A lupus anticoagulant detection system and method based on conventional coagulation reagents

Abstract

The present invention relates to a lupus anticoagulant detection system and method based on conventional coagulation reagents, belonging to the technical field of detection, and comprising: collecting a user sample, performing anticoagulation detection on the user sample using conventional coagulation reagents to obtain a detection standardization ratio; obtaining historical detection data of the same type within a historical time, and performing fuzzy error analysis on the detection standardization ratio using a set of historical detection standardization ratios within the historical detection data to obtain a fuzzy error coefficient; calculating a reagent error correction coefficient according to the fuzzy error coefficient and the conventional coagulation reagent error; collecting anticoagulant treatment characteristic information of the user according to the fuzzy error coefficient, analyzing to obtain a user error correction coefficient, and combining the reagent error correction coefficient to perform error correction calculation on the detection standardization ratio to obtain a corrected detection standardization ratio interval as the detection result. The present invention solves the technical problems of lack of lupus anticoagulant detection conditions and low detection efficiency in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of detection technologies, and particularly to a lupus anticoagulant detection system and method based on conventional coagulation reagents. Background Art

[0002] Lupus anticoagulant belongs to anti-phospholipid antibodies. The general principle of lupus anticoagulant detection is that when lupus anticoagulant exists in plasma, it can competitively bind to phospholipids, resulting in an extension of the phospholipid-dependent clotting time (screening test). Adding an excessive amount of phospholipids can correct the extended clotting time (confirmation test). Therefore, when the ratio of the clotting time in the screening test to that in the confirmation test is greater than a judgment threshold, it can be considered that lupus anticoagulant exists. However, current lupus anticoagulant detection mainly relies on imported reagents, resulting in relatively scarce detection conditions and low detection efficiency in grass-roots laboratories. Summary of the Invention

[0003] Aiming at the technical problems of scarce detection conditions and low detection efficiency of lupus anticoagulant in grass-roots laboratories in the prior art, the present invention provides a lupus anticoagulant detection system and method based on conventional coagulation reagents.

[0004] The technical solutions of the present invention for solving the above technical problems are as follows:

[0005] In a first aspect, the present invention provides a lupus anticoagulant detection system based on conventional coagulation reagents, including: an anticoagulation test module for collecting a user sample, performing an anticoagulation detection on the user sample using a conventional coagulation reagent, and obtaining a detection standardization ratio;

[0006] An error analysis module for obtaining similar detection data within a historical time according to the detection standardization ratio, and performing a fuzzy error analysis on the detection standardization ratio using a set of historical detection standardization ratios within the historical detection data to obtain a fuzzy error coefficient;

[0007] A reagent error analysis module for calculating a reagent error correction coefficient according to the fuzzy error coefficient and the conventional coagulation reagent error;

[0008] An error correction module for collecting the anticoagulation treatment characteristic information of the user according to the fuzzy error coefficient, analyzing to obtain a user error correction coefficient, and combining the reagent error correction coefficient to perform an error correction calculation on the detection standardization ratio to obtain a corrected detection standardization ratio interval as a detection result.

[0009] In a second aspect, the present invention provides a lupus anticoagulant detection method based on conventional coagulation reagents, including:

[0010] Collecting a user sample, performing an anticoagulation detection on the user sample using a conventional coagulation reagent, and obtaining a detection standardization ratio;

[0011] According to the detected standardized ratio, obtain the same type of detection data within the historical time, and use the set of historical detected standardized ratios in the historical detection data to perform fuzzy error analysis on the detected standardized ratio to obtain a fuzzy error coefficient;

[0012] Calculate a reagent error correction coefficient according to the fuzzy error coefficient and the conventional coagulation reagent error;

[0013] According to the fuzzy error coefficient, collect the anticoagulant treatment characteristic information of the user, analyze and obtain a user error correction coefficient, and combine the reagent error correction coefficient to perform error correction calculation on the detected standardized ratio to obtain a corrected detected standardized ratio interval as the detection result.

[0014] The beneficial effects of the present invention are as follows: The present invention uses a conventional coagulation reagent for anticoagulant detection, and evaluates the presence of lupus anticoagulant by calculating the detected standardized ratio, so that the detection does not rely on imported reagents, improves the accessibility of primary medical institutions, improves the detection efficiency, and also performs fuzzy error analysis through historical detection data, calculates the reagent error correction coefficient and the individualized user error correction coefficient, can effectively identify the possible error fluctuations in the detection process, thereby reducing the detection errors caused by reagent batch differences and individual differences, and further improving the repeatability and reliability of the detection. By combining the reagent error correction coefficient and the user error correction coefficient, this method can perform error correction calculation on the detected standardized ratio, and finally obtain a corrected detected standardized ratio interval, further enhancing the robustness and credibility of the detection, ensuring that the detection result is more accurate, and providing more reliable data support for clinical diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic structural diagram of a lupus anticoagulant detection system based on a conventional coagulation reagent provided by the present invention;

[0016] Figure 2 It is a schematic flow chart of a lupus anticoagulant detection method based on a conventional coagulation reagent provided by the present invention.

