WO2013008449A1

WO2013008449A1 - Fat-checking method, fat-checking device, and fat-checking program

Info

Publication number: WO2013008449A1
Application number: PCT/JP2012/004434
Authority: WO
Inventors: 平野　雅嗣; 純三浦; 克人山崎
Original assignee: 国立大学法人豊橋技術科学大学
Priority date: 2011-07-08
Filing date: 2012-07-09
Publication date: 2013-01-17
Also published as: JPWO2013008449A1

Abstract

Provided are a fat-checking method, fat-checking device, and fat-checking program whereby fat can be determined with high accuracy within a tomographic image and whereby a trend toward obesity and a trend toward fatty liver can be checked. The fat-checking method comprises: an image acquisition step for acquiring a tomographic image of abdominal tissue in a medical tomographic image device; a discrimination step for discriminating between a subcutaneous fat layer, muscle layer, and visceral fat layer using CT values in the tomographic image; a distribution acquisition step for acquiring a correlation distribution of the CT values and the number of voxels in the subcutaneous fat layer and visceral fat layer thus discriminated between; and a determination step for making a determination with respect to a trend toward obesity, using the correlation distribution of the CT values. According to the present invention, in addition to the volume and proportion of the subcutaneous fat layer and visceral fat layer in the abdominal tissue, the quality of fat can be determined and liver fat can be determined.

Description

Fat inspection method, fat inspection device, and fat inspection program

The present invention relates to a fat inspection method, a fat inspection apparatus, and a fat inspection program for performing image processing of medical tomographic images and inspecting obesity constitution and fatty liver disease.

Fat is roughly classified into subcutaneous fat formed under the skin and visceral fat formed around the viscera. Of these, visceral fat is attracting attention in the diagnosis of lifestyle-related diseases and the like. Therefore, it is required not only to measure the amount of body fat that reflects the total amount of fat, but also to be able to measure the amount of fat by discriminating the types of subcutaneous fat and visceral fat.
Under such circumstances, a technique for discriminating fat in a tomographic image obtained from a medical tomographic image apparatus such as an X-ray CT apparatus is known (see Patent Document 1). According to the technical content disclosed in Patent Document 1, subcutaneous fat and visceral fat can be discriminated in a tomographic image of an X-ray CT apparatus.

On the other hand, with respect to the measurement of the amount of fat in the liver, a technique is disclosed in which an ultrasonic image is used to determine which tissue property corresponds to normal liver, fatty liver, or cirrhosis (for example, Patent Document 2). See).

JP 2003-339694 A JP 2005-110833 A

However, in the technique disclosed in Patent Document 1, the muscle pixel region representing the muscle layer (peritoneum) existing between the subcutaneous fat layer and the visceral fat layer is expanded, and the fat pixel region corresponding to the visceral fat is obtained. Processing is performed to automatically correct the tomographic image so that it completely surrounds. This is done because the CT value of the muscle layer (peritoneum) declines and there are cases where it cannot be distinguished from fat, but such cases appear locally in the muscle layer, The technique of Patent Document 1 has a problem that the muscle layer (peritoneum) is uniformly expanded.

Further, in the technique of Patent Document 1, a range of pixel values (CT values) to be determined as fat pixels and a range of pixel values to be determined as muscle pixels are set. Compared with each of these ranges, The fat layer is discriminated. Each range is set not to overlap each other. Therefore, when the CT value of the muscle layer locally decreases and overlaps with the CT value of the fat layer, there is a problem that the muscle layer is erroneously recognized as a fat layer.

Further, conventionally, for a plurality of tomographic images acquired by abdominal CT, the subcutaneous fat area and visceral fat area of each tomographic image are multiplied by the thickness of each layer, and the multiplication result is summed for all the layers, The total subcutaneous fat volume and visceral fat volume are obtained, or the obtained subcutaneous fat weight and visceral fat weight are calculated by multiplying the obtained volume by the specific gravity of fat. However, it is not known that an obesity tendency can be discriminated from the distribution of CT values of subcutaneous fat and visceral fat throughout the abdomen.

In view of such a situation, an object of the present invention is to provide a fat inspection method, a fat inspection apparatus, and a fat inspection program capable of determining fat with high accuracy in a tomographic image.
Moreover, this invention provides what can test | inspect an obesity tendency and a fatty liver tendency.

The inventors of the present invention have completed the present invention as a result of intensive research and development while aiming at highly accurate fat discrimination.
In order to achieve the above object, a fat inspection method of the present invention includes an image acquisition step of acquiring a tomographic image of an abdominal tissue in a medical tomographic image apparatus, and a subcutaneous fat layer, a muscle layer, and a visceral fat using a CT value of the tomographic image. A discrimination step for discriminating layers, a distribution obtaining step for obtaining a correlation distribution between the number of voxels in the discriminated subcutaneous fat layer and visceral fat layer and the CT value, and a discrimination step for discriminating an obesity tendency using the correlation distribution of the CT value Prepare.

According to the fat inspection method of the present invention, it is possible to measure the volume instead of the conventional area measurement, and to show the visceral fat amount and its distribution more easily to the subject. Specifically, in a tomographic image obtained from an X-ray CT apparatus that images the human body, the subcutaneous fat and the visceral fat are separated on the assumption that a muscle layer always exists between the subcutaneous fat layer and the visceral fat layer. And measure.
In general, it is known that the CT value (HU: Hounsfield Unit) of fat is in the range of −200 to −10 (negative value), and the CT value (H.U. .) Is known to be positive. By assuming that a muscle layer always exists between the subcutaneous fat layer and the visceral fat layer, the subcutaneous fat and the visceral fat are separated and measured.

