Error evaluation method, system, equipment and medium for medical test results

Abstract

The present application discloses an error evaluation method, system, electronic device and medium for medical test results, and the technical field to which it belongs is performance detection technology. The error evaluation method for medical test results includes: obtaining multiple test data of medical test equipment in each test item from inter-room quality evaluation data; performing linear regression calculation to obtain a linear function; determining the slope coefficient K and the intercept coefficient L; calculating the comprehensive evaluation index value of the test item according to a preset formula; wherein the preset formula is: bioy=A‑B×((1+L)×K)‑C×((1+K)×L)); generating an error evaluation result according to the comprehensive evaluation index value. The present application can quantitatively evaluate the error of medical test results and improve the accuracy of error detection. The present application also discloses an error evaluation system for medical test results, a storage medium and an electronic device, which have the above beneficial effects.

Description

Technical Field

[0001] The present application relates to the field of performance testing technology, and in particular to an error evaluation method, system, electronic equipment and medium for medical test results. Background Art

[0002] External Quality Assessment (EQA) is a process in which multiple laboratories analyze the same specimen, and an independent external organization collects and feedbacks the results reported by the laboratories to evaluate the laboratory operation. Through external quality assessment, the error of medical test results can be evaluated. Medical test results are also called medical test results, that is, medical test results obtained using medical test equipment. When the error of medical test results is large, it will lead to repeated testing, misdiagnosis, missed diagnosis, etc.

[0003] In the related art, the staff mainly relies on subjective data analysis of the inter-laboratory quality evaluation report to determine the error status of the medical test results. However, the above method cannot quantitatively evaluate the error of the medical test results and has low reliability.

[0004] Therefore, how to quantitatively evaluate the errors in medical test results and improve the accuracy of error detection is a technical problem that technical personnel in this field currently need to solve. Summary of the invention

[0005] The purpose of this application is to provide a method, system, electronic device and storage medium for error evaluation of medical test results, which can quantitatively evaluate the errors of medical test results and improve the accuracy of error detection.

[0006] In order to solve the above technical problems, the present application provides a method for error evaluation of medical test results, the method comprising:

[0007] Acquire multiple test data of each test item of the medical testing equipment from the inter-laboratory quality evaluation data; wherein the test data includes PT upper limit value, PT lower limit value, target value and result value; PT stands for capability verification;

[0008] Performing linear regression calculation on the multiple test data of the test items to obtain a linear function;

[0009] Determine a corresponding slope coefficient K according to the slope of the linear function;

[0010] Determine the corresponding intercept coefficient L according to the intercept of the linear function;

[0011] The comprehensive evaluation index value of the test item is calculated according to a preset formula; wherein the preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)) , bioy represents the comprehensive evaluation index value, and A, B and C are preset coefficients;

[0012] An error evaluation result is generated according to the comprehensive evaluation index value of the detection item; wherein the error evaluation result is used to describe the error degree of the medical test result of the medical testing equipment.

[0013] Optionally, determining a corresponding slope coefficient K according to the slope of the linear function includes:

[0014] Determine a maximum allowable error P0 of the inter-laboratory quality assessment data, and determine a first exemption error P1 according to the maximum allowable error P0; wherein 0<P1<P0<1;

[0015] If 1-P1<m<1+P1, the value of the slope coefficient K is set to 0; wherein m represents the slope of the linear function;

[0016] If m≤1- P0 or m≥1+ P0, the value of the slope coefficient K is set to 1;

[0017] If 1+P1≤m<1+P0, the value of the slope coefficient K is determined according to the first coefficient calculation formula; wherein the first coefficient calculation formula is ;

[0018] If 1-P0<m≤1-P1, the value of the slope coefficient K is determined according to the second coefficient calculation formula; wherein the second coefficient calculation formula is .

[0019] Optionally, determining a corresponding intercept coefficient L according to the intercept of the linear function includes:

[0020] Determine a second exemption error P2; wherein 0<P2<1;

[0021] Calculate reference coefficients ; Where i represents the number of test data in the test project, b represents the intercept of the linear function, B i Indicates the maximum allowable deviation value;

[0022] like <P2, then the value of the intercept coefficient L is set to 0;

[0023] like >1, then the value of the intercept coefficient L is set to 1;

[0024] If P2≤ ≤1, the value of the intercept coefficient L is determined according to the third coefficient calculation formula; wherein the third coefficient calculation formula is: .

[0025] Optionally, performing linear regression calculation on multiple test data of the test items includes:

[0026] The linear regression calculation object Simple Regression is used to perform linear regression calculation on the multiple detection data of the detection items.

[0027] Optionally, after performing linear regression calculation on the plurality of test data of the test items, the method further includes:

[0028] If the target value and the result value in the test data do not satisfy the linear regression condition, a linear regression calculation is performed on multiple test data of the next test item.

[0029] Optionally, after calculating the comprehensive evaluation index value of the test item according to a preset formula, the method further includes:

[0030] Performing weighted calculation on the comprehensive evaluation index values ​​of all the detection items of the medical testing equipment to obtain a weighted index value;

[0031] Determine whether the weighted index value is greater than a preset index value; if so, determine that the performance of the medical testing equipment is normal; if not, determine that the performance of the medical testing equipment is abnormal.

[0032] Optionally, after calculating the comprehensive evaluation index value of the test item according to a preset formula, the method further includes:

[0033] Obtaining an inter-laboratory quality evaluation value of the medical testing equipment in each of the testing items;

[0034] If the comprehensive evaluation index value is greater than the first value, the inspection item whose inter-room quality evaluation value is lower than the preset evaluation value is marked as an inspection item to be improved;

[0035] If the comprehensive evaluation index value is less than a second value, the inspection item whose inter-room quality evaluation value reaches the maximum value is marked as an inspection item to be improved; wherein the first value is greater than the second value.

[0036] The present application also provides a medical test result error evaluation system, the system comprising:

[0037] A data acquisition module is used to acquire multiple test data of the medical testing equipment in each test item from the inter-laboratory quality evaluation data; wherein the test data includes a PT upper limit value, a PT lower limit value, a target value and a result value; PT represents capability verification;

[0038] A linear regression calculation module, used for performing linear regression calculation on a plurality of test data of the test items to obtain a linear function;

[0039] A slope coefficient determination module, used to determine a corresponding slope coefficient K according to the slope of the linear function;

[0040] An intercept coefficient determination module, used to determine the corresponding intercept coefficient L according to the intercept of the linear function;

[0041] A comprehensive evaluation index determination module, used to calculate the comprehensive evaluation index value of the test item according to a preset formula; wherein the preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)) , bioy represents the comprehensive evaluation index value, and A, B and C are preset coefficients;

[0042] The error evaluation module is used to generate an error evaluation result according to the comprehensive evaluation index value of the detection item; wherein the error evaluation result is used to describe the error degree of the medical test result of the medical testing equipment.

[0043] The present application also provides a storage medium on which a computer program is stored, and when the computer program is executed, the steps of the above-mentioned error evaluation method for medical test results are implemented.

[0044] The present application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the steps of the error evaluation method for medical test results are implemented.

