JPH0518891A - Correcting method for chemical analyzer - Google Patents

Abstract

PURPOSE:To enable accurate correction to be conducted by weighting for primary feedback with high (low) weight in the photo-concentration range of small (large) acceptable optical concentration difference converted with an error factor. CONSTITUTION:The reflected light applied 2 to a chemical analysis slide 1 of a measured object is converted 3 to electric signal and a CPU 5 converts this to logarithm and then converts to an optical concentration value and a clinical value. A measurement accuracy is generally defined by the error factor to the clinical value, however; the acceptable optical concentration difference corresponding to an acceptable error factor along a measurement curve differs for the concentration range. To improve accuracy, the measured value is weighted accordingly to the concentration difference. That is, to a clinical value error Ni calculated by multiplying the clinical value Ci by a defined error ratio Mi%, the value Ci is added to get a value CEi. The CEi and Ci are inversely converted to optical concentration on the inverse function of the measurement curve to obtain a small difference dDi between the concentrations DEi and Di, which is an acceptable optical concentration difference. For a large (small) difference, a small (large) weight is put to a standard slide measurement value for primary feedback.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for correcting a chemical analysis device by optical density.

[0002]

2. Description of the Related Art For an analytical slide having at least one reagent layer on a transparent support, which causes a change in optical density due to suitable attachment of an analyte, an analyte such as blood or serum is appropriately adjusted to obtain an optical reflection density ( Alternatively, there is a chemical analyzer that measures the concentration (or permeation concentration) and converts the concentration value into a clinical value by calculation to chemically analyze the presence or absence, content, etc. of a specific component in a sample.

When the chemical analyzer is mass-produced, the concentration measurement value of each mass-produced device varies. Therefore, it is necessary to correct the density value of each mass production device. Therefore, it is conceivable to prepare a reference chemical analysis device and make the concentration value of another (to be corrected) chemical analysis device coincident with the reference chemical analysis device. For this reason, the concentration of a plurality of reference slides is measured by a reference chemical analysis device and another chemical analysis device, and the measured value of the above-mentioned "other chemical analysis device" is used as the reference chemical analysis device based on each measured concentration value. There is a method of obtaining a conversion formula for converting into a measurement value of an analyzer. This correction method is described in Japanese Patent Laid-Open No. 1-124751, and correction of a mass production device can be performed at a relatively low cost by a simple operation.

[0004]

In any of the correction methods described above, when the photometric characteristic of the chemical analyzer to be corrected is non-linear with respect to the photometric characteristic of the reference chemical analyzer, the correction effect is obtained. There is a limit.

Generally, the measurement accuracy required for a chemical analyzer is defined by an error ratio% with respect to a clinical value converted by calculation from optical density. However, this clinical value is obtained by converting the measured optical density by a conversion curve that is a non-linear function, that is, a calibration curve. For example, the accuracy required for the clinical value is 5% uniform over the entire clinical value. Even in such a case, the optical density measurement accuracy for ensuring the clinical value accuracy is different in each optical density range. In the method described in Japanese Patent Laid-Open No. 1-124751, the accuracy required for the optical density measurement is different for each optical density measurement area because the weights of the measured values are equalized in all the optical density areas and the primary regression is performed. The nature is not reflected. For example, there is a problem that the regression residual may be small in the optical density range where the required optical density measurement accuracy is relatively low, and conversely may be large in the high optical density range.

[0006]

SUMMARY OF THE INVENTION In a method of correcting a measurement value of an arbitrary chemical analysis device using a reference slide and a reference chemical analysis device, the present invention has been made for the purpose of solving the above problems. The accuracy of the chemical analyzer is expressed as an error ratio with respect to a clinical value, the error ratio is converted into an allowable optical density difference, and high weighting is performed in an optical density range in which the allowable optical density difference is small. This is a correction method for a chemical analysis device, characterized by performing low-weighting in the concentration range and performing correction by linear regression.

[0007]

EXAMPLE An analysis slide 1 according to the present invention is shown in FIGS.
It is configured as shown in FIG. 10 and prepared according to many items including GOT, GPT, glucose, urea nitrogen and the like.

In this analysis slide 1, an analytical element 51 having a reagent is mounted in a recess of a mount base 50 having a through hole 50a for photometry in the recess of the central part, and a transparent part for dropping a sample is mounted in the central part from above. The mount covers 52 having the holes 52a are overlapped with each other and bonded by a bonding means such as ultrasonic waves. On both sides of the mount base 50, stepped portions 50b that determine the insertion direction are formed, and on the surface of the mount cover 52, an arrow 53a indicating the insertion direction, a measurement item name 53b, and a measurement item for determining the measurement item. The identification code 54 is displayed.