Here, the discrimination step in the fat test method described above compares the measurement value of the thickness information of the subcutaneous fat layer of the previous line with the storage step of storing the thickness information of the subcutaneous fat layer of the previous line, When the difference is larger than a predetermined threshold, a change step for temporarily changing a CT value as a boundary value between the muscle layer and the subcutaneous fat layer, and a CT value of the changed boundary value between the muscle layer and the subcutaneous fat layer are used. And calculating the thickness information of the subcutaneous fat layer of the current line.

The CT value of the muscle layer is usually a positive value, but there are cases where the CT value of the muscle layer decreases and cannot be distinguished from fat. These cases occur locally rather than throughout the muscle layer. In the region where the CT value of the muscle layer is lowered, the CT value becomes a negative value, so that it cannot be distinguished from fat.
In order to deal with such a case, when the thickness information of the subcutaneous fat layer of the immediately preceding line is stored, the measured value of the thickness information of the subcutaneous fat layer of the previous line is compared, and the difference is larger than a predetermined threshold In addition, the CT value as the boundary value between the muscle layer and the subcutaneous fat layer was temporarily changed.
This presupposes that the thickness of the subcutaneous fat layer does not change rapidly. When the difference in the thickness of the subcutaneous fat layer between the current line and the immediately preceding line is larger than a predetermined threshold, the CT value as the boundary value between the muscle layer and the subcutaneous fat layer is temporarily reduced. Then, the thickness of the subcutaneous fat layer in the current line is measured again. The CT value as the boundary value between the muscle layer and the subcutaneous fat layer is decreased until the difference falls within the predetermined threshold.

For example, assume that the CT value as the boundary value between the muscle layer and the subcutaneous fat layer is −10, and the predetermined threshold is 7 pixels. If the thickness of the subcutaneous fat layer in the immediately preceding line is 5 pixels, but the thickness of the subcutaneous fat layer measured this time is 15 pixels, the thickness difference is 10 pixels. In this case, the CT value as the boundary value between the muscle layer and the subcutaneous fat layer is changed from -10 to -15, for example, and the thickness of the subcutaneous fat layer is measured. If the thickness of the subcutaneous fat layer is 10 pixels in the measurement again, the difference in the thickness of the subcutaneous fat layer between the current line and the previous line is within a predetermined threshold (within 7 pixels), so the current subcutaneous fat layer Measurement of the thickness of is finished. If the thickness of the subcutaneous fat layer is 13 pixels in the re-measurement, the difference in the thickness of the subcutaneous fat layer between the current line and the previous line is 8 pixels (13 pixels-5 pixels), and is within a predetermined threshold. Since it does not fall within (within 7 pixels), the CT value as the boundary value between the muscular layer and the subcutaneous fat layer is changed again, and the thickness is reduced from -15 to -20, for example, and the thickness of the subcutaneous fat layer is measured. In this way, rapid fluctuations in the thickness of the subcutaneous fat layer are suppressed.

Further, in the discrimination step in the fat inspection method described above, when a peak appears in the CT value of the subcutaneous fat layer, it is discriminated as an obesity tendency.
The inventors investigate the change in the peak value of the correlation distribution between the number of voxels in the subcutaneous fat layer and the CT value and the change in the peak value of the correlation distribution between the number of voxels in the visceral fat layer and the CT value from the data of many subjects. Thus, it was found that it is possible to detect at an early stage that the subject's obesity constitution and the subject are obese.

That is, from the CT-voxel number correlation distribution graph of the subcutaneous fat layer and the visceral fat layer, the constitutions such as those who are likely to become fat or those who are less likely to gain weight (skin) are known.
Further, from various data statistics, it was found that the peak of the correlation distribution of CT value of the subcutaneous fat layer-voxel number reached -120 faster than the peak of the visceral fat layer. Here, −120 is the average CT value (HU) of fat. Even if the person is thin, it can be inferred that the visceral fat layer starts to accumulate in the near future if the peak of the subcutaneous fat layer is close to −120. This makes it possible to quickly change the lifestyle habits of people called hidden obesity. That is, from the measurement of the correlation distribution graph of the CT value of the subcutaneous fat layer and the visceral fat layer-the number of voxels, when there is a subject whose subcutaneous fat layer is close to the peak of -120, the visceral organ is more or less visceral. An announcement can be made when the fat layer is about to accumulate.

Further, in the discrimination step of the fat test method, from the peak value of the CT value of the subcutaneous fat layer, the content of oleic acid that is an unsaturated fatty acid and palmitic acid or stearic acid that is a long chain fatty acid is estimated, It is determined whether adipose tissue is easy to burn.
When fatty acids are stored in the fat cells of the subcutaneous fat layer, the fatty acids first become 16 palmitic acids of carbon. Thereafter, two more carbons are added to the fatty acid to become stearic acid. Palmitic acid and stearic acid are long chain fatty acids. Stearic acid turns one carbon bond into an unsaturated bond and becomes oleic acid, which is an unsaturated fatty acid.

Here, comparing the properties of palmitic acid, stearic acid, and oleic acid, palmitic acid has a density of 0.853 (g / cm ³ ), a CT value of −147, a melting point of 62.9 ° C., and stearic acid. Has a density of 0.847 (g / cm ³ ), a CT value of −153, a melting point of 69.6 ° C., and oleic acid has a density of 0.89 (g / cm ³ ) and a CT value of −110. Melting point: 16.3 ° C. That is, when oleic acid is converted from palmitic acid or stearic acid, the melting point is greatly reduced from 60 ° C. or higher to 16.3 ° C. and becomes liquid at room temperature. Oleic acid has a low melting point and can be said to be easier to fuel in the body than palmitic acid or stearic acid.

