A method and system for determining the content of crude polysaccharide based on multi-standard comparison

Abstract

The application discloses a crude polysaccharide content determination method and system based on multi-standard comparison, and the method comprises the following steps: according to the type of a sample to be measured, matching applicable crude polysaccharide determination standard methods through a preset standard database, and forming a candidate standard list; synchronously performing detection according to the standard methods in the candidate standard list respectively, and obtaining original detection data under each standard method; according to an absorbance sequence, calculating kinetic characteristic parameters under each standard method, and matching through a preset interference characteristic library to obtain a matrix interference type of the sample to be measured; through a preset mutual calibration function, mapping crude polysaccharide content determination values of each standard method to the same reference, calculating calibration deviations of each standard method; based on the calibration deviations and the matrix interference type, obtaining comprehensive quality scores under each standard method, and determining an optimal recommended method according to the comprehensive quality scores. Through mutual calibration and elimination of matrix interference, the optimal detection standard can be automatically recommended.

Description

Technical Field

[0001] This application relates to the field of pharmaceutical analysis and testing technology, and in particular to a method and system for determining crude polysaccharide content based on multi-standard comparison. Background Technology

[0002] In recent years, with the increasing awareness of health among people, there has been a strong market demand for health products rich in crude polysaccharides, such as Ganoderma lucidum spore powder and edible fungi extracts. Crude polysaccharides, as key active ingredients, have a variety of biological activities such as lowering blood sugar, anti-oxidation, anti-inflammation, and immune regulation. Their content is a core indicator for measuring product quality.

[0003] Currently, there are several applicable standard methods for the determination of crude polysaccharide content (such as SN / T4260-2015, NY / T1676-2023, and the Chinese Pharmacopoeia). However, due to the systematic differences in the operating parameters of different standards, the measured values ​​for the same sample often differ significantly, making it difficult to determine which result is closer to the true value. Furthermore, the interference matrices of non-polysaccharide components in the sample vary for different standard methods, and a single standard cannot distinguish between polysaccharide signals and interference signals. As a result, existing technologies struggle to quickly and effectively determine the most suitable standard method for the sample being tested. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for determining crude polysaccharide content based on multi-standard comparison. By comprehensively utilizing parallel determination data from multiple standards, this method can automatically identify matrix interference and objectively recommend the optimal standard method for determining crude polysaccharide content.

[0005] In a first aspect, this application provides a method for determining crude polysaccharide content based on multi-standard comparison, comprising: Obtain basic information about the sample to be tested, including the sample type; Based on the sample type, applicable standard methods for crude polysaccharide determination are matched through a pre-set standard database, and a list of candidate standards is generated. Based on the candidate standard list, the sample to be tested is homogenized and tested according to the standard methods in the candidate standard list to obtain the original test data under each standard method. The original test data includes the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence. Based on the absorbance sequence, the kinetic characteristic parameters under each standard method are calculated, and based on the kinetic characteristic parameters, the matrix interference type and confidence level of the current sample are obtained by matching with a preset interference feature library. By using the mutual calibration function stored in the preset standard database, the crude polysaccharide content determination values ​​of each standard method are mapped to the same benchmark, and the calibration deviation of each standard method is calculated. Based on the original detection data, calibration bias, matrix interference type and confidence level, quality assessment indicators under each standard method are obtained; Based on quality assessment indicators, a comprehensive quality score is obtained under each standard method through weighted calculation. The optimal recommended method for the current sample to be tested is determined based on the comprehensive quality score.

[0006] The above technical solution involves obtaining a list of candidate standards for automatic matching of sample types, acquiring detection data through synchronous parallel detection, identifying matrix interference based on kinetic characteristic parameters, using a mutual calibration function to map different standard results to the same benchmark to calculate calibration deviation, and obtaining a comprehensive quality score by integrating multiple quality assessment indicators such as accuracy and reliability. This transforms the standard selection that originally relied on experience into an objective data-driven decision-making process, achieving rapid and accurate determination of the most suitable standard method and improving the accuracy and reliability of crude polysaccharide content measurement results.

[0007] Optionally, after homogenizing the sample to be tested and performing tests according to the standard methods in the candidate standard list to obtain the raw test data under each standard method, the method further includes: Calculate the monotonicity index for the absorbance sequences under each standard method; Determine whether the monotonicity index is lower than a preset threshold; If so, add an abnormal status marker to the absorbance sequence under the standard method.

[0008] Optionally, the kinetic characteristic parameters include reaction rate, proportion of rapid reactions, plateau stability, and initial absorbance. The step of matching these kinetic characteristic parameters against a preset interference feature library to obtain the matrix interference type and confidence level of the current sample includes: The dynamic characteristic parameters are standardized to form dynamic characteristic vectors; Based on the candidate standard list, the standard feature vector set corresponding to each standard method is matched by a preset interference feature library. The standard feature vector set contains the standard feature vectors corresponding to each interference type. By matching the dynamic feature vectors under each standard method with the corresponding set of standard feature vectors, the highest similarity value and the corresponding standard feature vector are obtained. If the highest similarity value reaches the preset similarity threshold, the interference type corresponding to the standard feature vector will be used as the matrix interference type under each standard method. Based on the matrix interference types under each standard method, the matrix interference type and confidence level of the current test sample are obtained through a voting method.

[0009] Optionally, the step of obtaining the matrix interference type and confidence level of the current sample under test through a voting method based on the matrix interference types under each standard method includes: Based on the matrix interference types under each standard method, the number of occurrences of the matrix interference type is used as the voting standard. The number of votes for each matrix interference type is counted, and the matrix interference type with the highest number of votes is obtained. Based on the matrix interference type with the highest number of votes, obtain the highest and second highest similarity obtained by similarity matching under the corresponding standard method, and calculate the average matching similarity of the matrix interference type based on the highest similarity. If the highest number of votes is unique, the matrix interference type with the highest number of votes is taken as the matrix interference type of the current sample to be tested, and the confidence level is calculated based on the highest number of votes and the average matching similarity. If the highest number of votes is not unique, the relative discrimination is calculated based on the highest similarity and the second highest similarity, and the weighted weight is calculated based on the highest similarity and the relative discrimination. Based on the weighted weights, the weighted score of the matrix interference type with the highest number of votes is calculated. The matrix interference type with the highest weighted score is taken as the matrix interference type of the current sample to be tested. The confidence level is calculated based on the weighted score of the matrix interference type with the highest number of votes and the average matching similarity.

[0010] Optionally, the step of mapping the crude polysaccharide content determination values ​​of each standard method to the same benchmark using a mutual calibration function stored in a preset standard database, and calculating the calibration deviation of each standard method, includes: Using each standard method in the candidate standard list as a benchmark, the crude polysaccharide content determination values ​​of other standard methods are mapped to the current benchmark through a mutual calibration function; Calculate the calibration deviation of each standard method under the current reference; For each standard method, calculate its average calibration deviation under all rotating references, and use the average calibration deviation as the calibration deviation under that standard method.

