JP4524642B2 - How to check the suction status - Google Patents

Description

  The present invention relates to a method for confirming a suction state when a predetermined amount of liquid is sucked through a tip of a suction tube connected to a negative pressure source, and more particularly, sampling in automatic analysis for a biological fluid sample such as blood. The present invention relates to a method for confirming the suction state.

  In an automatic analysis for a liquid sample, particularly a biological liquid sample such as blood, a part of the sample is generally sucked and held by a sampling probe and dispensed into a reaction container for analysis. When adopting an analysis method such as an immunoassay in which the amount of dispensing affects the analysis result, the accuracy and precision of sampling are important. When a blood sample such as serum or plasma is aspirated by a sampling probe, solid matter such as fibrin that may be mixed in the blood sample blocks the tip of the sampling probe or the inside of the suction pipe, adversely affecting the accuracy and precision of sampling. Often occurs. In addition, air bubbles may be mixed in or liquid leakage may occur due to a suction operation failure or nozzle tip mounting failure.

  When using a sampling probe connected to a negative pressure source such as a suction pump, means for confirming the suitability of sampling by monitoring the sampling probe or the internal pressure of the flow passage over time, or a dispensing apparatus equipped with the means Many have been disclosed. For example, there are an index using a first derivative value with respect to time of pressure inside the suction pipe during liquid suction (see Patent Document 1) and an index using an integral value with respect to time of pressure (see Patent Document 2). Moreover, what uses as an index a normal suction pressure frame composed of an upper limit and a lower limit of a pressure change curve when normal suction is performed is also disclosed (see Patent Document 3). These indices depend on the viscosity of the sample, and for samples such as serum whose viscosities vary greatly from sample to sample, there is a risk of making an error in determining normal suction as abnormal suction. That is, when a sample with high viscosity is sucked into the sampling probe, for example, there is a time delay in moving the sample with respect to the movement of the syringe pump, so the negative pressure inside the suction tube tends to temporarily increase. If a necessary amount can be sucked within a predetermined time, this should not be determined as abnormal suction.

  As a method for determining abnormal suction that does not depend on the viscosity of a sample, Patent Document 4 proposes to use a second-order differential value with time of a pressure change curve as an index. According to this method, it is possible to obtain a stable occlusion threshold and an insufficient suction threshold that do not depend on viscosity, but it is difficult to intuitively grasp the degree of abnormal suction from the second derivative curve. There is. In Patent Document 5, when a high viscosity liquid is sucked, the pressure in the pipe line greatly fluctuates to the negative pressure side during suction, but when sucked normally, when the low viscosity liquid is sucked after the suction operation is stopped. It is proposed that the pressure at a predetermined time after suction is used as an index for determining clogging by utilizing the return to the same pressure level as in the above. However, even if the suction amount of the high-viscosity sample is finally secured, if the incomplete blockage is sucked, the sucked blockage may adversely affect the subsequent dispensing operation or analytical reaction. There can be.

Japanese Patent Publication No. 6-19361 Japanese Patent Publication No. 6-19362 Patent No. 3065100 Japanese Patent No. 2711215 Japanese Patent Laid-Open No. 11-83868

  The present invention moderately allows the influence of the viscosity of the sample, and comprehensively detects abnormal suction resulting from various factors such as complete blockage, incomplete blockage, insufficient suction amount during suction, and the sampling probe suction state An object of the present invention is to provide a method for confirming a high accuracy.

  When the present inventors comprehensively evaluate the suction state based on the similarity between the pressure change curve in which normal suction is performed and the pressure change curve when the sample is sucked, the sample affects the pressure in the suction pipe. Assuming that the influence of viscosity and the like depends on the elapsed time of suction, the present invention was completed through extensive studies.