[0017] Reference numerals: anticoagulation test module 11, error analysis module 12, reagent error analysis module 13, error correction module 14. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the said features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0021] Example 1, as Figure 1 shown, an embodiment of the present invention provides a lupus anticoagulant detection system based on a conventional coagulation reagent, and the system includes an anticoagulation test module 11, an error analysis module 12, a reagent error analysis module 13, and an error correction module 14.

[0022] The anticoagulation test module 11 is used to perform:

[0023] Collect a user sample, perform anticoagulation detection on the user sample using a conventional coagulation reagent, and obtain a detection standardized ratio;

[0024] In the embodiments of the present application, first, anticoagulation detection is performed on the blood sample of the user using a conventional coagulation reagent to obtain a detection standardized ratio. Among them, by performing detection with a conventional coagulation reagent, the difficulty of obtaining the detection reagent can be reduced, thereby improving the accessibility and detection efficiency of anticoagulation detection in primary laboratories.

[0025] In the embodiments of the present application, the anticoagulation test module 11 is specifically used to perform:

[0026] Redissolve the prothrombin time determination reagent, dilute it in accordance with a first ratio and a second ratio in combination with pure water, and mix it with a calcium chloride solution to obtain a confirmation reagent and a screening reagent;

[0027] Collect user samples and normal samples, process them, and perform coagulation tests with confirmation reagents and screening reagents to obtain screening ratios and confirmation ratios.

[0028] Calculate the ratio of the screening ratio to the confirmation ratio to obtain a detection standardization ratio.

[0029] In the embodiments of the present application, first, user samples are collected, that is, whole blood specimens of the patients to be tested are collected by venous blood sampling. During the blood sampling process, sodium citrate (3.2%) anticoagulant tubes are used to ensure the stability of the coagulation system during the coagulation test of the blood samples, and to avoid detection errors caused by improper anticoagulant concentration or non-standard blood sampling operations. After blood sampling, the samples need to be centrifuged within 1 hour at 2000g for 10 minutes to remove platelets and obtain platelet-poor plasma, thereby reducing the interference of platelets on phospholipid-dependent coagulation reactions and improving the accuracy and repeatability of the detection.

[0030] Then, anticoagulation tests are performed on the user samples using conventional coagulation reagents. The conventional coagulation reagent used in the embodiments of the present application is a prothrombin time (PT) determination reagent, which contains tissue factor and phospholipids and can be used as an activation medium for phospholipid-dependent coagulation reactions. In order to distinguish the competitive binding effect of lupus anticoagulant (LA) on phospholipids, the reagent needs to be diluted to different degrees to change the phospholipid concentration in the reagent. In this method, the PT reagent is diluted according to the first ratio (specifically, 1:5) and the second ratio (specifically, 1:50), and is mixed with 0.025 mol / L calcium chloride solution according to a ratio of 2:1 to form a confirmation reagent (high phospholipid) and a screening reagent (low phospholipid), respectively.

[0031] Furthermore, normal blood samples are collected, that is, blood samples of users who have been determined to be negative for lupus anticoagulant, and are subjected to the above-mentioned centrifugation treatment as normal samples. Both the user samples and the normal samples are mixed with the screening reagent and the confirmation reagent at a volume ratio of 1:2, and the clotting time is measured on an automatic coagulation analyzer. By calculating the screening ratio of the user samples (i.e., the ratio of the clotting time of the user samples and the screening reagent to the clotting time of the normal samples and the screening reagent) and the confirmation ratio (the ratio of the clotting time of the user samples and the confirmation reagent to the clotting time of the normal samples and the confirmation reagent), the detection standardization ratio is further calculated, that is, the ratio of the screening ratio to the confirmation ratio. This standardization ratio can reflect the influence of lupus anticoagulant on phospholipid-dependent coagulation reactions, thereby providing a quantitative basis for the determination of lupus anticoagulant.

[0032] By using conventional coagulation reagents for detection, the detection accessibility and detection efficiency can be improved.

[0033] In the embodiments of the present application, by determining whether the detected standardized ratio is greater than or equal to the threshold or less than the threshold, a positive or negative test determination result can be obtained, and the accuracy is relatively high. Among them, the threshold can be obtained by testing the average standardized ratio of lupus anticoagulant negative users of the reagent using the above method as the threshold.

[0034] However, when using conventional coagulation reagents for testing, due to the differences between different batches of conventional coagulation reagents and the individual differences caused by the possible coagulation treatments of the users themselves, errors may occur in the testing. Therefore, it is necessary to analyze and correct the errors to improve the testing accuracy based on conventional coagulation reagents.

[0035] The error analysis module 12 is used to execute:

[0036] S20: According to the detected standardized ratio, obtain the same type of test data within the historical time, and use the set of historical detected standardized ratios in the historical test data to perform fuzzy error analysis on the detected standardized ratio to obtain a fuzzy error coefficient;

[0037] In the embodiments of the present application, according to the detected standardized ratio, obtaining the same type of test data within the historical time means calling the past lupus anticoagulant (LA) test records stored in the laboratory information management system or database to ensure that the data covers a sufficient time span and sample diversity, and improve the reliability of error analysis.

