JP4521339B2 - Data processing method with check function in integral data operation method - Google Patents

Description

  The present invention relates to an automatic analyzer that performs qualitative and quantitative analysis of biological samples such as blood and urine, and in particular, performs qualitative and quantitative analysis based on detection of a labeling reagent that specifically binds to an analysis target component (analyte). The present invention relates to an automatic analyzer.

  As one method of measuring components to be measured contained in body fluids such as blood and urine, a labeling reagent that reacts specifically with the target chemical component to be analyzed (a radioactive substance as a label, a substance that emits light by a chemical reaction, an excitation light There is known a method of revealing using a substance that emits fluorescence or phosphorescence upon irradiation. The chemical characteristics obtained by this reaction, that is, a phenomenon such as light absorption or light emission, is converted into an electric signal using a light receiving element (photomultiplier tube, solid-state image pickup device such as CCD) utilizing the photoelectric effect, Amplify to. Today, it is common to digitize such information by analog / digital conversion so that it can be processed by a computer. In such numerical processing, a method of time-integrating optical intensity values on a computer is common. In general, the integration value is processed afterwards, and for the data used for integration, this improves the resolution by integrating the limits of the measurement range of the sensor itself in a spatial and temporal manner. it can.

  However, particularly when measuring a small amount of light emission using a photomultiplier tube, there is a concern that an error may occur in the measurement value due to external noise or the like. In order to deal with such a concern, a method described in Patent Document 1 has been proposed.

JP 2001-91466 A

  The method described in Patent Document 1 is based on the assumption that the distribution of emission intensity measurement values is basically a normal distribution, and measurement values that do not lie on the normal distribution are determined to have noise superimposed on them. The data is corrected. However, actual measured values are those in which the emission intensity varies with time. For example, when the reagent adheres to the inner wall of the container and the adhering reagent drops onto the reaction solution during the measurement, the measured value itself is a data including an error even if it follows a normal distribution. I found out that

  An object of the present invention is to provide an automatic analyzer that can obtain measurement data with higher reliability by detecting an error factor of the measurement as described above.

  The configuration of the present invention for achieving the above object is as follows.

  A means for providing a trigger from outside the system; a detection means for detecting an optical signal generated from within the system based on the trigger; and a recording means for recording the intensity of the optical signal generated from within the system over time. Further, the comparison means for comparing the temporal position of the peak of the optical signal with a preset temporal position, and the evaluation result is associated with the information of the measured optical signal based on the comparison result of the comparison means. And an automatic analyzer having storage means for storing.

  The trigger given from outside the system is specifically a reagent for starting a specific reaction (including a luminescence reaction), an electric field, a magnetic field, or a combination thereof. Such a trigger causes a luminescent reaction in the system (reaction vessel, flow cell, etc.). Such a luminescence reaction is detected by a detecting means for converting an optical signal into an electric signal, such as a photomultiplier tube, a CCD, or a MOS sensor. Such a luminescence reaction is generally continued on the order of msec, and the change over time of such a reaction is stored in the storage means. As the storage means, general storage means such as a semiconductor memory or a hard disk can be used. After giving the trigger, it is possible to compare the position at which the emission signal reaches a peak with information on a predetermined peak position, and determine whether the reaction has proceeded normally or whether the reaction has become abnormal for some reason. In specifying the peak position, a known data processing means can be used.

  In this case, it may be determined whether or not to perform peak position determination using the integrated value of the optical signal. That is, when the integral value is small, it may be difficult to specify the peak position, or the error of the peak position may be large, and the merit of performing the peak position determination may be small. In such a case, if the peak position is forcibly determined, an incorrect determination may be made.For example, by giving suggestions to perform the same test again, a determination based on the analysis result is made. May be clearly marked as dangerous.

  Further, a function of adding an alarm to the measurement data and displaying it based on the information stored in the storage means for storing the evaluation results in association with each other may be provided. Specifically, when displaying a list of measurement results, one column may be provided to display the presence or absence of an alarm, or the display color may be changed to alert the operator.

  In addition, the measurement result may include a display unit that displays a change over time in the intensity of the optical signal recorded in the recording unit in a graph or the like.

  In an automatic analyzer, a highly reliable result can be obtained.

  The chemical reaction takes place depending on diffusion and the probability of reaction between molecules. In principle, in the measurement form used for chemical analysis, stirring is sufficiently performed and measurement is performed with a uniform measurement system. In this case, if there is a predetermined concentration, the frequency often follows a normal distribution. However, in an automatic analysis system used for a clinical examination, there are time restrictions and structural restrictions, and therefore, it is general to design within a range that can be regarded as practically uniform. In such an apparatusized reaction system, the reaction occurs within a certain range as long as the reaction is performed under the expected conditions even if the unique conditions of each reaction system are taken into consideration. It is expected.

  However, it is beyond the expectation that deviates from this certain range, that is, when the pH of the reaction solution has changed significantly, or when the stirring process for making the reaction uniform has not been performed sufficiently. An abnormal reaction may occur. These may include hardware problems such as mechanism position adjustment and flow path contamination, reagent deterioration, malfunctions due to reagent handling, and accidental process abnormalities. If these abnormalities occur, they must be reported promptly and corrective actions taken.

