An autoimmune disease antibody detection and clinical diagnosis support system

Abstract

This invention relates to a system for supporting the detection and clinical diagnosis of autoimmune diseases using antibodies. The system performs parallel detection of multiple autoantibody items in a subject's bodily fluid sample, obtaining the detection results for each antibody item. It then combines these results with the subject's clinical correlation information, historical testing records, and sample testing process information to form a multidimensional data foundation for auxiliary interpretation of autoimmune diseases. Based on antibody combination patterns associated with different autoimmune diseases, the system calculates the degree of matching between the subject's sample and candidate diseases, generating candidate disease categories and their initial risk scores. For boundary samples with weak positives, single positives, cross-positives, or antibody results inconsistent with clinical manifestations, the system further refines the initial risk score by considering antibody combination consistency, disease-specific conflicts, individual historical baseline changes, and detection reliability, resulting in a risk level more suitable for clinical reference.

Description

Technical Field

[0001] This invention belongs to the field of medical testing technology, and more specifically, relates to an antibody detection and clinical diagnostic support system for autoimmune diseases. Background Technology

[0002] Autoimmune diseases are a group of diseases caused by abnormal immune responses of the body's own tissues or cellular components. Common types include systemic lupus erythematosus, rheumatoid arthritis, Sjögren's syndrome, systemic sclerosis, inflammatory myopathy, antiphospholipid syndrome, and autoimmune vasculitis. Autoantibody testing is an important tool for screening, assisting in diagnosis, and assessing the condition of these diseases. Clinically, it is commonly used to assess a patient's immune abnormality through tests such as antinuclear antibody profiles, anti-double-stranded DNA antibodies, anti-Sm antibodies, anti-SSA antibodies, anti-SSB antibodies, anti-CCP antibodies, rheumatoid factor, ANCA-related antibodies, and antiphospholipid antibodies.

[0003] However, the clinical interpretation of existing autoantibody test results still has certain limitations. Some autoantibody items exhibit cross-positive or non-specific positivity, and a single antibody positivity does not necessarily correspond to a specific disease. Some patients in the early stages of the disease only present with weak positives, gray zone results, or atypical symptoms, making it difficult to directly determine their disease risk based on fixed thresholds. In addition, different testing platforms, sample conditions, testing batches, and quality control conditions may also affect the reliability of test results. Existing testing systems mostly focus on outputting positive or negative or quantitative results for individual antibody items, and do not adequately utilize the combined use of antibody combination relationships, clinical manifestations, historical testing trends, and test reliability. This results in low stability of auxiliary interpretation of boundary samples, cross-positive samples, and samples where antibody results are inconsistent with clinical manifestations. Therefore, it is necessary to propose an autoimmune disease antibody detection and clinical diagnostic support system that can comprehensively analyze multiple antibody test results, clinical correlation information, historical testing records, and sample test reliability.

[0004] A review of relevant publicly available technologies reveals the following: WO2017062946A1 proposes an immunoassay device to simplify the procedure for determining hemoglobinopathies; CN116338159B proposes a fully automated paper-based microfluidic system for local smartphone detection, enabling an immunoassay system based on a mobile phone; and JP2014215210A proposes a detection system to shorten the time required for whole blood immunoassay testing. By optimizing the testing procedure, the system reduces the overall testing time compared to existing methods.

[0005] The above technical solutions each propose a technical approach to optimize the immune detection process, but there are currently few technical solutions proposed to assist medical personnel in interpreting immune results.

[0006] The foregoing description of the background art is intended only to facilitate understanding of the invention. This description does not endorse or acknowledge any common general knowledge in the materials mentioned. Summary of the Invention

[0007] The purpose of this invention is to provide a system for supporting the detection and clinical diagnosis of autoimmune diseases using antibodies. This system performs parallel detection of multiple autoantibody items in a subject's bodily fluid sample, obtaining the detection results for each antibody item. It then combines this with the subject's clinical correlation information, historical testing records, and sample testing process information to form a multidimensional data foundation for the auxiliary interpretation of autoimmune diseases. Based on antibody combination patterns associated with different autoimmune diseases, the system calculates the degree of matching between the subject's sample and candidate diseases, generating candidate disease categories and their initial risk scores. For boundary samples with weak positives, single positives, cross-positives, or antibody results inconsistent with clinical manifestations, the system further refines the initial risk score by considering antibody combination consistency, disease-specific conflicts, individual historical baseline changes, and detection reliability, resulting in a risk level more suitable for clinical reference.

[0008] The present invention adopts the following technical solution: an autoimmune disease antibody detection and clinical diagnosis support system, the support system comprising: a multi-channel antibody detection module, an information management module, a feature vector construction module, an antibody combination pattern recognition module, a boundary sample correction module, and a diagnostic support output module; The multi-channel antibody detection module is used to perform parallel detection of multiple antibody items on the body fluid sample of the subject and obtain the detection results of each antibody item. The information management module is used to acquire the examinee's clinical correlation information, historical test records, sample correlation information, and testing process information, and to perform basic quality screening on the test samples based on the sample correlation information and testing process information; The feature vector construction module is used to construct a multidimensional diagnostic feature vector based on the detection results, clinical correlation information, and historical test records; the multidimensional diagnostic feature vector includes an antibody strength sub-vector, an antibody combination feature vector, a clinical support sub-vector, and a historical trend sub-vector; The antibody combination pattern recognition module is used to identify the degree of matching between the subject sample and different autoimmune disease-related antibody combination patterns based on the multidimensional diagnostic feature vector, generate candidate disease categories and their initial risk scores, and determine whether the subject sample is a boundary sample based on preset boundary sample determination conditions. The boundary sample correction module is used to correct the initial risk score of a boundary sample when the detected sample is identified as a boundary sample, based on the antibody combination consistency parameter, disease pointing conflict parameter, historical offset characteristics, and detection confidence set of the boundary sample, to obtain a corrected risk score. The diagnostic support output module is used to output diagnostic support results based on the initial risk score when the test sample is not identified as a borderline sample; and to output diagnostic support results based on the revised risk score when the test sample is identified as a borderline sample; wherein the diagnostic support results include the suspected disease category, risk level, and further examination recommendations.

[0009] Preferably, the information management module further includes performing the following operations: when the test sample does not meet the minimum usability requirements, preventing the test sample from entering the subsequent antibody combination pattern recognition process; when the test sample meets the minimum usability requirements, attaching a primary confidence mark to the antibody test result of the corresponding test sample.