[0011] Optionally, the quality assessment indicators include accuracy indicators and reliability indicators. The process of obtaining quality assessment indicators under each standard method based on raw detection data, calibration bias, matrix interference type, and confidence level includes: Based on the calibration deviation, an accuracy score is obtained through a preset first scoring function; Based on the absorbance sequence, the reaction completion rate is obtained, and the data quality score is calculated based on the reaction completion rate and the linearity of the standard curve. Based on the type of matrix interference, the interference influence coefficients corresponding to each standard method are matched using a preset interference coefficient table, and the interference influence score is calculated based on the interference influence coefficients. A reliability score is calculated based on data quality score, interference impact score, and confidence level.

[0012] Optionally, the calculation of the reliability score based on data quality score, interference impact score, and confidence level includes: Extract the maximum interference impact coefficient from the interference impact coefficients corresponding to each standard method; Calculate the reliability allocation weights based on the maximum interference impact coefficient and confidence level; Based on data quality score and interference impact score, a reliability score is calculated by assigning reliability weights.

[0013] Optionally, the step of obtaining a comprehensive quality score under each standard method based on quality assessment indicators through weighted calculation includes: Based on the sample type, the corresponding complexity coefficient is obtained through a preset sample-complexity mapping table; Calculate the standard deviation of the calibration deviation for each standard method, and based on the standard deviation, calculate the consistency coefficient using a preset piecewise linear function; Dynamically assigned weights are generated based on complexity and consistency coefficients. Based on quality assessment indicators, a weighted calculation is performed by dynamically allocating weights to obtain the comprehensive quality score under each standard method.

[0014] Optionally, the raw detection data also includes the total time consumed, and the optimal recommended method for determining the current sample to be tested based on the comprehensive quality score includes: The overall quality scores under each standard method are ranked, and the highest and second highest overall quality scores are obtained. Calculate the score difference between the highest overall quality score and the second highest overall quality score; If the score difference is greater than the preset difference threshold, the standard method corresponding to the highest comprehensive quality score will be used as the optimal recommended method. If the score difference is not greater than the preset difference threshold, the efficiency score is determined based on the total time spent, and the standard method corresponding to the highest efficiency score is taken as the optimal recommended method.

[0015] Secondly, this application provides a crude polysaccharide content determination system based on multi-standard comparison, comprising: Data acquisition module 101 is used to acquire basic information of the sample to be tested, including sample type; The standard matching module 102 is used to match applicable crude polysaccharide determination standard methods according to the sample type through a preset standard database and form a candidate standard list. The parallel detection module 103 is used to homogenize the sample to be tested based on the candidate standard list, and to perform detection according to the standard methods in the candidate standard list, and to obtain the original detection data under each standard method. The original detection data includes the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence. The interference identification module 104 is used to calculate the kinetic characteristic parameters under each standard method based on the absorbance sequence, and to obtain the matrix interference type and confidence level of the current sample under test by matching the kinetic characteristic parameters through a preset interference feature library. The mutual calibration module 105 is used to map the crude polysaccharide content determination values ​​of each standard method to the same reference through the mutual calibration function stored in the preset standard database, and to calculate the calibration deviation of each standard method. The comprehensive evaluation module 106 is used to obtain quality evaluation indicators for each standard method based on the original detection data, calibration deviation, matrix interference type and confidence level, and to obtain the comprehensive quality score for each standard method through weighted calculation based on the quality evaluation indicators, and to determine the optimal recommended method for the current sample to be tested based on the comprehensive quality score.

[0016] In summary, this application first utilizes the kinetic characteristic parameters of the colorimetric reaction to match a pre-set interference feature library, thereby identifying matrix interference types and providing an objective basis for the selection of standard methods. Furthermore, by establishing a mutual calibration function and employing a rotating benchmark method to map the results of each standard to the same reference system, calibration deviations are calculated, making the results of different standards comparable. In addition, by integrating accuracy and reliability scores and employing a dynamic weight allocation based on sample matrix complexity and method consistency to calculate a comprehensive quality score, an objective quantitative comparison of multiple standard methods is achieved, improving the credibility of the selected standard method. Attached Figure Description

[0017] Figure 1 This is a flowchart of a method for determining crude polysaccharide content based on multi-standard comparison provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the process of obtaining the matrix interference type and confidence level of the current sample to be tested, as provided in an embodiment of this application. Figure 3 This is a flowchart illustrating the calculation of calibration deviations for various standard methods provided in the embodiments of this application; Figure 4 This is a flowchart provided in the embodiments of this application for obtaining quality assessment indicators under various standard methods based on raw detection data, calibration deviation, matrix interference type and confidence level; Figure 5 This is a schematic diagram of a crude polysaccharide content determination system based on multi-standard comparison provided in an embodiment of this application. Detailed Implementation

[0018] The following is in conjunction with the appendix Figure 1 - Appendix Figure 5 This application will be described in further detail below.

[0019] This application provides a method for determining crude polysaccharide content based on multi-standard comparison, see [link to relevant documentation]. Figure 1 This includes the following steps: S100. Obtain basic information about the sample to be tested.

[0020] S200. Based on the sample type, match applicable standard methods for crude polysaccharide determination through a pre-set standard database and generate a candidate standard list.

[0021] S300: Based on the candidate standard list, homogenize the sample to be tested and perform detection according to the standard methods in the candidate standard list to obtain the original detection data under each standard method.

[0022] S400. Based on the absorbance sequence, calculate the kinetic characteristic parameters under each standard method, and based on the kinetic characteristic parameters, match them through a preset interference feature library to obtain the matrix interference type and confidence level of the current sample to be tested.

[0023] S500: Using the mutual calibration function stored in the preset standard database, the crude polysaccharide content determination values ​​of each standard method are mapped to the same reference, and the calibration deviation of each standard method is calculated.

[0024] S600: Based on the original test data, calibration deviation, matrix interference type, and confidence level, obtain quality assessment indicators under various standard methods.

[0025] S700: Based on quality assessment indicators, a weighted calculation is performed to obtain the comprehensive quality score under each standard method, and the optimal recommended method for the current sample to be tested is determined based on the comprehensive quality score.

[0026] In this embodiment, the basic information of the sample to be tested is first obtained. The basic information includes the sample type, which can be divided into edible fungi, plant-derived foods, traditional Chinese medicine, and health foods. Each sample to be tested may have more than one sample type. For example, Ganoderma lucidum spore powder belongs to both edible fungi and health foods.