That is, the present invention is a method for confirming a suction state when sucking a predetermined amount of liquid through the tip of a suction pipe connected to a negative pressure source, and the suction elapsed time ti (i = 1, 2,..., N) Acquiring the pressure data yi (i = 1, 2,..., N) inside the suction pipe and the suction process in the suction state of the object to be confirmed Obtaining the pressure data Yi (i = 1, 2,..., N) inside the suction pipe with respect to time ti (i = 1, 2,..., N); Pressure data yci (i = 1, 2,..., N) converted by a correction coefficient depending on the suction elapsed time, or pressure data Yci (i) converted from the pressure data Yi by a correction coefficient depending on the suction elapsed time. = 1, 2,..., N), Yi and y and a i Tokara or yi and Yci, residual square sum Σ ^(Yi-yci) the minimum value of ² or Σ ^(yi-Yci) and calculating the minimum value of ^2, the residual squared calculated Comparing the minimum value of the sum or the square root of the calculated minimum value of the residual sum of squares with respect to the threshold value and confirming whether the suction state is correct or not according to the comparison result.

  Since the pressure data of the present invention is acquired for a plurality of suction elapsed times, the pressure data includes a plurality of values and can be expressed as a pressure change curve with respect to the suction elapsed time. Therefore, the pressure data of the reference suction state can be compared with the pressure change curve to be confirmed as a reference pressure change curve, and the suction state can be evaluated by their similarity. In this specification, the pressure change curve refers to a set or a subset of pressure data that can change over time. The pressure change curve is not limited to a smooth curve that changes in a continuous curve, but also includes a pressure-time data and a polygonal line shape.

  The pressure data acquired when normal suction without clogging of the suction probe or empty suction is performed can be used as the reference pressure data of the suction state. The pressure data of the suction state as a reference may be a pressure change curve obtained from a single experiment, or it may be obtained from a plurality of experiments, and an average value of pressure data for each elapsed suction time may be taken. It may be a pressure change curve obtained by using a general method.

  In the experimental conditions in which the method of the present invention is actually applied, if there are plans to change factors that greatly change the suction state, for example, the time required for suction, the suction speed, the suction amount, the shape of the sampling probe, etc. It is desirable to obtain a reference pressure change curve. On the other hand, if it is a factor that has the possibility of changing the suction state, but it is actually difficult to obtain a standard pressure change curve for each factor, experiment with various factors to change the pressure change It is desirable to know in advance how much the curve will be affected. For example, when a sampling probe of an automatic analyzer is used, a plurality of reference pressure change curves are obtained using a plurality of devices or sampling probes which are the same kind of separate bodies. Thus, obtaining many reference pressure change curves under various conditions under which normal suction is performed is useful for setting a threshold value for determining whether or not the suction state is correct. In addition, if the standard pressure change curve and threshold value are obtained in advance, when performing actual measurement, the target pressure change curve to be confirmed is acquired, and the already obtained standard pressure change curve and threshold value are used. The method of the invention can be carried out.

  In the present invention, when comparing the reference pressure change curve with the pressure change curve of the object whose suction state is to be confirmed, a correction coefficient depending on the suction elapsed time is used. This correction coefficient that depends on the elapsed time of suction reflects the assumption that the influence of the sample viscosity and the like on the pressure inside the suction tube depends on the elapsed time of suction. For example, when the reference suction state is a liquid having a low viscosity such as water and the target whose suction state is to be confirmed is a liquid having a high viscosity such as serum, the above correction coefficient has the following form. The correction coefficient for converting the pressure data yi in the reference suction state to yci only needs to be monotonically increased in the section of the elapsed suction time to be observed. For example, a linear function, a quadratic function, or more with respect to the elapsed suction time It may be a higher order function. On the other hand, the correction coefficient for converting the pressure data Yi in the suction state of the target to be confirmed to Yci only needs to be monotonically decreased in the section of the elapsed suction time to be observed, and the form of the function with respect to the elapsed suction time is used as a reference. It may be the same as the correction coefficient for converting the pressure data yi in the suction state into yci. Comparing yci obtained by converting pressure data of a suction state as a reference and Yi and comparing Yci obtained by converting pressure data of a suction state to be confirmed with yi are mathematical Therefore, explaining one of them will explain the other at the same time. In short, the reference suction state and the target suction state to be confirmed differ in the degree of influence on the pressure in the suction pipe due to the difference in the viscosity of the liquid, etc. It is.