The external appearance of the chemical analyzer is shown in FIG. Further, FIG. 12 is a schematic view of the chemical analysis device, FIG. 13 is a plan view of the measuring part thereof, and FIG. 14 is a sectional view thereof.

This chemical analyzer is a device for dropping a sample onto an analysis slide 1 and measuring a change in color density due to the reaction to chemically analyze the presence or absence of a specific component in the sample. A section A, a constant temperature section B, a sample dropping section C, a photometric section D, and a slide discharge section E are provided.

Slide Supply Section A The slide supply section A supplies the analysis slide 1 to the constant temperature section B, and the analysis slide 1 is inserted in a fixed direction. The analysis slides 1 may be inserted one by one, or may be automatically and continuously supplied.

Constant Temperature Part B The constant temperature part B is for keeping the analysis slide 1 on which the sample is dropped at a predetermined temperature so that the color density change due to the reaction of the sample is appropriately performed. This constant temperature section B has a disc
22 is rotatably provided, and a plurality of slide storage chambers 23 each having an insertion opening are formed outside the disc 22. The analysis slide 1 supplied from the slide supply unit A is stored in the slide storage chamber 23, and the analysis slide 1 is conveyed by the rotation of the disk 22 in the order of the sample dropping unit C, the photometric unit D, and the slide discharge unit E.

Sample Dropping Section C To the sample dropping section C, the analysis slide 1 is moved from the slide accommodating chamber 23 of the constant temperature section B, and the shutter 24 is opened here so that the inspector drops the sample on the analysis slide 1. You can When the dropping of the sample is completed, the analysis slide 1 is again stored in the slide storage chamber 23 of the constant temperature part B and conveyed.

Photometric unit D The photometric unit D measures the change in color density due to the reaction of the sample dropped on the analysis slide 1. The photometric unit D is provided with a photometric device 7 below a photometric window 26 formed on a support plate 25 that supports the analysis slide 1, and a permanent magnet 29 is provided in a housing 28 of the constant temperature unit B above the photometric window 26. It is provided.

A window 22a is formed on the upper wall of the slide storage chamber 23 of the disk 22, and a permanent magnet 30 is supported at a position facing the permanent magnet 29 via a pressing plate 31. This pressing plate
31 is pressure-bonded to the analysis slide 1 through the elastic body 32 to cover it,
The sample dropped on the analytical slide 1 is prevented from drying.

When the analysis slide 1 stored in the slide storage chamber 23 of the disk 22 is conveyed to the photometric section D by the rotation of the disk 22, the analysis slide 1 is connected to the permanent magnet 29 of the photometric section D,
By the repulsive force of the permanent magnet 30 of the slide storage chamber 23, the permanent magnet 30 of the slide storage chamber 23 pushes the pressing plate 31 to which it is attached downward, so that the analysis slide 1 is supported by the support plate 25.
It is pressure-bonded to the measurement reference plane X of.

As described above, since the analysis slide 1 is held in pressure contact with the measurement reference plane X by utilizing the repulsive force of the permanent magnets 29, 30, the photometry unit is prevented from vibrating due to the pressing and holding of the analysis slide 1. Moreover, the structure for holding the analysis slide 1 in pressure contact with the measurement reference plane X is simple.

The analysis slide 1 is also held in this pressure-contact holding state.
Since the concentration is measured, the measurement accuracy can be further improved.

Further, although the pressing plate 31 is in direct contact with the analysis slide 1 in this embodiment, another member may be disposed between them to indirectly press-bond the analysis slide 1.

Further, the pressing plate 31 is an elastic member on the analysis slide 1.
It is closely attached via 32, and also has the effect of preventing evaporation of the dropped sample as described above.

Further, when the chemical analysis apparatus has a transport mechanism for a plurality of analysis slides 1 and the measurement is performed with the transport slides inserted in the transport mechanism, a magnet is provided for each transport mechanism. Just place it.

The analysis slide 1 is constructed so as to be pressure-bonded and held only at a position where photometry is performed, and not to be pressure-bonded when transported to another position.

Further, the permanent magnet 29 of the photometry section D is replaced with an electromagnet, and a magnetic field is generated only during measurement, the analysis slide 1 is press-held, and no magnetic field is generated when the disk 22 rotates and moves. In this case, since the analysis slide 1 is not crimped when it is conveyed, the analysis slide 1 can be conveyed more smoothly.