[0045] The present application provides an error evaluation method for medical test results, the method obtains multiple test data of medical test equipment in each test item from inter-laboratory quality evaluation data, and performs linear regression analysis on the test data to obtain a linear function. The present application also determines the slope coefficient and the intercept coefficient according to the slope and intercept of the linear function, respectively, and the slope coefficient and the intercept coefficient directly reflect the degree of deviation between the target value and the result value. The present application also calculates the comprehensive evaluation index value of the test item according to a preset formula, so as to generate an error evaluation result according to the comprehensive evaluation index value; the above preset formula comprehensively considers the influence of the slope coefficient and the intercept coefficient to comprehensively reflect the overall quality and consistency of the data. The present application realizes the error evaluation of medical test results through linear regression analysis and comprehensive evaluation index value calculation, and can quantitatively evaluate the error of medical test results, thereby improving the accuracy of error detection. The present application also provides an error evaluation system for medical test results, a storage medium and an electronic device, which have the above beneficial effects and are not repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 A flowchart of a method for evaluating the error of a medical test result provided in an embodiment of the present application;

[0048] Figure 2 A schematic diagram of the structure of a medical test result error evaluation system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] See below Figure 1 , Figure 1 A flowchart of a method for error evaluation of medical test results provided in an embodiment of the present application.

[0051] Specific steps may include:

[0052] S101: Acquire multiple test data of each test item of the medical testing equipment from the inter-laboratory quality evaluation data;

[0053] Among them, this embodiment can be applied to electronic equipment with data analysis and computing capabilities, so as to obtain error evaluation results of medical test results by analyzing and calculating inter-laboratory quality evaluation data. The above-mentioned medical testing equipment may include fully automatic biochemical analyzers, blood analyzers, urine analyzers, blood gas analyzers, mass spectrometers, molecular diagnostic instruments, immunoassay analyzers and other equipment.

[0054] Before this step, there may be an operation of obtaining an external quality evaluation report, and extracting data from the external quality evaluation report to obtain the external quality evaluation data. The generation process of the above external quality evaluation report includes: each medical testing device in multiple laboratories tests the same or similar tested items according to pre-defined conditions, and the quality control center performs statistical analysis on the test data to obtain the external quality evaluation report.

[0055] Specifically, after receiving the samples, each laboratory will conduct the test according to the prescribed test methods and operating procedures. During the test process, the medical testing equipment will generate multiple test data, including target values ​​and result values. The target value refers to the expected result or standard value of a test item under ideal conditions. The result value refers to the actual result obtained after the laboratory uses the medical testing equipment to test the sample.

[0056] This embodiment can read the inter-laboratory quality evaluation data from the database, and determine multiple test data of the medical testing equipment in each test item based on the inter-laboratory quality evaluation data; the above test data include PT upper limit value, PT lower limit value, target value and result value; PT represents proficiency verification (i.e., proficiency verification result).

[0057] External quality assessment (EQA) is also known as proficiency testing (PT). The PT upper limit in external quality assessment refers to the maximum acceptable deviation limit that the test results of the laboratory need to reach when participating in proficiency testing; the PT lower limit in external quality assessment refers to the minimum acceptable deviation limit that the test results of the laboratory need to reach when participating in proficiency testing. If the laboratory's test results are higher than the PT upper limit or lower than the PT lower limit, the laboratory will be considered unqualified in this round of proficiency testing.

[0058] S102: performing linear regression calculation on a plurality of test data of the test items to obtain a linear function;

[0059] Among them, for multiple test data of the test items, this step can analyze the linear relationship between the data through statistical methods and fit the corresponding linear function. The above method can determine the change trend between the target value and the result value, that is, how the other variable changes accordingly as a certain variable increases or decreases. Linear regression calculation can quantify this relationship and express it in the form of a linear function, providing a basis for subsequent performance evaluation.

[0060] This embodiment can perform linear regression calculation on multiple test data of each test item one by one. After performing linear regression calculation on multiple test data of the test item, if the target value and result value in the test data do not meet the linear regression conditions, linear regression calculation is performed on multiple test data of the next test item.

[0061] S103: Determine a corresponding slope coefficient K according to the slope of the linear function;

[0062] Among them, on the basis of obtaining the linear function of the medical testing equipment in the test item, this step can determine the corresponding slope coefficient K according to the slope of the linear function. In this embodiment, the corresponding relationship between the slope and the slope coefficient can be preset, and the corresponding slope coefficient K can be determined according to the slope of the linear function based on the corresponding relationship. The slope coefficient K reflects the sensitivity or response degree of the test result, that is, the amount of change in the result reflected by the medical testing equipment to the change of the test item.

[0063] S104: Determine a corresponding intercept coefficient L according to the intercept of the linear function;

[0064] Wherein, on the basis of obtaining the linear function of the medical testing equipment in the detection item, this step can determine the corresponding intercept coefficient L according to the intercept of the linear function. In this embodiment, the corresponding relationship between the intercept and the intercept coefficient can be preset, and based on the corresponding relationship, the corresponding intercept coefficient L can be determined according to the intercept of the linear function. The intercept coefficient L is used to represent the background noise or the inherent deviation of the medical testing equipment.

[0065] S105: Calculating the comprehensive evaluation index value of the detection item according to a preset formula;

[0066] Among them, based on the slope coefficient K and intercept coefficient L of the medical testing equipment in the test items that have been determined, this step uses a preset formula to calculate the comprehensive evaluation index value bioy. The above preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)), bioy represents the comprehensive evaluation index value, A, B and C are preset coefficients;

[0067] The above preset formula combines the influence of the slope coefficient K and the intercept coefficient L, and takes into account the weights of the preset coefficients A, B, and C. Specifically, the calculation of bioy takes into account the linear and nonlinear combination effects of the slope K and the intercept L. By adjusting the values ​​of A, B, and C, it can flexibly reflect the contribution of different detection characteristics to the comprehensive evaluation index. The comprehensive evaluation index value calculated by the preset formula helps to comprehensively and objectively evaluate the accuracy and reliability of medical testing equipment in the detection items.

[0068] Specifically, this embodiment does not limit the values ​​of A, B and C; the values ​​of A, B and C can be greater than 0, A is greater than B, and A is greater than C.

[0069] S106: generating an error evaluation result according to the comprehensive evaluation index value of the detection item;

[0070] Among them, the comprehensive evaluation index value takes into account the influence of the slope coefficient and the intercept coefficient, as well as the adjustment of the preset coefficient, and can accurately reflect the accuracy and stability of the device during the detection process. Therefore, the error evaluation result can be generated according to the comprehensive evaluation index value of the detection item. The above error evaluation result is used to describe the error degree of the medical test result of the medical test device, and the comprehensive evaluation index value is negatively correlated with the error degree; a higher comprehensive evaluation index value usually means that the error of the medical test result is smaller and the medical test result is more reliable. This embodiment can perform error correction on the medical test device according to the error evaluation result.

[0071] As a feasible implementation method, this embodiment can set a corresponding weight value for each detection item. After obtaining the comprehensive evaluation index value of all the detection items, the comprehensive evaluation index value of all the detection items of the medical testing equipment is weightedly calculated to obtain a weighted index value; determine whether the weighted index value is greater than a preset index value; if so, determine that the performance of the medical testing equipment is normal; if not, determine that the performance of the medical testing equipment is abnormal.