  In order to detect abnormal suction comprehensively based on the minimum value of the residual sum of squares or the square root of the minimum value of the residual sum of squares, a threshold value serving as a criterion for determination is used. In order to set the threshold value, a situation in which abnormal suction occurs is created, its pressure change curve is acquired, and the minimum value or the square root of the above-mentioned residual sum of squares is obtained. For example, it is also preferable to determine the limit at which normal suction is performed by preparing samples with various concentrations of serum albumin dissolved in water or serum and changing the viscosity of the solution, and determining the pressure change curve accompanying suction. It is. Moreover, various pressure change patterns of abnormal suction can be collected using various blood samples that are likely to clog the suction probe, and can be used for setting a threshold value. Whether normal suction has been performed can be determined on the premise of the reproducibility of the analysis value of the same suction sample, or can be confirmed by directly measuring the weight of the suction sample. It is practical to set the threshold value for each factor that greatly changes the suction state, for example, the amount of suction, in the same manner as when the pressure data of the suction state as a reference is acquired.

Regarding the correction coefficient depending on the suction elapsed time, the calculation is particularly simple and effectively preferable when the correction coefficient is expressed as a linear function b + a × ti of the suction elapsed time. Here, the parameters b and a are the residual sum of squares Σ {Yi− (b + a × ti) yi} ² or Σ {yi− (b + a × ti) Yi} ² for each suction elapsed time ti to be observed. This is the smallest constant. The numerical values or signs of the parameters a and b differ between the residual sum of squares Σ {Yi− (b + a × ti) yi} ² and Σ {yi− (b + a × ti) Yi} ² . The parameter a includes a contribution such as a suction elapsed time, that is, a viscosity, and the parameter b includes a function of adjusting the level of the pressure value in the suction state to be confirmed and the suction state to be confirmed.

  The suction elapsed time t1 for starting observation may be an arbitrary predetermined time after the start of suction, may be the same as the start of suction, or may be the latter half of the suction operation. The last suction elapsed time tn to be observed may be an arbitrary predetermined time after t1, and may be before or after the operation of the negative pressure source is stopped. However, the suction elapsed time ti for acquiring the reference suction state pressure data and the suction elapsed time ti for acquiring the target suction state pressure data to be confirmed need to be the same. Also, when sucking a highly viscous sample, the pressure in the suction tube changes even after the operation of the negative pressure source is stopped, so it is certain that tn is set to a predetermined time after the operation of the negative pressure source is stopped. It may be useful for detecting abnormal suction.

  There is no limitation on the number of suction elapsed times ti (i = 1, 2,..., N) for acquiring pressure data in the suction pipe, and it may be set as appropriate according to the implementation conditions. The number of n is preferably at least 3 points and at most about 100 points, and is particularly preferably set between 5 and 15. Further, the pressure data acquisition interval is not particularly limited, and may be an equal time interval, or it is also preferable to acquire a time zone effective for detecting abnormal suction depending on the implementation conditions. . However, in order to confirm the suction state accurately, it is preferable to obtain the suction elapsed time section in which the pressure data of the entire section in which the negative pressure fluctuates greatly is observed. For example, even if pressure data just after suction is acquired, abnormal suction that occurs in the second half of suction cannot be detected, and even if pressure data is acquired only after the operation of the negative pressure source is stopped, abnormal suction that occurs in the first half of suction is not This is because it becomes difficult to detect.

