JPS639461B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to a method for evaluating X-ray images for measuring the bone age of children. For more details,
The present invention relates to a method of measuring the age of bones by measuring the degree of darkening at the midpoint of the metacarpal bone in a hand X-ray image of a child using a densitometer. Traditionally, to determine the state of bone growth in children,
A widely used method is to take an X-ray image of the hand and compare it with a dorsal palmar X-ray atlas to determine the so-called bone age (see Clin.Orthop., 75:248, 1971). With Bone Age, it is easy to incorporate the subjectivity of the judge, and since it is a semi-quantitative judgment method,
It was insufficient to quantitatively determine whether children's bone growth was delayed or advanced by several months. On the other hand, as a standard for the degree of darkening in the X-ray image, an aluminum staircase (Gray Scale) was installed and an X-ray image of the hand was taken.
The degree of blackening is measured using a densitometer at the midpoint between the proximal and distal ends of the first metacarpal, and the peak height is converted to the number of steps on an aluminum staircase using a computer. Width (D), bone marrow width (d), metacarpal index [MCI (=D-d/D)], maximum bone density (GSmax), minimum bone density (GSmin), maximum/minimum bone density ratio (E= GSmin/GSmax), average bone density (ΣGS/D, where ΣGS is the integral value of the peak area)
and a method of determining the degree of bone atrophy from these indicators by determining bone patterns (hereinafter abbreviated as MD method)
is known (Tetsuro Inoue et al.: Bone Metabolism, 13:
187, 1980 and 14:91, 1981). However, this method is a method for determining the degree of bone atrophy in patients with artificial dialysis or osteoporosis, and is suitable for determining the degree of bone atrophy in adults aged 20 years or older and tracking its changes over time. , it cannot be used as is as a method for determining the state of bone growth in children. For example, in adults,
As bone atrophy progresses, bone resorption occurs from the marrow side, so the marrow width (d) increases and becomes a powerful indicator of the degree of bone atrophy, and as bone atrophy progresses, the metacarpal index (MCI) increases. Although the bone width (D) remains constant, the bone marrow width (d) increases, so the MCI decreases and is widely used as an index to determine the degree of bone atrophy.
In the case of children, the marrow width (d) increases as the bone width (D) increases, so d increases even when the bone is in a good developmental state, and even if bone resorption occurs from the marrow side, d increases. On the other hand, if the state of bone development is poor, D may also be small, and d may also be small, and d may also be small if bone resorption has not occurred from the bone marrow side.
The state of bone growth cannot be determined by the size of d. Furthermore, as mentioned above, since bone width (D) and bone marrow width (d) change simultaneously, MCI, which is calculated by D-d/D, is also used as an index to understand the state of bone growth in children. is not appropriate. Therefore, the present inventors used an objective method that does not involve the subjectivity of the judge to understand the growth status of children's bones using important indicators (parameters) and quantified them. As a result of intensive research into a method for quantitatively determining age, we used five indicators as parameters in the X-ray image evaluation method: bone width, maximum bone density, minimum bone density, average bone density, and bone pattern. By calculating the correlation with age and creating a regression equation for each regression equation, finding each index age from each regression equation, and calculating the arithmetic average value, we can objectively determine the child's bone age. The present invention was achieved by discovering that the method can be measured more accurately. That is, the present invention calculates the X-ray value of the first metacarpal bone of a healthy child using a linear regression equation obtained using a large number of normal X-ray images of the second metacarpal bone of healthy children of the same sex and various ages as the child being tested. An X-ray image evaluation method for evaluating a ray image, the method comprising: Obtain the darkening degree pattern in the direction, and (ii) calculate the bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), average bone density (ΣGS/D) and numerical values from the darkening degree pattern. (iii) find the correlation with the age of the healthy child for each of the five parameters by least squares method; The following formula (1)
Yi=ai+bi×(age)...(1) (However, Yi, ai, and bi each represent the parameter value, constant, and coefficient for type i.) (B) The subject X-ray Evaluation of the image is performed by (i) determining a darkening degree pattern in a direction perpendicular to the axis at substantially the midpoint of the bone using a densitometer from the X-ray image to be examined; and (ii) determining a darkening degree pattern from the darkening degree pattern. D, GSmax,
Determine five parameters: GSmin, ΣGS/D, and the corresponding numerical value of the bone pattern, and (iii) determine the index age for each of the five parameters using the following formula (2), Xi = (Yi -ai)/bi...(2) (However, Xi is the index age for the parameter of type i, and Yi, ai and bi are the same as in equation (1).) (iv) Indices for the five types A method for evaluating an X-ray image, characterized in that the arithmetic mean value of ages is taken as the bone age from the X-ray image to be examined. In the present invention, first, the degree of darkening at the midpoint of the first metacarpal bone in a hand X-ray image of a healthy child is measured using a densitometer, and (a ₁ ) bone width (D), maximum bone density ( GSmax), minimum bone density (GSmin), average bone density (ΣGS/D), and bone pattern. (a ₂ ) For each parameter, calculate the correlation with age for each gender and perform regression. Create an expression. In addition, X mentioned in the present invention
A ray image means an image on an X-ray film obtained by X-ray photography. Here, healthy children are usually healthy boys under the age of 15.
I'm talking about healthy girls. The number of participants usually ranges from 30 to 100 people. For example, the main body X-ray image of a healthy child is 1 step high.