[0010] Preferably, the feature vector construction module, when constructing a multidimensional diagnostic feature vector, includes performing the following: Based on the original detection value A of the i-th antibody item, calculate the standardized antibody intensity value xi of the i-th antibody item, and construct the antibody intensity sub-vector XA from multiple standardized antibody intensity values ​​xi; Based on the set of antibody items related to the j-th candidate disease and their corresponding standardized antibody strength values, an antibody combination feature vector XC is formed. The clinical association information is converted into a clinical support subvector XL; Based on the offset of the current detection value of the same detection item relative to the individual's historical baseline, a historical trend sub-vector XH is formed; The antibody strength sub-vector, antibody combination feature vector, clinical support sub-vector, and historical trend sub-vector are concatenated or fused to obtain the multidimensional diagnostic feature vector.

[0011] Preferably, the feature vector construction module is further configured to: generate a detection confidence set Q={q1,q2,…,q} based on one or more data information output by the information management module, including the initial confidence marker, sample status, detection batch information, quality control signal, repeated detection difference, and channel background signal. i ,…,q m}, q i denoted as the confidence coefficient of the detection result of the i-th antibody item, and m represents the number of antibody items included in the detection.

[0012] Preferably, the boundary sample correction module is configured as follows: The system receives the candidate disease category, initial risk score, antibody combination consistency parameter, disease pointing conflict parameter, standardized antibody strength value, historical offset features, and detection confidence set corresponding to the sample determined as a boundary sample by the antibody combination pattern recognition module. A boundary correction factor is generated based on the relevant antibody set corresponding to the j-th candidate disease category, the antibody combination consistency parameter, the disease pointing conflict parameter, the historical offset features, and the detection confidence set. The initial risk score for the j-th candidate disease category is corrected based on the boundary correction factor to obtain the corrected risk score; Specifically, when the antibody combination consistency parameter of the j-th candidate disease category reaches a preset threshold and the relevant antibody item shows a credible upward trend relative to the individual's historical baseline, the boundary correction factor is used to raise or maintain the risk score of the corresponding candidate disease category; when the disease-pointing conflict parameter reaches a preset threshold, or the relevant antibody item is within the boundary range and the detection credibility is lower than the preset threshold, the boundary correction factor is used to lower the risk score of the corresponding candidate disease category.

[0013] Preferably, the boundary sample determination criteria include at least one of the following: At least one antibody item associated with a candidate autoimmune disease has a standardized antibody strength value that is between the upper limit of negative and the threshold for positive judgment, or is in the vicinity of weak positive. The presence of only a single positive antibody, with insufficient auxiliary antibody features or clinical supporting features associated with the same candidate autoimmune disease; Two or more candidate autoimmune disease categories have similar initial risk scores, or the disease-pointing conflict parameter reaches a preset conflict threshold. The antibody test result was positive or weakly positive, but the clinical symptoms, inflammatory markers, complement markers, or organ involvement indicators did not provide corresponding support.

[0014] The beneficial effects achieved by this invention are: This technical solution does not rely solely on the positive or negative results of a single antibody test. Instead, it combines the combination relationships between multiple autoantibody tests, clinical correlation information, and historical testing trends to identify the degree of matching between the subject's sample and the antibody combination patterns associated with different autoimmune diseases, thereby reducing misjudgments caused by single antibody positivity or nonspecific positivity.

[0015] This technical solution corrects the initial risk score for samples with weak positive, single positive, cross-positive, or antibody test results inconsistent with clinical manifestations by further combining antibody combination consistency, disease-specific conflicts, individual historical baseline changes, and test reliability, thus avoiding the direct interpretation of gray zone results or occasional weak positive results as high-risk disease indications.

[0016] This technical solution can output suspected disease category, risk level and further examination suggestions based on the initial risk score or the revised risk score, thereby improving the clinical usability of the test results.

[0017] The supporting system described in this technical solution adopts a modular design for each working part. The system can be maintained and upgraded by optimizing and replacing the working modules individually, thereby reducing the subsequent usage and upgrade costs. Attached Figure Description

[0018] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0019] Reference numerals: 10-Support System; 110-Multi-channel Antibody Detection Module; 120-Information Management Module; 130-Feature Vector Construction Module; 140-Antibody Combination Pattern Recognition Module; 150-Boundary Sample Correction Module; 160-Diagnostic Support Output Module; 12-Sample Input Unit; 114-Reagent Reaction Unit; 116-Detection Channel Array; 118-Signal Acquisition Unit; 119-Detection Control Unit; 122-Patient Information Entry Unit; 124-Clinical Related Information Acquisition Unit; 126-Historical Test Record Retrieval Unit; 128-Sample Related Information Management Unit; 129-Quality Verification Unit; 500-Computing System; 502-Bus; 504-Processor; 506-Main Memory; 508-Read-Only Memory; 510-Storage Device; 512-Display; 514-Input Device; 516-Cursor Control Device; 518-Network Device; Figure 1 This is a schematic diagram of the framework of the support system described in an embodiment of the present invention; Figure 2 This is a schematic diagram of the framework of the multi-channel antibody detection module described in this embodiment of the invention; Figure 3 This is a schematic diagram of the framework of the information management module described in this embodiment of the invention; Figure 4 This is a schematic diagram of the user interface of the support system described in an embodiment of the present invention; Figure 5 This is a schematic diagram of the computer system architecture used in the support system described in this embodiment of the invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to its embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Other systems, methods, and / or features of this embodiment will become apparent to those skilled in the art after reviewing the following detailed description. All such additional systems, methods, features, and advantages are intended to be included within this specification, within the scope of the invention, and protected by the appended claims. Further features of the disclosed embodiments are described in the following detailed description, and these features will become apparent from the following detailed description.

[0021] In the accompanying drawings of this invention, the same or similar reference numerals correspond to the same or similar components. In the description of this invention, it should be understood that if terms such as "upper," "lower," "left," and "right" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation. Because the invention is constructed and operated in a specific orientation, the terms describing positional relationships in the drawings are for illustrative purposes only and should not be construed as limiting this patent. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0022] Example 1: For example, an autoimmune disease antibody detection and clinical diagnosis support system is proposed. The support system includes: a multi-channel antibody detection module, an information management module, a feature vector construction module, an antibody combination pattern recognition module, a boundary sample correction module, and a diagnostic support output module. The multi-channel antibody detection module is used to perform parallel detection of multiple antibody items on the body fluid sample of the subject and obtain the detection results of each antibody item. The information management module is used to acquire the examinee's clinical correlation information, historical test records, sample correlation information, and testing process information, and to perform basic quality screening on the test samples based on the sample correlation information and testing process information; The feature vector construction module is used to construct a multidimensional diagnostic feature vector based on the detection results, clinical correlation information, and historical test records; the multidimensional diagnostic feature vector includes an antibody strength sub-vector, an antibody combination feature vector, a clinical support sub-vector, and a historical trend sub-vector; The antibody combination pattern recognition module is used to identify the degree of matching between the subject sample and different autoimmune disease-related antibody combination patterns based on the multidimensional diagnostic feature vector, generate candidate disease categories and their initial risk scores, and determine whether the subject sample is a boundary sample based on preset boundary sample determination conditions. The boundary sample correction module is used to correct the initial risk score of a boundary sample when the detected sample is identified as a boundary sample, based on the antibody combination consistency parameter, disease pointing conflict parameter, historical offset characteristics, and detection confidence set of the boundary sample, to obtain a corrected risk score. The diagnostic support output module is used to output diagnostic support results based on the initial risk score when the test sample is not identified as a borderline sample; and to output diagnostic support results based on the revised risk score when the test sample is identified as a borderline sample; wherein the diagnostic support results include the suspected disease category, risk level, and further examination recommendations.