The CT value of oleic acid is -110, whereas the CT values of palmitic acid and stearic acid are -147 and -153, and the difference value of the CT value is large. Focusing on this, the greater the oleic acid content, the closer the CT value peak in the adipose tissue based on histogram analysis approaches the oleic acid CT value of −110. For example, lard, which is pork fat, and head, which is bovine fat, contain nearly 50% of oleic acid in the total fat, and the CT value is -130.
That is, the content of oleic acid versus palmitic acid or stearic acid can be estimated by measuring the peak of the CT value by histogram analysis in the adipose tissue.

In the determination step, when the CT value peak of the subcutaneous fat layer is lower than a predetermined threshold, it is determined that the content of palmitic acid or stearic acid, which is a long chain fatty acid, is high and the fatty tissue is difficult to burn. If it is high, it is determined that the content of oleic acid, which is an unsaturated fatty acid, is high and the adipose tissue is likely to burn. Here, the predetermined threshold is in the vicinity of -130, which is an intermediate value between the CT value of oleic acid -110 and the CT value of palmitic acid or stearic acid (-147, -153), or a value of -125 to -135. Set.

The fat inspection method includes a liver region extraction step of extracting a liver region from a three-dimensional region obtained by superimposing the tomographic images obtained in the image acquisition step, and the number of voxels and CT values of the extracted liver region. A liver region distribution acquisition step for acquiring a correlation distribution and a fatty liver determination step for determining fatty liver using the correlation distribution of CT values are further provided.
By further including a liver region extraction step, a liver region distribution acquisition step, and a fatty liver discrimination step, it is possible to estimate the amount of fat in the liver based on the degree of decrease in CT value and discriminate fatty liver. .

In the fatty liver discrimination step, the fat percentage of the extracted liver region is calculated from the average value of the CT values of the extracted liver region, the average value of the CT values of the healthy subject's liver region, and the average value of the CT value of fat. Is calculated.
For example, after extracting the liver region, if the average CT value of the region is usually 60 (H.U.), for example, if it has decreased to 42 (H.U.), Assuming that the average CT value is −120 (HU), the fat percentage is 10% based on the calculation of (60−42) / (60 − (− 120)) = 18/180 = 10. You can understand that.

Next, the fat inspection apparatus of the present invention discriminates a subcutaneous fat layer, a muscle layer, and a visceral fat layer using an image acquisition unit that acquires a tomographic image of an abdominal tissue in a medical tomographic image apparatus and a CT value of the tomographic image. A configuration including a discrimination unit, a distribution acquisition unit that acquires a correlation distribution between the number of voxels in the discriminated subcutaneous fat layer and visceral fat layer and the CT value, and a determination unit that discriminates an obesity tendency using the correlation distribution of the CT value; Is done.
According to the fat inspection apparatus of the present invention, it is possible to measure the volume, not the conventional area measurement, and show the visceral fat amount and distribution thereof more easily to the subject. Similar to the above-described fat examination method, in the tomographic image obtained from the X-ray CT apparatus for imaging the human body, it is assumed that there is always a muscle layer between the subcutaneous fat layer and the visceral fat layer. Separate and measure visceral fat.

In addition, the discrimination unit in the above fat test apparatus compares the measured value of the thickness information of the subcutaneous fat layer of the previous line with the storage unit that stores the thickness information of the subcutaneous fat layer of the previous line, and the difference Is larger than a predetermined threshold, using a change unit that temporarily changes the CT value as the boundary value between the muscle layer and the subcutaneous fat layer, and the CT value of the boundary value between the changed muscle layer and the subcutaneous fat layer, It is preferable to include a calculation unit that calculates thickness information of the subcutaneous fat layer of the current line.
Similar to the fat test method described above, on the assumption that the thickness of the subcutaneous fat layer does not change abruptly, if the difference in the thickness of the subcutaneous fat layer between the current line and the previous line is greater than a predetermined threshold, the muscle layer The CT value as the boundary value between the skin layer and the subcutaneous fat layer is temporarily reduced. Then, after the change, the thickness of the subcutaneous fat layer in the current line is measured again, and the CT value as the boundary value between the muscle layer and the subcutaneous fat layer is decreased until the difference falls within a predetermined threshold.

Further, the discrimination unit in the fat testing apparatus discriminates an obesity tendency when a peak appears in the CT value of the subcutaneous fat layer.
By examining the change in the peak value of the correlation distribution between the number of subcutaneous fat layers and the CT value, and the change in the peak value of the correlation distribution between the number of voxels in the visceral fat layer and the CT value, Determine that there is.

Further, in the discrimination unit of the fat test apparatus, from the CT value of the subcutaneous fat layer, the content of oleic acid that is an unsaturated fatty acid and palmitic acid or stearic acid that is a long-chain fatty acid is estimated, It is determined whether adipose tissue is easy to burn.
In the determination unit, when the CT value peak of the subcutaneous fat layer is lower than a predetermined threshold, it is determined that the content of palmitic acid or stearic acid, which is a long chain fatty acid, is high and the fatty tissue is difficult to burn, If it is high, it is determined that the content of oleic acid, which is an unsaturated fatty acid, is high and the adipose tissue is likely to burn.

Further, the fat inspection apparatus includes a liver region extraction unit that extracts a liver region from a three-dimensional region obtained by superimposing the obtained tomographic images in the image acquisition unit, and the number of voxels and CT values of the extracted liver region. A liver region distribution acquisition unit that acquires the correlation distribution of the liver and a fatty liver determination unit that determines fatty liver using the correlation distribution of CT values.
By further including a liver region extraction unit, a liver region distribution acquisition unit, and a fatty liver determination unit, the fatty amount in the liver can be estimated from the degree of decrease in CT value for fatty liver, and fatty liver can be determined.