[0072] This embodiment obtains multiple test data of the medical testing equipment in each test item from the inter-laboratory quality evaluation data, and performs linear regression analysis on the test data to obtain a linear function. This embodiment also determines the slope coefficient and the intercept coefficient according to the slope and intercept of the linear function, respectively, and the slope coefficient and the intercept coefficient directly reflect the degree of deviation between the target value and the result value. This embodiment also calculates the comprehensive evaluation index value of the test item according to a preset formula, so as to generate an error evaluation result according to the comprehensive evaluation index value; the above preset formula comprehensively considers the influence of the slope coefficient and the intercept coefficient to comprehensively reflect the overall quality and consistency of the data. This embodiment realizes the error evaluation of the medical test results through linear regression analysis and comprehensive evaluation index value calculation, and can quantitatively evaluate the errors of the medical test results, thereby improving the accuracy of error detection.

[0073] As for Figure 1 As a further introduction to the corresponding embodiment, the corresponding slope coefficient K can be determined in the following manner: determine the maximum allowable error P0 of the inter-laboratory quality assessment data, and determine the first exemption error P1 according to the maximum allowable error P0; wherein 0<P1<P0<1. The value of the slope coefficient K is determined according to the slope m representing the linear function, the maximum allowable error P0 and the first exemption error P1.

[0074] If 1-P1<m<1+P1, the value of the slope coefficient K is set to 0;

[0075] If m≤1- P0 or m≥1+ P0, the value of the slope coefficient K is set to 1;

[0076] If 1+P1≤m<1+P0, the value of the slope coefficient K is determined according to the first coefficient calculation formula; wherein the first coefficient calculation formula is ;

[0077] If 1-P0<m≤1-P1, the value of the slope coefficient K is determined according to the second coefficient calculation formula; wherein the second coefficient calculation formula is .

[0078] Specifically, in this embodiment, P0 and P1 can be any value greater than 0 and less than 1.

[0079] As for Figure 1 For further description of the corresponding embodiment, the corresponding intercept coefficient L may be determined in the following manner: Determine the second exemption error P2; 0<P2<1;

[0080] Calculate reference coefficients ; Where i represents the number of test data in the test project, b represents the intercept of the linear function, B i Indicates the maximum allowable deviation value;

[0081] like <P2, then the value of the intercept coefficient L is set to 0;

[0082] like >1, then the value of the intercept coefficient L is set to 1;

[0083] If P2≤ ≤1, the value of the intercept coefficient L is determined according to the third coefficient calculation formula; wherein the third coefficient calculation formula is: .

[0084] Through the above method, the value range of L can be controlled within 0~1.

[0085] Specifically, in this embodiment, P2 can be any value greater than 0 and less than 1. The practical mathematical meaning is: the composite ratio of the constant error brought by the intercept and the allowable range; if the final ratio of the error brought by the intercept and the allowable range is less than P2:1, that is, P2, there is no need to calculate the error; if the final ratio of the constant error brought by the intercept and the allowable range is 1:1, that is, 100%, it exceeds the critical value, and full marks are given at this time.

[0086] As a feasible implementation, this embodiment can use the linear regression calculation object SimpleRegression to perform linear regression calculation on multiple test data of the test items. In addition to using the SimpleRegression object in the Apache CommonsMath library (a mathematical and statistical library) for linear regression analysis, other statistical software packages can also be used for linear regression calculation, such as the scikit-learn library of R language or Python. These tools provide rich statistical analysis functions and more flexible data processing capabilities.

[0087] As a feasible implementation method, after calculating the comprehensive evaluation index value of the detection item according to a preset formula, the inter-laboratory quality evaluation value of the medical testing equipment in each of the detection items can also be obtained; if the comprehensive evaluation index value is greater than a first value, the detection item whose inter-laboratory quality evaluation value is lower than a preset evaluation value is marked as a detection item to be improved; if the comprehensive evaluation index value is less than a second value, the detection item whose inter-laboratory quality evaluation value reaches the maximum value is marked as a detection item to be improved; wherein, the first value is greater than the second value.

[0088] The process described in the above embodiment is explained below through an embodiment in actual application.

[0089] In the fields of biomedicine and clinical diagnosis, it is crucial to accurately evaluate the reliability and consistency of test results. With the advancement of technology, the demand for laboratory testing and quality control is growing, and the need for in-depth analysis and comprehensive evaluation of data is also increasing. Traditional data analysis methods, such as linear regression analysis, although they play an important role in evaluating the linear relationship between data points, still have limitations when dealing with complex data sets. Therefore, developing a method that can evaluate project data more comprehensively and in-depth has become an important direction of current technological development.

[0090] At present, technicians mainly use linear regression analysis, error analysis and data preprocessing techniques when evaluating project data. These techniques include calculating slope, intercept and correlation coefficient, and evaluating the mean and standard deviation of data. Although these methods can provide the fit and error range of data to a certain extent, they often seem to be unable to cope with complex error patterns, such as proportional error and constant error. In addition, the existing technology lacks sufficient depth and precision when evaluating the deviation of a single data point or batch number. Although the existing technology has played a certain role in data analysis and quality control, it still has some significant defects. First, when evaluating errors, the existing technology usually only focuses on a single type of error, such as systematic error or random error, and ignores the mixed effects of multiple errors. Secondly, when dealing with outliers, the existing technology may rely too much on simple elimination or correction methods without in-depth analysis of the reasons behind the outliers. In addition, when evaluating the accuracy and reliability of test results, the existing technology relies on fixed thresholds and lacks the ability to dynamically adjust. These defects limit the effectiveness of the existing technology in a wider range and more complex application scenarios. Therefore, it is necessary to develop a new method that can evaluate project data more comprehensively and in-depth and improve the accuracy and reliability of data analysis.

[0091] The present invention relates to a method for calculating a comprehensive evaluation index value. This method significantly improves the accuracy and reliability of project data evaluation through precise data analysis and statistical means. Based on the comprehensive evaluation index value, equipment performance can be accurately and reliably evaluated. The specific technical implementation includes the following key steps:

[0092] Data preprocessing: Use the high-precision java.math.BigDecimal data type to preprocess the data queried from the data storage server or manually entered to ensure the accuracy and consistency of the data. In addition to using the java.math.BigDecimal data type for high-precision calculations, other high-precision calculation libraries, such as BigFraction or BigRational in Apache Commons Math, can also be used to ensure accurate data processing.

[0093] Linear regression analysis: The SimpleRegression object in the Apache Commons Math library was used to perform linear regression analysis on the target value and result value of each batch number, and the slope m, intercept b and correlation coefficient R were calculated.

[0094] Calculation of comprehensive evaluation index value: Based on the slope coefficient K and the intercept coefficient L, the comprehensive evaluation index value is obtained through the specific calculation formula bioy=AB×((1+L)×K)-C×((1+K)×L)). This index comprehensively reflects the measurement quality and consistency of the project.

[0095] Result output and storage: The calculated comprehensive evaluation index values ​​and related analysis results are stored in the data storage system for subsequent query and analysis. Web-side interactive design: A user-friendly web interface is developed to allow users to upload files, customize report content, view analysis results, and provide feedback and corrections.