[0023] Preferably, the information management module further includes performing the following operations: when the test sample does not meet the minimum usability requirements, preventing the test sample from entering the subsequent antibody combination pattern recognition process; when the test sample meets the minimum usability requirements, attaching a primary confidence mark to the antibody test result of the corresponding test sample.

[0024] Preferably, the feature vector construction module, when constructing a multidimensional diagnostic feature vector, includes performing the following: Based on the original detection value A of the i-th antibody item, calculate the standardized antibody intensity value xi of the i-th antibody item, and construct the antibody intensity sub-vector XA from multiple standardized antibody intensity values ​​xi; Based on the set of antibody items related to the j-th candidate disease and their corresponding standardized antibody strength values, an antibody combination feature vector XC is formed. The clinical association information is converted into a clinical support subvector XL; Based on the offset of the current detection value of the same detection item relative to the individual's historical baseline, a historical trend sub-vector XH is formed; The antibody strength sub-vector, antibody combination feature vector, clinical support sub-vector, and historical trend sub-vector are concatenated or fused to obtain the multidimensional diagnostic feature vector.

[0025] Preferably, the feature vector construction module is further configured to: generate a detection confidence set Q={q1,q2,…,q} based on one or more data information output by the information management module, including the initial confidence marker, sample status, detection batch information, quality control signal, repeated detection difference, and channel background signal. i ,…,q m}, q i denoted as the confidence coefficient of the detection result of the i-th antibody item, and m represents the number of antibody items included in the detection.

[0026] Preferably, the boundary sample correction module is configured as follows: The system receives the candidate disease category, initial risk score, antibody combination consistency parameter, disease pointing conflict parameter, standardized antibody strength value, historical offset features, and detection confidence set corresponding to the sample determined as a boundary sample by the antibody combination pattern recognition module. A boundary correction factor is generated based on the relevant antibody set corresponding to the j-th candidate disease category, the antibody combination consistency parameter, the disease pointing conflict parameter, the historical offset features, and the detection confidence set. The initial risk score for the j-th candidate disease category is corrected based on the boundary correction factor to obtain the corrected risk score; Specifically, when the antibody combination consistency parameter of the j-th candidate disease category reaches a preset threshold and the relevant antibody item shows a credible upward trend relative to the individual's historical baseline, the boundary correction factor is used to raise or maintain the risk score of the corresponding candidate disease category; when the disease-pointing conflict parameter reaches a preset threshold, or the relevant antibody item is within the boundary range and the detection credibility is lower than the preset threshold, the boundary correction factor is used to lower the risk score of the corresponding candidate disease category.

[0027] Preferably, the boundary sample determination criteria include at least one of the following: At least one antibody item associated with a candidate autoimmune disease has a standardized antibody strength value that is between the upper limit of negative and the threshold for positive judgment, or is in the vicinity of weak positive. The presence of only a single positive antibody, with insufficient auxiliary antibody features or clinical supporting features associated with the same candidate autoimmune disease; Two or more candidate autoimmune disease categories have similar initial risk scores, or the disease-pointing conflict parameter reaches a preset conflict threshold. The antibody test result was positive or weakly positive, but clinical symptoms, inflammatory markers, complement markers, or indicators of organ involvement did not provide corresponding support. As attached Figure 1 The diagram shown is a schematic representation of the architecture of the support system 10. The support system 10 includes a multi-channel antibody detection module 110, an information management module 120, a feature vector construction module 130, an antibody combination pattern recognition module 140, a boundary sample correction module 150, and a diagnostic support output module 160.

[0028] For example, the multi-channel antibody detection module 110 is used to perform parallel multi-item detection on serum, plasma, or other bodily fluid samples suitable for autoantibody detection from the subject to obtain basic detection data for subsequent antibody combination pattern recognition and boundary sample correction. The multi-channel antibody detection module 110 can be implemented using one or more of the following: multiplex immunochromatographic assay card, multi-channel microfluidic immunoassay chip, antigen-coated microarray chip, magnetic particle chemiluminescence detection component, fluorescence immunoassay component, or enzyme-linked immunosorbent assay component. The detection platform used by the multi-channel antibody detection module 110 is not limited to a single form. In preferred exemplary embodiments, any detection method capable of obtaining quantitative, semi-quantitative, or positive / negative detection results for multiple autoantibody items within the same sample or batch can be used as a specific implementation of the multi-channel antibody detection module 110.

[0029] In exemplary embodiments, as shown in the appendix Figure 2 As shown, the multi-channel antibody detection module 110 includes a sample input unit 112, a reagent reaction unit 114, a detection channel array 116, a signal acquisition unit 118, and a detection control unit 119. The sample input unit 112 receives the sample to be tested and distributes it to multiple detection channels according to a preset volume. The reagent reaction unit 114 enables the autoantibodies in the sample to specifically bind with immobilized antigens, labeled antigens, labeled secondary antibodies, or other immunoreaction reagents. The detection channel array 116 carries the detection reaction sites corresponding to different autoantibody items. The signal acquisition unit 118 acquires optical signals, electrochemical signals, chemiluminescent signals, fluorescence signals, or colorimetric signals generated by each detection channel. The detection control unit 119 controls the sample addition, incubation, washing, color development, reading, and data upload processes.

[0030] Preferably, the detection channel array 116 is provided with multiple antigen reaction sites, each corresponding to a different autoantibody item. Depending on the target disease screening scope, the autoantibody items may include one or more of the following: antinuclear antibody-related antibody profile, anti-double-stranded DNA antibody, anti-Sm antibody, anti-SSA antibody, anti-SSB antibody, anti-U1-RNP antibody, anti-Scl-70 antibody, anti-Jo-1 antibody, anti-CCP antibody, rheumatoid factor, anti-neutrophil cytoplasmic antibody-related antibody, anticardiolipin antibody, anti-β2 glycoprotein I antibody, or lupus anticoagulant-related detection indicators. For connective tissue disease auxiliary screening scenarios, it preferably includes at least anti-dsDNA, anti-Sm, anti-SSA, anti-SSB, anti-RNP, anti-Scl-70, anti-Centromere, anti-Jo-1, anti-CCP, and rheumatoid factor, to form an antibody combination basis capable of distinguishing various diseases, such as systemic lupus erythematosus, Sjögren's syndrome, systemic sclerosis, inflammatory myopathy, and rheumatoid arthritis.

