A method and system for automatic information recognition of fluorescent immunoassay analyzer

Abstract

The present invention discloses an automatic information recognition method and system for a fluorescent immunoassay analyzer, which relates to the field of data recognition technology. The automatic information recognition method for a fluorescent immunoassay analyzer includes the following steps: simulating the path of the original light propagating to the surface of the detection sample card, adjusting the original light path layout parameters based on the path simulation results; obtaining the light loss generated by the original light reflected by the optical element, and performing light loss compensation processing to generate an optimized light path after light loss compensation; identifying the sample information of the detection sample card according to the optimized light path, and matching the sample information with the central database. The present invention uses multi-particle simulation technology to evaluate the dispersion of secondary photons, providing a scientific basis for optimizing the light path design, and at the same time obtaining and compensating for the light intensity loss of the light after multiple reflections, thereby improving the uniformity of sample excitation.

Description

Technical Field

[0001] The present invention relates to the technical field of data recognition, and in particular to an automatic information recognition method and system for a fluorescent immunoassay analyzer. Background Art

[0002] Fluorescence immunoassay analyzer is a laboratory device used to detect and quantify various biomarkers, such as proteins, viruses, bacteria, and other trace components in the human body. This device uses immunochemical methods of specific antibody-antigen binding, combined with fluorescent labels to achieve the detection of target molecules.

[0003] When using the fluorescent immunoassay analyzer, the target antigen in the sample is first mixed with antibodies that have been pre-labeled with fluorescent dyes. These antibodies can specifically recognize and bind to the target antigen. Subsequently, the fluorescent immunoassay analyzer uses its excitation light source to excite the fluorescent marker in the sample to make it glow, and the instrument then detects the light signal emitted by the fluorescent dye.

[0004] The fluorescence intensity detected by the instrument is proportional to the concentration of the target antigen in the sample, and this relationship can be used to quantify the concentration of the antigen. However, the measurement accuracy of fluorescence intensity is extremely dependent on the stability of the excitation light source. If the intensity of the excitation light source fluctuates during use, it may affect the consistency of the light and the accuracy of the detection. In addition, when the light is reflected multiple times and passes through the swing mirror, each reflection may cause light intensity loss and scattering. These factors may cause uneven sample excitation, thereby affecting the accuracy of data recognition.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes an automatic information recognition method and system for a fluorescent immunoassay analyzer to overcome the above technical problems existing in the existing related art.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, there is provided a method for automatic information recognition for a fluorescent immunoassay analyzer, the method for automatic information recognition for a fluorescent immunoassay analyzer comprising the following steps:

[0009] S1, simulating the path of the original light propagating to the surface of the detection sample card, and adjusting the original light path layout parameters based on the path simulation results;

[0010] S2, obtaining the light loss caused by the original light reflected by the optical element, and performing light loss compensation processing to generate an optimized light path after light loss compensation;

[0011] S3, identifying the sample information of the test sample card according to the optimized light path, and matching the sample information with the central database;

[0012] S4. Based on the matching results between the sample information and the central database, the test samples are evaluated and classified in turn to generate a sample test report.

[0013] Preferably, obtaining the light loss caused by the original light reflected by the optical element, performing light loss compensation processing, and generating the optimized light path after light loss compensation includes the following steps:

[0014] S21, obtaining light loss parameters generated by the original light of the fluorescent immunoassay analyzer passing through the optical element, and preprocessing the light loss parameters;

[0015] S22, dividing the preprocessed light loss parameters into a training set and a test set according to a preset ratio, constructing a Gaussian process regression model with the training set, and verifying the performance of the Gaussian process regression model with the test set;

[0016] S23, constructing a light loss parameter combination according to the range and step size in the light loss parameter space, and outputting a light loss compensation value by taking a predefined evaluation function as an input of the light loss parameter combination;

[0017] S24. The Gaussian process regression model is optimized in combination with the light loss compensation numerical value, and an optimized light path is generated after the light loss compensation and the original light path layout parameters are adjusted.