[0096] This embodiment adopts high-precision data types and accurate linear regression analysis, and the present invention significantly improves the accuracy of data preprocessing and analysis. The comprehensive evaluation index value introduced in this embodiment provides a quantitative evaluation index, which fully reflects the overall quality and consistency of the data and enhances the reliability of the data. This embodiment optimizes the data processing flow and reduces unnecessary testing and waste of resources by refining the analysis of data points and introducing multi-dimensional error analysis. The accurate data evaluation and optimized data processing flow of this embodiment can significantly reduce the processing time of laboratories and production departments on data issues and improve production efficiency. This embodiment can significantly reduce the cost of production and testing by reducing repeated tests and waste of resources, and the cost savings are expected to reach 15% to 25%. The optimized data processing flow of this embodiment reduces the consumption of energy and resources, and the energy consumption is expected to be reduced by about 20%. This embodiment improves data quality, which can not only improve the quality of the final product or service, but also enhance customer trust and market competitiveness, especially in industries that require high-precision data support.

[0097] This embodiment uses technical means such as linear regression analysis, error analysis and data preprocessing to accurately evaluate and analyze project data. Through this method, the "source index" bioy, that is, the comprehensive evaluation index value, can be calculated to quantitatively evaluate the accuracy and reliability of the project. This technology has significant advantages in processing complex data sets and conducting scientific analysis. This embodiment has broad application prospects in the fields of biomedicine and clinical diagnosis. Especially in laboratory testing and quality control, it can help researchers and clinicians evaluate the accuracy and consistency of test results. By calculating the source index, possible systematic errors and random errors can be identified and corrected, thereby improving the reliability of diagnostic tests.

[0098] The following is a detailed introduction to the implementation process of this solution:

[0099] Step 1. Query and read the data from the data storage server, which is derived from the parsed data of the inter-laboratory quality assessment report PDF file or the manually entered inter-laboratory quality assessment result data, and has been manually corrected, and is used for error analysis and source index calculation. It mainly includes the following information related to inter-laboratory quality assessment testing: project name, project unit, project test batch number, batch test result value, batch target value, bias value, standard deviation, batch test value allowable range, project pt (proficiency testing, capability verification) results, and the information such as the instruments, methods, reagents, and calibrants used in the project test, and the project collection object storage for the report is opened in the server memory.

[0100] Step 2: Define a set of evaluation rules, and store the project evaluation rules predefined manually in the background obtained from the data storage server. For example, the fitting straight line is approximately equal to the standard line, the proportional deviation (slope) is insensitive, and there is no proportional error.

[0101] Step 3. Since the evaluation rule set contains multiple types of data such as analysis, conclusions, and suggestions, in order to facilitate subsequent matching and searching for a variety of different types of analysis, conclusions, suggestions, etc., the evaluation rule set of the project conclusions in step 2 is automatically grouped by computer according to four classification types (1. analysis, 2. conclusion, 3. suggestion, 4. data problem) to obtain a grouped project evaluation rule.

[0102] In addition to using the Stream API introduced in Java 8 and higher versions combined with the Collectors tool class to implement data grouping processing, data grouping can also be implemented using several technologies including but not limited to: 1. SQL querying the collection stone in the database to specify the grouping field implementation; 2. Using the Java 8 Map interface and its implementation class; 3. Using a third-party library such as Apache Commons Collections' MultiMap. In addition, this embodiment can also use a distributed computing framework, such as Apache Hadoop or Spark, to process large-scale data sets to improve the efficiency and scalability of data processing.

[0103] Step 4: traverse the objects grouped in step 3, and define an analysis set from the grouped evaluation rule objects to store the analysis class evaluation rule data of type 1.

[0104] Step 5. In order to further refine the analysis data set objects obtained in step 4 and facilitate the subsequent rapid matching to obtain a specific type of analysis script, the analysis data set obtained in step 4 is automatically grouped by computer according to the analysis type (1, proportional error analysis, 2, constant error analysis, 3, precision analysis), and the grouped objects are traversed to obtain three types of analysis and judgment script objects of proportional error, constant error, and precision.

[0105] Step 6: traverse the objects grouped in step 3, and define a conclusion set storage type 2 conclusion class judgment rule data from the grouped judgment rule objects.

[0106] Step 7. In order to further refine the conclusion data set objects obtained in step 6 and facilitate the subsequent rapid matching to obtain a specific type of conclusion language, the conclusion class set obtained in step 6 is automatically grouped by computer according to the conclusion type (1. proportional error conclusion, 2. constant error conclusion, 3. precision conclusion, 4. mixed error conclusion), and the grouped objects are traversed to obtain the conclusion language objects of proportional error class, constant error class, precision class, and mixed error class respectively.

[0107] Step 8: Traverse the objects grouped in step 3, and define a suggestion set from the classified evaluation rule data to store all the suggestion category evaluation rule data.

[0108] Step 9. In order to further refine the suggestion class data set objects obtained in step 8 and facilitate the subsequent rapid matching to obtain a specific type of suggestion words, the suggestion class set obtained in step 8 is grouped according to the suggestion type (1, proportional error class suggestions, 2, constant error class suggestions, 3, mixed error class suggestions, 5, precision class suggestions) and then traversed to obtain each type of suggestion object of proportional error class, constant error class, mixed error class, and precision class respectively.

[0109] Step 10: The different classification objects refined in the above steps 5, 7, and 9 are used as parameters of the source index calculation method, and are brought into the source index calculation method for matching and use in the corresponding specific situations in steps 23, 32, 41, and 48.

[0110] Step 11. Since each item in the inter-laboratory quality assessment test needs to be calculated for the source index separately and are not related to each other, the query results of step 1 are looped through, and the data of one inter-laboratory quality assessment test item is obtained each time, and the project data object is defined to store the data of a single inter-laboratory quality assessment test item.

[0111] Step 12, starting the source index correlation calculation for the single detection item data object in step 11, including the following steps:

[0112] Step 13: Since non-scoring items do not participate in the calculation, for a single item data object, you need to first obtain its scoring item attributes (the user manually sets whether to include it in the calculation scope when uploading the inter-laboratory quality assessment report or manually entering the data), and determine whether the item participates in the scoring based on the attribute value. Only perform data calculations on the scoring items.

[0113] Step 14, define the linear regression calculation object, using Apache's SimpleRegression object, loop the single project data object in step 11, get the project batch data, take the target value data of each batch as the X value, and the result value of the batch as the Y value, substitute them into the linear regression calculation object SimpleRegression, and use its built-in linear regression analysis method to perform linear regression calculation.

[0114] Step 15: After substituting all the batch number data in the current project into the calculation, obtain the slope m, intercept b, and correlation coefficient R of the linear regression of the batch number of the project from the linear regression calculation object SimpleRegression.

[0115] Step 16: Determine the linear regression of each batch of project data, including the following steps:

[0116] (1) If the slope does not exist or is 0, add the conclusion: the data of this project does not satisfy the linear regression; end the calculation of this project, continue to loop through step 11, and start the calculation of the next project;

[0117] (2) If the intercept does not exist, add the conclusion: the data for this project does not satisfy linear regression; end the calculation for this project, continue to loop through step 11, and start the calculation for the next project;

[0118] (3) If the correlation coefficient does not exist, add the conclusion: the data of this project does not satisfy the linear regression; end the calculation of this project, continue to loop through step 11, and start the calculation of the next project.

[0119] Step 17: For the items that meet the linear regression requirements after the determination in step 16, calculate the slope coefficient K and the intercept coefficient L.