Take a photo of an aluminum step wedge with 20 steps (minimum height 1mm, maximum height 20mm) or an aluminum slope. The degree of darkening of the obtained X-ray image at the midpoint between the proximal end and the distal end of the first metacarpal is measured using a microdensitometer. Next, five parameters are determined from the degree of blackening: bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), average bone density (ΣGS/D), and bone pattern. Such parameters are determined as follows. The degree of blackening determined by a microdensitometer was magnified 10 times and recorded on paper with an optical density (OD) scale. The metacarpal bone pattern as shown in Figure 1 is drawn by converting and correcting the number of steps, and the bone width (D) and minimum bone density (GSmin) are calculated from the figure.
seek. Calculate the maximum bone density (GSmax = 1/2 (GSmax1 + Gmax2) from GSmax1 and GSmal2 in Figure 1. Also, from the pattern area (ΣGS) obtained by integrating each GS value, the average bone density (ΣGS / D)
seek. The bone pattern is shown in Figure 2 (Reference “Bone Metabolism
13:187-195, 1980)).
~I), and the one that most closely approximates the metacarpal bone pattern in Figure 1 among the nine bone models A~I is the desired bone pattern (A~I), and pattern A is 1, Pattern B is made to correspond to pattern 2 in order and quantified to obtain an index (parameter) of the bone pattern. Next, for each parameter, the correlation with age is determined and a regression equation is created for each gender. The regression equation is created by plotting the measured values of each parameter with age on the horizontal axis and each parameter value on the vertical axis, and performing a correlation analysis using the least squares method. Since the degree of bone growth differs depending on gender, regression equations are created separately for men and women. Graphs showing regression equations for each parameter are illustrated in FIGS. 3 and 4. In each figure, the solid line represents the regression equation, and the dotted line represents the range of the regression equation ±σ. The above parameters used in the present invention have been revealed through research by the present inventors to show particularly remarkable changes with age. It is extremely effective in measuring. In the present invention, next, from the hand X-ray image of the child whose bone age is to be measured, five parameters are obtained in the same manner as in (A), and (b ₁ ) from the parameter values, (a) of (A) is determined. Among the regression equations created in _{step 2} ), use the regression equation for the same sex as the child to find the index age for each parameter, and (b ₂ ) calculate the arithmetic mean of those index ages to determine the child's Bone age. In exactly the same manner as described above, a hand X-ray image of a child whose bone age is to be measured is taken, and the bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), and average bone density of the pediatric patient are determined. Five parameters of bone density (ΣGS/D) and bone pattern are determined. Next, from this parameter value, an index age is determined for each parameter using a regression formula for the same sex as the child among the regression formulas created by the method described above. When the pediatric patient is a boy, a regression equation determined from a hand X-ray image of a healthy boy is used, and when the child is a girl, a regression equation for a healthy girl is used. Here, the method for determining the index age will be explained in detail with reference to Figure 13. FIG. 13 is a schematic diagram of a graph showing a regression equation for a healthy child. Plot the relationship between the actual age of the pediatric patient and the measured values of each parameter, and as shown in Figure 13, draw a perpendicular line from the measurement point to the vertical axis, and draw a perpendicular line from the intersection with the regression equation to the horizontal axis. to find the index age for each parameter. Next, the arithmetic mean value of the five parameters of bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), average bone density (ΣGS/D), and bone pattern determined in this way is determined from the index age. By calculating the bone age of the child, the bone age of the child can be determined. In the present invention, the gender of the child whose bone age is to be determined, and the gender of the healthy child whose hand X-ray image is to be taken when creating a regression equation for determining the bone age. must match. Therefore, the regression equation determined from hand X-ray images of healthy girls is used to measure the bone age of female children. Bone age obtained by the method detailed above is highly comparable to bone age, which is conventionally widely used for diagnosing the degree of bone growth, as shown in the reference example below. There is a correlation. Therefore, the method for evaluating X-ray images of the present invention is extremely useful as a method for quantitatively determining the degree of bone growth in children, and can also be used to treat rickets, growth hormone deficiency, cretinism, and administration of steroids and anticonvulsants. It is very advantageously used for diagnosing bone growth retardation due to poor growth, tracking therapeutic effects, etc. Hereinafter, the present invention will be explained in more detail with reference to Reference Examples and Examples. Example 1 (a) Creation of regression equation For each of 56 healthy boys and girls under the age of 15, 20 steps with each step height of 1 mm (minimum height 1
An X-ray image of the inserted hand was taken using the aluminum step (mm, maximum height 20 mm) as a reference, and its darkening was measured using a densitometer at the midpoint between the proximal and distal ends of the first metacarpal. Measure the optical density (OD) and record it on a scaled sheet of paper at a magnification of 10 times. Using a computer, calculate each index using the method described above. Plot the measured values of each index for men and women on the vertical axis, and calculate the regression equation by correlation analysis,
It is illustrated in FIGS. 3 to 12 together with the width of σ. Details of the regression formula, σ, etc. for each index are shown in Table 1. The correlation coefficients between age and each index determined by correlation analysis are shown in Table 2.