[0038] According to the historical detected standardized ratios within the historical time, analyze the concentrated distribution level of the detected standardized ratios when positive or negative appears, and then analyze the error range between the current detected standardized ratio and the concentrated distribution level, and calculate to obtain a fuzzy error coefficient. Since this error is only used for reference analysis to facilitate further test determination by experimenters, fuzzy error analysis is performed to provide support for subsequent error correction.

[0039] The error analysis module 12 is specifically used to execute:

[0040] According to the detected standardized ratio, obtain the same type of test data within the historical time, and extract the set of historical detected standardized ratios;

[0041] Randomly select a first centroid detected standardized ratio from the set of historical detected standardized ratios, and calculate the first centroid cost;

[0042] Continue to perform iterative centroid selection to obtain the centroid detected standardized ratio with the minimum centroid cost;

[0043] Calculate the error range between the detected standardized ratio and the centroid detected standardized ratio as the fuzzy error coefficient.

[0044] In the embodiments of the present application, it is determined whether the detected standardized ratio is greater than or equal to a threshold value or less than the threshold value to obtain a detection judgment result, specifically a preliminary positive or negative judgment. Then, among the detection data within the historical time, similar detection data is obtained. For example, the detection data finally determined to be positive or negative is obtained, and the detected standardized ratio therein is extracted to obtain a set of historical detected standardized ratios of similar detection results.

[0045] In order to reflect the centralized distribution level of the detected standardized ratios of similar detection results, in the embodiments of the present application, appropriate centroid data points are selected within the set of historical detected standardized ratios to reflect the centralized distribution points of the detected standardized ratios.

[0046] Randomly select a historical detected standardized ratio within the set of historical detected standardized ratios as the first centroid detected standardized ratio, and calculate the first centroid cost. The smaller the first centroid cost, the more suitable the first centroid detected standardized ratio is as the centroid data point of the set of historical detected standardized ratios.

[0047] In the embodiments of the present application, randomly select the first centroid detected standardized ratio within the set of historical detected standardized ratios and calculate the obtained first centroid cost, including:

[0048] Randomly select a historical detected standardized ratio within the set of historical detected standardized ratios as the first centroid detected standardized ratio;

[0049] Calculate the difference amplitude between other historical detected standardized ratios and the first centroid detected standardized ratio, and calculate the sum to obtain the first centroid cost.

[0050] In the embodiments of the present application, randomly select a historical detected standardized ratio within the set of historical detected standardized ratios as the first centroid detected standardized ratio.

[0051] Further, calculate the difference amplitude between other historical detected standardized ratios in the set of historical detected standardized ratios and the first centroid detected standardized ratio. Exemplarily, calculate the absolute value of the difference between each other historical detected standardized ratio and the first centroid detected standardized ratio, and then calculate the ratio of the absolute value to the first centroid detected standardized ratio as the difference amplitude.

[0052] Calculate the sum of the difference amplitudes between all other historical detected standardized ratios and the first centroid detected standardized ratio as the first centroid cost. The larger the first centroid cost, the greater the overall difference between the first centroid detected standardized ratio and other historical detected standardized ratios, and the less suitable it is as the centroid data point, and it cannot reflect the centralized distribution level within the set of historical detected standardized ratios. On the contrary, the smaller the first centroid cost, the more it can reflect the centralized distribution level within the set of historical detected standardized ratios.

[0053] Based on the same steps, continue with iterative random centroid selection and calculate the centroid cost until convergence. For example, convergence occurs after randomly selecting 100 centroid detection normalization ratios, and the detection normalization ratio with the minimum centroid cost is output as the finally selected centroid detection normalization ratio, which can reflect the centralized distribution level of the historical detection normalization ratios of the same type of detection results. Accurate detection normalization ratios are relatively close and are distributed near the centroid point. The closer the detection normalization ratio of the current detection is to the centroid detection normalization ratio, the more likely the detection result is accurate. Conversely, the greater the difference between the detection normalization ratio of the current detection and the centroid detection normalization ratio, the less likely the detection result is accurate.

[0054] Calculate the error margin between the detection normalization ratio of the current detection and the centroid detection normalization ratio. For example, calculate the ratio of the absolute value of the difference between the two to the centroid detection normalization ratio as the fuzzy error coefficient. The fuzzy error coefficient reflects the extent to which the current detection normalization ratio deviates from the center of the detection normalization ratios of the same type of detection results. The larger it is, the more likely the error of the detection normalization ratio of the current coagulation detection is.

[0055] By selecting the centroid point within the set of historical detection normalization ratios of the same type of detection results, the error between the current detection normalization ratio and the detection normalization ratios of the same type of detection results can be quantified, and then the possible error coefficient of the current detection normalization ratio can be obtained for subsequent error correction.

[0056] The reagent error analysis module 13 is used to execute:

[0057] S30: Calculate and obtain a reagent error correction coefficient based on the fuzzy error coefficient and the conventional coagulation reagent error.