  In order to acquire pressure data of a suction state as a reference, it is convenient and preferable to use water as a reference sample so that normal suction is performed. On the other hand, for precise detection of abnormal suction, it is desirable to use, as a reference sample, the same type of solvent as that of the sample whose suction state is to be confirmed, from which the cause of clogging has been removed. The same type of solvent means that the object to be checked is such that the physical properties that are thought to affect the pressure change curve such as viscosity, density, surface tension, etc. show the same values as the sample to be checked for suction. In the case of serum, serum can be used as a reference sample. In addition, among the physical properties of the solvent, the viscosity has a large influence on the pressure change curve. Therefore, if the solvent whose viscosity is close, for example, the sample to be checked for suction, is serum, a well-controlled standard as a reference sample It is preferred to use serum, control serum or serum albumin aqueous solution. In addition, as the solvent with the same physical properties, the sample itself for which the suction state is to be confirmed may be used. However, in this case, it is necessary to ensure that normal suction is performed when acquiring pressure data for the reference suction state. Therefore, it is desirable to remove the cause of the blockage or to set the pressure data in the suction state based on the average value of the pressure data acquired from a plurality of experiments.

  Various suction pumps can be used as the negative pressure source. Assuming that the method of the present invention is applied to a sampling probe of a dispensing apparatus, it is particularly preferable to use a syringe pump that can easily control suction and discharge operations.

  Various sampling probes can be used as the suction tube connected to the negative pressure source, but the liquid may be sucked directly from the tip of the sampling probe, or a disposable tip may be used. Further, the connection between the suction pipe and the negative pressure source may be directly connected, or may be connected via a trap or the like therebetween.

  According to the present invention, when comprehensively evaluating the suction state based on the similarity between the pressure change curve in which normal suction is performed and the pressure change curve when the sample is sucked, a correction coefficient that depends on the elapsed time of suction. Thus, it is possible to effectively suppress the influence of the sample viscosity and the like on the pressure inside the suction tube, and to perform accurate evaluation. For example, an error in determining a liquid having a high viscosity as abnormal suction can be prevented. Further, the accuracy of distinguishing between normal suction and abnormal suction can be increased, and the reliability in setting the threshold can be increased. In the present invention, the pressure change curve in which normal suction is performed and the pressure change curve when the sample is sucked are statistically applied, and the minimum value of the residual square sum or the minimum of the residual square sum is obtained. Since the square root of the value is used as an index of the suction state, the shapes of both curves are intuitively compared. Therefore, high reproducibility with respect to the pressure value level itself in the reference suction state and the suction state of each object to be confirmed is not required. That is, it is not necessary to precisely adjust the gain of the pressure sensor when acquiring the reference pressure change curve and when acquiring the target pressure change curve. Therefore, the present invention can be said to be a more stable method than the conventional method in which the pressure at a certain point in time during or after the suction is stopped is compared with a threshold value.

  Furthermore, the method of the present invention can also be implemented as a method for confirming the suction state in an automatic sample suction means mounted on a biochemical automatic analyzer or the like. In this case, for example, a reference pressure change curve and a threshold value acquired in advance in the storage means of the biochemical automatic analyzer, for example, are stored, and the pressure change curve to be confirmed measured by the sensor is output to the arithmetic unit. It is possible to automatically display whether the suction state is correct or not on a display means such as a display or a printer.

From the pressure data Yi in the suction state to be checked and the yci obtained by converting the pressure data yi in the reference suction state by the correction coefficient b + a × ti, the residual sum of squares Σ (Yi−yci) ² , The best mode for carrying out the present invention will be described by taking as an example the case where the minimum value of Σ {Yi− (b + a × ti) yi} ² is obtained and the square root is used as an index for confirming the correctness of the suction state.

In order to obtain the parameters a and b that minimize the residual sum of squares Σ {Yi− (b + a × ti) yi} ^{2, the} partial derivatives relating to b and a are set to 0, and the following equations (1) and (2 Get):

∂Σ {Yi− (b + a × ti) yi} ² / ∂a = −ΣYiiti + bΣyi ² ti + aΣyi ² ti ² = 0 (1)
∂Σ {Yi− (b + a × ti) yi} ² / ∂b = −ΣYii + bΣyi ² + aΣyi ² ti = 0 (2)