[0120] Step 18. In linear regression analysis, it is defined that a slope greater than 1 + P0 belongs to the standard line with an obvious deviation from y = x, that is, the laboratory value deviates from the target value by more than P0. P0 is a threshold for judging proportional error. When the slope exceeds P0, the influence of the slope on the standard source index reaches the maximum. Then, when the limit value is P0, it is considered that each experiment cannot be absolutely accurate. Therefore, a part of the proportional error of P0, namely P1, is used as the exemption error and not included in the calculation range of the standard source index. Thus, the positive slope of the K value is between (1 + P1) and (1 + P0), and the negative slope is between (1 - P0) and (1 - P1). The positive slope indicates how much the laboratory value increases based on the target value, and the negative slope indicates how much the laboratory value decreases based on the slope. To make the increase of 1% and the decrease of 1% consistent in the calculation of the standard source index, the negative slope needs to be converted into a positive slope for calculation. Therefore, there is a definition of the true slope M, that is, when the slope is greater than 1, M = m; when the slope is less than 1, M = 2 - m. To ensure that M obtains a linear K value between (1 + P1) and (1 + P0), when M = 1 + P1, the slope coefficient K is scored 0; when M = 1 + P0, the slope coefficient K is scored 1. Since two points determine a straight line, substituting the two point coordinates of M = 1 + P1, K = 0 and M = 1 + P0, K = 1, the K value equation obtained based on the M value is . Based on the above calculation logic, according to the slope m obtained in Step 15, define the slope coefficient object K for calculation and judgment. When 1 - P1 < m < 1 + P1, it is defaulted to an excellent slope and the K value is not calculated, K = 0; when 1 - P0 ≥ m or m ≥ 1 + P0, it is defaulted to a non - qualified slope, K = 1; when 1 + P1 ≤ m < 1 + P0, M = m, ; when 1 - P0 < m ≤ 1 - P1, M = 2 - m, .

[0121] Step 19. Since the magnitude of a single intercept in the project linear regression calculation cannot directly explain the problem, for a project (under a fixed intercept), the smaller the batch target value, the greater the influence of the intercept; the larger the batch target value, the smaller the influence of the intercept. Therefore, the calculation of the intercept adopts a composite calculation, and there is a calculation formula for the power - weighted coefficient n of the intercept. The intercept coefficient L is calculated from the influence degree of the intercept on the allowable bias value for each batch number, that is, the power - weighted coefficient n of the intercept. When the influence of the intercept reaches 100%, the influence rate of the intercept reaches the maximum. In specific experimental operations, since the existence of errors cannot be completely avoided, an exemption error of P2 is given and not included in the calculation range of the standard source index. Since two points determine a straight line, substituting the coordinate values of two points: n = P2, L = 0; n = 1, L = 1, the equation for the L value is According to the slope b obtained in step 15, define the intercept coefficient object L for calculation and substitute it into the calculation formula: The specific calculation formula of n value is as follows, where i is the number of batches under the project, B i is the unilateral allowable deviation value of a single batch number, taking the difference between the allowable upper limit value of the batch number and the target value:

[0122] ;

[0123] Step 20. Substitute the slope coefficient K in step 18 and the intercept coefficient L in step 19 into the calculation formula of the source index: bioy=AB×((1+L)×K)-C×((1+K)×L)) to calculate the source index; based on the source index value and the pt score of the project, experts recommend: when the source index is greater than or equal to 85, focus on test items with low pt scores (less than or equal to 60); when the source index is greater than or equal to 75 and less than 85, focus on projects with very low or even 0 PT scores; when the source index is less than 75, pay more attention to projects with full PT scores. The calculation formula of the standard source index is divided into three parts. The first part A is the upper limit of the standard source index, that is, when L / K are both equal to 0, the score of the standard source index; the second part is (1+L)K, where the score of the slope coefficient K is the score of the slope part; the reason why K has a coefficient of 1+L is that it is considered that the univariate coefficient of the slope coefficient K alone cannot reflect the effect of data discreteness, that is, the intercept and slope corresponding to the intercept coefficient L and the slope coefficient K have a common influence on the straight line. Therefore, the influence coefficient of the corresponding intercept coefficient L is added before the slope coefficient K. The third part is (1+K)L in the same way. The coefficient values ​​of the second and third parts are B and C respectively; if the values ​​of L and K are 1, the constant is A-2B-2C.

[0124] Step 21. Combined with the judgment criteria in steps 18 and 19, the thresholds of the proportional (slope) error and the constant error are divided into 7 parts, the smallest part is the error exemption area, and the largest part is beyond the acceptable range. Between the minimum acceptable and maximum acceptable thresholds, it is divided into 5 stages in combination with the opinions of industry experts. The upper limit of the first stage is 25% of the limit threshold. Experts believe that although there is an error at this level, it has not reached the level that can be evaluated and needs prompts and corrections; the upper limit of the second stage is 35% of the limit threshold. Experts suggest that when the error is at this starting point, it can be evaluated and corrected. The end threshold of this stage is obtained based on data analysis; the end threshold of the third stage is 50% of the limit threshold; the last two stages are an average distribution of the remaining 50% of the limit threshold, 25% each. That is, the intervals are: 0~10%, 10%~25%, 25%~35%, 35%~50%, 50%~75%, 75%~100%. Therefore, the overall segmentation is 15%, 10%, 15%, 25%, and 25% of the limit threshold, which can reflect the precise control of small errors in the medical testing industry. In the stage with small errors (the first 50%), it is smoothly divided into 3 segments for evaluation. In the stage with large errors (the last 50%), the error stage setting is relatively open.

[0125] Step 22: Perform error analysis and conclusion and suggestion summary of the proportional (slope) error class according to the value of the true slope M calculated in step 18.

[0126] Step 23. According to the segmented judgment standard recommended by the experts in step 21, combined with the proportional error threshold of 1.2 (i.e. 1+ P0) and the error exemption value of 1.02 (i.e. 1+ P1) in step 18, 15%, 10%, 15%, 25%, and 25% of the threshold are taken as the endpoint values ​​of different intervals, and the proportion (slope) is divided into seven levels according to its influence degree. The corresponding preset template statements related to the "proportional error class" are obtained from the speech object sets of steps 5, 7, and 9, and the seven situations described in the following steps 24 to 30 can be obtained.

[0127] Step 24. When M<1.02, it is an excellent slope. The analysis shows that the proportion (slope) of this project has no deviation. The conclusion is: the fitting straight line is approximately equal to the standard line, the proportion deviation (slope) is insensitive, and there is no proportion error. No recommendations are given.

[0128] Step 25. When M≥1.02 and M<1.05, the analysis shows that the proportion (slope) of the project has basically no deviation; the conclusion is: the fitting straight line is close to the standard line, the proportion deviation (slope) is not very sensitive, and the proportion error is slight; no recommendations are available.

[0129] Step 26. When M≥1.05 and M<1.07, analyze the slight deviation of the proportion (slope) of the project; conclude that the fitting straight line has a small deviation from the standard line, the deviation of the proportion (slope) is relatively sensitive, and the proportion error is small; suggestion: check all calibration values ​​and check the method specificity.

[0130] Step 27. When M≥1.07 and M<1.10, analyze the proportion (slope) deviation of this item; conclude that the fitting straight line deviates from the standard line, the proportion (slope) deviation is sensitive, and there is a proportion error; recommended content: check all calibration values ​​and check method specificity.

