showing to which region on the CT image the process of FIG. 13 is applied;
FIG. 15 is a diagram showing details of unnecessary fat pixel removal processing in step S311 of FIG. 13;
[Explanation of symbols]
11 ... Spine standing muscle
12, 13 ... Horizontal abdominal muscles
14 ... Rectus abdominis
72 ... Body surface area of interest
80: Internal organs of interest
91, 92, 93, 95 ... small circle
81, 82, 85 ... Points of interest
83,86 ... Recent
30 ... Central processing unit (CPU)
31 ... Main memory
32 ... Magnetic disk
33 ... Display memory
34 ... CRT display
35 ... Mouse
36 ... Controller
37 ... Keyboard
38 ... Speaker
39 ... Common bus

Claims (4)

A body surface region-of-interest extracting means for extracting a body surface region of interest surrounding the entire living body based on the contour of the body surface in the tomographic image relating to the biological information;
A whole fat region extracting means for extracting a whole fat region in the body surface region of interest;
A visceral region of interest extracting means for extracting a visceral region of interest surrounding the internal organs by tracing the outline of the abdominal wall muscle layer inside the subcutaneous fat layer in the tomographic image based on the pixel value of the tomographic image;
Visceral fat region extracting means for extracting a visceral fat region in the visceral region of interest;
Mean value and standard deviation calculating means for calculating an average value and standard deviation of pixel values of fat regions from the body surface region of interest and the visceral region of interest;
A fat threshold range calculation means for calculating an image-specific fat threshold range from the fat pixel average value and the standard deviation;
A fat region re-extracting unit that extracts the whole fat region and the visceral fat region again based on the fat threshold range unique to the image obtained by the fat threshold range calculating unit and the body surface region of interest and the visceral region of interest;
Unnecessary fat region deletion means for deleting unnecessary fat regions from the whole fat region extracted by the fat region re-extraction means;
In an image diagnostic apparatus comprising: body fat calculation means for calculating body fat based on a whole fat region and a visceral fat region from which unnecessary regions have been deleted by the unnecessary fat region deletion unit ,
The visceral region-of-interest extraction means tracks the outline of the abdominal wall muscle layer while removing pixels of the whole fat region in the tomographic image and filling a gap in the abdominal wall muscle layer region obtained by binarization processing An image diagnostic apparatus characterized by that. 2. The diagnostic imaging apparatus according to claim 1 , wherein the body fat calculating means calculates at least one of visceral fat: total fat value, subcutaneous fat: total fat value, and visceral fat: subcutaneous fat value. A diagnostic imaging apparatus. 3. The diagnostic imaging apparatus according to claim 1 , further comprising display means for displaying a value indicating the body fat calculated by the body fat calculating means in the tomographic image. 4. The diagnostic imaging apparatus according to claim 3 , wherein the display unit displays whole fat, visceral fat and subcutaneous fat on a tomographic image so as to be identifiable.

Images