[0058] In the embodiments of the present application, due to the component errors of conventional coagulation reagents caused by factors such as reagent batch differences and detection environment changes, the detection normalization ratios obtained by detection may also have errors.

[0059] Based on the above-mentioned fuzzy error coefficient and the analyzed conventional coagulation reagent error, calculate and obtain a reagent error correction coefficient for subsequent correction of the detection normalization ratio. The conventional coagulation reagent error is the error margin of the detection normalization ratio caused by testing with different batches of conventional coagulation reagents, and can be obtained by calculating the error through testing the same user samples with different batches of conventional coagulation reagents within a historical time.

[0060] The reagent error analysis module 13 is specifically used to execute:

[0061] Obtain a set of detection normalization ratio errors for anticoagulation detection using the conventional coagulation reagent;

[0062] Calculate the mean of the detected standardized ratio error set to obtain the error of the conventional coagulation reagent;

[0063] Calculate the reagent error correction coefficient based on the fuzzy error coefficient and the obtained error of the conventional coagulation reagent.

[0064] In the embodiment of the present application, first, a detected standardized ratio error set obtained by performing anticoagulation detection using a conventional coagulation reagent is acquired. The detected standardized ratio error set includes a deviation set of detected standardized ratios obtained by testing the same blood sample under the same detection conditions using different batches of conventional coagulation reagents. For example, if the detected standardized ratio obtained by testing a sample with a batch of conventional coagulation reagent is X, and the detected standardized ratio obtained by testing the same sample with another batch is Y, then the detected standardized ratio error is |X - Y| / X. In this way, the detected standardized ratio error set is collected.

[0065] Furthermore, calculate the mean of all the detected standardized ratio errors in the detected standardized ratio error set as the error of the conventional coagulation reagent, that is, the average error magnitude of the detected standardized ratio errors caused by different batches of conventional coagulation reagents during detection.

[0066] Furthermore, calculate the reagent error correction coefficient based on the fuzzy error coefficient and the obtained error of the conventional coagulation reagent. Herein, multiply the fuzzy error coefficient by the error of the conventional coagulation reagent as the reagent error correction coefficient. The larger the current fuzzy error coefficient, the larger the error of the detected standardized ratio, and the greater the probability that the error of the detected standardized ratio caused by the reagent batch is also larger, and the larger the reagent error correction coefficient. Conversely, the smaller the error of the detected standardized ratio caused by the reagent batch, the smaller the reagent error correction coefficient. For example, if the fuzzy error coefficient is 10% and the error of the conventional coagulation reagent is 10%, then the calculated reagent error correction coefficient is 1%.

[0067] Use the reagent error correction coefficient for subsequent error correction of the detected standardized ratio to eliminate the error of the detected standardized ratio caused by changes in different batches of reagents, and obtain the interval of the detected standardized ratio that is not affected by the error caused by reagent changes for the reference of experimenters and subsequent detection verification.

[0068] The error correction module 14 is used to execute:

[0069] S40: According to the fuzzy error coefficient, collect the anticoagulation treatment characteristic information of the user, analyze and obtain the user error correction coefficient, and combine with the reagent error correction coefficient to perform error correction calculation on the detected standardized ratio to obtain the corrected detected standardized ratio interval as the detection result.

[0070] In the embodiments of the present application, in addition to the errors caused by different batches of reagents, the anticoagulant therapy performed by the user will also affect the results of the coagulation test, resulting in errors in the detected standardized ratio. Therefore, the anticoagulant therapy characteristic information of the user is collected, and according to the fuzzy error coefficient, the user error correction coefficient for the error of the detected standardized ratio caused by the anticoagulant therapy is analyzed and obtained.

[0071] Furthermore, combining the reagent error correction coefficient and the user error correction coefficient that cause errors due to the characteristics of different batches of reagents and the user individual, the error correction calculation is performed on the detected standardized ratio obtained from the current detection, and the corrected corrected detection standardized ratio interval is obtained, which reflects the distribution interval of the actually detected standardized ratio after error correction, and is used for further detection or judgment by the experimenter, etc., to improve the detection accuracy.

[0072] The error correction module 14 is specifically used to execute:

[0073] Collect the anticoagulant therapy parameters and anticoagulant therapy time of the user as the anticoagulant therapy characteristic information;

[0074] Adopt ensemble learning to train a user anticoagulant detection error analyzer including an integrated anticoagulant detection error analysis path, wherein the number of paths of the integrated anticoagulant detection error analysis path is the integration number;

[0075] Multiply the fuzzy error coefficient by the integration number to obtain the selected path number;

[0076] Randomly select the selected number of anticoagulant detection error analysis paths in the user anticoagulant detection error analyzer, input the anticoagulant therapy characteristic information, obtain a set of user error correction coefficients, and calculate the mean value to obtain the user error correction coefficient;

[0077] According to the user error correction coefficient and the reagent error correction coefficient, calculate the error correction coefficient, and perform positive and negative error correction calculations on the detected standardized ratio to obtain the corrected detection standardized ratio interval as the detection result.