[0031] For example, during the detection process, the information collected by the multi-channel antibody detection module 110 includes at least the original detection signal of each antibody item, the converted antibody detection value, the detection unit, the negative threshold, the weak positive threshold, the positive threshold, the detection batch number, the sample number, the detection time, the reagent batch number, the channel number, and the corresponding quality control signal. For immunochromatography or fluorescence immunoassay, it may also include the acquisition of the detection line signal intensity, the quality control line signal intensity, and the ratio of the detection line to the quality control line; for chemiluminescence or enzyme-linked immunosorbent assay, it may acquire the luminescence intensity, absorbance value, standard curve conversion result, and duplicate well differences; for microarray chip, it may acquire the image grayscale value, background subtraction value, spot consistency parameter, and array positioning information of each antigen spot.

[0032] The multi-channel antibody detection module 110 outputs information including a test result table for each autoantibody item, raw signal data before standardization, and test process data for quality assessment. The test result table includes at least the antibody item name, test value, result grade, and reference range, where the result grade can be set to negative, gray zone, weak positive, positive, or strong positive. For test results requiring subsequent algorithm processing, the multi-channel antibody detection module 110 can also convert the results of different test items into a unified format data packet and send it to the information management module 120 and the feature vector construction module 130. Thus, the support system 10 can further determine whether a sample is weakly positive, single positive, cross-positive, or other boundary sample based on the intensity differences, combination relationships, and gray zone status of multiple antibody items in subsequent steps, providing basic data for clinical diagnostic support output.

[0033] In an exemplary embodiment, the information management module 120 is used to uniformly receive, organize, verify, and structure the examinee's clinical information, historical test records, sample association information, and testing process information, so as to provide a reliable data foundation for subsequent feature vector construction, antibody combination pattern recognition, and boundary sample correction. The information management module 120 is not only used to input basic patient information, but also to standardize the management of auxiliary interpretation information other than antibody test results and to make a preliminary judgment on the usability of samples and testing processes.

[0034] For example, see attached Figure 3 As shown, the information management module 120 may include a patient information entry unit 122, a clinical related information collection unit 124, a historical test record retrieval unit 126, a sample related information management unit 128, and a quality verification unit 129.

[0035] For example, the patient information entry unit is used to input basic information such as the examinee's age, gender, department visited, sample number, test application time, and test purpose.

[0036] The clinical association information collection unit is used to receive clinical manifestations and laboratory indicators related to autoimmune diseases. Clinical manifestations may include symptom tags such as joint pain, morning stiffness, rash, photosensitivity, dry mouth, dry eyes, Raynaud's phenomenon, muscle weakness, thrombotic events, recurrent miscarriage, proteinuria, hematuria, fever, and fatigue. Laboratory indicators may include one or more of the following: C-reactive protein, erythrocyte sedimentation rate, complement C3, complement C4, immunoglobulin levels, complete blood count, urinalysis, liver and kidney function indicators, and coagulation-related indicators.

[0037] The historical test record retrieval unit is used to obtain the subject's historical test records during past tests and medical procedures, including test results / data. This data may include autoantibody results, inflammatory markers, complement markers, urine test results, previous diagnostic records, follow-up records, and medication information. These historical test records can originate from the hospital's laboratory information system, electronic medical record system, historical test database, or data tables manually entered by medical staff. The ultimate goal of the historical test record retrieval unit is to establish an individual historical baseline, such as the changes in the test values ​​of the same antibody item at multiple historical test time points, the downward trend of the same complement marker, and the fluctuation of the same inflammatory marker. An individual historical baseline is formed using the test values ​​of the same test item at multiple historical test time points. The current test value is then compared with the individual historical baseline to determine the direction and magnitude of change of the test item relative to its historical state, thereby obtaining a test trend. This test trend may include status markers such as continuous increase, continuous decrease, fluctuating changes, basically stable, or no historical record. The individual historical baseline is an important basis for subsequent boundary sample correction, especially applicable for distinguishing between different situations such as long-term stable weak positives, short-term continuous increases, or decreased antibody levels after treatment.

[0038] The sample association information management unit records information related to the sample to be tested, including sample type, sampling time, delivery time, receipt time, sample storage conditions, sample volume, centrifugation status, hemolysis status, lipemia status, jaundice interference status, sample barcode, batch number, reagent batch number, and testing equipment number. This information is used to determine whether the sample meets the testing requirements, to assess sample quality, and to trace the cause when test results are abnormal. For example, when multiple channels of a batch of samples show low signals or abnormal quality control signals, batch association analysis can be performed on the relevant test results based on the reagent batch number, equipment number, and testing time.

[0039] The quality verification unit is used to determine the reliability of sample association information and antibody detection process information. The quality verification may include one or more of the following: sample status verification, batch calibration, negative control validity assessment, positive control validity assessment, control line validity assessment, repeated detection difference assessment, channel background signal assessment, and abnormal signal marking. For immunochromatography or fluorescence immunoassay, the quality verification unit can determine whether the control line reaches an effective threshold, whether there is background contamination on the detection line, and whether the ratio of the detection line to the control line is within a reasonable range. For chemiluminescence or enzyme-linked immunosorbent assay (ELISA), the quality verification unit can determine whether the standard curve is effective, whether the difference between repeated wells exceeds the allowable range, and whether the positive and negative controls meet preset conditions. For antigen microarray chips, the quality verification unit can determine whether antigen sites are missing, whether the background grayscale is abnormal, and whether the consistency between repeated sites of the same antigen meets the requirements.

[0040] The information management module 120 ultimately outputs a structured information set. This structured information set includes at least the subject's basic information, clinical symptom labels, inflammation and complement-related indicators, test trends, sample status identifiers, batch identifiers, and quality verification results. Quality verification results may include status labels such as "available," "requires verification," and "not recommended for interpretation." The information management module 120 can also attach a primary confidence marker to the corresponding antibody test results; when the sample status or quality control results do not meet the minimum usability requirements, the information management module 120 prevents the sample's data from entering the subsequent antibody combination pattern recognition process and outputs a resampling or retesting prompt; when the sample meets the minimum usability requirements but has mild quality abnormalities, the primary confidence marker is retained and provided for reference by subsequent modules.

[0041] Through the above settings, the information management module 120 can associate the test results obtained by the multi-channel antibody detection module with the patient's clinical background, historical trends, and sample quality status, avoiding mechanical interpretation based solely on a single positive antibody result. Simultaneously, the information management module 120 can provide the boundary sample correction module with individual historical baselines, supporting clinical characteristics, and test reliability information, thereby improving the reliability of auxiliary interpretations for weakly positive, single-positive, and cross-positive samples.