[0027] Then, based on the sample type, a pre-set standard database is used to match applicable standard methods for crude polysaccharide determination, forming a candidate standard list. This pre-set standard database contains a mapping relationship between crude polysaccharide determination standards and their applicable sample types. For example, standard SN / T4260-2015 uses the phenol-sulfuric acid method and is applicable to plant-derived foods and health foods; standard NY / T1676-2023 uses a spectrophotometric method and is applicable to edible fungi; and the Chinese Pharmacopoeia uses a crude polysaccharide determination method applicable to traditional Chinese medicine materials and health foods. The final candidate standard list must contain at least two standard methods.

[0028] Once the candidate standard list is determined, the sample to be tested can be homogenized based on the candidate standard list, and then tested according to the standard methods in the candidate standard list to obtain the raw test data under each standard method.

[0029] Homogenization refers to the process of treating the sample to be tested to make it a uniform and consistent dispersion system. For example, if the candidate standard list contains three standard methods, the sample to be tested can be homogenized and divided into three parallel samples, which can then be tested separately according to the three standard methods. If conditions permit, the tests can be performed simultaneously to shorten the total testing time and ensure the consistency of sample conditions.

[0030] The detection here, in addition to measuring the sample according to standard methods, also includes establishing a standard curve. For example, a series of crude polysaccharide standard solutions of known concentrations are prepared, and colorimetric reactions are performed according to each standard method. The endpoint absorbance is read, and a linear regression is performed with concentration as the dependent variable and absorbance as the independent variable to obtain the regression coefficients and generate the regression equation, which is the standard curve. Of course, standard curves can also be pre-established and stored, and the standard curves for each standard method can be directly extracted based on a list of candidate standards.

[0031] By measuring the sample according to each standard method, the corresponding absorbance sequence can be obtained. The absorbance sequence is generated by collecting absorbance at multiple time points according to the colorimetric reaction of each standard method. The absorbance sequence can be expressed as: ,in, Indicates the first A standard method, For the reaction endpoint time specified by each standard method, the collection time points should include at least the early reaction (e.g., 5-10 minutes), the middle reaction (e.g., 20-40 minutes), the late reaction (e.g., 50-70 minutes), and the reaction endpoint (e.g., 60-120 minutes).

[0032] For example, the sample to be tested was Ganoderma lucidum spore powder, which was tested according to the standard method of SN / T4260-2015. The collected absorbance sequence data are shown in Table 1.

[0033] Table 1 absorbance 0.065 0.120 0.210 0.260 0.390 0.460 0.475 Based on the absorbance sequence, the absorbance values ​​at the endpoint of each standard method reaction are taken, i.e. The corresponding standard curve (regression equation) can be called to calculate the crude polysaccharide content, because only when the reaction is complete will the absorbance have a stable linear relationship with the concentration.

[0034] For example, taking Ganoderma lucidum spore powder as an example, the crude polysaccharide content determination values ​​under various standard methods are shown in Table 2.

[0035] Table 2 Ganoderma lucidum spore powder 1 3.11% 4.03% 3.64% Ganoderma lucidum spore powder 2 2.61% 3.58% 3.15% Ganoderma lucidum spore powder 3 2.45% 3.35% 2.97% In this embodiment, for ease of explanation, a single measurement result is used. In practical applications, multiple measurement results can be used and the average value can be taken.

[0036] Furthermore, the linearity of the standard curve can be calculated from the standard curve. The linearity of the standard curve is usually expressed by the correlation coefficient R. 2 This indicates that it reflects the fitting quality of the standard curve, that is, the degree of agreement between the measured data points and the fitted straight line, and can be used to evaluate the measurement data quality and reliability of each standard method.

[0037] Therefore, for each standard method, the original detection data obtained include the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence.

[0038] Considering that different standard methods have different sensitivities to the same type of interference, and that interference can cause characteristic deviations in the measured values ​​of each standard, it is necessary to first determine the type of matrix interference of the current sample to be tested, that is, the specific category of interference caused by non-crude polysaccharide components in the sample to the crude polysaccharide content measurement results. These interferences will react with the colorimetric reagent or affect the colorimetric process, causing the measured values ​​to deviate from the true values.

[0039] Once the type of matrix interference is determined, it helps to select the standard method that is least affected by the interference, or in other words, the standard method that provides the most reliable measurement data.

[0040] In this embodiment, the kinetic characteristic parameters under each standard method are calculated based on the absorbance sequence, and the matrix interference type and confidence level of the current test sample are obtained by matching the kinetic characteristic parameters through a preset interference feature library.

[0041] Because different interfering substances have different reaction kinetic characteristics, the so-called reaction kinetic characteristics refer to the law of absorbance change with reaction time. Therefore, the kinetic characteristic parameters under each standard method can be calculated based on the absorbance sequence. The kinetic characteristic parameters include the proportion of rapid reaction, reaction rate, plateau stability and initial absorbance.

[0042] The percentage of rapid reactions refers to the proportion of color development completed in the early stages of the reaction relative to the total color development. It mainly reflects the relative content of rapidly reacting components such as small-molecule reducing sugars, and can be expressed as... The reaction rate refers to the average rate of color development during the linear growth phase, reflecting the speed of the colorimetric reaction, and can be expressed as... ,in , These are two time points in the linear phase, typically taken as 20 minutes and 45 minutes; the plateau stability period refers to the absorbance stability in the later stages of the reaction, which can be expressed as... ,in For example, the point in time preceding the endpoint, as shown in the example above. =90min, The initial absorbance is 60 min; it refers to the absorbance value at the first time point after the reaction starts, reflecting the color contribution of the sample itself and the rapidly developing substances in the early stage of the reaction, and can be expressed as... .

[0043] However, considering that in the colorimetric reaction of crude polysaccharides, the absorbance should increase monotonically with the increase of reaction time until it reaches the plateau period, that is, the absorbance is almost unchanged. If the absorbance decreases, it indicates an abnormality, such as product degradation, reagent failure or operation error. Therefore, after obtaining the absorbance sequence, the absorbance sequence will be verified.

[0044] Specifically, after homogenizing the sample to be tested and performing tests according to the standard methods in the candidate standard list, and obtaining the raw test data under each standard method, the following steps are also included: S310. Calculate the monotonicity index for the absorbance sequences under each standard method.

[0045] S320. Determine whether the monotonicity index is lower than the preset threshold.

[0046] S330. If so, add an abnormal status marker to the absorbance sequence under the standard method.

[0047] First, for the absorbance sequences under each standard method, the monotonicity index M is calculated. This monotonicity index is a quantitative measure of whether the absorbance sequence monotonically increases with reaction time, and can be expressed as: in, This represents the number of time points in the absorbance sequence. For the first absorbance values ​​at each time point For the first +1 absorbance value at time point This is an indicator function that, when the condition is true, i.e. The value is 1 if it is true, and 0 otherwise.

[0048] Then, it is determined whether the monotonicity index is lower than the preset threshold. The preset threshold is usually set to a range of 0.7 to 0.9. If the monotonicity index of a certain standard method is lower than the preset threshold, an abnormal state mark is added to the absorbance sequence of that standard method, that is, the measurement result of the standard method is marked as unreliable and will not be included in the subsequent calibration deviation calculation and comprehensive score.