From the equations (1) and (2), the following equation for obtaining the parameters a and b is obtained:

a = (ΣYiii × Σyi ² ti−Σyi ² × ΣYiiti) / {(Σyi ² ti) ² −Σyi ² × Σyi ² ti ² } (3)
b = (ΣYiii−aΣyi ² ti) / Σyi ² (4)

The pressure data Yi in the suction state to be confirmed and the pressure data yi in the reference suction state are substituted into formulas (3) and (4) to obtain numerical values of a and b. Using the numerical values of a and b thus obtained, the residual square sum Σ {Yi− (b + a × ti) yi} ² from each actually measured pressure data Yi and the pressure data yi in the reference suction state. Is calculated. The positive square root of the minimum value of the residual sum of squares is compared against a threshold value.

Example 1
As an application example of the best mode for carrying out the method of the present invention, normal suction, clogging tendency, and idling phenomenon will be sequentially described. The value of the pressure data is shown as negative pressure (arbitrary unit) when the pressure before suction is 0.

  FIG. 1 is an example of normal suction in which pressure data representing a normal suction state as a reference and pressure data representing a specimen suction state change in substantially the same manner with respect to the elapsed suction time. A disposable pipette tip was attached to the tip of the suction tube connected to the syringe pump. The output of a pressure sensor that measures the pressure of air inside the suction pipe was monitored. The average value of the pressure data obtained by repeating the water suction in the open container five times was used as the pressure data in the suction state as a reference. Serum specimens were aspirated as subjects to be confirmed. The suction amount is 100 μL per one time, and the operation time of the syringe pump is about 550 ms. The pressure data at a time point of about 100 ms after the start of suction was taken, and 9 points of pressure data were plotted, including 8 points during the suction operation and 200 ms after the suction operation was stopped. In FIG. 1, the pressure data (black circle plot marker) in the suction state as a reference and the pressure data (white square plot marker) associated with the actually measured serum suction are connected by broken lines. By substituting these pressure-time data into the above equations (3) and (4), parameters a and b are obtained, and the correction data b + a × ti (in this example, b = 1.0, The change in the pressure data yci corrected by multiplying by a = 0.2) is shown by a broken line. As shown in FIG. 1, the corrected pressure data almost overlaps with the plot representing the actually measured serum aspiration state. In the case of this normal suction, the positive square root of the minimum value of the residual sum of squares was 3.3 (arbitrary unit representing negative pressure).

  FIG. 2 is an example of abnormal suction showing a tendency of clogging, in which the increase in negative pressure accompanying the suction of serum (progress of pressure reduction) is remarkable. The experimental conditions such as the reference pressure data, the suction amount, and the acquisition time of the pressure data are the same as those in FIG. This time, unlike FIG. 1, the negative pressure is kept high from 400 ms before to about 550 ms, which is just before the suction operation stops. However, at 200 ms after stopping the suction operation, the negative pressure level of the serum suction has dropped to almost the same level as the reference pressure data. This is also considered to be a result of, for example, a loosely cross-linked fibrin solidified body stagnating at the narrowed portion at the tip of the pipette and then sucked in against the suction pressure. The sample collected by suction in such a state must be regarded as having a factor that adversely affects the analysis. The conventional method of detecting abnormal suction using the pressure at a predetermined time after stopping the suction operation as an index cannot detect the incomplete clogging phenomenon as in this example. In the data of FIG. 2, the correction parameters were b = 0.5 and a = 2.1, and the positive square root of the minimum value of the residual sum of squares was 42.7. This value 42.7 strongly suggests abnormal suction.

  FIG. 3 is an example of abnormal suction showing a tendency of idle suction, in which the negative pressure shows an abnormal drop during suction. Experimental conditions such as reference pressure data, suction amount, and pressure data acquisition time are the same as in FIGS. In the actual measurement data of the serum sample, the negative pressure has suddenly decreased after about 300 ms. This is probably because the sample was insufficient during aspiration due to an error in detecting the liquid level or the presence of bubbles. In this data, the correction parameters were b = 1.5 and a = −2.0, and the positive square root of the minimum value of the residual sum of squares was 30.7. This example also shows that the method of the present invention can detect abnormal suction that could not be detected by the conventional method using the pressure at a predetermined time after stopping the suction operation as an index.