  Furthermore, in clinical tests, specimens used for reactions contain a wide variety of compounds prescribed or inoculated by the patient from whom the specimen originates, or the concentration of a specific component in the specimen varies significantly depending on the disease. There is a concern that the reaction system may deviate from the initially prepared tolerance. However, it is practically impossible to cover all the possibilities in developing the reagents constituting the system and the reaction system. Hemolysis, bilirubin, and chyle have been established as evaluation items that have significant effects on biochemical items. However, interfering substances have not been established for immunization.

  In this situation, when examining the reaction in microanalysis, the final reaction uses a single and highly sensitive reaction in the measurement system that includes separation by Bound / Free. Can be defined in advance. Therefore, it can be considered that an expected reaction can be obtained on the contrary from the integrated value finally output. In other words, even if the same output value by the integrated value is obtained, there are cases where there is a significant difference in the contents, and these are positively excluded to guarantee the quality of the obtained data. It is considered possible.

  Embodiments of the present invention will be described below with reference to the drawings.

  An example of a method for measuring a change in light emission amount with respect to time will be described. The chemiluminescence reaction system described here is a method generally called a chemiluminescence method. For example, chemical issuance reaction by mixing luminol and hydrogen peroxide. The reaction system used in the actual clinical examination is described in, for example, Japanese Patent Application Laid-Open No. 2003-50204 (Acridinium ester label having a hydrophilic modifier). Edited by Katsumi Tokumaru, ISBN 4-477-00357-9 Dai Nippon Books.

  In these chemiluminescent reactions, the luminescent substrate is said to emit light in the process of being oxidized by an oxidizing agent such as hydrogen peroxide in a solution at an optimum pH.

  A configuration example of the apparatus is shown in FIG. The reaction vessel 101 is a magnetic particle 102 that has been B / F separated in the previous process, and contains the adsorbed luminescent material. Here, as a first step, the pH adjustment liquid stored in the pH adjustment liquid tank 105 is discharged from the nozzle 103 by the pump 104. By this step, the magnetic particles 102 are evenly dispersed in the solution. After the time necessary for sufficient dispersion, the emission initiator is further discharged from the emission initiator tank 108 by the nozzle 107 by the nozzle 106. The pump 104 and the pump 107 are controlled by a control device 109. By mixing the emission initiator, the shutter 110 is opened synchronously, and the photomultiplier tube 111 amplifies the photocurrent by light emission, electrically amplifies the photocurrent via the LOG amplifier 112, and an analog-digital converter The digital conversion is performed by the (ADC) 113 and the digital value is stored by the data processing unit 114. Here, the photomultiplier tube 111 and the log amplifier operate passively, and the shutter 110 and the analog-digital converter (ADC) 113 are controlled synchronously by the control device. In the analog-to-digital converter (ADC) 113, an average value of the light emission signal is calculated every 10 msec, for example, and the value is stored in the data processing unit 114. That is, 200 points of data can be obtained for a measurement of 2 seconds.

An example of the data obtained is shown in FIG. The horizontal axis is time, and time 0 is the injection timing of the luminescence initiator. At this time, first, before injection of the luminescence initiator, for example, at −80 msec, the shutter 110 is opened to obtain a background before reaction. Data is acquired from a time point of 50 msec before the start of light emission, and a background level 202 is obtained. The background level is determined by a method for obtaining an average value in a section from −50 msec to 0 msec. Next, by injecting the light emission starting material, light emission occurs abruptly and a signal 211 is generated. After reaching the peak 212, the light emission gradually decreases, and data is acquired up to 2000 msec.
When data is acquired in units of 10 msec, 2000 points of data can be acquired. The background value 202 is removed from each point, and the total value is an integrated value 213 indicated by a vertical line. The following are examples of reactions that exhibit different behavior. Reaction 221 is an example in which 500 msec is required to reach peak 222 after the start of the reaction. In this case, the integrated value is equivalent to the integrated value compared with the reaction 221, and the data after the integrated value is treated equally. However, in this reaction, it is considered that the reaction was delayed due to several factors, and separate evaluation is required to show the same reliability as data.

In this embodiment, a logic for detecting a peak is provided. Further, a peak delay allowable range is determined in advance, and alarm information is attached to output data when this range is exceeded. A basic flow is shown in FIG. The signal data 310 is input, and first, an integral value 320 is calculated. Further, feature value extraction 330 is performed in parallel with the signal data. With the feature value as an input, a comment is read out from a determination table prepared in advance and attached to output data. In this example, it is characterized by a delay. For example, if there is a delay of 50 to 100 msec,
If there is a delay from msec to 200 msec, be careful. If there is a delay of 200 msec or more, comment as a measurement failure. In the case of measurement failure, the output data itself may be an error.

