JP4276894B2 - Anomaly detection system and anomaly detection method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a technique and an automatic analyzer for measuring the concentration or activity value of a target component in a sample, and software and a system using data at the time of measurement.
[0002]
[Prior art]
The automatic analyzer dispenses a sample and a reagent into a reaction vessel to generate a chemical reaction, measures the absorbance of the mixed solution, and measures the time series data of the absorbance due to the chemical reaction (in the following explanation, the reaction process data and This is a device for calculating the concentration or activity value of the target component in the sample by calculating the amount of change in absorbance and the rate of change in absorbance.
[0003]
Such automatic analyzers are mainly used in medical institutions for biochemical tests, immunological tests, etc., and the results of the tests are used by doctors to make various diagnoses, such as understanding the patient's medical condition, determining treatment effects, and monitoring the prognosis. It plays an extremely important role in doing it. For this reason, in the analysis by the automatic analyzer, accuracy management for assuring that the measurement has been performed correctly is important.
[0004]
As one of the quality control methods, there is a method of measuring a quality control sample whose concentration of each component is known. This is because the quality control sample was measured before and after the measurement of the patient sample sample or between the measurement of multiple patient sample samples, and if the test result of the quality control sample was within the allowable error range for the known concentration, the measurement was performed correctly. This is a method for judging and guaranteeing the test result of a patient specimen sample. Therefore, if the inspection result of the quality control sample deviates from the allowable error range with respect to the known concentration, it means that some abnormality has occurred during the measurement, and the test result of the patient specimen sample cannot be guaranteed. It is necessary to stop the analysis, investigate the cause of the abnormality, and take countermeasures.
[0005]
As a conventional technique for detecting an abnormality at the time of measurement, the “analysis process confirmation method of an analyzer and an automatic analyzer” described in Patent Document 1 calculates the concentration or activity value of a target component within a specific range. By observing a change pattern at a plurality of preset measurement times (referred to as photometry points in the following description) for the reaction process data of the sample and comparing it with a change pattern at the time of occurrence of an abnormality registered in advance, A technique for estimating the presence or absence of an abnormality is described.
[0006]
[Patent Document 1]
JP 2000-275252 A
[0007]
[Problems to be solved by the invention]
However, in the technique described in Patent Document 1, when estimating the presence / absence of an abnormality in a sample whose concentration is unknown, such as a patient specimen sample, changes in reaction process data according to all possible values of the concentration or activity value of the target component Since it is necessary to register the pattern in advance, there is a problem that it is practically impossible.
[0008]
In addition, abnormalities that occur at the time of measurement include accidental abnormalities such as noise. Such an abnormality does not necessarily change at a predetermined photometric point in the reaction process data, but changes in the reaction process data at irregular timing. There is a problem that it is difficult to estimate the presence or absence of an abnormality that occurs.
[0009]
An object of the present invention is to provide an abnormality detection system and an abnormality detection method for detecting an abnormality that occurs when measuring a target component in a sample.
[0010]
[Means for Solving the Problems]
The problem is to detect anomalies in time-series data of photometric values obtained by optically measuring a mixture of one or more reagents used to measure the concentration or activity value of a target component in a sample and the sample. Reference time-series data storage means for storing time-series data corresponding to a chemical reaction model when a sample is normally analyzed as reference time-series data, and a time-series data and a reference time-series. This can be solved by an abnormality detection system characterized by comprising an abnormality determination means for comparing data and determining the presence or absence of an analysis abnormality of a sample.
[0011]
In addition, the above-mentioned problems include theoretical parameter storage means for storing a parameter including a coefficient corresponding to a chemical reaction rate for each target component used in a chemical reaction model, and a coefficient corresponding to the concentration or activity value of a sample component, and the parameter Time series data calculation means for calculating time series data of a chemical reaction model by using, and reference time series data setting means for setting theoretical time series data calculated by the theoretical time series data calculation means as reference time series data With the abnormality detection system characterized by having a plurality of chemical reaction models different for each target component can be expressed by parameters, it can be solved easily.
[0012]
In addition, the above-described problems include known concentration sample time-series data storage means for storing one or a plurality of time-series data obtained by measuring a known concentration sample whose concentration or activity value of the target component is known, and known concentration An abnormality characterized by having theoretical parameter calculation means for calculating parameters using time series data stored in the sample time series data storage means, and theoretical parameter setting means for setting the calculated parameters in the theoretical parameter storage means Since the detection system can calculate parameters suitable for each measurement condition and measurement environment such as an automatic analyzer, it can be solved while further improving the reliability of the determination result.
[0013]
DETAILED DESCRIPTION OF THE INVENTION
The present invention provides an anomaly detection system and an anomaly detection method for detecting an anomaly that occurs during measurement, regardless of the timing at which changes appear in reaction process data for samples of any concentration.