[0042] Furthermore, in an exemplary embodiment, the feature vector construction module 130 receives the autoantibody detection results output by the multi-channel antibody detection module 110 and, in conjunction with the clinical correlation information and testing trends output by the information management module 120, constructs a multi-dimensional diagnostic feature vector for subsequent antibody combination pattern recognition and boundary sample correction. Since different antibody items have different detection units, reference intervals, positive and negative thresholds, and detection platforms, the feature vector construction module 130 first standardizes the original detection values ​​of each antibody item, enabling different detection items to participate in subsequent calculations on a unified scale.

[0043] For example, let the original detection value of the i-th autoantibody item be Org. i The corresponding upper limit threshold for negative is Ne. i The corresponding positive threshold is Pt. i Pt i Greater than Ne i The feature vector construction module 130 can be based on Org i Compared to Ne i and Pt i Calculate the normalized antibody strength value x for the i-th antibody item based on its position. i In one implementation, x i It can be determined in the following way: ; When x i A value less than 0 indicates that the antibody level is below the negative reference range; when x i When the value is between 0 and 1, it indicates that the antibody item is in the gray area or the weakly positive range between the upper limit of negative and the positive threshold; when x i A value greater than or equal to 1 indicates that the antibody item has reached or exceeded the positive threshold. In some exemplary embodiments, to avoid abnormally high values ​​from having an excessive impact on subsequent scoring, x can also be set... i An upper limit cutoff value can be set, or a piecewise function can be used to convert it into a level feature corresponding to negative, gray zone, weak positive, positive, and strong positive. In some exemplary embodiments, for antibody projects using semi-quantitative detection results, the feature vector construction module 130 can convert detection levels such as "-, ±, +, ++, +++" into preset numerical levels; for antibody projects that only output positive and negative results, negative can be set to 0, and positive can be set to 1. Thus, this support system can be compatible with different detection methods such as quantitative detection, semi-quantitative detection, and qualitative detection.

[0044] Furthermore, the feature vector construction module 130 forms an antibody strength sub-vector XA based on the standardized antibody strength values ​​of multiple antibody items, which is represented as follows: ; In the above formula, m represents the number of autoantibody items included in the test, and x1 to x m These represent the standardized antibody intensity values ​​for each autoantibody item. The antibody intensity subvector reflects the overall positive level and relative strength of each autoantibody item in the current sample of the subject.

[0045] Based on this, the feature vector construction module 130 further constructs antibody combination features. These antibody combination features represent whether multiple antibody items form a combination pattern associated with a specific autoimmune disease. For example, for systemic lupus erythematosus, anti-dsDNA antibody, anti-Sm antibody, decreased complement C3 / C4 ratio, and abnormal urinalysis can be considered as a set of combination features; for Sjögren's syndrome, anti-SSA antibody, anti-SSB antibody, ANA-related antibody, and symptoms such as dry mouth and dry eyes can be considered as a set of combination features; for rheumatoid arthritis, anti-CCP antibody, rheumatoid factor, inflammatory markers, and joint symptoms can be considered as a set of combination features. For the j-th candidate disease, a corresponding antibody combination set G can be set. j The combined response value C is calculated based on the normalized intensity values ​​of multiple antibody items within the set. j Combined response value C j It can be obtained using weighted average, weighted summation, or rule-based scoring methods, for example: ; Among them, G j Let G represent the set of antibody items associated with the j-th candidate disease, and i represent the set of antibody items G. j For any antibody item in x i ω represents the standardized antibody strength value of the i-th antibody item. ij C represents the combined weight of the i-th antibody item for the j-th candidate disease. j This represents the antibody combination response value corresponding to the j-th candidate disease.

[0046] After obtaining the antibody combination response values ​​corresponding to multiple candidate diseases, the feature vector construction module 130 combines them into an antibody combination feature vector XC, which is represented as follows: .

[0047] Simultaneously, the feature vector construction module 130 is also used to numerically process clinically relevant information. For example, clinical symptoms such as joint pain, rash, dry mouth, dry eyes, Raynaud's phenomenon, thrombotic events, and proteinuria can be converted into symptom label features; laboratory indicators such as C-reactive protein, erythrocyte sedimentation rate, complement C3, complement C4, urinary protein, and complete blood count can be converted into clinical indicator features based on whether they exceed the reference range and the degree of deviation. This forms the clinical support sub-vector XL, represented as follows: ; Where n represents the number of clinical association features included in the analysis, l1 to l n It can be used to represent numerical features corresponding to indicators such as symptom labels, inflammatory activity indicators, complement change indicators, or organ involvement indicators.

[0048] For historical test records, the feature vector construction module 130 is used to construct a historical trend sub-vector XH. Specifically, for the same antibody item or the same clinical indicator, the feature vector construction module 130 can read the test values ​​of the subject at multiple historical test time points and calculate the offset of the current test value relative to the individual's historical baseline. For example, let the historical baseline value of the i-th antibody item be Base. i The current standardized antibody strength value is x. i Then the historical offset feature h can be calculated. i : h i =x i -Base i ; When h i When h is positive and exceeds a preset change threshold, it indicates that the antibody item shows an increasing trend relative to the individual's historical baseline; when h i When the value is close to 0, it indicates that the antibody project is relatively stable; when h i A negative value indicates a decrease in the antibody level compared to historical levels. This historical shift characteristic is useful in distinguishing between weak positives that are long-term stable or have recently shown a sustained increase.

[0049] Furthermore, after obtaining the historical offset features corresponding to multiple antibody projects, the feature vector construction module 130 combines them into a historical trend sub-vector XH, which is expressed as follows: ; Where r represents the total number of antibody items included in the historical trend analysis, h1 to h r These represent the historical offset features corresponding to different antibody items. The historical trend sub-vector XH is used to characterize the overall change of the current test result relative to the individual's historical baseline.

[0050] Finally, the feature vector construction module 130 concatenates or fuses the antibody strength sub-vector XA, the antibody combination feature vector XC, the clinical support sub-vector XL, and the historical trend sub-vector XH to obtain the multidimensional diagnostic feature vector F, which is expressed as follows: F = [XA, XC, XL, XH]; Furthermore, in an exemplary embodiment, the antibody combination pattern recognition module 140 is used to receive the multidimensional diagnostic feature vector F output by the feature vector construction module 130, and to identify the degree of matching between the subject sample and different autoimmune disease-related antibody combination patterns based on the multidimensional diagnostic feature vector F. The antibody combination pattern recognition module 140 does not only judge the positive or negative results of a single antibody item, but is used to comprehensively analyze the co-occurrence relationship, intensity relationship, disease-directing relationship, and correspondence between multiple autoantibody items and their clinical supporting features, thereby obtaining an initial risk score for one or more candidate disease categories.