[0049] If the monotonicity index under each standard method is higher than the preset threshold, the kinetic characteristic parameters can be calculated based on the absorbance sequence. Based on the kinetic characteristic parameters, the matrix interference type and confidence level of the current sample can be obtained by matching with the preset interference feature library.

[0050] Specifically, see Figure 2 Based on kinetic characteristic parameters, the matrix interference type and confidence level of the current test sample are obtained by matching through a preset interference feature library, including the following steps: S410. Standardize the dynamic characteristic parameters and form a dynamic characteristic vector.

[0051] S420. Based on the candidate standard list, match the standard feature vector set corresponding to each standard method through a preset interference feature library.

[0052] S430. By matching the dynamic feature vectors under each standard method with the corresponding set of standard feature vectors, the highest similarity value and the corresponding standard feature vector are obtained.

[0053] S440. If the highest similarity value reaches the preset similarity threshold, the interference type corresponding to the standard feature vector will be used as the matrix interference type under each standard method.

[0054] S450. Based on the matrix interference types under each standard method, the matrix interference type and confidence level of the current test sample are obtained through voting.

[0055] First, the kinetic characteristic parameters under each standard method are standardized to eliminate the dimensional differences between different standard methods. This can be achieved through preset benchmark parameters, which include the mean and standard deviation of the kinetic characteristic parameters under undisturbed standard samples. After standardization, kinetic characteristic vectors under each standard method can be formed.

[0056] Then, based on the candidate standard list, the standard feature vector set corresponding to each standard method is matched by a preset interference feature library. The standard feature vector set contains standard feature vectors corresponding to each interference type, including no interference, oil interference, protein interference, reducing sugar interference, pigment interference, etc.

[0057] Next, the dynamic feature vectors under each standard method are matched with the corresponding standard feature vector sets through similarity matching to obtain the highest similarity value and the corresponding standard feature vector. It is then determined whether the highest similarity value reaches the preset similarity threshold, which is set to, for example, 0.85.

[0058] If the highest similarity value reaches the preset similarity threshold, the interference type corresponding to the standard feature vector is taken as the matrix interference type under each standard method; if the highest similarity value does not reach the preset similarity threshold, the matrix interference type under the corresponding standard method is considered to be "unidentified" or "no obvious interference".

[0059] For example, according to the above example, the standard method of SN / T matches the matrix interference type as "oil interference" with a maximum similarity value of 0.89; the standard method of NY / T matches the matrix interference type as "no interference" with a maximum similarity of 0.92; and the standard method of the Chinese Pharmacopoeia has a maximum similarity of 0.79, with a matching result of "unidentified".

[0060] After determining the matrix interference type under each standard method, it is necessary to further determine the matrix interference type of the current test sample. The matrix interference type and confidence level of the current test sample will be obtained through a voting method.

[0061] Specifically, based on the matrix interference types under each standard method, the matrix interference type and confidence level of the current sample to be tested are obtained through a voting method, including the following steps: S451. Based on the matrix interference types under each standard method, the number of occurrences of the matrix interference type is used as the voting standard to count the number of votes for each matrix interference type and obtain the matrix interference type with the highest number of votes.

[0062] S452. Based on the matrix interference type with the highest number of votes, obtain the highest similarity and the second highest similarity obtained by similarity matching under the corresponding standard method, and calculate the average matching similarity of the matrix interference type based on the highest similarity.

[0063] S453. If the highest number of votes is unique, the matrix interference type with the highest number of votes shall be taken as the matrix interference type of the current sample to be tested, and the confidence level shall be calculated based on the highest number of votes and the average matching similarity.

[0064] S454. If the highest number of votes is not unique, the relative discrimination is calculated based on the highest similarity and the second highest similarity, and the weighted weight is calculated based on the highest similarity and the relative discrimination.

[0065] S455. Calculate the weighted score of the matrix interference type with the highest number of votes based on the weighted weights. Take the matrix interference type with the highest weighted score as the matrix interference type of the current sample to be tested. Calculate the confidence level based on the weighted score of the matrix interference type with the highest number of votes and the average matching similarity.

[0066] First, using the frequency of occurrence of matrix interference types as the voting criterion, the number of votes for each matrix interference type can be counted, and the matrix interference type with the most votes can be obtained. For example, following the example above, oil interference, SN / T (1 vote); no interference, NY / T (1 vote).

[0067] Then, based on the matrix interference type with the highest number of votes, the highest similarity and the second highest similarity obtained by similarity matching under the corresponding standard method can be obtained, and the average matching similarity of the matrix interference type can be calculated based on the highest similarity. The average matching similarity represents the average matching quality of all standards that support the matrix interference type. For example, although there are two standards that point to grease interference, if the matching quality of one of them is not high (the highest similarity is low), the overall credibility should also be affected.

[0068] In the example above, the matrix interference types with the most votes are grease interference and no interference, each with one vote. For grease interference, i.e., the standard method of SN / T, the highest similarity obtained by similarity matching is 0.89, and the second highest similarity is 0.35. Since there is only one vote, the average matching similarity is also 0.89, calculated according to the highest similarity.

[0069] Similarly, without interference, i.e., the standard method of NY / T, the highest similarity obtained by similarity matching is 0.92, the second highest similarity is 0.32, and the average matching similarity is also 0.92.

[0070] Then, it is determined whether the highest number of votes has shifted. If the highest number of votes is unique, the matrix interference type with the highest number of votes is taken as the matrix interference type of the current sample to be tested, and the result is determined according to the highest number of votes. and average matching similarity Calculate the confidence level, i.e., the confidence score. It can be represented as: Where N is the total number of standard methods of valid voting; in the example above, N=2.

[0071] If the highest number of votes is not unique, then the highest similarity will be used. Second highest similarity The relative discrimination score is calculated, and a weighted weight is calculated based on the highest similarity score and the relative discrimination score. The relative discrimination score is an indicator used to measure the advantage of the optimal match over the second-best match, reflecting the clarity of the interference identification result. The higher the relative discrimination score, the more obvious the advantage of the optimal match over the second-best match, and the more reliable the identification result. , can be represented as: Weighted weight It can be represented as, Based on the weighted average, the weighted score for the matrix interference type with the highest number of votes can be calculated. The matrix interference type for each tie can be represented as: That is, the sum of the weighted weights of each standard method that identifies the interference type.

[0072] Finally, the matrix interference type with the highest weighted score is taken as the matrix interference type of the current sample to be tested, and the confidence level is calculated based on the weighted score of the matrix interference type with the highest number of votes and the average matching similarity. It can be represented as: in, This represents the highest weighted score among the tie-breaker interference types. It is the second highest weighted score among the interference types of tied votes.