In the fatty liver discriminating unit, the fat percentage of the extracted liver region is calculated from the average value of the CT values of the extracted liver region, the average value of the CT values of the healthy subject's liver region, and the average value of the CT value of fat. Is calculated.

The fat test program of the present invention is a program for causing a computer to execute each step in the above-described fat test method.

According to the present invention, there is an effect that an obesity constitution or a fatty liver disease can be examined. According to the present invention, even if the CT value of the muscle layer is low and the fat layer may be recognized, it can be dealt with by changing the threshold in consideration of the thickness of the surrounding subcutaneous fat. Furthermore, according to the present invention, in addition to the volume and ratio of the subcutaneous fat layer and the visceral fat layer of the abdominal tissue, there is an effect that it is possible to determine fat quality and liver fat.

As for fatty liver, CT imaging is frequently performed in recent years, but for example, it is also possible to test for fatty liver disease by measuring fat mass as an optional test for screening.

Abdominal fault schematic diagram Schematic configuration diagram of the fat test apparatus of the present invention Schematic process flow diagram of a fat testing method of one embodiment of the present invention Process flow diagram of fat test method of embodiment 1 Schematic process flow diagram of a fat testing method of another embodiment of the present invention Illustration of how to calculate the fat percentage of the extracted liver region Functional block diagram of the fat test apparatus according to the second embodiment Functional block diagram of the discriminating part of the fat test apparatus of Embodiment 2 Correlation distribution chart of voxel number and CT value CT value distribution map of subcutaneous fat and visceral fat (in the case of obesity) CT value distribution map of subcutaneous fat and visceral fat (normal case) CT value distribution map of subcutaneous fat and visceral fat (in case of skinnyness) Distribution chart of voxel number and CT value in visceral fat Distribution of voxel number and CT value in subcutaneous fat CT value distribution map of subcutaneous fat and visceral fat (BMI: 18.7) CT value distribution map of subcutaneous fat and internal fat (BMI: 21.2) CT value distribution map of subcutaneous fat and internal fat (BMI: 32.3) Results of single regression analysis of the ratio of fat mass to fat mass and liver volume and each test item Internal configuration diagram of computer hardware

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The scope of the present invention is not limited to the following examples and illustrated examples, and many changes and modifications can be made.

Body mass index (BMI) is a value calculated by weight (kg) / (height (m) × height (m)), and is widely used as an index for knowing the degree of obesity. The criteria for determining a BMI value are generally less than 18.5, “lossy”, 18.5 and less than 25, “standard”,
It is determined that “obesity” is 25 or more and less than 30 and “high obesity” is 30 or more. A body fat percentage value indicating what percentage of body weight is fat is also useful for examining obesity trends.
Fats that cause obesity include subcutaneous fat on the subcutaneous tissue and visceral fat on the visceral tissue. Of these, visceral fat is said to be a major factor in lifestyle-related diseases, and therefore, a method capable of discriminating and measuring subcutaneous fat and visceral fat is required instead of simply measuring body fat mass.

Fig. 1 shows a tomographic schematic diagram of the abdomen of a subject such as a mammal such as a human or a bird. As schematically shown in FIG. 1, there is a subcutaneous fat layer 12 inside the outermost skin 10, a muscle layer 14 inside, and a visceral fat layer 16, a visceral tissue 18, a bone inside the muscle layer 14. 19 is present. In the case of X-ray CT, the muscle layer and the skin show positive CT values, and are negative CT values of the subcutaneous fat layer and the visceral fat layer, which are basically clearly distinguishable.

FIG. 2 shows a schematic configuration diagram of the fat test apparatus of the present invention. The medical tomographic image apparatus 20 is an apparatus that performs X-ray CT imaging of a subject. An existing X-ray CT apparatus can be used. CT tomographic image data obtained by the medical tomographic image apparatus 20 is stored in the tomographic image memory 22. The CT tomographic image data stored in the tomographic image memory 22 from the medical tomographic image apparatus 20 may be a map of CT values at each point in the tomographic region, or a processed image after processing the CT values.

The CT tomographic image data stored in the tomographic image memory 22 is processed by the arithmetic processing unit 24 to generate correlation distribution data 26 between the number of voxels and the CT value. Based on the correlation distribution data 26, an obesity tendency, a fatty liver disease tendency, and the like are determined. The discrimination result can be displayed by the discrimination result display unit 28 as a slice image or a three-dimensional image of the subcutaneous fat layer and the visceral fat layer together with physical quantities (area, volume, mass, etc.) of the subcutaneous fat and the visceral fat.

FIG. 3 shows a schematic processing flow of the fat test method according to the embodiment of the present invention. As shown in FIG. 3, the fat inspection method according to an embodiment of the present invention includes an image acquisition step (S10) for acquiring a tomographic image of an abdominal tissue in a medical tomographic image apparatus, and subcutaneous fat using a CT value of the tomographic image. A discrimination step (S12) for discriminating layers, muscle layers and visceral fat layers; a distribution acquisition step (S14) for obtaining a correlation distribution between the voxel numbers of the discriminated subcutaneous fat layers and visceral fat layers and CT values; A discrimination step (S16) for discriminating an obesity tendency using the correlation distribution is provided.
In addition, the fat test program according to the embodiment of the present invention causes the computer to execute the same steps (S10 to S16) as the flow of the fat test method of FIG.

Here, in the discrimination step (S12), it is assumed that the CT value (HU) of the fat layer of the subcutaneous fat layer and the visceral fat layer is in the range of −200 to −10 (negative value). The CT value (H.U.) of No. 1 is in the range of 0 (corresponding to the CT value of water) to a positive value, and each is separated.
Then, by assuming that a muscle layer always exists between the subcutaneous fat layer and the visceral fat layer, the subcutaneous fat and the visceral fat are separated and measured.