[0131] Step 28. When M≥1.10 and M<1.15, the deviation of the proportion (slope) of this project is relatively large. Conclusion: There is a large deviation between the fitting straight line and the standard line, the deviation of the proportion (slope) is relatively sensitive, and the proportion error is large. Recommendation: Check all calibration values ​​and check the method specificity.

[0132] Step 29. When M≥1.15 and M<1.20, the analysis shows that the proportion (slope) of this project deviates greatly. It is concluded that the fitting straight line deviates greatly from the standard line, the proportion (slope) deviation is very sensitive, and the proportion error is close to the critical value (20%). Recommendation: Check all calibration values ​​and check the method specificity.

[0133] Step 30. When M≥1.20, analyze the large deviation of the proportion (slope) of this project; conclude that there is a very large deviation between the fitting straight line and the standard line, the deviation of the proportion (slope) is very sensitive, and the proportion error exceeds the critical value (20%); Recommendation: Check all calibration values ​​and check the method specificity.

[0134] Step 31: Based on the value of the intercept weighting coefficient n calculated in step 19, perform error analysis and conclusion and suggestion summary for the constant error (intercept) class.

[0135] Step 32. According to the expert-suggested segmented judgment criteria in step 21, combined with the constant error threshold 1 and the error exemption value 0.1 (i.e. P2) in step 19, 15%, 10%, 15%, 25%, and 25% of the threshold are taken as the endpoint values ​​of different intervals, and the impact of the constant error is divided into seven levels. The corresponding preset template statements related to the "constant error class" are obtained from the speech object sets in steps 5, 7, and 9, and the seven situations described in steps 33 to 39 can be obtained.

[0136] Step 33. When n<0.1, the analysis shows that the constant (intercept) of this project has no offset; it is concluded that the constant error (intercept) has no effect on all levels; no recommendations are given.

[0137] Step 34. When n≥0.1 and n<0.25, the analysis shows that the constant (intercept) of this project has basically no deviation; it is concluded that the constant error (intercept) has a slight impact on the low level and has basically no impact on all levels; no recommendations are given.

[0138] Step 35. When n≥0.25 and n<0.35, the analysis shows that the project constant (intercept) is slightly offset. The conclusion is: the fitting straight line has a small deviation from the standard line, the proportion (slope) deviation is relatively sensitive, and the proportion error is small. Recommendations: Check the blank and the method specificity.

[0139] Step 36. When n≥0.35 and n<0.50, analyze the project constant (intercept) offset; conclude that the constant error (intercept) has an impact on the low level, a small impact on the medium and high levels, and a certain impact on all levels combined; recommended content: check the blank and check the method specificity.

[0140] Step 37. When n≥0.50 and n<0.75, the analysis shows that the constant (intercept) deviation of this project is relatively large. It is concluded that the constant error (intercept) has a greater impact on the low and medium levels, has an impact on the high level, and has a more obvious impact on all levels combined. Recommendations: Check the blank and check the method specificity.

[0141] Step 38. When n≥0.75 and n<1, the analysis shows that the constant (intercept) of this project has a large offset; it is concluded that the constant error (intercept) has a large impact on the low and medium levels, a greater impact on the high level, and a significant impact on all levels combined; recommended content: check the blank and check the method specificity.

[0142] Step 39. When n≥1, analyze that the project constant (intercept) has a large offset; conclude that the constant error (intercept) has a large impact on each level and has a significant impact on all levels combined; recommended content: check blanks and method specificity.

[0143] Step 40: After completing the single error analysis of steps 21 to 39, it is necessary to combine the proportional error and the constant error to comprehensively consider the influence of the mixed error caused by the two.

[0144] Step 41. According to the slope m and intercept b obtained in step 15, when m>1 and b<0, or when m<1 and b>0, the mixed error condition is met; at this time, combined with the true slope M in step 18 and the power weighted average coefficient n in step 19 and the expert suggestion segmentation judgment standard in step 21, the corresponding preset template statement related to the "mixed error class" is obtained from the suggestion speech object set in step 9, and the four situations described in steps 42 to 45 are obtained.

[0145] Step 42: When M<1.05 and n<0.25, the mixed error of this project can be ignored and no recommendations are given.

[0146] Step 43: When M < 1.05 and n ≥ 0.25, recommended contents: single-point calibration, check blank, check method specificity.

[0147] Step 44. When M ≥ 1.05 and n ≥ 0.25, recommended content: single-point calibration, check all calibrator values, check blank.

[0148] Step 45. When M ≥ 1.05 and n < 0.25, recommended content: single-point calibration, check all calibrator values, and check method specificity.

[0149] Step 46. The suggestions in steps 24 to 30, 33 to 39, and 42 to 45 above involve four types of inspection and calibration techniques related to testing instruments, including checking blanks, checking method specificity, checking all calibrator values, and single-point calibration. Their specific practical significance in external quality assessment testing is as follows:

[0150] (1) Check all calibrator values: Calibrators are substances used to ensure that a measuring device or method provides accurate results. They contain known concentrations of analytes. By comparing the results of a measuring device with the known values ​​of the calibrators, the device can be calibrated.

[0151] (2) Test method specificity: Method specificity refers to the ability of a test method to detect a specific analyte, including its ability to accurately distinguish the target analyte from other substances. In medical testing, method specificity is crucial to reducing cross-reactions and improving test accuracy. Test method specificity usually involves evaluating the performance of the test method under different sample types, operating conditions or reagent batches;

[0152] (3) Check blank: Check blank means that during the test process, a blank sample is analyzed using the same processing method as the sample to evaluate the contamination or background signal that may be introduced during the test. The blank sample usually does not contain the substance to be tested, or only contains a low concentration of the substance with a known concentration. By comparing the results of the blank sample with the results of the sample, errors caused by contamination of reagents, containers or other non-sample sources can be estimated and corrected;

[0153] (4) Single-point calibration: In the inspection industry, mixed error refers to a systematic error in which the slope and intercept in the regression analysis show opposite directions, that is, there is a certain proportion of proportional error and constant error. Under such error conditions, single-point calibration has become a commonly used calibration method due to its simplicity and cost-effectiveness. It uses a standard of known concentration to adjust the measurement system and is suitable for situations where the measurement system error is mainly proportional error and the error is relatively constant. Single-point calibration not only simplifies the calibration process and reduces the time and resources required, but is also suitable for situations where a quick response is required.

[0154] Step 47: Based on the correlation coefficient R obtained in step 15, conduct error analysis of project precision and summarize conclusions and suggestions.

[0155] Step 48. Combined with the precision error judgment standard for the correlation coefficient R provided by the inter-laboratory quality assessment agency, the value range of R is divided into three levels. The corresponding preset template statements related to the "precision error category" are obtained from the speech object sets of steps 5, 7, and 9. Error analysis and conclusion suggestion matching are performed to obtain the three situations described in the following steps 49 to 51.

[0156] Step 49. When R≥0.999, the linear regression fit of the project is high. It is concluded that the precision of the test project is good. No recommendations are needed.

[0157] Step 50. When R≥0.985 and R<0.999, the linear regression fit of the project is acceptable. It is concluded that the precision of the test project is good. No recommendations are needed.

[0158] Step 51. When R<0.985, the linear regression fit of the project is relatively low. Conclusion: the precision of the test project needs attention. Suggestions: check or replace reagents, and check the compatibility of reagents and instruments.