[0014]
The abnormality detection method of the present invention includes a step of using reaction process data corresponding to a chemical reaction model when analysis is normally performed as reference reaction process data, and a step of comparing reaction process data with reference reaction process data. And detecting the presence or absence of abnormality during measurement by comparing the reaction process data with the reference reaction process data.
[0015]
That is, the abnormality detection method of the present invention is a time series of photometric values obtained by optically measuring a mixed solution obtained by mixing one or more types of reagents used for measuring the concentration or activity value of a target component in a sample and the sample. An anomaly detection method for detecting anomalies in data, the step of storing time series data corresponding to a chemical reaction model when a sample is normally analyzed as reference time series data in a reference time series data storage means; The method includes a step of comparing the time series data with the reference time series data and determining whether there is an abnormality in the analysis of the sample by the abnormality determination means.
[0016]
In the abnormality detection method, a parameter including a coefficient corresponding to the chemical reaction rate for each target component used in the chemical reaction model and a coefficient corresponding to the concentration or activity value of the sample component is stored in the theoretical parameter storage means. Using the process and parameters, the time series data of the chemical reaction model is calculated by the theoretical time series data calculation means, and the theoretical time series data calculated by the theoretical time series data calculation means is used as the reference time series data. And a step of setting by time-series data setting means.
[0017]
In the abnormality detection method, one or a plurality of time series data obtained by measuring a known concentration sample whose concentration or activity value of the target component is known is stored in a known concentration sample time series data storage means. Using the time series data stored in the known concentration sample time series data storage means, calculating the parameters by the theoretical parameter calculation means, and setting the calculated parameters in the theoretical parameter storage means by the theoretical parameter setting means It has the characteristics.
[0018]
FIG. 1 is a diagram illustrating a configuration example of an abnormality detection system 100 according to an embodiment of the present invention.
[0019]
A system 100 includes a control unit 101, an in-device reaction process data extraction unit 102, a temporary storage device 103, a known concentration reaction process data storage device 104, a theoretical parameter calculation unit 105, a theoretical parameter DB writing unit 106, Theoretical parameter DB 107, theoretical parameter DB reading unit 108, theoretical reaction process data calculation unit 109, reference reaction process data setting unit 110, reference reaction process data storage device 111, abnormality determination unit 112, and determination result output unit 113.
[0020]
Although the system 100 is described as a hardware configuration, the function of the system 100 may be configured by software. Further, the system 100 can communicate with the automatic analyzer 130 via the network 120. In addition, the system 100 can communicate with the input / output terminal 140 via the network 120. The network 120 is premised on a network in an examination facility, but may be a network in a medical facility having an examination department.
[0021]
The automatic analyzer 130 stores the reaction process data in the reaction process data storage device 131. The reaction process data stored in the reaction process data storage device 131 can be extracted by the in-device reaction process data extraction unit 102 via the network 120. The reaction process data stored in the reaction process data storage device 131 can be viewed on the input / output terminal 140 via the network 120. Although the automatic analyzer 130 is described as hardware different from the system 100, the configuration of the system 100 may be configured in the automatic analyzer 130.
[0022]
It is assumed that the reaction process data storage device 131 is configured in the automatic analyzer 130, but the automatic analyzer 130 and the reaction process data storage device 131 may be configured by different hardware. Further, the reaction process data storage device 131 may be configured in the system 100. Further, the reaction process data storage device 131 may be configured in the input / output terminal 140.
[0023]
The input / output terminal 140 is assumed to be an information device such as a personal computer having an input function of a keyboard and a mouse and an output function of a CRT display, but may have other input / output functions. The input / output terminal 140 may be a terminal equipped with a web browser function. Further, although the input / output terminal 140 is described as hardware different from the system 100, the input / output function of the input / output terminal 140 may be installed in the system 100. Further, although the input / output terminal 140 is described as hardware different from the automatic analyzer 130, the input / output function of the input / output terminal 140 may be mounted on the automatic analyzer 130. The input / output terminal 140 is assumed to be a personal computer, but may be a clinical examination system. The user of the input / output terminal 140 is assumed to be an operator of the automatic analyzer 130 such as an inspection engineer, but may be another user such as a maintenance person of the automatic analyzer 130. The system 100, the automatic analyzer 130, and the input / output terminal 140 may be configured as a single piece of hardware.
[0024]
FIG. 2 is a diagram illustrating a configuration example of the automatic analyzer 130 according to the embodiment of this invention.