  From the above results, by setting the threshold value to about 10 in this measurement system, it was possible to detect clogging or abnormal suction due to idle suction, and to confirm the correctness of the suction state.

Example 2
One of the features of the present invention is that, in order to consider the influence of the sample viscosity on the pressure inside the suction tube, a correction coefficient depending on the elapsed time of suction is used, and the reference pressure change curve and the pressure change curve of the inspection object are used. Is to compare. That is, if b + a × ti is used as a correction coefficient, the parameter a plays an important role. This will be explained with an example.

If the time dependency of the correction coefficient is not taken into consideration, that is, when the parameter a of the correction coefficient b + a × ti is 0, b = ΣYiii / Σyi ^{2 according} to the equation (4). Substituting the acquired pressure data into this equation, the numerical value of b is obtained. The residual sum of squares Σ (Yi−b × yi) ² is calculated from each actually measured pressure data Yi and the pressure data yi in the reference suction state by using the numerical value of b thus obtained. The positive square root of the residual square sum was compared with the positive square root of the residual square sum Σ {Yi− (b + a × ti) yi} ² according to the present invention.

For illustration, FIG. 4 shows an example of a pressure change curve for a human serum sample. Experimental conditions such as a pressure change curve, suction amount, and pressure data acquisition time are the same as in FIGS. In FIG. 4, the pressure data of the suction state as a reference (black circle plot marker) and the pressure data (white square plot marker) associated with the actually measured serum are connected by a broken line. By substituting these pressure-time data into the above equations (3) and (4), parameters a and b are obtained, and the correction data b + a × ti (in this example, b = 1.14, The change of the pressure data yci corrected by multiplying by a = 1.94) is indicated by a broken line (correction (a, b)). On the other hand, by substituting the same pressure-time data into b = ΣYiii / Σyi ² to obtain parameter b, pressure data corrected by multiplying the reference pressure data by a correction coefficient b (in this example, b = 1.75). The change in yci is indicated by a one-dot chain line (correction (a = 0)).

  Looking at the shape of the measured serum pressure change curve (white square plot marker), the negative pressure bulge in the vicinity of 550 ms in the latter half of the suction as shown in FIG. 2 and the negative pressure during the suction as shown in FIG. There is no sudden drop in the price. In addition, the pressure after stopping the suction operation is also reduced to almost the same level as the reference pressure change curve. Furthermore, since there is no report of an abnormality in the serum analysis value after the suction operation, no abnormality is found in the serum suction shown in FIG. When FIG. 4 is observed in detail, it can be seen that the peak position of the actually measured pressure change curve is slightly shifted to the right as compared with the peak position (time axis) of the reference pressure change curve. The pressure change curve (one-dot chain line) by correction with the parameter a set to 0 has a shape in which the reference pressure change curve is proportionally moved upward, whereas the correction coefficient b + a × ti is used. It can be seen that the pressure change curve (broken line) by the method well follows the peak shifted in the time direction of the actually measured pressure change curve. In fact, the positive square root of the residual sum of squares according to the present invention using the correction coefficient b + a × ti is 8.2, whereas the correction is made with a = 0, that is, the correction coefficient independent of the suction elapsed time. The positive square root of the residual sum of squares was 44.1.