[0073] Following the example above, since the highest number of votes is not unique, the weighted score for interference from grease can be calculated as follows: 0.89 × (0.89 - 0.35) / (1 - 0.35) = 0.739; and the weighted score for no interference is 0.92 × (0.92 - 0.32) / (1 - 0.32) = 0.811.

[0074] Since 0.811 > 0.739, the matrix interference type of the current sample to be tested is "no interference". The average matching similarity of "no interference" is 0.92, and the confidence level is (0.811 / (0.811+0.739))×0.92×100%=0.523×0.92×100%=48.1%.

[0075] Since the results of the various standard methods are not comparable, it is difficult to determine which standard method's results are closer to the true value. Therefore, it is necessary to convert the crude polysaccharide content determination results of the various standard methods to the same reference system to make them comparable.

[0076] Therefore, in this embodiment of the application, the crude polysaccharide content determination values ​​of each standard method are mapped to the same benchmark by using the mutual calibration function stored in the preset standard database, and the calibration deviation of each standard method is calculated.

[0077] Specifically, see Figure 3 By using the mutual calibration function stored in the preset standard database, the crude polysaccharide content determination values ​​of each standard method are mapped to the same benchmark, and the calibration deviation of each standard method is calculated, including the following steps: S510. Using each standard method in the candidate standard list as a benchmark, map the crude polysaccharide content determination values ​​of other standard methods to the current benchmark through a mutual calibration function.

[0078] S520. Calculate the calibration deviation of each standard method under the current reference.

[0079] S530. For each standard method, calculate its average calibration deviation under all rotating references, and use the average calibration deviation as the calibration deviation under that standard method.

[0080] A mutual calibration function is a function that describes the mathematical relationship between the results of two different standard methods. Since there are systematic differences in the results of different standard methods for the same crude polysaccharide sample, the mutual calibration function can convert the measured value of one standard method into an equivalent value under the reference of another standard method.

[0081] The mutual calibration function can be established by preparing a series of crude polysaccharide standard solutions of different concentrations. Since the true value of the standard solution is known, a mapping relationship between the two measurement results can be established by measuring the same set of standard solutions using different standard methods and stored in a preset standard database.

[0082] First, using each standard method in the candidate standard list as a benchmark, the crude polysaccharide content determination values ​​of other standard methods are mapped to the current benchmark through a mutual calibration function.

[0083] For example, using NY / T as the reference, the calibration function for SN / T->NY / T is: The calibration function for Pharmacopoeia -> NY / T is: .

[0084] Then, calculate the calibration deviation of each standard method under the same current reference, denoted as the current reference. The calibration deviation of each standard method can then be expressed as: .in, For the first The standard method is the current benchmark. Calibration deviation at time, For the first Standard methods mapped to benchmarks The converted value below, Based on The crude polysaccharide content was determined.

[0085] Following the example above, , ; Based on NY / T , Standard method When it is SN / T, Calibration deviation Standard method When it was included in the pharmacopoeia, Calibration deviation .

[0086] Similarly, using SN / T and the pharmacopoeia as benchmarks, the calibration deviations in each standard direction can be calculated.

[0087] Finally, for each standard method, the average calibration deviation under all rotating references is calculated, and the average calibration deviation is taken as the calibration deviation under that standard method. For example, for the SN / T standard method, the calibration deviations based on NY / T, SN / T and Pharmacopoeia are 22.1%, 0%, and 14.0%, respectively. The average calibration deviation is 12.0%, that is, the calibration deviation of the SN / T standard method is 12.0%.

[0088] After determining the calibration bias and matrix interference type, the quality assessment index can be calculated based on the previous raw test data, combined with the calibration bias and matrix interference type. The quality assessment index is further divided into accuracy index and reliability index, that is, the accuracy and reliability of each standard method for the current test sample are evaluated by the quality assessment index.

[0089] Specifically, see Figure 4 Based on the original detection data, calibration bias, matrix interference type, and confidence level, quality assessment indicators for each standard method are obtained, including the following steps: S610. Based on the calibration deviation, obtain the accuracy score through a preset first scoring function.

[0090] S620. Based on the absorbance sequence, obtain the reaction completion rate, and calculate the data quality score according to the reaction completion rate and the linearity of the standard curve.

[0091] S630. Based on the matrix interference type, match the interference influence coefficients corresponding to each standard method through a preset interference coefficient table, and calculate the interference influence score based on the interference influence coefficients.

[0092] S640. Calculate the reliability score based on the data quality score, interference impact score, and confidence level.

[0093] First, based on the calibration deviation, an accuracy score can be obtained using a preset first scoring function. This preset first scoring function maps the calibration deviation to an accuracy score. For example, a piecewise linear function can be used: when the calibration deviation is ≤5%, the accuracy score is 100 points; when the calibration deviation is between 5% and 15%, 2 points are deducted for every 1% increase in deviation; when the calibration deviation is between 15% and 35%, 1 point is deducted for every 1% increase in deviation; and when the calibration deviation is >35%, the accuracy score is 0 points.

[0094] Then, based on the absorbance sequence, the reaction completion rate is obtained. A data quality score is calculated based on the reaction completion rate and the linearity of the standard curve. The reaction completion rate (P) reflects whether the sample reacted completely under this standard method; the linearity of the standard curve (R)... 2 It is used to reflect the fitting quality of the standard curve and indirectly reflects the current state of the instrument and reagents.

[0095] The response completion rate and the linearity of the standard curve are then converted into response completion rate scores using preset scoring rules. And curve linearity score For example, using a piecewise function as described above, when P ≥ 98%, =100; 95%≤P<98%, =85 points; 90%≤P<95%, =60; P<90%, the reaction is incomplete, and the corresponding standard method is not included in the quality assessment.

[0096] When R 2 ≥0.995, =100; 0.990≤R 2 <0.995, =85 points; 0.950≤R 2 <0.995, =60; R 2 If the value is less than 0.950, the standard curve is unqualified, and the corresponding standard method is not included in the quality assessment.

[0097] Data quality score It can be represented as: Here, the two indicators are assumed to have equal weights, but the weights can be adjusted according to the actual situation.

[0098] Next, based on the type of matrix interference, the interference influence coefficients corresponding to each standard method are matched using a preset interference coefficient table, and the interference influence score is calculated based on the interference influence coefficients.

[0099] The preset interference coefficient table here contains the interference influence coefficients of each interference type under each standard method. It can be obtained by preparing simulated samples containing known interfering substances (such as oils, proteins, reducing sugars, pigments, etc.), measuring them using each standard method, calculating the deviation of each standard method from the true value, and normalizing the result. The interference influence coefficient represents the degree of influence of the interference type on a specific standard method, and its value ranges from 0 to 0.5. The larger the value, the more severe the influence. When the identification result of the matrix interference type is "no interference", the interference influence coefficients of all standard methods are 0.

[0100] Determine the interference effect coefficient Then, the interference impact score can be calculated. It can be represented as: .