[0159] Step 52: Update the calculation results, suggestions, conclusions, analysis, etc. obtained in steps 13 to 51 to the project data object in step 11, and loop through all the project sets of the entire report in step 1 in turn, performing the source index related calculations and error analysis, as well as error conclusions and suggestions of steps 13 to 51 for each project.

[0160] Step 53: After completing the calculation of the standard source index and the error analysis of all items, the recommended adjustment or inspection content of the detection instruments of all items is obtained, and an analysis report is generated using the results, and the results are saved to the data storage server.

[0161] This embodiment queries and reads the inter-laboratory quality assessment report data from the data storage server, manually enters or automatically parses the data, and ensures the accuracy and completeness of the original data, which is the basis for any subsequent analysis and calculation. This embodiment defines a set of judgment rules, and automatically groups the rules according to the types of analysis, conclusions, suggestions and data problems, so as to quickly retrieve and match, which is the key to ensure the standardization and normalization of the evaluation process. This embodiment uses the ApacheSimpleRegression object to perform linear regression analysis, calculates the slope (m), intercept (b) and correlation coefficient (R), and determines whether the data meets the linear regression conditions. This is the mathematical basis for the calculation of the standard source index. This embodiment converts the slope into the true slope M, and then calculates the slope coefficient K according to the M value, ensuring that the positive and negative changes of the slope are consistently processed in the calculation of the standard source index. This embodiment calculates the influence of the intercept on the allowable bias value of the batch number, obtains the power weighting coefficient n of the intercept, and then calculates the intercept coefficient L, which is an important step in evaluating the constant error. This embodiment brings the slope coefficient K and the intercept coefficient L into the standard source index calculation formula, and ensures the quantification of the joint influence of the slope and the intercept through the weighted calculation of (1+L)K and (1+K)L. Before calculation, this embodiment screens out the items involved in the calculation through the scoring item attributes, excludes non-scoring items, and ensures the pertinence and effectiveness of the calculation. This embodiment defines the error exemption range, conducts a graded assessment of the proportional error and the constant error, and divides the error into multiple stages in order to analyze the error degree in more detail and provide corresponding suggestions. This embodiment further refines the judgment rules into analysis, conclusions and suggestions for proportional error, constant error and precision error, so as to match them with the calculation results and provide specific analysis and suggestions. This embodiment summarizes the calculation results, analysis, conclusions and suggestions into the project data object, generates a complete analysis report, and saves it to the data storage server, which is a necessary link to realize the automation process and result archiving.

[0162] In this embodiment, the evaluation rules are subdivided into four categories: analysis, conclusion, suggestion and data problems, and each type is further grouped by error type (proportional error, constant error, precision error) to ensure the targeted nature of the evaluation process. This embodiment uses the Apache Simple Regression object to perform linear regression analysis to quantify the relationship between laboratory measurement values ​​and target values. The source index is calculated by accurately calculating the slope, intercept and correlation coefficient. This embodiment introduces the concept of true slope M to standardize the impact of the slope and ensure consistency in the calculation of positive and negative deviations. Through a specific formula , the slope coefficient K is able to change linearly within a limited range, reflecting the degree of proportional error. Taking into account the limitations of a single intercept size, this embodiment proposes the concept of a square weighted coefficient n of the intercept to measure the degree of influence of the intercept on the allowable bias values ​​of different batches, and then calculates the intercept coefficient L. The calculation formula of the source index comprehensively considers the interaction of the slope coefficient K and the intercept coefficient L, and through the weighted calculation of (1+L)K and (1+K)L, ensures the common influence of the slope and the intercept on the final index, reflecting the comprehensiveness and scientificity of the assessment. In this embodiment, the proportional error and the constant error are subdivided into seven levels, from the error exemption zone to beyond the acceptable range, providing a detailed evaluation standard, making the error assessment more refined, and providing corresponding analysis, conclusions and suggestions according to the error level, enhancing the guiding significance of the assessment.

[0163] In conventional medical tests, the existence of proportional error, constant error, and mixed error is very common. In general, proportional error, constant error, and mixed error are judged based on the slope and intercept of linear regression. However, the judgment of mixed error is more subjective. Therefore, the error level or error distribution of the inspection industry cannot be fully reflected by the simple least squares linear regression. The comprehensive evaluation index value in this embodiment reflects the proportional error, constant error, and mixed error in this inspection and expresses them in numerical form, so as to locate the error level of the instrument and calibrate the instrument according to the error level.

[0164] See also Figure 2 , Figure 2 A schematic diagram of the structure of a medical test result error evaluation system provided in an embodiment of the present application;

[0165] The system may include:

[0166] The data acquisition module 201 is used to acquire multiple test data of the medical testing equipment in each test item from the inter-laboratory quality evaluation data; wherein the test data includes a PT upper limit value, a PT lower limit value, a target value and a result value; PT represents capability verification;

[0167] A linear regression calculation module 202 is used to perform linear regression calculation on a plurality of test data of the test items to obtain a linear function;

[0168] A slope coefficient determination module 203, configured to determine a corresponding slope coefficient K according to the slope of the linear function;

[0169] An intercept coefficient determination module 204, configured to determine a corresponding intercept coefficient L according to the intercept of the linear function;

[0170] The comprehensive evaluation index determination module 205 is used to calculate the comprehensive evaluation index value of the detection item according to a preset formula; wherein the preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)) , bioy represents the comprehensive evaluation index value, and A, B and C are preset coefficients;

[0171] The error evaluation module 206 is used to generate an error evaluation result according to the comprehensive evaluation index value of the detection item; wherein the error evaluation result is used to describe the error degree of the medical test result of the medical testing equipment.

[0172] This embodiment obtains multiple test data of the medical testing equipment in each test item from the inter-laboratory quality evaluation data, and performs linear regression analysis on the test data to obtain a linear function. This embodiment also determines the slope coefficient and the intercept coefficient according to the slope and intercept of the linear function, respectively, and the slope coefficient and the intercept coefficient directly reflect the degree of deviation between the target value and the result value. This embodiment also calculates the comprehensive evaluation index value of the test item according to a preset formula, so as to generate an error evaluation result according to the comprehensive evaluation index value; the above preset formula comprehensively considers the influence of the slope coefficient and the intercept coefficient to comprehensively reflect the overall quality and consistency of the data. This embodiment realizes the error evaluation of the medical test results through linear regression analysis and comprehensive evaluation index value calculation, and can quantitatively evaluate the errors of the medical test results, thereby improving the accuracy of error detection.

[0173] Further, the process of the slope coefficient determination module determining the corresponding slope coefficient K according to the slope of the linear function includes: determining the maximum allowable error P0 of the inter-room quality assessment data, and determining the first exemption error P1 according to the maximum allowable error P0; wherein, 0<P1<P0<1; if 1-P1<m<1+P1, the value of the slope coefficient K is set to 0; wherein, m represents the slope of the linear function; if m≤1-P0 or m≥1+P0, the value of the slope coefficient K is set to 1; if 1+P1≤m<1+P0, the value of the slope coefficient K is determined according to the first coefficient calculation formula; wherein, the first coefficient calculation formula is ; If 1- P0<m≤1- P1, the value of the slope coefficient K is determined according to the second coefficient calculation formula; wherein the second coefficient calculation formula is .