[0018] Preferably, simulating the path of the original light propagating to the surface of the detection sample card and adjusting the original light path layout parameters based on the path simulation result includes the following steps:

[0019] S11, dynamically simulating the light propagation trajectory of the original laser in the fluorescent immunoassay analyzer reflecting the light to the surface of the detection sample card;

[0020] S12, calculating the state change of the light when it contacts the optical element during the propagation process according to the light propagation trajectory, and obtaining the secondary photons when it contacts the optical element;

[0021] S13. Use multi-particle simulation technology to characterize the collision reaction of secondary photons with various optical elements in the light propagation trajectory and evaluate the dispersion of secondary photons;

[0022] S14. Evaluate the fitness of the original optical path layout based on the dispersion of the secondary photons, and adjust the parameters of the original optical path layout.

[0023] Preferably, using multi-particle simulation technology to characterize the collision reaction of secondary photons with each optical element in the light propagation trajectory, and evaluating the dispersion of secondary photons includes the following steps:

[0024] S131, simulating and generating a plurality of photon chains in the light propagation trajectory, and setting an initial position and initial light intensity for each photon chain;

[0025] S132, taking the photon chain with the longest chain length as the central reference chain, calculating the distance between the central reference chain and the remaining photon chains, and tracking the position change of each photon chain in the light propagation trajectory;

[0026] S133, while tracking the position change of the photon chain, collecting each collision of the secondary photons with each optical element, and using a physical model to calculate the energy loss and direction deviation of the secondary photons during the collision;

[0027] S134. Comprehensively evaluate the dispersion of secondary photons based on their energy loss and direction deviation during collision.

[0028] Preferably, identifying the sample information of the detection sample card according to the optimized light path and matching the sample information with the central database comprises the following steps:

[0029] S31, propagating the light generated by the fluorescent immunoassay analyzer to the surface of the test sample card according to the optimized light path, ensuring that the fluorescent label on the surface of the test sample card is excited;

[0030] S32, identifying and extracting sample electronic data of the test sample card, and identifying sample information from the sample electronic data using keyword extraction technology;

[0031] S33. Use correlation matching technology to match the sample information with the pre-built central database.

[0032] Preferably, identifying and extracting the sample electronic data of the test sample card, and identifying the sample information from the sample electronic data using a keyword extraction technique comprises the following steps:

[0033] S321, segmenting the sample electronic data into regions, and selecting candidate words from the sample electronic data in each region;

[0034] S322, performing association rule mining on the sample words, defining two sample words contained in the same region as an association relationship, and generating a frequent word item set;

[0035] S323, generating sliding window parameters according to frequent vocabulary item sets of different sizes, and calculating the associated entropy value contained in the sample vocabulary;

[0036] S324, calculating the probability of occurrence of each vocabulary item set in the frequent vocabulary item set, and calculating the associated entropy weight value in combination with the associated entropy value, and taking the associated entropy value of the sample vocabulary as the damping coefficient;

[0037] S325, traversing the link matrix of the frequent vocabulary item set, and iteratively calculating the sample vocabulary until the sample vocabulary converges to a credible stable state;

[0038] S326: Based on the iterative calculation results of the sample vocabulary, generate sample information by fusion.

[0039] Preferably, segmenting the sample electronic data into regions and selecting candidate words from the sample electronic data in each region comprises the following steps:

[0040] S3211, using segmentation technology to segment the sample electronic data into regions, and converting the sample electronic data into text data through a word segmenter;

[0041] S3212, extracting sample words from the text data of each region, and screening out repeated words in the sample words of each region, and establishing an initialization word link matrix;

[0042] S3213, marking sample words in the same window in the text data, and calculating weights of words at corresponding positions in the traversal graph and the word link matrix;

[0043] S3214: Take the words at the corresponding positions in the traversal graph as nodes, sort all the node weights, and select the word nodes whose node weights are within a preset range as candidate words for the current area.

[0044] Preferably, matching the sample information with the pre-built central database using the correlation matching technology includes the following steps:

[0045] S331, extracting sample features from sample information and encoding the sample features;

[0046] S332. Define an index for measuring the relevance of sample features to each record in the central database;

[0047] S333. Calculate the correlation between each sample feature and each record in the central database based on the correlation index, and select the record with the highest correlation as the matching item from the correlation results.

[0048] Preferably, the calculation formula of the damping coefficient is:

[0049] ;

[0050] In the formula, W represents the associated entropy weight; Indicates i Sample words; Indicates j Sample words; Indicates iThe sample vocabulary is used as the damping coefficient of the text data.

[0051] Preferably, the calculation formula of the correlation degree is:

[0052] ;

[0053] In the formula, a Represents sample characteristics; b Represents a record in the central database; t represents a constant; I It represents the correlation index; n Indicates the number; Indicates the correlation between each sample feature and each record in the central database.