[0101] Finally, a reliability score can be calculated based on the data quality score, interference impact score, and confidence level.

[0102] Specifically, based on data quality score, interference impact score, and confidence level, a reliability score is calculated, including the following steps: S641. Extract the maximum interference influence coefficient from the interference influence coefficients corresponding to each standard method.

[0103] S642. Calculate the reliability allocation weight based on the maximum interference impact coefficient and confidence level.

[0104] S643. Based on the data quality score and the interference impact score, calculate the reliability score by assigning reliability weights.

[0105] The reliability score reflects the reliability of the measurement results of the sample under test by the standard method. It combines the data quality score and the interference effect score, and reduces the uncertainty of interference identification through confidence level.

[0106] Considering that the data quality score and the interference impact score have different degrees of influence on the reliability score, it is necessary to first determine the weight allocation. Taking the interference impact as the benchmark, the more severe the interference and the higher the confidence level of interference identification, the greater the weight of the interference impact score should be. The severity of the interference can be characterized by the maximum interference impact coefficient.

[0107] Therefore, in this embodiment, the maximum interference influence coefficient is first extracted from the interference influence coefficients corresponding to each standard method. The maximum interference impact coefficient here refers to the largest interference impact coefficient among all standard methods under the currently identified interference type, and is used to measure the severity of the interference.

[0108] Then, based on the maximum interference impact coefficient and confidence level, the reliability allocation weight is calculated, which is divided into the weight of the interference impact score. Weighting of data quality scores , It can be represented as: ,in, The baseline weights range from 0.4 to 0.6, and α is an adjustment coefficient used to control the degree to which the severity of the interference affects the weights, with a value ranging from 0.2 to 0.5. The maximum confidence level is set to 100%. .

[0109] Then, based on data quality scoring And interference affects the score By assigning weights to reliability metrics, a reliability score can be calculated, denoted as [reliability score]. ,but It can be represented as: .

[0110] Once the quality assessment indicators are determined, a comprehensive quality score under each standard method can be obtained through weighted calculation.

[0111] Specifically, based on quality assessment indicators, a weighted calculation is performed to obtain the comprehensive quality score under each standard method, including the following steps: S710. Based on the sample type, obtain the corresponding complexity coefficient through a preset sample-complexity mapping table.

[0112] S720. Calculate the standard deviation of the calibration deviation of each standard method, and calculate the consistency coefficient based on the standard deviation using a preset piecewise linear function.

[0113] S730. Generate dynamically allocated weights based on the complexity coefficient and consistency coefficient.

[0114] S740. Based on quality assessment indicators, a weighted calculation is performed by dynamically allocating weights to obtain the comprehensive quality score under each standard method.

[0115] The quality assessment indicators include accuracy score and reliability score. Considering that different sample types have different matrix complexities and different impacts on the test results, the more complex the matrix, the more interfering substances may exist in the sample, and the greater the uncertainty of the test results. In this case, the reliability score has a high degree of discrimination and can effectively guide the selection of standard methods. Therefore, the reliability score should be given a higher weight when calculating the comprehensive score.

[0116] Therefore, firstly, the corresponding complexity coefficient can be obtained by using a pre-defined sample-complexity mapping table, which stores the complexity levels associated with different sample types and their corresponding complexity coefficients. Used to reflect the complexity of the sample matrix, the value range is set to 0.9~1.2. For example, edible fungi extracts and single-product health products have a complexity level of "medium" and a complexity coefficient of 1.0; Ganoderma lucidum spore powder and plant-derived crude extracts have a complexity level of "high" and a corresponding complexity coefficient of 1.1.

[0117] Furthermore, since the accuracy score is calculated based on the calibration deviation, if the calibration deviations of each standard method are highly consistent, it indicates that the mutual calibration function is effective and the accuracy score is reliable. If the calibration deviations are seriously divergent, it indicates that there is unknown interference or the mutual calibration function is ineffective, and the accuracy score is unreliable. Therefore, the reliability of the accuracy score can be judged by quantifying the dispersion of the standard deviations of each standard method.

[0118] Therefore, the standard deviation of the calibration deviation for each standard method can be calculated. Based on the standard deviation, a consistency coefficient can be calculated using a preset piecewise linear function. The consistency coefficient quantifies the dispersion of the standard deviation for each standard method. The calibration deviation for each standard method can be expressed as: , ,…, , where n is the number of standard methods in the candidate standard list, and the standard deviation of the calibration bias is... , can be represented as: in, The average and standard deviation of the calibration deviations for each standard method. The standard deviation reflects the dispersion of calibration deviations of each standard method. The smaller the standard deviation, the closer the calibration deviations of each standard method are, and the better the consistency of the measurement results of each standard method. The larger the standard deviation, the greater the discrepancy between the measurement results of each standard method.

[0119] Based on the standard deviation, the consistency coefficient can be calculated using a pre-defined piecewise linear function. For example, when <3%, =0.9; 3%≤ ≤10%, =1.0; 10% < <20%, =1.1; >20%, =1.2.

[0120] Next, based on the complexity coefficient and consistency coefficient, dynamically assigned weights can be generated, which are then used to assign weights to the accuracy score. Weighting of reliability score , It can be represented as: ,in This is the dynamic baseline weight, usually set to 0.5. If the value is greater than 1, the matrix complexity is high and it relies more on reliability. A value greater than 1 indicates lower consistency and a greater reliance on reliability.

[0121] Additionally, boundary constraints can be set, that is, upper and lower limits for the weights. For example, ... The constraint is in the range of 0.3 to 0.7, that is... In this way, even if the matrix is ​​the simplest and the consistency of the measurement results is very high, the reliability score still retains a 30% weight; similarly, even if the matrix is ​​the most complex and the consistency of the measurement results is very low, the accuracy score still retains a 30% weight.

[0122] Finally, based on the quality assessment metrics, that is, the accuracy score. and reliability score By dynamically assigning weights and performing weighted calculations, the overall quality score under each standard method can be obtained. This overall quality score is denoted as... ,but It can be represented as: Once the overall quality score is determined, the optimal recommended method for the current sample to be tested can be determined based on the overall quality score.

[0123] Specifically, determining the optimal recommended method for the current sample to be tested based on the comprehensive quality score includes the following steps: S810. Sort the overall quality scores under each standard method and obtain the highest overall quality score and the second highest overall quality score.

[0124] S820: Calculate the score difference between the highest overall quality score and the second highest overall quality score.

[0125] S830. If the score difference is greater than the preset difference threshold, the standard method corresponding to the highest comprehensive quality score shall be used as the optimal recommended method.

[0126] S840. If the score difference is not greater than the preset difference threshold, the efficiency score is determined based on the total time spent, and the standard method corresponding to the highest efficiency score is taken as the optimal recommended method.