In the distribution acquisition step (S14), not only the area and volume of subcutaneous fat and visceral fat are measured, but also the distribution of CT values is measured. That is, as before, the subcutaneous fat area and visceral fat area of each tomographic image are multiplied by the thickness of each layer, and the multiplication result is summed for all layers to obtain the subcutaneous fat volume and visceral fat in the entire abdomen. The volume is calculated, but at the same time, the distribution of CT values is also measured. The CT value of each section (voxel) is different in the subcutaneous fat layer and the visceral fat layer. The number of voxels is counted for each CT value, and correlation distribution data between the number of voxels and the CT value is acquired.
Then, in the discrimination step (S16) for discriminating the obesity tendency, depending on whether or not there is a peak value in the correlation distribution between the number of voxels and the CT value, in particular, whether or not a peak has appeared in the CT value of the subcutaneous fat layer, It is determined whether or not the subject is obese.

FIG. 5 shows a schematic processing flow of a fat test method according to another embodiment of the present invention.
The fat inspection method shown in FIG. 5 includes an image acquisition step (S30) for acquiring a tomographic image of an abdominal tissue in a medical tomographic image apparatus, and a liver for extracting a liver region from a three-dimensional region obtained by superimposing the obtained tomographic images. Region extraction step (S32), liver region distribution acquisition step (S34) for acquiring the correlation distribution between the number of voxels in the extracted liver region and the CT value, and liver fat determination for determining fatty liver using the CT value correlation distribution Step (S36) is provided. According to the fat test method of the present embodiment, for fatty liver, the amount of fat in the liver can be estimated from the degree of decrease in CT value, and fatty liver can be determined.

In the liver region extraction step (S32), the liver tissue of each tomographic image is multiplied by the thickness of each layer, and the multiplication result is summed for all layers to specify the three-dimensional region of the liver tissue. In the liver fat discrimination step (S36), the average CT value of the liver region of a healthy person is usually 60 (HU), whereas the average CT value of the liver region of fatty liver disease is a low value. Become. Fatty liver generally refers to a fat percentage of 30% or more in the entire liver region.

If the average CT value of the measured liver region is reduced to 42 (H.U.), the average CT value of fat is -120 (H.U.). That is, the average CT value of a normal liver region is usually 60 (HU), indicating that the fat percentage of the subject's liver region is 10%. In this case, since the fat percentage of the liver region is 10%, it can be determined that there is a fatty liver tendency although it is not a fatty liver disease.

(Fat test method for discrimination of obesity tendency)
FIG. 4 shows a processing flow of the fat inspection method of the first embodiment.
In the fat inspection method of the first embodiment, an image acquisition step (S10) for acquiring a tomographic image of an abdominal tissue in a medical tomographic image apparatus, and a subcutaneous fat layer, a muscle layer, and a visceral fat layer are discriminated using CT values of the tomographic image. Discriminating step (S120), storing step (S122) for storing the thickness information of the subcutaneous fat layer of the immediately preceding line, and comparing step (S124) for comparing the measured values of the thickness information of the subcutaneous fat layer of the current and immediately preceding line ) And the comparison step (S124), and if the difference is larger than a predetermined threshold, the change step (S126) for temporarily changing the CT value as the boundary value between the muscle layer and the subcutaneous fat layer, A calculation step (S128) for calculating thickness information of the subcutaneous fat layer of the current line using the CT value of the boundary value between the changed muscle layer and subcutaneous fat layer, and a comparison step (S 24) a distribution acquisition step (S14) for acquiring a correlation distribution between the number of voxels in the subcutaneous fat layer and the visceral fat layer and the CT value when the difference falls within a predetermined threshold, and a correlation distribution of the CT value And determining step (S16) for determining obesity tendency.

Locally, there are cases in which the CT value of the muscle layer decreases and cannot be distinguished from fat. In order to deal with such a case, when the thickness information of the subcutaneous fat layer of the immediately preceding line is stored, the measured value of the thickness information of the subcutaneous fat layer of the previous line is compared, and the difference is larger than a predetermined threshold In addition, the CT value as the boundary value between the muscle layer and the subcutaneous fat layer is temporarily changed. Considering the thickness of surrounding subcutaneous fat, the problem of misrecognizing a muscle layer as a fat layer is solved by changing the threshold value. For example, if the thickness of the subcutaneous fat in the immediately preceding line is 5 pixels, but the current thickness is 15 pixels, the boundary value between muscle and fat is changed to suppress the fluctuation.

Next, in the discrimination step in the fat test method of the first embodiment, a method for discriminating an easily fattened constitution and a less faty constitution using a CT-voxel number correlation distribution graph of the subcutaneous fat layer and the visceral fat layer will be described. .
FIG. 9 shows a correlation distribution diagram between the number of voxels in the subcutaneous fat layer and the visceral fat layer and the CT value. The X axis in the figure represents the CT value, and the Y axis represents the number of voxels, that is, the number of fats. FIG. 9 (1) shows the form of a correlation distribution graph of CT value-voxel number of the general subcutaneous fat layer and visceral fat layer of obese people. FIG. 9 (2) shows the form of a correlation distribution graph of CT value-number of voxels of the general subcutaneous fat layer and visceral fat layer of the lean side. In the figure, A indicates the subcutaneous fat layer, and B indicates the visceral fat layer. As shown in FIG. 9 (1), the correlation distribution diagram between the number of voxels and the CT value for obese people basically has a shape in which a CT value peak exists around -100 to -120. On the other hand, as shown in FIG. 9 (2), the correlation distribution diagram between the number of voxels and the CT value of the thinner one basically exhibits a shape in which no peak of the CT value exists.