[0054] According to another aspect of the present invention, there is also provided an automatic information recognition system for a fluorescent immunoassay analyzer, the automatic information recognition system for a fluorescent immunoassay analyzer comprising an optical path adjustment module, an optical path production module, a sample information recognition module and a sample report generation module;

[0055] The light path adjustment module is sequentially connected to the light path production module, the sample information identification module and the sample report generation module;

[0056] An optical path adjustment module is used to simulate the path of the original light propagating to the surface of the detection sample card, and adjust the original optical path layout parameters based on the path simulation results;

[0057] A light path production module is used to obtain the light loss caused by the original light reflected by the optical element, and to perform light loss compensation processing to generate an optimized light path after light loss compensation;

[0058] A sample information recognition module is used to identify the sample information of the detection sample card according to the optimized light path and match the sample information with the central database;

[0059] The sample report generation module is used to evaluate and classify the test samples in turn based on the matching results of the sample information and the central database, and generate a sample test report.

[0060] The beneficial effects of the present invention are:

[0061] 1. The present invention can accurately understand the behavior of light in the actual device by dynamically simulating the light propagation trajectory of the original laser in the fluorescent immunoassay analyzer, thereby improving the optical efficiency and detection accuracy of the entire system, and using multi-particle simulation technology to evaluate the dispersion of secondary photons, providing a scientific basis for optimizing the optical path design, while obtaining and compensating for the light intensity loss after multiple reflections, thereby improving the uniformity of sample excitation.

[0062] 2. The present invention automatically identifies and extracts sample information through data processing technology, reduces human errors, improves data processing speed, and deeply explores potential connections in sample data by calculating associated entropy values ​​and iterative vocabulary, providing richer data insights. At the same time, decision support based on iterative calculation and data fusion provides users with clear data-based suggestions and guidance.

[0063] 3. By defining and calculating the correlation index, the present invention can measure the similarity and correlation between sample features and database records in detail, thereby significantly improving the accuracy of matching, and directly using the matching results to assist decision-making, such as diagnostic support, patient management, etc., thereby increasing the basis for medical or research decisions and improving the scientificity and effectiveness of decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0065] Figure 1 is a flow chart of an automatic information recognition method for a fluorescent immunoassay analyzer according to an embodiment of the present invention;

[0066] Figure 2 The invention is a principle block diagram of an automatic information recognition system for a fluorescent immunoassay analyzer according to an embodiment of the invention.

[0067] In the figure:

[0068] 1. Light path adjustment module; 2. Light path production module; 3. Sample information identification module; 4. Sample report generation module. DETAILED DESCRIPTION

[0069] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. By referring to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0070] According to an embodiment of the present invention, a method and system for automatic information recognition for a fluorescent immunoassay analyzer are provided.

[0071] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1As shown, according to the automatic information recognition method for a fluorescent immunoassay analyzer according to an embodiment of the present invention, the automatic information recognition method for a fluorescent immunoassay analyzer comprises the following steps:

[0072] S1. Simulate the path of the original light propagating to the surface of the detection sample card, and adjust the original light path layout parameters based on the path simulation results.

[0073] The steps of simulating the path of the original light propagating to the surface of the detection sample card and adjusting the original light path layout parameters based on the path simulation result include the following steps:

[0074] S11, dynamically simulate the light propagation trajectory of the original laser in the fluorescent immunoassay analyzer reflecting the light to the surface of the detection sample card.

[0075] It should be noted that the light propagation trajectory of the original laser in the dynamic simulation fluorescent immunoassay analyzer reflecting the light to the surface of the test sample card includes:

[0076] Define the initial parameters of the light: set the initial intensity, wavelength and emission angle of the light.

[0077] Design the optical path: Design the optical path layout according to the structure of the fluorescent immunoassay analyzer, including the position and angle of the lens and other optical components.

[0078] Simulated components: Detailed definition of the physical properties of each optical component, such as reflectivity, transmittance, and refractive index.

[0079] Ray Tracing: Use ray tracing technology to simulate the complete path of light from the laser to the test sample card, taking into account reflection, refraction and scattering when it contacts each optical component.

[0080] Record light status: record the state changes of light at each key point in the propagation process, including light intensity, direction and phase changes.