[0051] Specifically, the antibody combination pattern recognition module 140 pre-sets multiple candidate disease categories and their corresponding antibody combination templates. The candidate disease categories may include one or more of systemic lupus erythematosus, rheumatoid arthritis, Sjögren's syndrome, systemic sclerosis, inflammatory myopathy, antiphospholipid syndrome, and vasculitis-related autoimmune diseases. Each candidate disease category corresponds to one or more antibody combination templates, which represent characteristic antibody items, auxiliary antibody items, clinical support items, and exclusion indication items related to the disease.

[0052] For example, for systemic lupus erythematosus (SLE), antibody combination templates may include features such as anti-dsDNA antibodies, anti-Sm antibodies, ANA-related antibodies, decreased complement C3 / C4 ratio, proteinuria, or hematuria; for rheumatoid arthritis, antibody combination templates may include features such as anti-CCP antibodies, rheumatoid factor, elevated erythrocyte sedimentation rate, elevated C-reactive protein, and joint pain or morning stiffness; for Sjögren's syndrome, antibody combination templates may include features such as anti-SSA antibodies, anti-SSB antibodies, ANA-related antibodies, dry mouth, and dry eyes; for systemic sclerosis, antibody combination templates may include anti-Scl-70 antibodies, anti-Centromere antibodies, Raynaud's phenomenon, and skin sclerosis-related symptoms; for inflammatory myopathy, antibody combination templates may include anti-Jo-1 antibodies, muscle weakness symptoms, and abnormal muscle enzymes; for antiphospholipid syndrome, antibody combination templates may include features such as anticardiolipin antibodies, anti-β2 glycoprotein I antibodies, lupus anticoagulant-related markers, thrombotic events, or a history of recurrent miscarriage.

[0053] In some exemplary embodiments, let the antibody combination template corresponding to the j-th candidate disease category be denoted as . The antibody combination template includes several feature items, which may be derived from the antibody strength sub-vector XA, the antibody combination feature vector XC, the clinical support sub-vector XL, and the historical trend sub-vector XH. The antibody combination pattern recognition module 140 can extract the feature values ​​corresponding to the antibody combination template from the multidimensional diagnostic feature vector F to calculate the pattern matching score M between the subject sample and the j-th candidate disease category. j .

[0054] In one exemplary calculation method, the pattern matching score M j It can be calculated using the following formula: ; Among them, f k α represents the k-th feature value in the multidimensional diagnostic feature vector F that is related to the j-th candidate disease category. jk This represents the matching weight of the feature value in the j-th candidate disease category, z. j This represents the number of feature terms involved in the matching calculation in the antibody combination template corresponding to the j-th candidate disease category. The matching weight α jk The weighting can be determined based on medical rules, historical sample database statistics, expert-annotated samples, or trained classification models. Higher weights can be assigned to antibody items with high disease specificity, while lower weights can be assigned to non-specific antibody items or items prone to background positivity.

[0055] Furthermore, to avoid a single strong positive antibody having an excessive impact on the overall recognition result, the antibody combination pattern recognition module 140 is also used to set the combination consistency parameter U. j The combined consistency parameter U j This is used to indicate whether multiple related features jointly point to the same candidate disease category. For the j-th candidate disease, when its core antibody project, auxiliary antibody project, and clinical support project simultaneously meet preset conditions, U... j Take the higher value; when only a single antibody is positive and there is a lack of isoantibody or clinical characteristics to support it, U j Take the lower value. Thus, the support system can distinguish between single, occasional positive results and consistent positive results across multiple indicators.

[0056] In some exemplary implementations, the combined consistency parameter U j It can be calculated separately based on core antibody characteristics, helper antibody characteristics, and clinical supporting characteristics. The following formula is used: ; in, ; In the above formula, function A(S) represents the proportion of feature terms in set S that meet the preset response conditions. Function I() is an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise; |S| represents the number of features in set S.

[0057] U j P represents the combinatorial consistency parameter for the j-th candidate disease; j Let Q represent the set of core antibody features corresponding to the j-th candidate disease. j This represents the set of auxiliary antibody features corresponding to the j-th candidate disease, and under normal circumstances, G can be considered as... j =P j ∪Q j ;RC j This represents the set of clinical supporting features corresponding to the j-th candidate disease. The selection of each feature can be determined based on the clinical diagnostic criteria, guidelines, consensus, antibody specificity and sensitivity, statistical results of historical confirmed samples, and expert annotation rules for the j-th candidate disease. Generally, antibody items with high disease specificity and strong distinguishing effect on disease classification are included in the core antibody feature set P, antibody items with auxiliary suggestive significance or high sensitivity but relatively low specificity are included in the auxiliary antibody feature set Q, and symptoms and test indicators related to the clinical manifestations, inflammatory activity, or organ involvement of the candidate disease are included in the clinical supporting feature set RC.

[0058] θ i The effect threshold corresponding to the i-th feature can be set based on the medical reference range, positive determination threshold, gray zone threshold, statistical distribution of historical confirmed samples, or expert annotation rules for the i-th feature; for example, for antibody projects, θ i It can correspond to a weak positive threshold or a positive threshold. For clinical supporting features, θ i It can correspond to the presence of symptoms, abnormal thresholds of test indicators, or thresholds indicating organ involvement.

[0059] λ1, λ2, and λ3 represent the weighting coefficients of core antibody features, auxiliary antibody features, and clinical supporting features in the combination consistency calculation, respectively. These three weighting coefficients can be determined based on the medical diagnostic rules for the target candidate disease, the statistical differences between confirmed and non-confirmed cases in the historical sample database, and the sensitivity, specificity, positive predictive value, or importance of each feature set or model feature. In a preferred embodiment, λ1, λ2, and λ3 satisfy λ1 > λ2 > λ3, ensuring that the core antibody features with high disease specificity contribute significantly to the combination consistency calculation, the auxiliary antibody features contribute secondarily, and the clinical supporting features are used to assist in verifying the antibody combination results. For early screening, review and interpretation, or specific disease categories, the weighting coefficients can also be adjusted based on historical sample training results or clinical rules, and are not limited to fixed values.

[0060] The antibody combination pattern recognition module 140 can also calculate the disease-directed conflict parameter V. j The disease-related conflict parameter V j This parameter is used to characterize the degree to which different antibody items or combinations in the current sample competitively point to multiple candidate disease categories. In other words, if the main positive antibody, weak positive antibody, or gray zone antibody in the current sample collectively supports the same candidate disease, the disease-pointing conflict parameter corresponding to that candidate disease is low. If multiple antibody items in the current sample support different candidate diseases, and the initial risk scores of each candidate disease are relatively similar, it indicates that the sample may have cross-positive results, overlap syndromes, non-specific antibody positivity, or fluctuations in the detection boundary, in which case the disease-pointing conflict parameter increases.