Explain with specific data. 10 to 12 are distribution diagrams of CT values of subcutaneous fat and visceral fat. FIG. 10 shows distribution data when obesity is present, FIG. 11 is normal, and FIG. 12 is thin. Since the data are centered on visceral fat and subcutaneous fat, the X-axis is -200 to 0. As shown in FIG. 10, in the case of obesity, the peak of CT value appears in both subcutaneous fat and visceral fat. On the other hand, as shown in FIG. 12, in the case of skinnyness, no peak of CT value is observed in both subcutaneous fat and visceral fat. In addition, as shown in FIG. 11, in the normal case, a peak clearly appears in the CT value of subcutaneous fat, but it can be seen that a peak is being formed in visceral fat.

FIGS. 13 and 14 show composite graphs in which the Y-axis scales of the three graphs of FIGS. 10 to 12 are aligned. FIG. 13 is a distribution graph of voxel number and CT value in visceral fat, and FIG. 14 is a distribution graph of voxel number and CT value in subcutaneous fat.
In the graphs of FIGS. 13 and 14, in the case of skinnyness, no peak of the CT value is seen in both graphs, and it can be seen that the number of voxels is counted at the boundary of the CT value of about −100. Further, in the graph of FIG. 13, in the normal case, a peak of the CT value is seen in the vicinity of −74, and it can be seen that the number of voxels is increased as compared with the case of the skinnyness. Further, in the graph of FIG. 14, in the normal case, a peak of the CT value is seen in the vicinity of −92, and it can be seen that the number of voxels is increased as compared with the case of the thin case. In the graphs of FIGS. 13 and 14, in the case of obesity, in both graphs, the peak of the CT value is clearly seen in the vicinity of −110 with a half width of about 25-30.

From the tendency of ordinary people, it can be seen that the subcutaneous fat in FIG. 14 forms a CT value peak faster than the visceral fat in FIG. That is, the distribution graph of the number of voxels in subcutaneous fat and the CT value can be used as judgment data for predicting obesity. Even if it seems thin, if the peak of subcutaneous fat begins to appear, it can be estimated that visceral fat accumulation will begin in the near future, and it will be possible to quickly change the lifestyle habits of people called hidden obesity You will understand.

15 to 17 show CT value distribution charts of subcutaneous fat and visceral fat for three representative examples. FIG. 15 is a distribution map of subjects who are male, age 52, and BMI 18.7. FIG. 16 is a distribution diagram of subjects who are male, age 43, and BMI 21.2. FIG. 17 is a distribution diagram of subjects who are male, age 45, and BMI 32.3.

In the case of FIG. 15, the CT value of subcutaneous fat rises rapidly from −100 to −80, and shows a gradual downward curve from −62. On the other hand, the CT value of visceral fat shows a monotonous increase curve having no peak from -100 to -10. The peak CT values for subcutaneous fat and visceral fat were -62 for subcutaneous fat and -10 for visceral fat.
In the case of FIG. 16, the CT value of subcutaneous fat shows a unimodal curve having a peak of −102, and has a peak at a lower CT value than that of the lean skin. On the other hand, the CT value of visceral fat is a unimodal curve having a peak at −97, and the CT value higher than the peak shows a gentle curve.
In the case of FIG. 17, the CT value of subcutaneous fat shows a unimodal curve with a peak at −111. On the other hand, the CT value of visceral fat is a unimodal curve having a peak at −110, which is similar to the CT value of subcutaneous fat. However, as compared with the CT value curve of the subject of BMI 21.2 shown in FIG.

15 to 17, the subcutaneous fat has a peak at a CT value lower than that of the built-in fat. From this, it can be seen that as fat becomes obese, fat accumulates in the subcutaneous adipose tissue and then gradually accumulates in the visceral adipose tissue. Also, from FIGS. 15 to 17, it can be confirmed that the curve tends to become sharper as it becomes obese. It is presumed that the fat purity of fat cells tends to be higher as they become obese.

(Fat testing device)
Next, the fat test apparatus of the present invention will be described. FIG. 7 is a functional block diagram of the fat test apparatus according to the second embodiment, and FIG. 8 is a functional block diagram of the discrimination unit.
The fat inspection apparatus 1 according to the second embodiment is configured to discriminate a subcutaneous fat layer, a muscle layer, and a visceral fat layer using an image acquisition unit 3 that acquires a tomographic image of an abdominal tissue in the medical tomographic image apparatus 2 and a CT value of the tomographic image. A discriminating unit 4, a distribution acquiring unit 5 for acquiring a correlation distribution between the number of voxels in the discriminated subcutaneous fat layer and visceral fat layer and the CT value, and a discriminating unit 6 for discriminating an obesity tendency using the correlation distribution of the CT value. Prepare. The image acquisition unit 3, the discrimination unit 4, the distribution acquisition unit 5, and the determination unit 6 are each subjected to functional processing by an arithmetic processing unit of a computer. The correlation distribution data of the number of voxels and the CT value is stored in the memory.

Further, as shown in FIG. 8, the discriminating unit 4 compares the measured value of the thickness information of the subcutaneous fat layer of the previous line with the storage unit 42 that stores the thickness information of the subcutaneous fat layer of the previous line. When the difference is larger than the threshold, the changing unit 44 that temporarily changes the CT value as the boundary value between the muscle layer and the subcutaneous fat layer, and the CT value of the boundary value between the changed muscle layer and the subcutaneous fat layer And a calculation unit 46 for calculating the thickness information of the subcutaneous fat layer of the current line.
The storage unit 42, the change unit 44, and the calculation unit 46 are each subjected to functional processing by an arithmetic processing unit of a computer. The measurement value and threshold value of the thickness information of the subcutaneous fat layer of the immediately preceding line are stored in the memory.