[0081] S12. Calculate the state change of the light when it contacts the optical element during the propagation process according to the light propagation trajectory, and obtain the secondary photons when it contacts the optical element.

[0082] S13. Use multi-particle simulation technology to characterize the collision reaction of secondary photons with various optical elements in the light propagation trajectory and evaluate the dispersion of secondary photons.

[0083] Among them, the multi-particle simulation technology is used to characterize the collision reaction of secondary photons with various optical elements in the light propagation trajectory, and the dispersion of secondary photons is evaluated, which includes the following steps:

[0084] S131, simulating and generating a plurality of photon chains in the light propagation trajectory, and setting an initial position and initial light intensity for each photon chain;

[0085] S132, taking the photon chain with the longest chain length as the central reference chain, calculating the distance between the central reference chain and the remaining photon chains, and tracking the position change of each photon chain in the light propagation trajectory;

[0086] S133. While tracking the position change of the photon chain, each collision between the secondary photons and each optical element is collected, and the energy loss and direction deviation of the secondary photons during the collision are calculated using a physical model.

[0087] It should be noted that in tracking the position change of the photon chain, each collision of the secondary photon with each optical element is collected, and the energy loss and direction deviation of the secondary photon during the collision are calculated using the physical model, including:

[0088] During the propagation of the photon chain, each time a photon comes into contact with an optical element (such as a lens, a mirror, or a filter), the specific circumstances of the collision are recorded, including the time, position, energy, and angle of incidence of the collision. At each collision, the conditions for producing secondary photons are calculated based on the physical laws of the interaction between light and matter. The energy loss of the photon in each collision is calculated using a physical model, including consideration of the interaction between the photon and the material of the optical element.

[0089] It should be noted that the energy loss of photons in each collision calculated using the physical model includes:

[0090] Determine the physical scattering model, which includes Rutherford scattering, or models that are more suitable for photon scattering, such as Rayleigh scattering or Mie scattering; obtain the relevant physical properties of the optical component material; combine the energy of the photon and the refractive index of the target material to generate the cross-section of the interaction between the photon and the material, and simulate the scattering event of the photon with a single atom or molecule to obtain the scattering angle and the energy after scattering; if the photon undergoes multiple scatterings when passing through the optical component, it is necessary to iteratively calculate the state of the photon after each scattering, and use the energy conservation and momentum conservation laws in the existing technology to calculate the energy loss of the photon during the scattering process.

[0091] S134. Comprehensively evaluate the dispersion of secondary photons based on their energy loss and direction deviation during collision.

[0092] S14. Evaluate the fitness of the original optical path layout based on the dispersion of the secondary photons, and adjust the parameters of the original optical path layout.

[0093] S2. Obtain the light loss caused by the original light reflected by the optical element, and perform light loss compensation processing to generate an optimized light path after light loss compensation.

[0094] The steps of obtaining the light loss caused by the original light reflected by the optical element and performing light loss compensation processing to generate the optimized light path after light loss compensation include the following steps:

[0095] S21, obtaining light loss parameters generated by the original light of the fluorescent immunoassay analyzer passing through the optical element, and preprocessing the light loss parameters;

[0096] S22, dividing the preprocessed light loss parameters into a training set and a test set according to a preset ratio, constructing a Gaussian process regression model with the training set, and verifying the performance of the Gaussian process regression model with the test set.

[0097] It should be noted that the pre-processed light loss parameters are divided into a training set and a test set according to a preset ratio, and the training set is used to construct a Gaussian process regression model. The test set verifies the performance of the Gaussian process regression model, including:

[0098] The light loss parameters are divided according to a preset ratio (e.g., 70% training set, 30% test set), and data points are randomly selected and assigned to the training set and the test set to ensure the uniformity of data distribution. A kernel function (such as the radial basis function) is selected, which is the key to defining the relationship between data points in the Gaussian process regression model. The Gaussian process regression model is trained using the training set data, including parameter tuning of the kernel function and learning the relationship between data. The optical system is optimized according to the prediction results of the Gaussian process regression model.