Next, the basic concept of the present invention will be described.
FIG. 1 is an example of a dry chemical analyzer utilizing optical density. As shown in the configuration diagram of FIG. 1, a chemical analysis slide 1 which is a horizontally placed object to be measured is irradiated with light from a light emitting element as a light source 2 at an angle of 45 ° from four directions, and the reflected light is received by a light receiving element. In step 3, it is converted into an electric symbol and converted into a logarithm to obtain an optical density value. Then, the concentration value is converted into a clinical value by a conversion formula and the target clinical value is displayed, and these conversion operations are performed by the CPU 5. As an example of the conversion formula, Y = [B / (X−A)] + C (1) where Y is a clinical value, X is an optical density, and A, B, and C are both unique to an analysis slide and an analyzer. There is a constant. The conversion curve (calibration curve) based on these conversion formulas is shown in FIG.
In FIG. 2, the vertical axis represents the clinical value and the horizontal axis represents the optical density value. If the permissible error range (ratio) in the clinical values Ci and Cj is defined as a%, for example, each density value Di,
The permissible error range optical density difference Δb and Δc at Dj is Δ
Not b = Δc but different values. The present invention pays attention to the difference in the allowable error range in the optical density for each density range, weights the measured values in the same range in order to improve the optical density measurement accuracy in the narrower allowable error range, and then By performing regression, it is intended to easily achieve the accuracy required for the chemical analysis device.

The procedure for carrying out the present invention will be described below. The measurement accuracy of a chemical analyzer is generally several clinical values C
Focusing on i, the error ratio Mi% in each clinical value Ci
Stipulate. Then, a clinical value error Ni (Ni = Ci × Mi) ÷ 100 corresponding to the error ratio Mi% is calculated, and the clinical value Ci is added to the clinical value error Ni to obtain CEi. That is, CEi = [clinical value error Ni] + [clinical value Ci] is calculated.

Then, both CEi and the clinical value Ci are inversely converted into optical densities by using the inverse function of the calibration curve function. That is, the density DEi corresponding to CEi and the density Di corresponding to the clinical value Ci are obtained. Further, a small difference dDi between the density DEi and the density Di is obtained, and when the small difference dDi is large, a small weight is attached to it, and when the small difference dDi is small, a large weight is attached to the measured value of the reference slide, and the primary regression Is basically done.

The error ratio is defined as an equal amount for both positive and negative, and the calibration curve is often regarded as a substantially straight line in the minute range, so the absolute value of the positive minute difference + dDi and the negative minute difference −dDi. Are often almost equal, but when they are not equal, the smaller absolute value of positive and negative minute difference ± dDi, whichever is smaller, is used as the minute concentration value dDi for determining the weight on regression. The difference of ± dDi becomes large when the allowable error ratio is different between positive and negative and when the calibration curve is largely separated from the straight line in the allowable error ratio region.

For example, the weight T is based on the maximum value MAX (dDi) of the minute difference concentration dDi in the measurable clinical value region of the chemical analyzer, that is, within the measurement dynamic range, and T = MAX (dDi) / dDi (2) You can decide with.

In another method, T = MAX (dDi) / dDi + 1 (3) in the optical density range corresponding to the clinical value region of the measurement item, and T = 1 (4 ) Is good.

The latter method is effective in securing statistical regression accuracy when the number of data points is relatively small, and the weight determination method other than this is also effective for the weight of the minute difference density.
As long as dDi is reflected, it is in accordance with the gist of the present invention.

The clinical value Ci for performing the above calculation may be set by converting from the optical density value measured by the standard measuring device of the standard slide using a calibration curve, and in this case, the measurable clinical value region It is desirable to determine the weight T according to the formula (2) or the formula (3) for the reference slide that is necessary and sufficient to include the optical density range corresponding to.

Further, there are two types of chemical analysis slides which are measured by the chemical analysis device, which have almost the same color density range at the same wavelength and which are used to obtain a clinical value with a similar calibration curve.
If there are more than one species, these items can be treated as one group. That is, the regression operation in this case may be performed for that group, and does not have to be performed for each item.

As an example, the case of glucose will be described. The glucose calibration curve has the coefficients of A; 3.3072, B; 1610.1, C; 561.7 according to the equation (1).

Further, the minute difference density dDi as an allowable error
Is obtained as shown in FIG.

Further, FIG. 3 shows a calibration curve of glucose, and FIG. 4 shows an error ratio defined for clinical values.
The clinical value measurement range (dynamic range) is 20 mg / dL.
At ~ 500 mg / dL, the error ratio range is 5% uniform in this region. In order to ensure a 5% uniform accuracy in the clinical value, the optical density measurement error of about 0.007 is observed in the vicinity of the optical density value of 0.6 from FIG.
Within, optical density error around 0.034 near optical density value 1.7
I know that I have to keep it within. The maximum value MAX (dDi) of the minute difference concentration dDi within the clinical value measurement range is 0.
It is 035.