  Further, the determination may be made by combining a plurality of feature amounts, for example, an integral value and a peak position. The basic flow is shown in FIG. Following the signal data acquisition 410, first, an integral value calculation 420 is performed. Here, the integrated value is determined. For example, when the range of the integrated value is from 0 to 10 ^ 7, in the range of 0 to 1000, the peak position is not checked, and the calculated value (integrated value) is output 440. . On the other hand, when the integral value is larger than 1000, the peak position calculation 430 is performed, and the peak position determination acquisition 440 is performed by accessing the peak position determination table 490 using the peak position as a key. This is because, in the measurement of the low concentration region, in addition to the target reaction, extremely weak light emission may occur, and there is a phenomenon that a peak does not appear clearly.

  Even if the same integrated value is obtained by such a method, it is possible to determine whether the response is an expected response or an unexpected response, and to output more reliable data. It becomes possible.

  A method for processing data by separating reactions into at least two main components will be described. FIG. 5 shows the two main expected reactions. The horizontal axis is the time axis, and the vertical axis is the relative signal amount normalized by the peak. The handling of the background needs to be adjusted in each measurement system.

  The reference reaction is acquired in advance according to the conditions indicating the respective characteristics. For example, what is necessary is just to acquire on ideal conditions with a sample with high purity. The response curve 511 is a reaction that ends in a short time with a steep peak 512, and the signal value 513 at the end of observation for the peak 512 is relatively small. This reaction is often a reaction with a compound specifically selected to perform strong amplification. This is because, in immunoassay, the emission intensity per unit molecule is large, and the lifetime is required to be somewhat short as a characteristic that is significantly different from the background. On the other hand, the reaction curve 521 is light emission that is not originally intended, and may be a light-emitting substance that is strongly bound to a substance that is eventually included in the reaction system, such as an antibody, or light emission of the substrate itself. In these luminescences, since the time until the reaction is relatively slow, the timing at which the peak 522 is observed is late and the time until the luminescence stops is long. Therefore, compared with the peak 522 at the end of the observation. There is a large 523 emission.

  The ratio of the reaction curves 511 and 521 obtained in advance is determined by linear combination from the actually obtained luminescence curve. Simply, as the linear linear combination, the time at the steep peak 512 and the light emission amount at the end time can be estimated by calculating with a binary linear equation.

  From the viewpoint of separating the reaction into at least two reactions, in addition to the above, attention is paid to the difference in reaction rate between the reaction from A to B and the reaction from B to C in the reaction in three stages of substances A to B and B to C However, a method is also conceivable in which the time course due to the reaction is obtained under the conditions where each material becomes apparent, and these are calculated in combination.

1 is a diagram for explaining an outline of a system configuration in Embodiment 1. FIG. FIG. 5 is a diagram illustrating an example of light emission with respect to time in Example 1; FIG. 3 is a flowchart illustrating a determination process according to the first embodiment. The flowchart which branches a determination process by using the integral value in Example 1 as a key. FIG. 6 is a diagram illustrating an example of light emission with respect to time in Example 2.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Reaction container, 102 ... B / F separated magnetic particle, 103 ... pH adjusting liquid nozzle, 104 ... pH adjusting liquid pump, 105 ... pH adjusting liquid tank, 106 ... Luminescence initiator nozzle,
DESCRIPTION OF SYMBOLS 107 ... Light emission initiator pump, 108 ... Light emission initiator tank, 109 ... Control apparatus, 110 ... Shutter, 111 ... Photomultiplier tube, 112 ... LOG amplifier, 113 ... Analog-digital converter (ADC), 114 ... Data processing unit , 202 ... background level, 211 ... signal, 212 ... peak, 213 ... integral value, 221 ... reaction showing different behavior, 222 ... peak after reaction starts, 511 ... reaction curve obtained under ideal conditions, 512 ... peak of reaction curve acquired under ideal conditions, 513 ... signal value at end of observation of reaction curve acquired under ideal conditions, 521 ... reaction curve of luminescence not intended originally 522 ... not intended originally Reaction peak of luminescence, 523... Signal at the end of observation of luminescence that was not originally intended.

Claims (3)

Means to give a trigger from outside the system,
Detection means for detecting an optical signal generated from the system based on the trigger;
Recording means for recording the intensity of the optical signal generated from the system over time ;
An integral value calculating means for calculating an integral value of a change with time of the intensity of the optical signal ;
Determining means for determining whether or not to execute a peak position comparison for comparing a temporal position of a peak of an optical signal with a preset temporal position based on the integral value calculated by the integral value calculating means; ,
Comparison means for performing the peak position comparison based on a determination result of the determination means;
Storage means for storing an evaluation result based on a peak position comparison result by the comparison means in association with the integral value;
An automatic analyzer characterized by comprising: The automatic analyzer according to claim 1, wherein
An automatic analyzer comprising display means for adding an alarm to the integral value and displaying based on the evaluation result stored in the storage means. In the automatic analyzer according to claim 1 or 2,
An automatic analyzer comprising display means for displaying a change over time in the intensity of an optical signal recorded in the recording means.

Images