[0078] In the embodiments of the present application, first, the anticoagulant therapy parameters and anticoagulant therapy time of the user are collected as the anticoagulant therapy characteristic information. The anticoagulant therapy parameters include the types of anticoagulant drugs used (such as warfarin, heparin, direct oral anticoagulants DOACs), the dosage used, and the anticoagulant therapy time includes the time from the last anticoagulant therapy to the present.

[0079] Further, by using ensemble learning, a user anticoagulation detection error analyzer including an integrated anticoagulation detection error analysis path is trained, where the number of paths of the integrated anticoagulation detection error analysis path is the number of integrations. By constructing the user anticoagulation detection error analyzer through ensemble learning, the user anticoagulation detection error analyzer can analyze the error influence coefficient of the user's anticoagulation treatment on the current detected normalized ratio according to the user's anticoagulation treatment characteristic information, and the multiple integrated anticoagulation detection error analysis paths can improve the analysis accuracy and the training efficiency of each anticoagulation detection error analysis path.

[0080] In the embodiment of the present application, by using ensemble learning, a user anticoagulation detection error analyzer including an integrated anticoagulation detection error analysis path is trained, including:

[0081] According to the anticoagulation test data record, a set of sample anticoagulation treatment characteristic information is collected, and the error coefficient of the detected normalized ratio under different sample anticoagulation treatment characteristic information is collected and labeled as a set of sample user error correction coefficients;

[0082] According to the number of integrations, the set of sample anticoagulation treatment characteristic information and the set of sample user error correction coefficients are divided to obtain multiple sets of user error training data;

[0083] Respectively using the multiple sets of user error training data, based on machine learning, the number of integrations of anticoagulation detection error analysis paths is trained to obtain a user anticoagulation detection error analyzer.

[0084] In the embodiment of the present application, this method trains a user anticoagulation detection error analyzer through ensemble learning to accurately evaluate the influence of the user's anticoagulation treatment characteristics on the detected normalized ratio, perform error correction, and improve the accuracy and individual adaptability of the detection.

[0085] First, according to the anticoagulation test data records of multiple users in the past time, the sample anticoagulation treatment characteristic information of each user is collected to obtain a set of sample anticoagulation treatment characteristic information, and then the detected normalized ratios of these users are collected, as well as the detected normalized ratios of these users when they do not receive anticoagulation treatment or the anticoagulation treatment has ended for a long time (such as half a year). The difference amplitude between the detected normalized ratios in the two states is calculated as the error coefficient of the detected normalized ratio under the influence of the sample anticoagulation treatment characteristic information of the user, and is labeled as a set of sample user error correction coefficients.

[0086] Then, according to the number of integrations of the integrated anticoagulation detection error analysis path, for example, 10, the set of sample anticoagulation treatment characteristic information and the set of sample user error correction coefficients are randomly divided into multiple sets of data with the same data volume as multiple sets of user error training data. Among them, the sample anticoagulation treatment characteristic information is used as the input data for training, and the sample user error correction coefficient is used as the output data for training.

[0087] Further, multiple sets of user error training data are respectively used to train an integrated number of anticoagulation detection error analysis paths based on integrated machine learning to obtain a user anticoagulation detection error analyzer. Since the training data for each anticoagulation detection error analysis path is different, the performance of each anticoagulation detection error analysis path is different. Integrating multiple anticoagulation detection error analysis paths for error analysis in anticoagulation treatment can improve the accuracy rate.

[0088] Taking the training process of one of the anticoagulation detection error analysis paths as an example, exemplarily, an anticoagulation detection error analysis path is constructed based on the feedforward neural network in machine learning, which includes an input layer, an output layer, and a hidden layer. The dimensions of the input layer and the hidden layer are 1, corresponding to the sample anticoagulation treatment feature information and the sample user error correction coefficient. During the training process, the sample anticoagulation treatment feature information in the user error training data is input, the output user error correction coefficient is obtained, the error from the corresponding true sample user error correction coefficient is calculated, and the mean square error loss function is used to calculate the loss. The network parameters of the anticoagulation detection error analysis path are randomly adjusted and optimized to reduce the loss until the requirements are met, such as the loss being less than 5%. Iterative training with multiple sets of training data is performed until the overall loss meets the requirements, completing the training and obtaining the trained anticoagulation detection error analysis path.

[0089] The integrated number of trained anticoagulation detection error analysis paths is combined to obtain a user anticoagulation detection error analyzer.

[0090] In the embodiments of the present application, the larger the fuzzy error coefficient, the greater the possible error in the detected normalized ratio. Therefore, more anticoagulation detection error analysis paths are selected for error analysis to improve the accuracy of error analysis. On the contrary, if the fuzzy error coefficient is smaller, the possible error in the detected normalized ratio is smaller, and the error correction amplitude for the detected normalized ratio is smaller. Therefore, fewer anticoagulation detection error analysis paths are selected for error analysis to save processing computing power.