[0099] S23, constructing a light loss parameter combination according to the range and step size in the light loss parameter space, and outputting a light loss compensation value by taking a predefined evaluation function as an input of the light loss parameter combination;

[0100] It should be noted that, according to the range and step size in the light loss parameter space, a light loss parameter combination is constructed, and according to the predefined evaluation function as the input of the light loss parameter combination, the output light loss compensation value includes:

[0101] Generate all possible parameter combinations based on the range and step size in the light loss parameter space by programming loops or using specific library functions; write a function that takes the parameter combination as input and returns the corresponding light loss compensation value. This function needs to predict the light loss based on the physical model; apply the evaluation function to each set of parameter combinations, analyze the light loss compensation values ​​under different parameter combinations, and find the best combination; adjust the range or step size of the parameters based on the analysis results.

[0102] S24. The Gaussian process regression model is optimized in combination with the light loss compensation numerical value, and an optimized light path is generated after the light loss compensation and the original light path layout parameters are adjusted.

[0103] S3. Identify the sample information of the test sample card according to the optimized light path, and match the sample information with the central database.

[0104] The steps of identifying the sample information of the sample card according to the optimized light path and matching the sample information with the central database include the following steps:

[0105] S31, propagating the light generated by the fluorescent immunoassay analyzer to the surface of the test sample card according to the optimized light path, ensuring that the fluorescent label on the surface of the test sample card is excited;

[0106] S32. Identify and extract the sample electronic data of the test sample card, and use keyword extraction technology to identify sample information from the sample electronic data.

[0107] Among them, identifying and extracting the sample electronic data of the test sample card, and using keyword extraction technology to identify sample information from the sample electronic data includes the following steps:

[0108] S321, segmenting the sample electronic data into regions, and selecting candidate words from the sample electronic data in each region.

[0109] The process of segmenting the sample electronic data into regions and selecting candidate words from the sample electronic data in each region includes the following steps:

[0110] S3211. Use segmentation technology to segment the sample electronic data into regions, and convert the sample electronic data into text data through a word segmenter.

[0111] It should be noted that using segmentation technology to segment sample electronic data into regions and converting the sample electronic data into text data through a word segmenter includes:

[0112] Select regional segmentation technology based on the characteristics of sample data. Regional segmentation technologies include threshold-based segmentation, cluster segmentation, edge detection segmentation, etc.; use segmentation technology to divide electronic data into physical or logical areas, each area contains a group of data points with similar characteristics; use word segmentation tools to convert the electronic data in each area into text data, and the choice of word segmentation depends on the language and characteristics of the data, such as Chinese word segmentation, English word segmentation, etc.; standardize the converted text, including unifying word forms, eliminating spelling errors, removing stop words, etc.; check whether the converted text data accurately reflects the original electronic data information.

[0113] S3212, extracting sample words from the text data of each region, and screening out repeated words in the sample words of each region, and establishing an initialization word link matrix;

[0114] S3213. Mark the sample words in the same window in the text data, and calculate the weights of the words at corresponding positions in the traversal graph and the word link matrix.

[0115] S3214: Take the words at the corresponding positions in the traversal graph as nodes, sort all the node weights, and select the word nodes whose node weights are within a preset range as candidate words for the current area.

[0116] S322: Perform association rule mining on sample words, define two sample words contained in the same region as an association relationship, and generate frequent word item sets.

[0117] It should be noted that association rules are mined for sample words, and two sample words in the same region are defined as association relationships. At the same time, frequent word item sets are generated, including:

[0118] Generate frequent item sets using technologies such as Apriori (a priori algorithm) and FP-Growth (frequent pattern growth algorithm); determine a minimum support threshold, and only when the support of the item set is greater than or equal to this threshold, it is considered a frequent item set; generate association rules from frequent item sets, these rules are expressed in the form of "A is greater than or equal to B", which means that if the transaction contains A, it is very likely to also contain B; determine the minimum confidence, and only when the confidence of the rule is higher than this threshold is the rule considered valid; evaluate the quality of the generated association rules, adjust the algorithm parameters (such as support and confidence thresholds) according to the evaluation results, and regenerate frequent item sets and association rules.

[0119] It should be noted that the Apriori algorithm is based on a core principle, that is, all non-empty subsets of frequent itemsets must also be frequent. It gradually increases the size of the itemset starting from the smallest candidate itemset, and each step will prune the itemsets that do not meet the minimum support threshold.

[0120] The FP-Growth algorithm uses an efficient data structure called a frequent pattern tree (FP-tree) to compress the data set. After the FP-tree is constructed, the tree structure is used to directly mine frequent item sets without generating candidate item sets.