Therefore, when the time dependency is not considered in the correction coefficient, the positive square root of the residual sum of squares Σ {Yi− (b + a × ti) yi} ² shows a high value even in normal suction. It was found that when the threshold was set to about 10 as in Example 1, it was impossible to distinguish between abnormal suction and normal suction. FIG. 5 summarizes 244 suction measurement data including the data shown in FIG. 4 and no reported abnormal suction. The aspirated specimen is human serum. FIG. 5 shows the positive square root (correction (a, b)) of the residual sum of squares when the correction coefficient b + a × ti according to the present invention is used, and the residual sum of squares when the parameter a = 0. The frequency distribution with respect to the positive square root (correction (a = 0)) is indicated by a broken line. The width of the class was 0.5 (negative pressure: arbitrary unit) for the positive square root of the residual sum of squares. The distribution of the positive square root of the residual sum of squares calculated by the method according to the present invention (correction (a, b)) clearly has an average value of 3.5 and a standard deviation of 0 in data reflecting a normal suction state. A shape close to a normal distribution of .8 is shown. On the other hand, the distribution (correction (a = 0)) when the parameter a = 0 is set, the average value excluding the data whose positive square root of the residual sum of squares is 14 or more is 5.4, and the standard deviation is 1. .7 shows a broad distribution spreading to the right. In particular, the breakdown of the three data points whose positive square root of the residual sum of squares exceeds 15 is {8.2, 44.1, respectively, as a combination of correction (a, b) and correction (a = 0) according to the present application. }, {11.4, 57.8} and {11.1, 27.4}, it can be seen that the present method is useful for avoiding an error in determining normal suction as abnormal suction.

FIG. 5 is a diagram showing normal suction, which is an example showing pressure change representing a sample suction state, pressure change representing a reference suction state, and pressure data converted by the correction coefficient of the present invention. It is an example showing the pressure change indicating the suction state of the specimen, the pressure change showing the reference suction state, and the pressure data converted by the correction coefficient of the present invention, and is a diagram showing the suction state suggesting clogging. It is an example showing the pressure change which represents the pressure change showing the suction state of the specimen, the pressure change showing the reference suction state, and the pressure data converted by the correction coefficient of the present invention, and is a diagram showing the suction state which suggests empty suction. . Changed by a pressure change indicating the suction state of the specimen and a pressure change showing the reference suction state, pressure data converted by the correction coefficient of the present invention (correction (a, b)), and a correction coefficient that does not depend on the elapsed time of suction. It is an Example showing pressure data (correction (a = 0)). When the correction coefficient b + a × ti according to the present invention is used for the aspiration measurement data of 244 serum samples, the positive square root of the sum of squares of the residual (correction (a, b)) and the parameter a = 0 It is a frequency distribution regarding the positive square root (correction (a = 0)) of the residual sum of squares.

Claims (5)

A method for confirming a suction state when a predetermined amount of liquid is sucked through a tip of a suction pipe connected to a negative pressure source,
・ Acquires pressure data yi (i = 1, 2,..., N) inside the suction pipe with respect to the suction elapsed time ti (i = 1, 2,..., N) in the reference suction state. To do
The pressure data Yi (i = 1, 2,..., N) inside the suction pipe with respect to the suction elapsed time ti (i = 1, 2,..., N) in the suction state to be confirmed. Getting,
The pressure data yci (i = 1, 2,..., N) converted from the pressure data yi by a correction coefficient depending on the suction elapsed time or the pressure data Yi was converted by a correction coefficient depending on the suction elapsed time. Obtaining pressure data Yci (i = 1, 2,..., N);
Calculating the minimum value of the residual sum of squares Σ (Yi−yci) ² or the minimum value of Σ (yi−Yci) ² from Yi and yci or from yi and Yci;
Comparing the minimum value of the calculated residual sum of squares or the square root of the minimum value of the calculated residual sum of squares against a threshold;
-Checking the correctness of the suction state according to the comparison result, the method. A correction coefficient depending on the suction elapsed time is expressed as a linear function b + a × ti of the suction elapsed time, and the parameters b and a are calculated residual sums of squares Σ {Yi− (b + a × ti) yi} ² or Σ The method according to claim 1, wherein {yi− (b + a × ti) Yi} ² is a constant that minimizes. The method according to claim 1 or 2, wherein n is any one of 5 to 15. The method according to claim 1, wherein water is sucked in a reference suction state. The method according to any one of claims 1 to 3, wherein, in a reference suction state, the same kind of solvent as that of the sample whose suction state is to be confirmed is sucked.

Images