[0091] Specifically, the fuzzy error coefficient is multiplied by the integrated number and rounded up to obtain the number of selected paths. The larger the fuzzy error coefficient, the larger the number of selected paths. For example, if the fuzzy error coefficient is 15%, the number of selected paths is 15% * 10 rounded up, which is 2.

[0092] Randomly select the number of anticoagulation test error analysis paths for the selection path within the user anticoagulation test error analyzer. Input the anticoagulation treatment characteristic information of the current user into the number of selected anticoagulation test error analysis paths respectively, and obtain the output user error correction coefficients for the number of selected paths, which are used as the user error correction coefficient set. Then integrate multiple outputs. For example, calculate the mean of the user error correction coefficient set to obtain the user error correction coefficient. Integrating multiple outputs is more accurate. The user error correction coefficient reflects the magnitude of the error in the detected standardized ratio caused by the user's anticoagulation treatment characteristic information.

[0093] Furthermore, according to the user error correction coefficient and the reagent error correction coefficient, calculate the total error correction coefficient for the error in the detected standardized ratio caused by the reagent batch and the user characteristics, which reflects the magnitude of the error in the detected error caused by the two dimensions. For example, calculate the sum of the user error correction coefficient and the reagent error correction coefficient as the error correction coefficient. For example, if the reagent error correction coefficient is 1% and the user error correction coefficient is 3%, then the error correction coefficient is 4%.

[0094] Then, use the error correction coefficient to perform positive and negative error correction calculations on the detected standardized ratio obtained from the test. For example, calculate the detected standardized ratio * (1 ± error correction coefficient) to obtain the endpoint values of the two corrected standardized ratios, and construct the corrected standardized ratio interval as the test result after detection and error correction.

[0095] The corrected standardized ratio interval reflects the interval of the actual distribution of the detected standardized ratio caused by the error, which can be used for the experimenter to make a judgment or further perform a test. For example, if the entire corrected standardized ratio interval is greater than or equal to the above threshold, it can be judged as positive. If a part of the corrected standardized ratio interval is greater than or equal to the above threshold and a part is less than the threshold, further testing can be performed.

[0096] By integrating the errors caused by different batches of reagents and the errors caused by the user's anticoagulation treatment characteristics, the accuracy and reliability of the test results can be improved, and the disadvantage that the detection accuracy of conventional coagulation reagents is lower than that of imported reagents can be compensated.

[0097] The lupus anticoagulant detection system based on conventional coagulation reagents provided by the embodiments of the present invention has at least the following technical effects:

[0098] The present invention uses conventional coagulation reagents for anticoagulation detection, and evaluates the presence of lupus anticoagulant by calculating the detection standard ratio, enabling the detection to be independent of imported reagents, improving the accessibility of primary medical institutions, enhancing the detection efficiency, and also performing fuzzy error analysis through historical detection data, calculating the reagent error correction coefficient and the individualized user error correction coefficient, which can effectively identify the possible error fluctuations during the detection process, thereby reducing the detection errors caused by reagent batch differences and individual differences, and further improving the repeatability and reliability of the detection. By integrating the reagent error correction coefficient and the user error correction coefficient, this method can perform error correction calculation on the detection standard ratio, and finally obtain the corrected detection standard ratio interval, further enhancing the robustness and credibility of the detection, ensuring that the detection results are more accurate, and providing more reliable data support for clinical diagnosis.

[0099] Example 2, as Figure 2 shown, with the same inventive concept as the lupus anticoagulant detection system based on conventional coagulation reagents in Example 1, the embodiment of the present invention also provides a method for detecting lupus anticoagulant based on conventional coagulation reagents. The explanation of the lupus anticoagulant detection system based on conventional coagulation reagents in Example 1 also applies to the method for detecting lupus anticoagulant based on conventional coagulation reagents. The method includes:

[0100] S10: Collect user samples, perform anticoagulation detection on the user samples using conventional coagulation reagents, and obtain the detection standard ratio;

[0101] S20: According to the detection standard ratio, obtain the same type of detection data within the historical time, and use the historical detection standard ratio set in the historical detection data to perform fuzzy error analysis on the detection standard ratio to obtain the fuzzy error coefficient;

[0102] S30: Calculate the reagent error correction coefficient according to the fuzzy error coefficient and the conventional coagulation reagent error;

[0103] S40: According to the fuzzy error coefficient, collect the anticoagulation treatment characteristic information of the user, analyze and obtain the user error correction coefficient, and combine the reagent error correction coefficient to perform error correction calculation on the detection standard ratio to obtain the corrected detection standard ratio interval as the detection result.

[0104] Further, collecting user samples and performing anticoagulation detection on the user samples using conventional coagulation reagents to obtain the detection standard ratio includes:

[0105] Reconstitute the prothrombin time determination reagent, dilute it with pure water according to the first ratio and the second ratio, and mix it with the calcium chloride solution to obtain the confirmation reagent and the screening reagent;

[0106] Collect user samples and normal samples, process them, and perform coagulation tests with confirmation reagents and screening reagents to obtain screening ratios and confirmation ratios;

[0107] Calculate the ratio of the screening ratio and the confirmation ratio to obtain a detection standardization ratio.