[0121] S323: Generate sliding window parameters according to frequent vocabulary item sets of different sizes, and calculate the associated entropy value contained in the sample vocabulary.

[0122] It should be noted that the sliding window parameters are generated according to the frequent vocabulary item sets of different sizes, and the associated entropy values ​​contained in the sample vocabulary are calculated, including:

[0123] Determine the size of the sliding window based on the frequent vocabulary item sets of different sizes. The size of the item set can directly affect the size of the window. Set specific parameters of the sliding window, such as window length and step size. Apply sliding window technology to text data. Each window covers a certain number of words. At this time, the windows can overlap and can be moved according to the step size parameter.

[0124] Correlation entropy is a measure of the mutual information between two or more variables. Correlation entropy can be used to measure the information complexity of word combinations within a window; analyze the correlation entropy values ​​of different windows to find out word combinations with high information content, and adjust the window size and step size according to the analysis results of the correlation entropy values ​​to better capture key information.

[0125] S324, calculating the probability of occurrence of each vocabulary item set in the frequent vocabulary item set, and calculating the associated entropy weight value in combination with the associated entropy value, and taking the associated entropy value of the sample vocabulary as the damping coefficient.

[0126] The calculation formula of the damping coefficient is:

[0127] ;

[0128] In the formula, W represents the associated entropy weight; Indicates i Sample words; Indicates j Sample words; Indicates i The sample vocabulary is used as the damping coefficient of the text data.

[0129] S325, traversing the link matrix of the frequent vocabulary item set, and iteratively calculating the sample vocabulary until the sample vocabulary converges to a credible stable state;

[0130] S326: Based on the iterative calculation results of the sample vocabulary, generate sample information by fusion.

[0131] S33. Use correlation matching technology to match the sample information with the pre-built central database.

[0132] The process of matching sample information with a pre-built central database using correlation matching technology includes the following steps:

[0133] S331, extracting sample features from sample information and encoding the sample features;

[0134] S332. Define an index to measure the correlation between sample characteristics and each record in the central database.

[0135] It should be noted that the indicators defined to measure the correlation between sample characteristics and each record in the central database include:

[0136] First, important features should be identified. Important features should be able to represent the key attributes of samples and database records, such as numerical, categorical or textual features. The selected features should be preprocessed, such as normalizing numerical features, vectorizing text features, encoding categorical features, etc. Samples and records are composed of multiple different types of features, and a combination of multiple measurement methods is needed to comprehensively evaluate the correlation.

[0137] S333. Calculate the correlation between each sample feature and each record in the central database based on the correlation index, and select the record with the highest correlation as the matching item from the correlation results.

[0138] The calculation formula of correlation is:

[0139] ;

[0140] In the formula, a Represents sample characteristics; b Represents a record in the central database; t represents a constant; I It represents the correlation index; n Indicates the number; Indicates the correlation between each sample feature and each record in the central database.

[0141] S4. Based on the matching results between the sample information and the central database, the test samples are evaluated and classified in turn to generate a sample test report.

[0142] It should be noted that based on the matching results of sample information and the central database, the test samples are evaluated and classified in turn, and the sample test report generated includes:

[0143] Using the correlation index defined above, match each test sample with the records in the central database; filter out records with high correlation according to the set threshold and identify them as successful matches; conduct a comprehensive assessment of the status, category or nature of the sample based on the matching results. For example, infer the category of the sample based on the degree of match between the sample and the known categories in the database; if applicable, conduct a risk assessment of the sample to determine whether the sample meets safety standards or contains dangerous components; classify the samples into different categories or levels based on business needs and assessment results. For example, classify the samples into normal, abnormal, and those requiring further testing, and generate a sample test report based on the classification results; the generated sample test report should contain the basic information of the sample, matching details, assessment results, classification results, and any recommended follow-up measures.

[0144] like Figure 2 As shown, according to another embodiment of the present invention, there is also provided an automatic information recognition system for a fluorescent immunoassay analyzer, the automatic information recognition system for a fluorescent immunoassay analyzer comprises an optical path adjustment module 1, an optical path production module 2, a sample information recognition module 3 and a sample report generation module 4;

[0145] The light path adjustment module 1 is connected to the light path production module 2, the sample information identification module 3 and the sample report generation module 4 in sequence;

[0146] The optical path adjustment module 1 is used to simulate the path of the original light propagating to the surface of the detection sample card, and adjust the original optical path layout parameters based on the path simulation result;

[0147] The light path production module 2 is used to obtain the light loss caused by the original light reflected by the optical element, and to perform light loss compensation processing to generate an optimized light path after light loss compensation;

[0148] The sample information identification module 3 is used to identify the sample information of the detection sample card according to the optimized light path and match the sample information with the central database;

[0149] The sample report generation module 4 is used to evaluate and classify the test samples in turn based on the matching results between the sample information and the central database, and generate a sample test report.