On the other hand, Table 1 shows an example of the approximate density of the reference slide used at the time of mass production and shipping adjustment.

[0037] Table 2 shows the minute difference concentrations dDi of glucose examples corresponding to these approximate concentrations.

[0038] The values measured with the standard measuring machine on the above reference slide are shown in column S of Table 3, and the values measured with the mass production machine are shown in column R.

(Table 3) Slide number RS wavelength 546 1 0.518 0.515 2 0.522 0.559 3 0.743 0.772 4 0.735 0.738 5 0.894 0.902 6 0.907 0.922 7 1.051 1.094 8 1.085 1.145 9 1.301 1.376 10 1.347 1.397 11 1.474 1.589 12 1.535 1.698 13 1.746 1.982 14 1.828 2.185 15 1.873 2.249 Furthermore, the weight value of glucose determined according to the equation (2) is shown in Table 4, and this value is converted into an integer by rounding off to the first decimal place.

[0040] Table 5 shows the correction values of the glucose mass-production machine obtained from these Tables 3 and 4.

[0041] As another example, the case of urea is shown.

FIG. 6 is a calibration curve, and FIG. 7 is an error ratio defined for clinical values. In this case, the coefficient of the equation (1) is A; 2.8446 B; 84.98 C; 35, and the clinical value measurement range (dynamic range) is 1 mg / dL to 10 mg /
The dL and error ratios are 12% at 1 mg / dL, 8% at 4.4 mg / dL, and 5% at 6.8 mg / dL and above, and the values are interpolated by a straight line. FIG. 8 shows the minute difference density dDi as the allowable error in this example. The maximum value MAX (dDi) of the minute difference concentration dDi within the clinical value measurement range is 0.020.
8

Table 6 shows the minute difference concentration dDi of urea with respect to the approximate concentration of the reference slide.

[0044] Further, the weight values of glucose determined according to the equations (2) and (3) are shown in Table 7. These values are rounded to one decimal place and converted into integers. Table 8 shows the correction values of urea for the mass production machine obtained from these Tables 3 and 7.

[0045]

[0046]

EFFECTS OF THE INVENTION When correcting a chemical analysis device, it is preferable if the conventional photometric characteristic has linearity with respect to the reference photometric characteristic, but if it has non-linearity, the correction effect is There is a limit, and for example, there is a problem that the regression residual may be small in the optical density range where the required optical density measurement accuracy is relatively low, and conversely the regression residual may be large in the high optical density range. However, in order to realize the accuracy specified on the clinical value by the present invention, the correction value of the chemical analysis device which makes the residual error particularly small in the optical density region where relatively high optical density measurement accuracy is required is simple. It has become possible to provide a highly accurate and inexpensive correction method.

[Brief description of drawings]

FIG. 1 is a diagram of an optical system of a chemical analyzer.

FIG. 2 is a diagram of a calibration curve.

FIG. 3 is a diagram showing a calibration curve of glucose.

FIG. 4 is a diagram of an error rate of glucose.

FIG. 5 is a diagram showing the minute difference concentration of glucose.

FIG. 6 is a diagram of a calibration curve of urea.

FIG. 7 is a diagram of an error ratio of urea.

FIG. 8 is a diagram showing a slight difference concentration of urea.

FIG. 9 is a block diagram of an analysis slide.

FIG. 10: Mounted analysis slides.

FIG. 11 is an external view of a chemical analyzer.

FIG. 12 is a schematic view of a chemical analyzer.

FIG. 13 is a plan view of a measurement unit.

FIG. 14 is a sectional view of a measurement unit.

[Explanation of symbols]

 1 slide 2 light source 3 light receiving element 4 liquid sample 5 CPU Ci clinical value Di concentration value dDi minute difference concentration Ni clinical value error M error ratio

Claims (1)

Claim: What is claimed is: 1. A method for correcting a chemical analysis device for correcting a measurement value of an arbitrary chemical analysis device using a reference slide and a reference chemical analysis device, wherein the measurement accuracy of the chemical analysis device is set to a clinical value. The error ratio is converted into a corresponding allowable optical density, and high weighting is performed in the optical density range in which the allowable optical density difference is small, and low weighting is performed in the optical density range in which the allowable optical density difference is large. A method for correcting a chemical analysis device, which comprises correcting by linear regression.