[0025]
The automatic analyzer 130 includes a light source lamp 201, a thermostatic chamber 202, a cell 203, a sample dispensing nozzle 204, a first reagent dispensing nozzle 205a, a second reagent dispensing nozzle 205b, a stirring rod 206, A spectroscope 207, a detector 208, an amplifier 209, and an A / D converter 210 are included. At the time of analysis, white light (all wavelengths) emitted from the light source lamp 201 passes through the cell 203 attached to the thermostat 202 and enters the spectroscope 207, and a specific single wavelength component that differs depending on the analysis item is detected by the detector 208. The light is received, amplified by the amplifier 209, A / D converted by the A / D converter 210, and output as absorbance.
[0026]
The cell is a reaction vessel, and a sample is dispensed from the sample dispensing nozzle 204, a first reagent is dispensed from the first reagent dispensing nozzle 205a, and a second reagent is dispensed from the second reagent dispensing nozzle 205b, and stirred. A chemical reaction occurs inside the cell by being stirred by the rod 206. The chemical reaction at this time can be converted into the concentration or activity value of the analyte in the sample by measuring the absorbance over time (photometry).
[0027]
Hereinafter, the operation of the automatic analyzer 130 during analysis will be described.
(1) First, the cell 203 is filled with water (pure water), and water blank photometry is performed. This value is a standard for the absorbance measured thereafter.
(2) The water in the cell 203 is discharged, and a sample is dispensed into the cell 203.
(3) Dispense the first reagent into the cell 203 and stir.
(4) Measure the mixed solution of the sample and the first reagent a plurality of times at regular time intervals.
(5) Dispense the second reagent into the cell 203 and stir.
(6) Measure the mixed solution of the sample, the first reagent, and the second reagent a plurality of times at regular time intervals.
(7) After performing photometry a total of 34 times in a reaction time of 10 minutes, the cell 203 is washed and the analysis is terminated.
[0028]
At this time, the reaction from the addition of the first reagent to the addition of the second reagent is called a first reaction, and the reaction from the addition of the second reagent to the end of photometry is called a second reaction. Some analysis items use only the first reagent and do not execute the operations (5) and (6). As described above, the reagent is classified into one-reagent system analysis and two-reagent system analysis according to the number of reagents used.
[0029]
The measured absorbance is converted into a concentration or activity value C by a conversion formula as shown in (Equation 1) based on the law (Labbert-Beer's law) in which the absorbance is proportional to the concentration of the substance. Saved in.
C = aA + b (Equation 1)
In (Equation 1), a and b are preset values and differ depending on the analysis item. A is an absorbance change amount at a certain photometry point or an absorbance change rate between a plurality of photometry points. As described above, as a method for converting the absorbance to the concentration or activity value of the target component in the sample, there are an end method using the absorbance change amount and a rate method using the absorbance change rate.
[0030]
FIG. 3 is a diagram showing a reaction process data example 300 of TP (total protein) as an example of an analysis item for which the endo method is used in the embodiment of the present invention.
[0031]
TP is a one-reagent analysis, and a peptide (—CO—NH—) in a sample and a copper ion (Cu ²⁺ ) Calculates a concentration of TP by measuring a change in the concentration of the chelate compound using a chemical reaction that changes to a chelate compound (a metal ion bonded to an organic molecule). Therefore, the photometric wavelength uses the absorption peak wavelength of the product chelate compound.
[0032]
As shown in the reaction process data example 300, immediately after the start of the reaction, the chelate compound increases and the absorbance rapidly increases. However, immediately after the end of the analysis, all the peptides in the sample are changed to chelate compounds, so that the reaction hardly proceeds and no increase in absorbance is observed. By using the amount of change in absorbance at the end of the analysis, the concentration of TP is calculated.
[0033]
FIG. 4 is a diagram showing a reaction process data example 400 of LDH (lactate dehydrogenase) as an example of an analysis item for which the rate method is used in the example of the present invention.
[0034]
LDH is a two-reagent analysis, as shown in chemical reaction formula 401, using the enzyme LDH in the sample as a catalyst, pyruvate in the second reagent is converted to lactic acid, and NADH (β-nicotinamide adenine diamine in the first reagent is used. The activity value of LDH is calculated by measuring the reduction rate of NADH using a chemical reaction in which the nucleotide reduced form is changed to NAD (β-nicotinamide adenine dinucleotide oxidized form). For this reason, the absorption peak wavelength of NADH is used as the photometric wavelength.
[0035]
As shown in reaction process data example 400, the absorbance in the first reaction is substantially constant, but in the second reaction, the absorbance decreases with a constant slope until the end of the analysis. This shows that NADH is gradually decreased according to the activity of LDH in the sample. The activity value of LDH in the sample is calculated by using the absorbance change rate during the second reaction.
[0036]
In the present embodiment, this chemical reaction process is modeled by an exponential function with time t as a variable as shown in (Equation 2), assuming that the sample, reagent, and apparatus are all measured in an abnormal state. The reaction process data expressed by (Equation 2) is called theoretical reaction process data.