[0174] Furthermore, the process of determining the corresponding intercept coefficient L according to the intercept of the linear function by the intercept coefficient determination module includes: determining the second exemption error P2; wherein 0<P2<1; calculating the reference coefficient ; Where i represents the number of test data in the test project, b represents the intercept of the linear function, B i Indicates the maximum allowable deviation value; if <P2, then set the value of the intercept coefficient L to 0; if >1, then set the intercept coefficient L to 1; if P2≤ ≤1, the value of the intercept coefficient L is determined according to the third coefficient calculation formula; wherein the third coefficient calculation formula is: .

[0175] Furthermore, the process of the linear regression calculation module performing linear regression calculation on the multiple detection data of the detection items includes: performing linear regression calculation on the multiple detection data of the detection items using the linear regression calculation object Simple Regression.

[0176] Furthermore, the linear regression calculation module is also used to perform linear regression calculation on multiple test data of the test item after performing linear regression calculation on the multiple test data of the test item, if the target value and the result value in the test data do not meet the linear regression conditions, then perform linear regression calculation on multiple test data of the next test item.

[0177] Furthermore, it also includes:

[0178] The performance analysis module is used to perform weighted calculation on the comprehensive evaluation index values ​​of the medical testing equipment in all the detection items to obtain a weighted index value; determine whether the weighted index value is greater than a preset index value; if so, determine that the performance of the medical testing equipment is normal; if not, determine that the performance of the medical testing equipment is abnormal.

[0179] Furthermore, it also includes:

[0180] A result analysis module is used to obtain the inter-laboratory quality evaluation value of the medical testing equipment in each of the test items after calculating the comprehensive evaluation index value of the test item according to a preset formula; it is also used to mark the test item whose inter-laboratory quality evaluation value is lower than a preset evaluation value as a test item to be improved if the comprehensive evaluation index value is greater than a first value; it is also used to mark the test item whose inter-laboratory quality evaluation value reaches a maximum value as a test item to be improved if the comprehensive evaluation index value is less than a second value; wherein the first value is greater than the second value.

[0181] Since the embodiments of the system part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the system part, which will not be repeated here.

[0182] The present application also provides a storage medium on which a computer program is stored, and when the computer program is executed, the steps provided in the above embodiment can be implemented. The storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0183] The present application also provides an electronic device, which may include a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps provided in the above embodiment may be implemented. Of course, the electronic device may also include various network interfaces, power supplies and other components.

[0184] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of this application.

[0185] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

Claims

1. A method for evaluating the error of medical test results, characterized in that: include: Acquire multiple test data of each test item of the medical testing equipment from the inter-laboratory quality evaluation data; wherein the test data includes PT upper limit value, PT lower limit value, target value and result value; PT stands for capability verification; Performing linear regression calculation on the multiple test data of the test items to obtain a linear function; Determine a corresponding slope coefficient K according to the slope of the linear function; Determine the corresponding intercept coefficient L according to the intercept of the linear function; The comprehensive evaluation index value of the test item is calculated according to a preset formula; wherein the preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)) , bioy represents the comprehensive evaluation index value, and A, B and C are preset coefficients; An error evaluation result is generated according to the comprehensive evaluation index value of the detection item; wherein the error evaluation result is used to describe the error degree of the medical test result of the medical testing equipment.

2. The error evaluation method for medical test results according to claim 1, characterized in that: Determining the corresponding slope coefficient K according to the slope of the linear function includes: Determine a maximum allowable error P0 of the inter-laboratory quality assessment data, and determine a first exemption error P1 according to the maximum allowable error P0; wherein 0<P1<P0<1; If 1-P1<m<1+P1, the value of the slope coefficient K is set to 0; wherein m represents the slope of the linear function; If m≤1- P0 or m≥1+ P0, the value of the slope coefficient K is set to 1; If 1+P1≤m<1+P0, the value of the slope coefficient K is determined according to the first coefficient calculation formula; wherein the first coefficient calculation formula is ; If 1-P0<m≤1-P1, the value of the slope coefficient K is determined according to the second coefficient calculation formula; wherein the second coefficient calculation formula is .

3. The error evaluation method for medical test results according to claim 1, characterized in that: Determining the corresponding intercept coefficient L according to the intercept of the linear function includes: Determine a second exemption error P2; wherein 0<P2<1; Calculate reference coefficients ; Where i represents the number of test data in the test project, b represents the intercept of the linear function, B i Indicates the maximum allowable deviation value; like <P2, then the value of the intercept coefficient L is set to 0; like >1, then the value of the intercept coefficient L is set to 1; If P2≤ ≤1, the value of the intercept coefficient L is determined according to the third coefficient calculation formula; wherein the third coefficient calculation formula is: .

4. The error evaluation method for medical test results according to claim 1, characterized in that: Performing linear regression calculation on multiple test data of the test items includes: The linear regression calculation object Simple Regression is used to perform linear regression calculation on the multiple detection data of the detection items.

5. The error evaluation method for medical test results according to claim 1, characterized in that: After performing linear regression calculation on the plurality of test data of the test items, the method further includes: If the target value and the result value in the test data do not satisfy the linear regression condition, a linear regression calculation is performed on multiple test data of the next test item.

6. The error evaluation method for medical test results according to claim 1, characterized in that: After calculating the comprehensive evaluation index value of the test item according to the preset formula, it also includes: Performing weighted calculation on the comprehensive evaluation index values ​​of all the detection items of the medical testing equipment to obtain a weighted index value; Determine whether the weighted index value is greater than a preset index value; if so, determine that the performance of the medical testing equipment is normal; if not, determine that the performance of the medical testing equipment is abnormal.

7. The error evaluation method for medical test results according to any one of claims 1 to 6, characterized in that: After calculating the comprehensive evaluation index value of the test item according to the preset formula, it also includes: Obtaining an inter-laboratory quality evaluation value of the medical testing equipment in each of the testing items; If the comprehensive evaluation index value is greater than the first value, the inspection item whose inter-room quality evaluation value is lower than the preset evaluation value is marked as an inspection item to be improved; If the comprehensive evaluation index value is less than a second value, the inspection item whose inter-room quality evaluation value reaches the maximum value is marked as an inspection item to be improved; wherein the first value is greater than the second value.

8. A medical test result error evaluation system, characterized in that: include: A data acquisition module is used to acquire multiple test data of the medical testing equipment in each test item from the inter-laboratory quality evaluation data; wherein the test data includes a PT upper limit value, a PT lower limit value, a target value and a result value; PT represents capability verification; A linear regression calculation module, used for performing linear regression calculation on a plurality of test data of the test items to obtain a linear function; A slope coefficient determination module, used to determine a corresponding slope coefficient K according to the slope of the linear function; An intercept coefficient determination module, used to determine the corresponding intercept coefficient L according to the intercept of the linear function; A comprehensive evaluation index determination module, used to calculate the comprehensive evaluation index value of the test item according to a preset formula; wherein the preset formula is: bioy=AB×((1+L)×K)-C×((1+K)×L)) , bioy represents the comprehensive evaluation index value, and A, B and C are preset coefficients; The error evaluation module is used to generate an error evaluation result according to the comprehensive evaluation index value of the detection item; wherein the error evaluation result is used to describe the error degree of the medical test result of the medical testing equipment.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, it implements the steps of the error evaluation method for medical test results as claimed in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores computer executable instructions, which, when loaded and executed by a processor, implement the steps of the error evaluation method for medical test results as described in any one of claims 1 to 7.

Images