[0061] Specifically, the antibody combination pattern recognition module 140 can first establish an antibody item-candidate disease mapping table based on the preset association between each antibody item and the candidate disease category. For example, anti-dsDNA antibody and anti-Sm antibody are highly associated with systemic lupus erythematosus; anti-CCP antibody and rheumatoid factor are highly associated with rheumatoid arthritis; anti-SSA antibody and anti-SSB antibody can be used to support the interpretation of Sjögren's syndrome or some systemic lupus erythematosus; anti-Scl-70 antibody and anti-Centromere antibody are highly associated with systemic sclerosis; anti-Jo-1 antibody is associated with inflammatory myopathy; and anti-cardiolipin antibody, anti-β2 glycoprotein I antibody, and lupus anticoagulant-related indicators are associated with antiphospholipid syndrome. The system uses this mapping table to determine whether the positive or weakly positive antibodies in the current sample fall into the same candidate disease template or are scattered across multiple candidate disease templates.

[0062] In one implementation, when the pattern matching score M of the j-th candidate disease... j The pattern matching scores for other candidate diseases were significantly lower than those for M. j When the main positive antibodies all belong to the core antibody set or auxiliary antibody set of the j-th candidate disease, the antibody combination pattern recognition module 140 determines that the sample has a low disease-targeting conflict for the j-th candidate disease. Conversely, when the pattern matching scores of two or more candidate diseases all reach the preset attention threshold, or when the difference between the highest pattern matching score and the second highest pattern matching score is less than the preset difference threshold, the system determines that the current sample has an unclear disease target and increases the disease-targeting conflict parameter of the corresponding candidate disease.

[0063] For example, when a subject is positive for both anti-SSA and anti-SSB antibodies, and also experiences dry mouth and dry eyes, while other disease-related antibodies such as anti-Scl-70, anti-CCP, and anti-dsDNA are negative or have low responses, the current sample primarily points to Sjögren's syndrome, with low disease-related conflict parameters. As another example, when a subject is positive for anti-SSA antibodies, accompanied by weakly positive anti-dsDNA antibodies and decreased complement levels, the sample may simultaneously support Sjögren's syndrome and systemic lupus erythematosus (SLE). Further differentiation is needed based on clinical symptoms, abnormal urinalysis results, and historical antibody trends; in this case, the disease-related conflict parameters are elevated. Furthermore, when a subject is positive for anti-SSA antibodies, and also shows weakly positive anti-Scl-70 and anti-CCP antibodies, but clinical symptoms are limited to fatigue or nonspecific joint discomfort, failing to clearly support any of Sjögren's syndrome, systemic sclerosis, or rheumatoid arthritis, the system can consider the sample to have significant disease-related conflict.

[0064] Disease-related conflict parameter V j Adjustments can also be made based on the intensity of antibody positivity. For core antibodies with strong positivity and high disease specificity, their targeting effect on the corresponding candidate disease is strong; for antibody items with weak positivity, gray zone positivity, or high non-specificity, their disease-targeting effect is weaker, but may increase the risk of cross-disease conflicts. Therefore, the antibody combination pattern recognition module 140 can set conflict contribution weights for different antibody items. When a weakly positive antibody does not belong to the same antibody combination template as the current highest-risk candidate disease, and the initial score of the other candidate disease it targets also reaches the attention threshold, the conflict contribution corresponding to the weakly positive antibody is included in the disease-targeting conflict parameter; when the weakly positive antibody lacks clinical feature support, its conflict contribution can be reduced to avoid over-interpreting occasional weak positivity as a real disease conflict.

[0065] In one alternative implementation, the disease-pointing conflict parameter V j The difference in candidate disease scores and the degree of heterologous antibody response can be jointly determined. For example, the system can calculate the similarity between the pattern matching scores of other candidate diseases (excluding the j-th candidate disease) and the pattern matching score of the j-th candidate disease, and combine this with the number or intensity of antibody items that do not belong to the j-th candidate disease template but meet the response threshold to generate disease-pointing conflict parameters. Therefore, the closer the scores of other candidate diseases are to the j-th candidate disease score, and the more or stronger the heterologous positive antibodies, the higher V becomes. j The higher the score, the better; when other candidate disease scores are low and there are few heterologous positive antibodies, the higher the V score. j The lower.

[0066] Therefore, the disease-pointing conflict parameter is not used to directly exclude candidate diseases, but rather serves as an important input to the subsequent boundary sample correction module. When Vj When the level is high, the system can mark the sample as a cross-positive sample or a sample to be retested, and in subsequent steps, combine the individual's historical baseline and the antibody combination concordance parameter U. j In addition, the system adjusts the initial risk score based on clinical characteristics and supporting evidence. By setting disease-specific conflict parameters, the system can avoid directly outputting a single high-risk disease category due to weak positive results for a single antibody or scattered positive results for multiple non-specific antibodies, thereby improving the stability and interpretability of auxiliary interpretation of autoimmune disease antibody test results.

[0067] In obtaining the pattern matching score M j Combinatorial consistency parameter U j Conflicting parameter V with disease indication j Subsequently, the antibody combination pattern recognition module 140 can generate an initial risk score R for the j-th candidate disease category. j In one implementation, the initial risk score R j It can be calculated as follows: ; Wherein, β1, β2, and β3 represent the adjustment coefficients corresponding to the pattern matching score, the combined consistency parameter, and the disease-directed conflict parameter, respectively. These adjustment coefficients can be statistically set based on confirmed samples, non-confirmed samples, and follow-up confirmed samples in a historical sample database, or determined through logistic regression, grid search, cross-validation, or expert rule assignment. In a preferred embodiment, with the goal of improving the consistency between candidate disease risk scores and the final clinical diagnosis, β1, β2, and β3 are trained or calibrated so that the pattern matching score and the combined consistency parameter contribute positively to the risk score, while the disease-directed conflict parameter contributes negatively to the risk score. The effect of the above formula is that when M... j The higher the value, the closer the current sample is to the antibody combination template of the j-th candidate disease; U j The higher the value, the stronger the concordance between the relevant antibody and clinical characteristics; V j A higher value indicates a more significant risk of multiple disease-related conflicts or cross-positive results in the current sample. Therefore, when calculating the initial risk score, the pattern matching score and the combined consistency parameter have a positive effect on the risk score, while the disease-related conflict parameter has a negative corrective effect on the risk score.

[0068] In another implementation, the antibody combination pattern recognition module 140 can also use a trained classification model to initially identify candidate disease categories. The classification model can be one or more of the following: logistic regression, random forest, gradient boosting tree, support vector machine, Naive Bayes, or neural network. In this case, the multidimensional diagnostic feature vector F is used as the model input, and the model outputs the initial probability value or risk score for each candidate disease category. To enhance the interpretability of the model output, the antibody combination pattern recognition module 140 preferably outputs features that contribute significantly to the risk score simultaneously, such as specific positive antibody items, supporting clinical features, complement abnormality features, or historical elevation trend features, rather than simply outputting the disease name or probability value.