[0127] Normally, the standard method with the highest overall quality score is recommended as the optimal method for the current sample to be tested. However, if the overall quality scores of two standard methods are very similar, an efficiency score can be introduced to assist in the decision-making process.

[0128] In this embodiment of the application, the comprehensive quality scores under each standard method are first sorted to obtain the highest comprehensive quality score and the second highest comprehensive quality score.

[0129] Then, the score difference between the highest overall quality score and the second highest overall quality score is calculated. If the score difference is greater than a preset difference threshold, the standard method corresponding to the highest overall quality score is taken as the optimal recommended method.

[0130] If the score difference is not greater than the preset difference threshold, it means that it is difficult to determine the optimal recommendation method by relying solely on the comprehensive quality score. In this case, the efficiency score will be determined based on the total time consumption, and the standard method corresponding to the highest efficiency score will be taken as the optimal recommendation method.

[0131] The total time includes the entire process from sample pretreatment, colorimetric reaction, detection, and result calculation. The efficiency score can be obtained through a preset scoring function, such as an exponential decay function or a piecewise function. Where T is the total time consumed. This represents the minimum total time taken under each standard method. A higher efficiency score indicates shorter time and higher detection efficiency.

[0132] Of course, if the efficiency scores are also similar, the measurement cost can be further considered for a comprehensive determination. After determining the optimal recommended method, the optimal recommended method and the associated detection data of the sample to be tested can be used together to generate an output report.

[0133] This application also provides a crude polysaccharide content determination system based on multi-standard comparison, see [link to relevant documentation]. Figure 5 The system includes: a data acquisition module 101, a standard matching module 102, a parallel detection module 103, an interference identification module 104, a mutual calibration module 105, and a comprehensive evaluation module 106.

[0134] The data acquisition module 101 is used to acquire basic information about the sample to be tested.

[0135] The standard matching module 102 is used to match applicable standard methods for crude polysaccharide determination according to the sample type through a preset standard database, and form a candidate standard list.

[0136] The parallel detection module 103 is used to homogenize the sample to be tested based on the candidate standard list, and to perform detection according to the standard methods in the candidate standard list to obtain the raw detection data under each standard method.

[0137] The interference identification module 104 is used to calculate the kinetic characteristic parameters under each standard method based on the absorbance sequence, and to obtain the matrix interference type and confidence level of the current sample under test by matching the kinetic characteristic parameters through a preset interference feature library.

[0138] The mutual calibration module 105 is used to map the crude polysaccharide content determination values ​​of each standard method to the same reference through the mutual calibration function stored in the preset standard database, and to calculate the calibration deviation of each standard method.

[0139] The comprehensive evaluation module 106 is used to obtain quality evaluation indicators for each standard method based on the original detection data, calibration deviation, matrix interference type and confidence level, and to obtain the comprehensive quality score for each standard method through weighted calculation based on the quality evaluation indicators, and to determine the optimal recommended method for the current sample to be tested based on the comprehensive quality score.

[0140] In this embodiment of the application, the data acquisition module 101 is specifically used to acquire basic information of the sample to be tested, wherein the basic information includes the sample type.

[0141] The standard matching module 102 is specifically used to match applicable crude polysaccharide determination standard methods according to the sample type obtained by the data acquisition module 101 through a preset standard database, and form a candidate standard list.

[0142] The parallel detection module 103 is specifically used to homogenize the sample to be tested based on the candidate standard list generated by the standard matching module 102, and to perform detection according to the standard methods in the candidate standard list, thereby obtaining the original detection data under each standard method. The original detection data includes the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence.

[0143] The interference identification module 104 is specifically used to calculate the kinetic characteristic parameters under each standard method based on the absorbance sequence generated by the parallel detection module 103, and to obtain the matrix interference type and confidence level of the current sample under test by matching the kinetic characteristic parameters through a preset interference feature library.

[0144] The mutual calibration module 105 is specifically used to map the crude polysaccharide content determination values ​​of each standard method to the same reference through the mutual calibration functions stored in the preset standard database, and to calculate the calibration deviation of each standard method.

[0145] The comprehensive evaluation module 106 is specifically used to obtain quality evaluation indicators for each standard method based on the original detection data, calibration deviation, matrix interference type and confidence level, and to obtain the comprehensive quality score for each standard method through weighted calculation based on the quality evaluation indicators, and to determine the optimal recommended method for the current sample to be tested based on the comprehensive quality score.

[0146] The embodiments described in this application are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the principles of this application should be included within the scope of protection of this application.

Claims

1. A method for determining crude polysaccharide content based on multi-standard comparison, characterized in that, include: Obtain basic information about the sample to be tested, including the sample type; Based on the sample type, applicable standard methods for crude polysaccharide determination are matched through a pre-set standard database, and a list of candidate standards is generated. Based on the candidate standard list, the sample to be tested is homogenized and tested according to the standard methods in the candidate standard list to obtain the original test data under each standard method. The original test data includes the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence. Based on the absorbance sequence, the kinetic characteristic parameters under each standard method are calculated, and based on the kinetic characteristic parameters, the matrix interference type and confidence level of the current sample are obtained by matching with a preset interference feature library. The kinetic characteristic parameters include reaction rate, rapid reaction percentage, plateau stability, and initial absorbance. The rapid reaction percentage refers to the proportion of color development completed in the early stages of the reaction relative to the total color development. Plateau stability refers to the absorbance stability in the later stages of the reaction, which can be expressed as the ratio of the final absorbance to the absorbance at the time point before the final endpoint. Based on these kinetic characteristic parameters, a preset interference feature library is used for matching to obtain the matrix interference type and confidence level of the current sample to be tested, including: The dynamic characteristic parameters are standardized to form dynamic characteristic vectors; Based on the candidate standard list, the standard feature vector set corresponding to each standard method is matched by a preset interference feature library. The standard feature vector set contains the standard feature vectors corresponding to each interference type. By matching the dynamic feature vectors under each standard method with the corresponding set of standard feature vectors, the highest similarity value and the corresponding standard feature vector are obtained. If the highest similarity value reaches the preset similarity threshold, the interference type corresponding to the standard feature vector will be used as the matrix interference type under each standard method. Based on the matrix interference types under each standard method, the matrix interference type and confidence level of the current test sample are obtained through voting. By using the mutual calibration function stored in the preset standard database, the crude polysaccharide content determination values ​​of each standard method are mapped to the same benchmark, and the calibration deviation of each standard method is calculated. The mutual calibration function is a function that describes the mathematical relationship between the determination results of two different standard methods. Based on the original detection data, calibration deviation, matrix interference type and confidence level, quality assessment indicators under each standard method are obtained. The quality assessment indicators include accuracy indicators and reliability indicators. Based on quality assessment indicators, a comprehensive quality score is obtained under each standard method through weighted calculation. The optimal recommended method for the current sample to be tested is determined based on the comprehensive quality score.

2. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, After homogenizing the sample to be tested and performing tests according to the standard methods in the candidate standard list to obtain the raw test data under each standard method, the process further includes: Calculate the monotonicity index for the absorbance sequences under each standard method; Determine whether the monotonicity index is lower than a preset threshold; If so, add an abnormal status marker to the absorbance sequence under the standard method.

3. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, The matrix interference type and confidence level of the current sample under test are obtained through a voting method based on the matrix interference type under each standard method, including: Based on the matrix interference types under each standard method, the number of occurrences of the matrix interference type is used as the voting standard. The number of votes for each matrix interference type is counted, and the matrix interference type with the highest number of votes is obtained. Based on the matrix interference type with the highest number of votes, obtain the highest and second highest similarity obtained by similarity matching under the corresponding standard method, and calculate the average matching similarity of the matrix interference type based on the highest similarity. If the highest number of votes is unique, the matrix interference type with the highest number of votes is taken as the matrix interference type of the current sample to be tested, and the confidence level is calculated based on the highest number of votes and the average matching similarity. If the highest number of votes is not unique, the relative discrimination is calculated based on the highest similarity and the second highest similarity, and the weighted weight is calculated based on the highest similarity and the relative discrimination. Based on the weighted weights, the weighted score of the matrix interference type with the highest number of votes is calculated. The matrix interference type with the highest weighted score is taken as the matrix interference type of the current sample to be tested. The confidence level is calculated based on the weighted score of the matrix interference type with the highest number of votes and the average matching similarity.

4. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, The process involves mapping the crude polysaccharide content determination values ​​of various standard methods to the same benchmark using a mutual calibration function stored in a preset standard database, and calculating the calibration deviation of each standard method, including: Using each standard method in the candidate standard list as a benchmark, the crude polysaccharide content determination values ​​of other standard methods are mapped to the current benchmark through a mutual calibration function; Calculate the calibration deviation of each standard method under the current reference; For each standard method, calculate its average calibration deviation under all rotating references, and use the average calibration deviation as the calibration deviation under that standard method.

5. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, The quality assessment indicators include accuracy indicators and reliability indicators. The acquisition of quality assessment indicators under various standard methods based on raw detection data, calibration deviation, matrix interference type, and confidence level includes: Based on the calibration deviation, an accuracy score is obtained through a preset first scoring function; Based on the absorbance sequence, the reaction completion rate is obtained, and the data quality score is calculated based on the reaction completion rate and the linearity of the standard curve. Based on the type of matrix interference, the interference influence coefficients corresponding to each standard method are matched using a preset interference coefficient table, and the interference influence score is calculated based on the interference influence coefficients. A reliability score is calculated based on data quality score, interference impact score, and confidence level.

6. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 5, characterized in that, The reliability score is calculated based on data quality score, interference impact score, and confidence level, including: Extract the maximum interference impact coefficient from the interference impact coefficients corresponding to each standard method; Calculate the reliability allocation weights based on the maximum interference impact coefficient and confidence level; Based on data quality score and interference impact score, a reliability score is calculated by assigning reliability weights.

7. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, The method of obtaining a comprehensive quality score under various standard methods based on quality assessment indicators and through weighted calculation includes: Based on the sample type, the corresponding complexity coefficient is obtained through a preset sample-complexity mapping table; Calculate the standard deviation of the calibration deviation for each standard method, and based on the standard deviation, calculate the consistency coefficient using a preset piecewise linear function; Dynamically assigned weights are generated based on complexity and consistency coefficients. Based on quality assessment indicators, a weighted calculation is performed by dynamically allocating weights to obtain the comprehensive quality score under each standard method.

8. The method for determining crude polysaccharide content based on multi-standard comparison according to claim 1, characterized in that, The raw detection data also includes the total time consumed, and the optimal recommended method for determining the current sample to be tested based on the comprehensive quality score includes: The overall quality scores under each standard method are ranked, and the highest and second highest overall quality scores are obtained. Calculate the score difference between the highest overall quality score and the second highest overall quality score; If the score difference is greater than the preset difference threshold, the standard method corresponding to the highest comprehensive quality score will be used as the optimal recommended method. If the score difference is not greater than the preset difference threshold, the efficiency score is determined based on the total time spent, and the standard method corresponding to the highest efficiency score is taken as the optimal recommended method.

9. A crude polysaccharide content determination system based on multi-standard comparison, characterized in that, include: The data acquisition module (101) is used to acquire basic information of the sample to be tested, including the sample type; The standard matching module (102) is used to match applicable crude polysaccharide determination standard methods according to the sample type through a preset standard database and form a candidate standard list. The parallel detection module (103) is used to homogenize the sample to be tested based on the candidate standard list, and to perform detection according to the standard methods in the candidate standard list, and to obtain the original detection data under each standard method. The original detection data includes the crude polysaccharide content determination value, the linearity of the standard curve, and the absorbance sequence. The interference identification module (104) is used to calculate the kinetic characteristic parameters under each standard method based on the absorbance sequence, and to obtain the matrix interference type and confidence level of the current test sample by matching the kinetic characteristic parameters through a preset interference feature library. The kinetic characteristic parameters include reaction rate, rapid reaction ratio, plateau stability, and initial absorbance. The rapid reaction ratio refers to the proportion of color development completed in the early stage of the reaction to the total color development. The plateau stability refers to the absorbance stability in the later stage of the reaction, which can be expressed as the ratio of the endpoint absorbance to the absorbance at the time point before the endpoint. The process of obtaining the matrix interference type and confidence level of the current test sample by matching the kinetic characteristic parameters through a preset interference feature library includes: The dynamic characteristic parameters are standardized to form dynamic characteristic vectors; Based on the candidate standard list, the standard feature vector set corresponding to each standard method is matched by a preset interference feature library. The standard feature vector set contains the standard feature vectors corresponding to each interference type. By matching the dynamic feature vectors under each standard method with the corresponding set of standard feature vectors, the highest similarity value and the corresponding standard feature vector are obtained. If the highest similarity value reaches the preset similarity threshold, the interference type corresponding to the standard feature vector will be used as the matrix interference type under each standard method. Based on the matrix interference types under each standard method, the matrix interference type and confidence level of the current test sample are obtained through voting. The mutual calibration module (105) is used to map the crude polysaccharide content determination values ​​of each standard method to the same reference through the mutual calibration function stored in the preset standard database, and to calculate the calibration deviation of each standard method. The mutual calibration function is a function that describes the mathematical relationship between the determination results of two different standard methods. The comprehensive evaluation module (106) is used to obtain quality evaluation indicators for each standard method based on the original detection data, calibration deviation, matrix interference type and confidence level. The quality evaluation indicators include accuracy indicators and reliability indicators. Based on the quality evaluation indicators, the comprehensive quality score for each standard method is obtained through weighted calculation. The optimal recommended method for the current sample to be tested is determined according to the comprehensive quality score.

Images