【table】

[Table] As shown in Table 2, bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), average bone density (ΣGS/D), and bone pattern are all related to age. There is a high degree of correlation between them, making them suitable as parameters for bone age measurement. (b) Measurement of bone age The bone age of pediatric patients was measured using the regression equation obtained in (a) above. The results are shown in Table 3.

[Table] From Table 3, for example, the bone age of No. 11, a 2.8-year-old child with rickets, is determined to be 1.3 years, and the bone growth is 1.5 years older than that of a healthy female (1.3-2.8=-
1.5) Rated as late. Reference Example 1 In the same manner as in Example (a), correlation analysis was performed on the bone marrow width (d), metacarpal bone index (MCI), and the maximum and minimum bone density (E), which are known as metacarpal bone parameters. The correlation coefficient between age and each parameter was determined. The results are shown in Table 4.

[Table] As shown in Table 4, bone marrow width (d), metacarpal index (MCI), and maximum to minimum bone density ratio (E) are highly correlated with age. This is difficult and inappropriate as a parameter for determining bone age in children. Reference Example 2 The correlation between bone age and MD age (measurement method of the present invention) for 112 healthy children and 20 abnormal children is as shown in Figure 13.
The regression equation is MD age = 0.44 + 0.931 [bone age], and the relationship is approximately MD age = bone age. Also, a very high correlation coefficient of r = 0.9163 was obtained, and the risk rate was 1. %, it became clear that there is a highly positive correlation between bone age and MD age.

[Brief explanation of the drawing]

Figure 1 shows the parameters bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), and average bone density (ΣGS/D), and Figure 2 shows the bone pattern. Figures 3 to 12 show bone width (D) (Figures 3 and 4), maximum bone density (GSmax) (Figures 5 and 6), and minimum bone density (GSmin).
(Figures 7 and 8), average bone density (ΣGS/D) (Figure 9,
Figure 10) and bone patterns (Figures 11 and 12) are shown for boys and girls.In each figure,
The solid line represents the regression equation, and the dotted line represents the range of the regression equation ±σ. Figure 13 shows the method of determining the index age in the present invention, and Figure 14 shows the measurement method of the present invention (MD age) and the conventional bone age (bone age).
This shows the correlation with age). In Figure 14, ○ marks are boys (n=71), ● marks are girls (n=71), and ● marks are girls (n=71).
= 61), solid line is MD age = 0.44 + 0.931 [bone age]
(r=0.9163, P<0.01), dotted line is MD age=Bone
It shows age.

Claims (1)

[Scope of Claims] 1. A linear regression equation obtained using a large number of normal X-ray images of the second metacarpal bones of healthy children of the same sex and various ages as the subject child. An X-ray image evaluation method for evaluating a subject X-ray image, the method comprising: Obtain the darkening degree pattern in the direction perpendicular to the axis, and (ii) from the darkening degree pattern, calculate bone width (D), maximum bone density (GSmax), minimum bone density (GSmin), and average bone density (ΣGS/D). and (iii) determine the correlation between each of the five parameters and the age of the healthy child at a minimum of two. The following formula (1) obtained by multiplication
Yi=ai+bi×(age)...(1) (However, Yi, ai, and bi each represent the parameter value, constant, and coefficient for type i.) (B) The subject X-ray Evaluation of the image is performed by (i) determining a darkening degree pattern in a direction perpendicular to the axis at substantially the midpoint of the bone using a densitometer from the X-ray image to be examined; and (ii) determining a darkening degree pattern from the darkening degree pattern. D, GSmax,
Determine five parameters: GSmin, ΣGS/D, and the corresponding numerical value of the bone pattern, and (iii) determine the index age for each of the five parameters using the following formula (2), Xi = (Yi -ai)/bi...(2) (However, Xi is the index age for the parameter of type i, and Yi, ai and bi are the same as in equation (1).) (iv) Indices for the five types A method for evaluating an X-ray image, characterized in that an arithmetic mean value of ages is taken as the bone age from the X-ray image to be examined.