A (t) = A ₀ + A ₁ (1-e ^-Kt ) (Equation 2)
e is the base of the natural logarithm. At this time, k is the rate of reaction, A ₀ Is the initial absorbance at the start of the reaction, A ₁ Is a parameter indicating the concentration of the target component (hereinafter referred to as a theoretical parameter). Where A ₀ , A ₁ Varies depending on the sample, but k is uniquely determined by the type of chemical reaction, that is, the combination of the analysis item and the reagent.
[0037]
In this way, by modeling the chemical reaction process with parameters, the chemical reaction process can be easily expressed with respect to samples of all concentrations for each target component.
[0038]
FIG. 5 is a flowchart 500 when setting theoretical parameters based on reaction process data obtained by measuring a quality control sample in the embodiment of the present invention.
[0039]
First, from the reaction process data stored in the reaction process data storage device 131, the reaction process data obtained by measuring the same quality control sample with respect to the same analysis item in a state in which there is no abnormality in all of the sample, reagent, and apparatus. Step 501 for selecting N (N ≧ 1) is executed.
Next, the control unit 101 activates the in-device reaction process data extraction unit 102, extracts the N reaction process data selected in step 501 from the reaction process data storage device 131, and stores it in the temporary storage device 103. Step 502 is executed.
[0040]
Next, the control unit 101 activates the theoretical parameter calculation unit 105 and determines whether the N reaction process data on the temporary storage device 103 are analysis items using the end method or analysis methods using the rate method. Step 503 is executed.
[0041]
When the rate method is determined in step 503, the control unit 101 activates the theoretical parameter calculation unit 105, and the N reaction process data on the temporary storage device 103 are used as the absorbance of the first reaction of each reaction process data. For each reaction process data, the absorbance at the photometric points of the first reaction and the second reaction is normalized and stored in the temporary storage device 103 so that the average value of all the N reaction process data becomes a constant value. Step 504 is executed.
[0042]
Next, a specific example of step 504 is shown. The average value of the absorbance of the first reaction after normalization is “c”, and the reaction process data on the temporary storage device 103 is the average value of the absorbance of the first reaction “c ′”, the amount of the reaction solution during the first reaction. "V ₁ , Reaction volume “V” in the second reaction ₂ ”, In step 504, the theoretical parameter calculation unit 105 adds the value“ c−c ′ ”to the absorbance at the photometric point of the first reaction, and adds the value“ (c −c ′) × V ₁ / V ₂ Is added. Step 504 makes it possible to correct the absorbance shift during the second reaction due to the influence of the absorbance shift during the first reaction in the rate method.
[0043]
Next, the control unit 101 activates the theoretical parameter calculation unit 105, and with respect to N reaction process data on the temporary storage device 103, an average value A for each photometric point. _AVE (T), standard deviation A _SD (T) is calculated, and step 505 of storing in the known concentration reaction process data storage device 104 is executed.
[0044]
Next, the control unit 101 activates the theoretical parameter calculation unit 105, and the A on the known concentration reaction process data storage device 104. _AVE From (t) to A of photometric points t to t + 1 as shown in (Equation 3) _AVE Change amount ΔA of (t) _AVE Step (506) is performed in which (t) is calculated and stored in the known concentration reaction process data storage device 104.
ΔA _AVE (T) = A _AVE (T + 1) -A _AVE (T + 1) (Equation 3)
FIG. 6 is a diagram illustrating an example 600 of the known concentration reaction process data storage device 104 after execution of step 506 of FIG.
[0045]
In Example 600, together with the number indicating the photometric point (t = 1, 2, 3,..., 33, 34), the time from the addition of the first reagent at each photometric point, A _AVE (T), A _SD (T), ΔA _AVE (T) is stored.
[0046]
Next, the control unit 101 activates the theoretical parameter calculation unit 105 to determine whether the N reaction process data on the temporary storage device 103 are the analysis items for the 1-reagent system analysis or the analysis items for the 2-reagent system analysis. Step 507 is executed.
When it is determined in step 507 that the two-reagent system analysis is performed, the control unit 101 activates the theoretical parameter calculation unit 105, and sets the photometric point at the first reaction on the known concentration reaction process data storage device 104 to the subsequent processing. Step 508 of assigning a flag indicating exclusion is executed.
[0047]
Next, the control unit 101 activates the theoretical parameter calculation unit 105 and adds a flag indicating that it is excluded in the subsequent processing to a preset photometry point on the known concentration reaction process data storage device 104. Execute.
[0048]
FIG. 7 is a diagram illustrating an example 700 of the known concentration reaction process data storage device 104 after the execution of step 509 in FIG. 5.
[0049]
In Example 700, a flag indicating that the photometric points t = 1 to 16 at the first reaction and three photometric points t = 17 to 19 immediately after the addition of the second reagent are excluded in the subsequent processing is given. .