(Statistical analysis of fat mass in the liver)
In the fat inspection method of the present invention, a liver region extraction step for extracting a liver region from a three-dimensional region obtained by superimposing tomographic images obtained in the image acquisition step, and a correlation distribution between the number of voxels in the extracted liver region and a CT value is obtained. The liver region distribution acquisition step to be acquired and the fatty liver determination step to determine fatty liver using the correlation distribution of CT values are used to estimate the amount of fat in the liver from the degree of decrease in CT values, and to determine fatty liver be able to.
Actually, the results of measuring the liver volume, the amount of fat in the liver, and the amount of fat (%) relative to the liver volume of 175 subjects (male, age 40 to 69 years, average age 56.4, standard deviation 10.6 years) Indicates. The respective measured values (average, standard deviation) are the liver volume (1062 cm ³ , 236 cm ³ ), the fat mass in the liver (149 cm ³ , 120 cm ³ ), and the ratio of the fat mass to the liver volume (5.4%, 1. 5%).

FIG. 18 shows the correlation between the fat amount (cm ³ ) and the ratio (%) of the fat amount to the liver volume by single regression analysis with each test item. Each inspection item is weight, body mass
index (BMI), waist circumference (WC), visceral fat area (VFA), subcutaneous fat area (SFA), visceral fat volume (VAT), subcutaneous fat volume (VAT) SAT: Subcutaneous adipose tissue), aspartate aminotransferase (AST), alanine aminotransferase (ALT), LDL cholesterol (LDL-C: low density lipoprotein cholesterol), HDL cholesterol (HDL-C: high density lipoprotein cholesterol), blood glucose level (BS: Blood Sugar), hemoglobin A1c (HbA1c: hemoglobinA1c), uric acid (UA), C-reactive protein (CRP), albumin (albumin), γ-glutamyl transpeptidase (γ- GTP), systolic blood pressure (SBP), diastolic blood pressure (DBP), neutral fat ( G: Triglyceride) is.
In FIG. 18, an item with a mark to the right of the correlation coefficient indicates that there is a correlation.

Fat mass (cm ³ ) in the liver showed a positive correlation with cholesterol (HDL) at a significance level of less than 1%. Moreover, the ratio (%) of fat mass to liver volume showed a strong positive correlation with weight, BMI, VFA, SFA, VAT, SAT, AST, ALT, and SBP at a significance level of less than 0.1%. LDL showed a positive correlation with a significance level of less than 5%.
That is, as weight, BMI, VFA, SFA, VAT, and SAT increase, obese people may have a higher proportion of fat contained in the liver volume than non-obese people.

In addition, as the ratio of the amount of fat contained in the liver volume increased, it was confirmed that the values of AST and ALT indicating the liver function increased, and as a result of accumulation of fat in the liver, hepatocytes It can be seen that it has some influence. Furthermore, the ratio of the amount of fat contained in the liver volume was confirmed to have a strong positive correlation with LDL. Since LDL is a risk factor for arteriosclerosis, long-term fat accumulation in the liver can lead to progression to arteriosclerosis.
By estimating the amount of fat in the liver from the degree of decrease in CT value by the fat test method of the present invention, it is possible to make a non-invasive quantitative determination of fat accumulation in the liver without performing a liver biopsy. It can be seen that liver fat and other diseases can be discriminated.

In the fourth embodiment, a fat test program for causing a computer to execute each step of the fat test method of the first embodiment will be described.
The fat test program according to the fourth embodiment uses the processing flow shown in FIG. 4, that is, an image acquisition step (S10) for acquiring a tomographic image of the abdominal tissue in the medical tomographic image apparatus, and a subcutaneous fat layer using CT values of the tomographic image A discrimination step (S120) for discriminating between the muscle layer and the visceral fat layer, a storage step (S122) for storing the thickness information of the subcutaneous fat layer of the previous line, and the thickness information of the subcutaneous fat layer of the current and previous line As a result of the comparison in the comparison step (S124) for comparing the measured values and the comparison step (S124), if the difference is larger than a predetermined threshold, a CT value as a boundary value between the muscle layer and the subcutaneous fat layer is temporarily set. Using the changing step (S126) to be changed and the CT value of the boundary value between the changed muscle layer and subcutaneous fat layer, the calculation step for calculating the thickness information of the subcutaneous fat layer in the current line is calculated. (S128) and a distribution acquisition step of acquiring a correlation distribution between the number of voxels of the subcutaneous fat layer and the visceral fat layer and the CT value when the difference is within a predetermined threshold as a result of the comparison in the comparison step (S124). (S14) and a determination step (S16) for determining an obesity tendency using the correlation distribution of CT values are executed by the computer.

FIG. 19 shows the internal configuration of computer hardware that executes a fat test program. 19, the internal configuration of the computer hardware includes a CPU 111, a ROM 112, a hard disk 113, a keyboard 114, a mouse 115, a display 116, an optical drive 117, and a RAM 118, and is connected to a system bus 119. The ROM 112 stores a program such as a boot up program for starting the computer. The RAM 118 temporarily stores fat test program instructions and provides a temporary storage space. The hard disk 113 stores a fat test program, a system program, and data. The keyboard 114 and the mouse 115 receive commands from a computer operator. The display 116 displays the discrimination result of the discrimination step in the fat test program. The computer may include a network card (not shown) that provides a connection to the network.

The fat test program only needs to include an instruction part that calls an appropriate module function and obtains a desired result for each step shown in FIG. It is well known how a computer operates, and a detailed description is omitted. The computer that executes the fat test program may perform centralized processing by one (stand-alone) or may perform distributed processing by a plurality of computers connected by a network. That is, each step shown in FIG. 4 may be realized by centralized processing by a single computer, or may be realized by distributed processing by a plurality of computers.