[0108] Furthermore, according to the detection standardization ratio, obtain similar detection data within a historical time period, and use the set of historical detection standardization ratios within the historical detection data to perform fuzzy error analysis on the detection standardization ratio to obtain a fuzzy error coefficient, including:

[0109] According to the detection standardization ratio, obtain similar detection data within a historical time period, and extract and obtain a set of historical detection standardization ratios;

[0110] Randomly select a first centroid detection standardization ratio within the set of historical detection standardization ratios, and calculate to obtain a first centroid cost;

[0111] Continue to perform iterative centroid selection to obtain the centroid detection standardization ratio with the minimum centroid cost;

[0112] Calculate the error range between the detection standardization ratio and the centroid detection standardization ratio as the fuzzy error coefficient.

[0113] Furthermore, randomly select a first centroid detection standardization ratio within the set of historical detection standardization ratios, and calculate to obtain a first centroid cost, including:

[0114] Randomly select a historical detection standardization ratio within the set of historical detection standardization ratios as the first centroid detection standardization ratio;

[0115] Calculate the difference range between other historical detection standardization ratios and the first centroid detection standardization ratio, and calculate the sum to obtain the first centroid cost.

[0116] Furthermore, calculate and obtain a reagent error correction coefficient according to the fuzzy error coefficient and the error of the conventional coagulation reagent, including:

[0117] Obtain a set of detection standardization ratio errors for anticoagulation detection using the conventional coagulation reagent;

[0118] Calculate the mean of the set of detection standardization ratio errors to obtain the error of the conventional coagulation reagent;

[0119] Calculate and obtain a reagent error correction coefficient according to the fuzzy error coefficient and the obtained error of the conventional coagulation reagent.

[0120] Further, according to the fuzzy error coefficient, collect the anticoagulation treatment characteristic information of the user, analyze and obtain the user error correction coefficient, and combine with the reagent error correction coefficient to perform error correction calculation on the detection normalized ratio to obtain a corrected detection normalized ratio interval as the detection result, including:

[0121] Collect the anticoagulation treatment parameters and anticoagulation treatment time of the user as the anticoagulation treatment characteristic information;

[0122] Adopt ensemble learning to train a user anticoagulation detection error analyzer including an ensemble anticoagulation detection error analysis path, wherein the number of paths of the ensemble anticoagulation detection error analysis path is the ensemble number;

[0123] Multiply the fuzzy error coefficient by the ensemble number to obtain the selected path number;

[0124] Randomly select the anticoagulation detection error analysis paths with the selected path number in the user anticoagulation detection error analyzer, input the anticoagulation treatment characteristic information, obtain a set of user error correction coefficients, and calculate the mean value to obtain the user error correction coefficient;

[0125] According to the user error correction coefficient and the reagent error correction coefficient, calculate the error correction coefficient, perform positive and negative error correction calculations on the detection normalized ratio, and obtain a corrected detection normalized ratio interval as the detection result.

[0126] Further, adopt ensemble learning to train a user anticoagulation detection error analyzer including an ensemble anticoagulation detection error analysis path, including:

[0127] According to the anticoagulation test data record, collect a set of sample anticoagulation treatment characteristic information, and collect the error coefficients of the detection normalized ratio under different sample anticoagulation treatment characteristic information, and label them as a set of sample user error correction coefficients;

[0128] Divide the set of sample anticoagulation treatment characteristic information and the set of sample user error correction coefficients according to the ensemble number to obtain multiple sets of user error training data;

[0129] Respectively adopt the multiple sets of user error training data to train the anticoagulation detection error analysis paths with the ensemble number based on machine learning to obtain a user anticoagulation detection error analyzer.

[0130] It should be noted that in the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0131] Those skilled in the art will understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0132] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0133] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0135] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concepts.

Claims

1. A lupus anticoagulant detection system based on conventional coagulation reagents, characterized in that: The system comprises: The anticoagulation test module is used to collect user samples, perform anticoagulation detection on the user samples using conventional coagulation reagents, and obtain a detection standardized ratio; The error analysis module is used to obtain similar detection data in the historical time according to the detection standardization ratio, and use the historical detection standardization ratio set in the historical detection data to perform fuzzy error analysis on the detection standardization ratio to obtain the fuzzy error coefficient, including: According to the detection standardization ratio, similar detection data in the historical time is obtained, and a historical detection standardization ratio set is extracted; Randomly selecting a first centroid detection normalization ratio in the historical detection normalization ratio set, and calculating a first centroid cost, including: Randomly selecting a historical detection normalization ratio from the historical detection normalization ratio set as the first centroid detection normalization ratio; Calculate the difference between other historical detection normalized ratios and the first mass center detection normalized ratio, and calculate the sum to obtain a first mass center cost; Continue to iterate the centroid selection to obtain the centroid detection normalization ratio with the minimum centroid cost; Calculating the error amplitude between the detection normalized ratio and the centroid detection normalized ratio as a fuzzy error coefficient; A reagent error analysis module, used for calculating a reagent error correction coefficient based on the fuzzy error coefficient and a conventional coagulation reagent error; The error correction module is used to collect the user's anticoagulant treatment characteristic information according to the fuzzy error coefficient, analyze and obtain the user error correction coefficient, combine the reagent error correction coefficient, perform error correction calculation on the detection standardized ratio, and obtain the corrected detection standardized ratio interval as the detection result.