[0050]
Next, the control unit 101 activates the theoretical parameter calculation unit 105, and the A on the known concentration reaction process data storage device 104. _AVE From (t), an A as shown in (Equation 4) is obtained for every photometric point to which no flag indicating exclusion is given. ₁ K in which the variance of (t) is below a preset range, and A at that time ₁ Step 510 of calculating the average of (t) is executed. In this case, k is a theoretical parameter k, and A ₁ The average of (t) is the theoretical parameter A ₁ And A ₁ The sign of is + when the reaction process data is monotonically increasing, and is − when the reaction process data is monotonically decreasing.
A ₁ (T) = ± √ ({ΔA _AVE (T)} ² /
{(-E ^{−k (t + 1)} )-(-E ^-Kt )} ² ) (Equation 4)
As a result, in the second reaction that is directly used for the conversion to the concentration or activity value of the analysis item, the absorbance of the first reaction and the photometric point at which the temperature in the reaction vessel changes temporarily, such as immediately after addition of the reagent, are excluded. An appropriate theoretical parameter can be calculated by using only a photometric point with a constant temperature in the reaction vessel.
[0051]
Next, the control unit 101 activates the theoretical parameter calculation unit 105 and calculates k and A calculated in step 510. ₁ Is substituted into (Equation 2), and the value of (Equation 2) and A on the known concentration reaction process data storage device 104 are determined for every photometric point that is not given a flag indicating exclusion. _AVE A that minimizes the sum of squared errors with (t) ₀ Step 511 of calculating is executed. A at this time ₀ The theoretical parameter A ₀ And
[0052]
Next, the control unit 101 activates the theoretical parameter DB writing unit 106, and stores the theoretical parameters k, A calculated in Step 510 and Step 511 in the theoretical parameter DB 107. ₀ , A ₁ Step 512 is stored.
[0053]
FIG. 8 is a diagram showing an example 800 of the theoretical parameter DB 107 after the execution of step 512 in FIG.
[0054]
The theoretical parameter DB 107 stores a field 801 for storing a quality control sample ID for identifying the type of the quality control sample, a field 802 for storing the analysis item name, and a reagent ID for identifying the type of the first reagent and the second reagent. After normalizing in step 504, the fields 803 and 804 for storing, the field 805 for storing the wavelength used for the analysis, the field 806 for storing the concentration or activity value of the analysis item, and the rate method analysis item are normalized. A field 807 for storing the mean value of the absorbance of the first reaction, and theoretical parameters k, A ₀ , A ₁ Are stored in fields 808, 809, and 810, respectively, and fields 811 and 812 that store a and b in Equation (1) for determining a conversion formula to a concentration or activity value, respectively.
[0055]
In Example 800, the reagent “R _A 1 ”and“ R _A 2 ”and the wavelength“ λ _A When the quality control sample identified by the quality control sample ID “A” having the activity value of LDH analyzed by “185” is normalized so that the average value of the absorbance of the first reaction is “10000”, k = 0.00061 ”,“ A ₀ = 14350 "," A ₁ = -3450 ". In Example 800, the reagent “R _B 1 ”and the wavelength“ λ _B For the quality control sample identified by the quality control sample ID “B” whose TP concentration is “6.8”, the theoretical parameters are “k = 0.00485”, “A ₀ = 1255 ”,“ A ₁ = 810 ".
[0056]
Through the above steps 501 to 512, the theoretical parameters k, A suitable for each measurement condition and measurement environment of the automatic analyzer or the like. ₀ , A ₁ Can be set.
[0057]
In the present embodiment, in step 509, a flag indicating that it is excluded in the subsequent processing is assigned to a preset photometry point. . As a result, a photometric point at which the temperature in the reaction vessel changes depending on the analysis item, or a photometric point at which the temperature in the reaction vessel is constant even when the temperature is different until the temperature becomes constant ₁ It is possible to calculate an appropriate theoretical parameter using only (t).
[0058]
Next, processing for determining the presence or absence of abnormality in reaction process data based on theoretical parameters obtained by measuring a quality control sample will be described.
[0059]
FIG. 9 shows a flowchart 900 for determining abnormality of reaction process data in the embodiment of the present invention.
[0060]
First, the control unit 101 activates the theoretical parameter DB reading unit 108 and executes Step 901 for storing information of each field of the theoretical parameter DB 107 in the temporary storage device 103.
[0061]
Next, the control unit 101 activates the in-device reaction process data extraction unit 102 and determines to-be-determined reaction process data A for determining whether the reaction process data storage device 131 is normal or abnormal. _Covered (T) and A _Covered Step 902 for extracting the measured value of (t) and storing it in the temporary storage device 103 is executed.