The present invention can be used as an optional device of X-ray CT apparatus or a fat mass measuring device.

DESCRIPTION OF SYMBOLS 1 Fat test | inspection apparatus 2 Medical tomographic image apparatus 3 Image acquisition part 4 Discrimination part 5 Distribution acquisition part 6 Discrimination part 10 Skin 12 Subcutaneous fat layer 14 Muscle layer 16 Visceral fat layer 18 Visceral tissue 19 Bone

Claims

An image acquisition step of acquiring a tomographic image of the abdominal tissue in the medical tomographic image apparatus;
A discrimination step of discriminating a subcutaneous fat layer, a muscle layer, and a visceral fat layer using a CT value of a tomographic image;
A distribution acquisition step of acquiring a correlation distribution between the discriminated subcutaneous fat layer and visceral fat layer and the CT value;
A determination step of determining an obesity tendency using the correlation distribution of the CT values,
A fat test method comprising:
In the discrimination step, a storage step of storing thickness information of the subcutaneous fat layer of the immediately preceding line;
A change that temporarily changes the CT value as the boundary value between the muscle layer and the subcutaneous fat layer when the measured value of the thickness information of the subcutaneous fat layer in the current and previous lines is compared and the difference is greater than a predetermined threshold Steps,
A calculation step of calculating the thickness information of the subcutaneous fat layer of the current line using the CT value of the boundary value of the changed muscle layer and subcutaneous fat layer;
The fat test method according to claim 1, comprising:
3. The fat inspection method according to claim 1, wherein, in the determination step, when a peak appears in the CT value of the subcutaneous fat layer, it is determined as an obesity tendency.
In the discrimination step, the content of oleic acid, which is an unsaturated fatty acid, and palmitic acid, or stearic acid, which is a long-chain fatty acid, is estimated from the peak CT value of the subcutaneous fat layer, and whether or not adipose tissue is easily burned. The fat testing method according to any one of claims 1 to 3, characterized in that:
In the determination step, when the CT value peak of the subcutaneous fat layer is lower than a predetermined threshold, it is determined that the content of long-chain fatty acid palmitic acid or stearic acid is high and the fatty tissue is difficult to burn, and is higher than the predetermined threshold. In this case, it is determined that the content of oleic acid, which is an unsaturated fatty acid, is high and the adipose tissue is easily combusted.
A liver region extraction step for extracting a liver region from a three-dimensional region obtained by superimposing the tomographic images obtained in the image acquisition step;
A liver region distribution acquisition step of acquiring a correlation distribution between the number of voxels in the extracted liver region and the CT value;
A fatty liver determination step for determining fatty liver using the correlation distribution of the CT values;
The fat test method according to any one of claims 1 to 4, further comprising:
In the fatty liver discrimination step, the fat percentage of the extracted liver region is calculated from the average value of the CT value of the extracted liver region, the average value of the CT value of the healthy subject's liver region, and the average value of the CT value of fat. The fat test method according to claim 6, wherein the fat test method is calculated.
An image acquisition unit for acquiring a tomographic image of the abdominal tissue in the medical tomographic image apparatus;
A discriminating unit for discriminating a subcutaneous fat layer, a muscle layer, and a visceral fat layer using a CT value of a tomographic image;
A distribution acquisition unit that acquires a correlation distribution of the CT value and the number of voxels in the discriminated subcutaneous fat layer and visceral fat layer;
A discriminator for discriminating an obesity tendency using the correlation distribution of the CT values;
A fat testing apparatus comprising:
In the discrimination unit, a storage unit for storing thickness information of the subcutaneous fat layer of the immediately preceding line;
A change that temporarily changes the CT value as the boundary value between the muscle layer and the subcutaneous fat layer when the measured value of the thickness information of the subcutaneous fat layer in the current and previous lines is compared and the difference is greater than a predetermined threshold And
Using the CT value of the boundary value between the changed muscle layer and subcutaneous fat layer, a calculation unit that calculates the thickness information of the subcutaneous fat layer of the current line;
The fat test | inspection apparatus of Claim 8 characterized by the above-mentioned.
10. The fat test apparatus according to claim 8, wherein the discrimination unit discriminates an obesity tendency when a peak appears in the CT value of the subcutaneous fat layer.
Whether the fatty tissue is likely to burn by estimating the content of unsaturated fatty acid oleic acid and long-chain fatty acid palmitic acid or stearic acid from the CT value peak value of the subcutaneous fat layer in the discrimination unit The fat test apparatus according to any one of claims 8 to 10, wherein:
In the determination unit, when the CT value peak of the subcutaneous fat layer is lower than a predetermined threshold value, it is determined that the content of palmitic acid or stearic acid, which is a long chain fatty acid, is high and the fatty tissue is difficult to burn, and is higher than the predetermined threshold value. The fat test apparatus according to claim 11, wherein in this case, it is determined that the content of oleic acid, which is an unsaturated fatty acid, is high and the adipose tissue is easily combusted.
A liver region extraction unit that extracts a liver region from a three-dimensional region obtained by superimposing the tomographic images obtained in the image acquisition unit;
A liver region distribution acquisition unit that acquires a correlation distribution between the number of voxels in the extracted liver region and the CT value;
A fatty liver discriminating unit for discriminating fatty liver using the correlation distribution of the CT values;
The fat test apparatus according to any one of claims 8 to 11, further comprising:
In the fatty liver discriminating unit, the fat percentage of the extracted liver region is calculated from the average value of the CT value of the extracted liver region, the average value of the CT value of the healthy subject's liver region, and the average value of the CT value of fat. The fat test apparatus according to claim 13, wherein the fat test apparatus calculates the fat test apparatus.
Each step in the fat testing method according to any one of claims 1 to 7,
A fat test program to be executed by a computer.