[0150] In summary, with the help of the above technical solutions of the present invention, the present invention can accurately understand the behavior of light in the actual device by dynamically simulating the light propagation trajectory of the original laser in the fluorescent immunoassay analyzer, thereby improving the optical efficiency and detection accuracy of the entire system, and using multi-particle simulation technology to evaluate the dispersion of secondary photons, providing a scientific basis for optimizing the optical path design, while obtaining and compensating for the light intensity loss after multiple reflections, thereby improving the uniformity of sample excitation; the present invention automatically identifies and extracts sample information through data processing technology, reduces human errors, and improves data processing speed, and deeply mines potential connections in sample data by calculating associated entropy values ​​and iterative vocabulary, providing richer data insights, and providing users with clear data-based suggestions and guidance based on decision support based on iterative calculations and data fusion; the present invention can exhaustively measure the similarity and correlation between sample features and database records by defining and calculating correlation indicators, thereby significantly improving the accuracy of matching, and directly using matching results for auxiliary decision-making, such as diagnostic support, patient management, etc., increasing the basis for medical or research decisions, thereby improving the scientificity and effectiveness of decisions.

[0151] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatic information recognition of a fluorescent immunoassay analyzer, characterized in that: The method for automatically identifying information for a fluorescent immunoassay analyzer comprises the following steps: S1, simulating the path of the original light propagating to the surface of the detection sample card, and adjusting the original light path layout parameters based on the path simulation results; S2, obtaining the light loss caused by the original light reflected by the optical element, and performing light loss compensation processing to generate an optimized light path after light loss compensation; S3, identifying the sample information of the test sample card according to the optimized light path, and matching the sample information with the central database; S4. Based on the matching results between the sample information and the central database, the test samples are evaluated and classified in turn to generate a sample test report; The step of obtaining the light loss caused by the original light reflected by the optical element, performing light loss compensation processing, and generating the optimized light path after light loss compensation comprises the following steps: S21, obtaining light loss parameters generated by the original light of the fluorescent immunoassay analyzer passing through the optical element, and preprocessing the light loss parameters; S22, dividing the preprocessed light loss parameters into a training set and a test set according to a preset ratio, constructing a Gaussian process regression model with the training set, and verifying the performance of the Gaussian process regression model with the test set; S23, constructing a light loss parameter combination according to the range and step size in the light loss parameter space, and outputting a light loss compensation value by taking a predefined evaluation function as an input of the light loss parameter combination; S24, optimizing the Gaussian process regression model in combination with the light loss compensation numerical value, and generating an optimized light path after light loss compensation and adjustment of the original light path layout parameters; The simulation of the path of the original light propagating to the surface of the detection sample card and adjusting the original light path layout parameters based on the path simulation result include the following steps: S11, dynamically simulating the light propagation trajectory of the original laser in the fluorescent immunoassay analyzer reflecting the light to the surface of the detection sample card; S12, calculating the state change of the light when it contacts the optical element during the propagation process according to the light propagation trajectory, and obtaining the secondary photons when it contacts the optical element; S13. Use multi-particle simulation technology to characterize the collision reaction of secondary photons with various optical elements in the light propagation trajectory and evaluate the dispersion of secondary photons; S14. Evaluate the fitness of the original optical path layout based on the dispersion of the secondary photons, and adjust the parameters of the original optical path layout.

2. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 1, characterized in that: The method of using multi-particle simulation technology to characterize the collision reaction of secondary photons with each optical element in the light propagation trajectory and evaluating the dispersion of secondary photons includes the following steps: S131, simulating and generating a plurality of photon chains in the light propagation trajectory, and setting an initial position and initial light intensity for each photon chain; S132, taking the photon chain with the longest chain length as the central reference chain, calculating the distance between the central reference chain and the remaining photon chains, and tracking the position change of each photon chain in the light propagation trajectory; S133, while tracking the position change of the photon chain, collecting each collision of the secondary photons with each optical element, and using a physical model to calculate the energy loss and direction deviation of the secondary photons during the collision; S134. Comprehensively evaluate the dispersion of secondary photons based on their energy loss and directional deviation during collision.

3. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 2, characterized in that: The method of identifying the sample information of the sample card according to the optimized light path and matching the sample information with the central database comprises the following steps: S31, propagating the light generated by the fluorescent immunoassay analyzer to the surface of the test sample card according to the optimized light path, ensuring that the fluorescent label on the surface of the test sample card is excited; S32, identifying and extracting sample electronic data of the test sample card, and identifying sample information from the sample electronic data using keyword extraction technology; S33. Use correlation matching technology to match the sample information with the pre-built central database.

4. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 3, characterized in that: The method of identifying and extracting sample electronic data of the test sample card and identifying sample information from the sample electronic data using keyword extraction technology includes the following steps: S321, segmenting the sample electronic data into regions, and selecting candidate words from the sample electronic data in each region; S322, performing association rule mining on the sample words, defining two sample words contained in the same region as an association relationship, and generating a frequent word item set; S323, generating sliding window parameters according to frequent vocabulary item sets of different sizes, and calculating the associated entropy value contained in the sample vocabulary; S324, calculating the probability of occurrence of each vocabulary item set in the frequent vocabulary item set, and calculating the associated entropy weight value in combination with the associated entropy value, and taking the associated entropy value of the sample vocabulary as the damping coefficient; S325, traversing the link matrix of the frequent vocabulary item set, and iteratively calculating the sample vocabulary until the sample vocabulary converges to a credible stable state; S326: Based on the iterative calculation results of the sample vocabulary, generate sample information by fusion.

5. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 4, characterized in that: The process of segmenting the sample electronic data into regions and selecting candidate words from the sample electronic data in each region comprises the following steps: S3211, using segmentation technology to segment the sample electronic data into regions, and converting the sample electronic data into text data through a word segmenter; S3212, extracting sample words from the text data of each region, and screening out repeated words in the sample words of each region, and establishing an initialization word link matrix; S3213, marking sample words in the same window in the text data, and calculating weights of words at corresponding positions in the traversal graph and the word link matrix; S3214: Take the words at the corresponding positions in the traversal graph as nodes, sort all the node weights, and select the word nodes whose node weights are within a preset range as candidate words for the current area.

6. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 5, characterized in that: The method of matching the sample information with the pre-built central database using the correlation matching technology includes the following steps: S331, extracting sample features from sample information and encoding the sample features; S332. Define an index for measuring the relevance of sample features to each record in the central database; S333. Calculate the correlation between each sample feature and each record in the central database based on the correlation index, and select the record with the highest correlation as the matching item from the correlation results.

7. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 6, characterized in that: The calculation formula of the damping coefficient is: ； In the formula, W represents the associated entropy weight; Indicates i Sample words; Indicates j Sample words; Indicates i The sample vocabulary is used as the damping coefficient of the text data.

8. The method for automatic information recognition for a fluorescent immunoassay analyzer according to claim 7, characterized in that: The calculation formula of the correlation degree is: ； In the formula, a Represents sample characteristics; b Represents a record in the central database; t represents a constant; I It represents the correlation index; n Indicates the number; Indicates the correlation between each sample feature and each record in the central database.

9. An automatic information recognition system for a fluorescent immunoassay analyzer, used to implement the automatic information recognition method for a fluorescent immunoassay analyzer according to any one of claims 1 to 8, characterized in that: The automatic information recognition system for fluorescent immunoassay analyzer includes a light path adjustment module, a light path production module, a sample information recognition module and a sample report generation module; The optical path adjustment module is sequentially connected to the optical path production module, the sample information identification module and the sample report generation module; The optical path adjustment module is used to simulate the path of the original light propagating to the surface of the detection sample card, and adjust the original optical path layout parameters based on the path simulation result; The light path production module is used to obtain the light loss caused by the original light reflected by the optical element, and perform light loss compensation processing to generate an optimized light path after light loss compensation; The sample information identification module is used to identify the sample information of the detection sample card according to the optimized light path and match the sample information with the central database; The sample report generation module is used to evaluate and classify the test samples in turn based on the matching results between the sample information and the central database, and generate a sample test report.

Images