[0062]
Next, the control unit 101 activates the abnormality determination unit 112, and A on the temporary storage device 103. _Covered Step 903 is executed to determine whether (t) is an analysis item using the end method or an analysis item using the rate method.
[0063]
If it is determined in step 903 that the rate method is used, the control unit 101 activates the abnormality determination unit 112, and A on the temporary storage device 103. _Covered (t) is normalized so that the average value of the absorbance of the first reaction matches the value stored in the field 807 of the theoretical parameter DB 107, respectively. Step 904 of saving in the temporary storage device 103 is executed.
[0064]
Next, a specific example of step 904 is shown. The value “c” is stored in the field 807 of the theoretical parameter DB 107, and A in the temporary storage device 103 is stored. _Covered For (t), the average value “c ′” of the absorbance of the first reaction and the reaction solution volume “V” during the first reaction. ₁ , Reaction volume “V” in the second reaction ₂ ”In step 904, the abnormality determination unit 112 determines that A _Covered The value “c−c ′” is added to the absorbance at the photometric point of the first reaction in (t), and A _Covered The value “(c−c ′) × V” represents the absorbance at the photometric point of the second reaction in (t). ₁ / V ₂ Is added. In step 904, in the rate method, it is possible to correct the absorbance shift during the second reaction due to the influence of the absorbance shift during the first reaction.
[0065]
Next, the control unit 101 activates the theoretical parameter calculation unit 105, and the measured value C of the quality control sample when the theoretical parameter is calculated on the temporary storage device 103. _Spirit A against _Covered Measured value C of (t) _Covered Ratio r = C _Covered / C _Spirit Based on this, A as shown in (Equation 5) _Covered Theoretical reaction process data A of (t) _Obedience For (t), A _Covered , A ₁ Step 905 is calculated and stored in the temporary storage device 103.
A _Obedience (T) = A _Covered + A ₁ (1-e ^-Kt ) (Equation 5)
The specific procedure of step 905 is shown below.
[0066]
Quality control sample and A _Covered The change in absorbance or the rate of change in absorbance used for conversion of the measured value of (t) is respectively expressed as A _Spirit , A _Covered Then, from (Equation 1), the relationships of (Equation 6) and (Equation 7) hold.
aA _Covered + B = r × (aA _Spirit + B) (Formula 6)
A _Covered = R * A _Spirit + (R-1) * b / a (Expression 7)
At this time, according to Lambert-Beer's law, (Expression 7) can be applied to all photometric points t. From this, A _Spirit By substituting (Equation 2) into _Obedience (T) can be expressed as (Equation 8).
A _Obedience (T) = r × A ₀ + (R-1) * b / a + r * A ₁ (1-e ^-Kt ) ... (Equation 8)
From the above, by comparing the right sides of (Equation 5) and (Equation 8), A _Covered = R * A ₀ + (R−1) × b / a, A ₁ = R * A ₁ Can be calculated.
[0067]
In this way, in step 905, theoretical reaction process data used to determine whether there is an abnormality in a sample whose concentration is unknown, such as a patient specimen sample, can be calculated regardless of the concentration or activity value of the target component.
[0068]
Next, the control unit 101 activates the theoretical reaction process data calculation unit 109, and the theoretical parameters k, A on the temporary storage device 103 are activated. _Covered , A ₁ For each photometric point to which a flag indicating that the value substituted for (Equation 5) is excluded on the known concentration reaction process data storage device 104 is not assigned, A _Covered Theoretical reaction process data A of (t) _Obedience Step 906 is calculated and stored in the temporary storage device 103 as (t).
[0069]
Next, the control unit 101 activates the reference reaction process data setting unit 110, and for the photometric points to which the flag indicating exclusion on the known concentration reaction process data storage device 104 is given, from (Equation 7) r × A _AVE (T) + (r−1) × b / a is set to A on the temporary storage device 103 for other metering points. _Obedience Step 907 of saving (t) as reference reaction process data in the reference reaction process data storage device 111 is executed.
[0070]
Next, the control unit 101 activates the reference reaction process data setting unit 110, and the A on the known concentration reaction process data storage device 104. _SD (T) is read, and the tolerance A for all photometric points _E (T) = 2 × A _SD Step 908 of calculating (t) and storing it in the reference reaction process data storage device 111 is executed.
[0071]
Next, the control unit 101 activates the abnormality determination unit 112, and A on the temporary storage device 103. _Covered (T) and A on the reference reaction process data storage device 111 _Obedience (T), A _E Using (t), A _Obedience (T) and A _Covered Difference from (t) D (t) = A _Obedience (T) -A _Covered For (t), “−A” for all photometric points t. _E (T) ≦ D (t) ≦ A _E Step 909 is executed to determine whether (t) "is true or false.