2. The lupus anticoagulant detection system based on conventional coagulation reagents according to claim 1, characterized in that: Collecting user samples, performing anticoagulation detection on the user samples using conventional coagulation reagents, and obtaining a detection standardized ratio, including: re-dissolving the prothrombin time assay reagent, diluting it with pure water according to a first ratio and a second ratio, and mixing it with a calcium chloride solution to obtain a confirmation reagent and a screening reagent; Collect user samples and normal samples, process them, and perform coagulation tests with confirmation reagents and screening reagents to obtain screening ratios and confirmation ratios; The ratio of the screening ratio and the confirmation ratio is calculated to obtain a detection standardized ratio.

3. The lupus anticoagulant detection system based on conventional coagulation reagents according to claim 1, characterized in that: According to the fuzzy error coefficient and the conventional coagulation reagent error, the reagent error correction coefficient is calculated, including: Obtaining a detection standardized ratio error set for anticoagulation detection using the conventional coagulation reagent; Calculating the mean of the detection standardized ratio error set to obtain a conventional coagulation reagent error; The reagent error correction coefficient is calculated based on the fuzzy error coefficient and the conventional coagulation reagent error.

4. The lupus anticoagulant detection system based on conventional coagulation reagents according to claim 1, characterized in that: According to the fuzzy error coefficient, the anticoagulation treatment characteristic information of the user is collected, the user error correction coefficient is obtained by analysis, and the detection standardized ratio is calculated by error correction in combination with the reagent error correction coefficient to obtain a corrected detection standardized ratio interval as a detection result, including: Collecting the user's anticoagulation therapy parameters and anticoagulation therapy time as anticoagulation therapy characteristic information; Using integrated learning, training a user anticoagulation detection error analyzer including an integrated anticoagulation detection error analysis path, wherein the number of paths of the integrated anticoagulation detection error analysis path is an integrated number; The fuzzy error coefficient is used and multiplied by the integration quantity to obtain the number of selected paths; Randomly selecting the anticoagulation detection error analysis paths of the selected number of paths in the user anticoagulation detection error analyzer, inputting the anticoagulation treatment characteristic information, obtaining a user error correction coefficient set, and calculating the mean to obtain the user error correction coefficient; According to the user error correction coefficient and the reagent error correction coefficient, the error correction coefficient is calculated, and the positive and negative error correction calculations are performed on the detection standardized ratio to obtain a corrected detection standardized ratio interval as a detection result.

5. The lupus anticoagulant detection system based on conventional coagulation reagents according to claim 4, characterized in that: Ensemble learning is used to train a user anticoagulation detection error analyzer including an integrated anticoagulation detection error analysis path, including: According to the anticoagulation test data record, a sample anticoagulation treatment characteristic information set is collected, and the error coefficient of the detection standardized ratio under different sample anticoagulation treatment characteristic information is collected, and marked as a sample user error correction coefficient set; Dividing the sample anticoagulation treatment feature information set and the sample user error correction coefficient set according to the integration quantity to obtain a plurality of user error training data; The plurality of user error training data are respectively used to train an integrated number of anticoagulation detection error analysis paths based on machine learning to obtain a user anticoagulation detection error analyzer.

6. A method for detecting lupus anticoagulants based on conventional coagulation reagents, characterized in that: The method comprises: Use conventional coagulation reagents to perform anticoagulation detection on the collected user samples to obtain a detection standardized ratio; According to the detection standardization ratio, similar detection data in historical time is obtained, and a historical detection standardization ratio set in the historical detection data is used to perform fuzzy error analysis on the detection standardization ratio to obtain a fuzzy error coefficient, including: According to the detection standardization ratio, similar detection data in the historical time is obtained, and a historical detection standardization ratio set is extracted; Randomly selecting a first centroid detection normalization ratio in the historical detection normalization ratio set, and calculating a first centroid cost, including: Randomly selecting a historical detection normalization ratio from the historical detection normalization ratio set as the first centroid detection normalization ratio; Calculate the difference between other historical detection normalized ratios and the first mass center detection normalized ratio, and calculate the sum to obtain a first mass center cost; Continue to iterate the centroid selection to obtain the centroid detection normalization ratio with the minimum centroid cost; Calculating the error amplitude between the detection normalized ratio and the centroid detection normalized ratio as a fuzzy error coefficient; Calculating a reagent error correction coefficient according to the fuzzy error coefficient and a conventional coagulation reagent error; According to the fuzzy error coefficient, the user's anticoagulant treatment characteristic information is collected, and the user error correction coefficient is obtained by analysis. In combination with the reagent error correction coefficient, the detection standardized ratio is subjected to error correction calculation to obtain the corrected detection standardized ratio interval as the detection result.

Images