[0072]
Next, the control unit 101 activates the abnormality determination unit 112, and A on the temporary storage device 103. _Covered Using (t), the absolute value ΔD (t) = | D (t + 1) −D (t) | of the difference between D (t) and D (t + 1) for each of the adjacent photometric points (t and t + 1) Step 910 is executed to determine the authenticity of “ΔD (t) ≦ a preset threshold value” for all photometric points t.
[0073]
Next, the control unit 101 activates the abnormality determination unit 112 and executes Step 911 for determining whether or not both are true in Step 909 and Step 910.
[0074]
As a result of step 911, when it is determined that both are true in steps 909 and 910, the control unit 101 activates the determination result output unit 113 and outputs the determination result “normal” to the input / output terminal 140. 912 is executed.
[0075]
In step 909 and step 910, when it is determined to be false in at least one of the steps, the control unit 101 activates the determination result output unit 113 and outputs the determination result “abnormal” to the input / output terminal 140. 913 is executed.
[0076]
From step 901 to step 913, the reference values for all photometric points are set and compared for each concentration according to the chemical reaction model. Regardless, it is possible to detect an abnormality that occurs during measurement.
[0077]
FIG. 10 is a diagram showing a screen example 1000 of the input / output terminal 140 when the result of step 909 in FIG. 9 is false.
[0078]
In the screen example 1000, D (t) is -A at some photometric points t due to an abnormality that affects the concentration of the reacting substance (abnormal dispensing amount of sample or reagent, thinning of the reagent, etc.). _E It is displayed that it is smaller than (t). As described above, it is possible to detect reaction process data including an abnormality that affects the concentration of the reacting substance by step 909.
[0079]
FIG. 11 is a diagram showing a screen example 1100 of the input / output terminal 140 when the result of step 910 in FIG. 9 is false.
[0080]
In the example screen 1100, it is displayed that ΔD (t) exceeds the threshold at some photometric points t due to abnormalities that cause variations in absorbance (photometer fluctuation, noise, etc.). As described above, the step 910 makes it possible to detect reaction process data including an abnormality that causes a variation in absorbance (photometer fluctuation, noise, etc.).
In this embodiment, in step 908, the tolerance A _E (T) = 2 × A _SD (T), but can be arbitrarily set by the user. As a result, permissible errors are set to optimize the F / P ratio (False Positive: an error that determines normal is abnormal) and F / N ratio (False Negative: an error that determines abnormal is normal) for each analysis item. It becomes possible to do.
[0081]
In step 910, the threshold for comparing ΔD (t) can be arbitrarily set by the user. As a result, permissible errors are set to optimize the F / P rate (False Positive: an error that determines normal is abnormal) and F / N rate (False Negative: an error that determines abnormal is normal) for each analysis item. It becomes possible to do.
[0082]
The parameters that represent the chemical reaction model are calculated by the abnormality detection system described above, and compared with the reaction process data using this chemical reaction model as a reference, so the timing at which changes appear in the reaction process data for samples of any concentration Regardless, it is possible to detect an abnormality that occurs during measurement.
[0083]
【The invention's effect】
ADVANTAGE OF THE INVENTION According to this invention, the abnormality detection system and abnormality detection method which detect abnormality which generate | occur | produces at the time of the measurement of the target component in a sample can be provided.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a configuration example of an abnormality detection system according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating a configuration example of an automatic analyzer according to an embodiment of the present invention.
FIG. 3 is a diagram showing an example of reaction process data of TP as an example of an analysis item for which the end method is used in the embodiment of the present invention.
FIG. 4 is a diagram showing an example of LDH reaction process data as an example of an analysis item for which the rate method is used in the embodiment of the present invention.
FIG. 5 is a diagram showing a flowchart when setting theoretical parameters based on reaction process data obtained by measuring a quality control sample in the embodiment of the present invention.
6 is a diagram showing an example of a known concentration reaction process data storage device after execution of step 506 in FIG. 5;
7 is a diagram showing an example of a known concentration reaction process data storage device after execution of step 509 in FIG. 5; The example of the known density | concentration reaction process data storage device after step 509 execution.
FIG. 8 is a diagram showing an example of a theoretical parameter DB after execution of step 512 in FIG. 5;
FIG. 9 is a flowchart for determining abnormality of reaction process data in the embodiment of the present invention.
10 is a diagram showing an example of a screen of the input / output terminal when the result of step 909 in FIG. 9 is false.
11 is a diagram showing an example of a screen of the input / output terminal when the result of step 910 in FIG. 9 is false.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 100 ... Abnormality detection system, 101 ... Control part, 102 ... In-device reaction process data extraction part, 103 ... Temporary storage device, 104 ... Known concentration reaction process data storage device, 105 ... Theoretical parameter calculation part, 106 ... Theoretical parameter DB writing 107: Theoretical parameter DB, 108: Theoretical parameter DB reading unit, 109: Theoretical reaction process data calculation unit, 110: Reference reaction process data setting unit, 111: Reference reaction process data storage device, 112 ... Abnormality determination unit, 113 DESCRIPTION OF SYMBOLS Output unit 120 ... Network in test facility 130 ... Automatic analyzer 131 ... Reaction process data storage device 140 ... Input / output terminal 201 ... Light source lamp 202 ... Constant temperature bath 203 ... Cell 204 ... Sample dispensing nozzle, 205a ... first reagent dispensing nozzle, 205b ... second reagent dispensing nozzle, 2 6 ... Stirring rod, 207 ... Spectroscope, 208 ... Detector, 209 ... Amplifier, 210 ... A / D converter, 300 ... Example of reaction process data of TP, 301 ... Chemical reaction formula used for calculation of concentration of TP , 400... Example of reaction process data of LDH, 401... Chemical reaction formula used for calculation of LDH activity value, 500... Flow chart when setting theoretical parameters, 501. Step of selecting N reaction process data, 502... Extracting N reaction process data from an automatic analyzer, 503. Step of determining whether it is an end method or a rate method, 504... N reaction process data. Normalizing step, 505... Average value A _AVE (T) and standard deviation A _SD Step of calculating (t), 506... ΔA _AVE (T) calculating step, 507... Determining whether one reagent system or two reagent system, 508... Giving an exclusion flag to the photometric point of the first reaction, 509. Assigning step 510 ... theoretical parameters k and A ₁ 511 ... A ₀ , 512... Storing the calculated theoretical parameter in the theoretical parameter DB, 600... Example of known concentration reaction process data storage device after execution of step 506, 700... Storing known concentration reaction process data after execution of step 509. Example of apparatus, 800 ... Example of theoretical parameter DB after execution of step 512, 801 ... Field for storing quality control sample ID for identifying the type of quality control sample, 802 ... Field for storing analysis item name, 803 ... First Field for storing reagent ID for identifying reagent type, 804... Field for storing reagent ID for identifying second reagent type, 805... Field for storing wavelength used for analysis, 806... Concentration or activity of analysis item A field for storing a measured value, 807... First after normalization in step 504 Field for storing the average value of absorbance of the response, fields for storing the 808 ... theoretical parameter k, 809 ... theoretical parameters A ₀ ,...,..., Theoretical parameter A ₁ ,..., 811... A field for storing a in the equation (1) for determining the conversion formula to the concentration or activity value, 812... B for determining the conversion equation to the concentration or activity value (the equation 1) 900, a flow chart for determining abnormality of reaction process data, 901, a step of reading theoretical parameter DB, 902, reaction process data A to be determined _Covered (T) and a step for extracting the measured value, 903... A step for determining whether the method is an end method or a rate method, 904. _Covered Normalizing (t), 905... R and A _Covered And A ₁ 906... A _Covered Theoretical reaction process data A of (t) _Obedience Step of calculating (t), 907... A _Obedience A step of setting (t) as reference reaction process data, 908... _E (T) calculating step, 909... Determining whether or not there is an abnormality affecting the concentration of the reacting substance, 910... Determining whether or not there is an abnormality causing variation in absorbance, 911. Step for determining whether or not there is an abnormality, 912 ... Step for outputting the determination result "normal" to the input / output terminal, 913 ... Step for outputting the determination result "abnormal" to the input / output terminal, 1000 ... Result of step 909 is false Screen example of input / output terminal at 1100 ... Screen example of input / output terminal when result of step 910 is false.

Claims (2)

An abnormality detection system for detecting anomalies in time-series data of photometric values obtained by optically measuring a mixed liquid obtained by mixing one or more types of reagents used for measuring the concentration or activity value of a target component in a sample and the sample. Because
Reference time series data storage means for storing the time series data corresponding to a chemical reaction model when the analysis of the sample is normally performed as reference time series data;
An abnormality determination means that compares the time series data with the reference time series data and determines the presence or absence of an analysis abnormality of the sample ;
Theoretical parameter storage means for storing a parameter including a coefficient according to a chemical reaction rate for each target component used in the chemical reaction model, and a coefficient according to a concentration or activity value of a component of the sample;
Theoretical time series data calculating means for calculating time series data of the chemical reaction model using the parameters;
An anomaly detection system comprising reference time series data setting means for setting theoretical time series data calculated by the theoretical time series data calculation means as the reference time series data .   2. The abnormality detection system according to claim 1, wherein one or a plurality of said time series data obtained by measuring a known concentration sample having a known concentration or activity value of said target component is stored. Data storage means, theoretical parameter calculation means for calculating the parameters using the time series data stored in the known concentration sample time series data storage means, and theory for setting the calculated parameters in the theoretical parameter storage means And a parameter setting means.

Images