[0069] The antibody combination pattern recognition module 140 ultimately outputs candidate disease identification results. These results include at least one or more candidate disease categories and an initial risk score R corresponding to each category. j Pattern matching score M j Combinatorial consistency parameter U j and the disease-indicating conflict parameter V j For samples with an initial risk score significantly higher than other candidate disease categories and strong combinatorial consistency, the system can mark them as samples with a clear propensity.

[0070] In an exemplary embodiment, on the other hand, when a test sample meets one or more of the following conditions, the antibody combination pattern recognition module 140 can identify the test sample as a boundary sample and transmit the relevant parameters detected by the sample to the boundary sample correction module 150: At least one antibody item associated with the candidate disease has a standardized antibody strength value x that is between the upper limit of negative and the threshold of positive, or is in the vicinity of weak positive. The presence of only a single positive antibody, with insufficient auxiliary antibodies or clinical supporting features associated with the same candidate disease; Two or more candidate disease categories have similar initial risk scores, or the disease-pointing conflict parameter V j The preset conflict threshold has been reached; The antibody test result was positive or weakly positive, but the clinical symptoms, inflammatory markers, complement markers, or organ involvement did not provide corresponding support.

[0071] Through the above settings, the antibody combination pattern recognition module 140 can convert the multidimensional diagnostic feature vector F into an initial recognition result with disease orientation. This module does not directly complete the final diagnosis, but rather extracts disease-related combination patterns from multiple autoantibody items and clinical supporting information to obtain candidate disease categories and their initial risk scores.

[0072] Example 2: This example should be understood as including at least all the features of any of the foregoing examples, and further improving upon them; Furthermore, in a preferred embodiment, the boundary sample correction module 150 is used to perform secondary interpretation and dynamic correction for samples with weak positive, single positive, cross-positive, or antibody results inconsistent with clinical manifestations, thereby improving the reliability of the overall auxiliary interpretation results of the system.

[0073] In an exemplary embodiment, the boundary sample correction module 150 is used to perform secondary correction on the initial identification results of candidate diseases output by the antibody combination pattern recognition module 140. The boundary sample correction module 150 does not need to perform the same risk adjustment on all samples. Instead, it receives the boundary samples identified by the antibody combination pattern recognition module 140 and their associated parameter data, and then initiates a boundary state correction step to avoid the system outputting unstable disease risk conclusions based solely on fixed thresholds or single antibody results.

[0074] Furthermore, in an exemplary embodiment, to provide the boundary sample correction module 150 with more confidence information about the detected samples, the feature vector construction module 130 includes calculating a confidence coefficient q for the corresponding antibody item based on the primary confidence marker, sample status information, detection batch information, quality control signal, repeated detection difference, and channel background signal provided by the information management module 120; for the detection confidence of the i-th antibody item, the feature vector construction module 130, after statistically analyzing the above confidence factors, records its confidence value as q. i And form a set Q, that is: Q={q1,q2,…,q i …}; The value of q can range from 0 to 1. The closer the q value is to 1, the more reliable the antibody test result is; the closer the q value is to 0, the less suitable the antibody test result is for disease risk assessment.

[0075] Furthermore, the boundary sample correction module 150 calculates the boundary correction factor B for the j-th candidate disease based on the combination consistency of candidate diseases, disease targeting conflicts, antibody historical shifts, and detection confidence. j And according to the boundary correction factor B j The initial risk score R is corrected to obtain the corrected risk score R0. j ´。 Among them, B j It can be calculated as follows: ; The corrected risk score is obtained as follows: ; In the above two equations, R j ´ represents the risk score of the j-th candidate disease after boundary sample correction; U j V represents the combinatorial consistency parameter for the j-th candidate disease; j G represents the disease-pointing conflict parameter for the j-th candidate disease; j Let G represent the set of antibodies corresponding to the j-th candidate disease. In the subsequent boundary sample correction process, let G be... j =P j ∪Q j ;∣G j | represents the number of antibody items in the set; q i h represents the reliability coefficient of the detection of the i-th antibody item; i This represents the historical offset characteristic of the i-th antibody item relative to the individual's historical baseline; x i Represents the standardized antibody strength value of the i-th antibody item; function I(0 <x i <1) is an indicator function, that is, the function takes the value of 1 when the i-th antibody item is in the boundary range between the negative upper limit and the positive threshold, and takes the value of 0 otherwise.

[0076] The above formula serves to determine the consistency parameter U of the combination of the j-th candidate disease. j When the level is high and the candidate disease-related antibody shows a credible upward trend relative to the individual's historical baseline, the boundary correction factor B... j Increase, thus correcting the risk score R j ´ Higher than or close to the initial risk score R j Conversely, when the disease points to the conflicting parameter V j When the value is high, or when the relevant antibody item is at the boundary and the detection reliability is low, the boundary correction factor B... j Decrease, thus correcting the risk score R j The value was lowered. Therefore, the system can distinguish between weakly positive samples with a historical upward trend and supporting evidence, and weakly positive samples lacking supporting evidence, insufficient detection reliability, or scattered disease indications.

[0077] Furthermore, the molecular part max(h) i (0) is used to extract only the portion that is elevated relative to the historical baseline. When an antibody item is elevated compared to an individual's historical baseline, it has a positive effect on the risk correction of the corresponding candidate disease; when the antibody item is not elevated or is lower than the historical baseline, the risk score is not improved by that item. This avoids incorrectly increasing the disease risk due to long-term stable weak positives or a downward trend after treatment. (1-q) i )×I(0 <x i <1) is used to represent the weakening effect of detection items that are at the boundary and lack credibility on the risk score, where q iThe lower the value, the greater the likelihood that the test result is affected by the sample condition or the testing process; if the item is also within the boundary range, its result should not be over-interpreted as a definitive positive.

[0078] Furthermore, in an exemplary embodiment, the diagnostic support output module 160 is used to receive the candidate disease identification results output by the antibody combination pattern recognition module 140, and the corrected risk score R output by the boundary sample correction module 150. j Based on the candidate disease identification results and the corrected risk score, diagnostic support results are generated for medical personnel or testing personnel to assist medical personnel in diagnosing the disease.

[0079] Specifically, the diagnostic support output module 160 can output the initial risk score R corresponding to each candidate disease. j Or revise the risk score R j The risk level of the candidate disease is determined. When the sample is not marked as a boundary sample, the diagnostic support output module 160 can directly use the initial risk score R output by the antibody combination pattern recognition module 140 to determine the risk level of the candidate disease. j Risk stratification is performed; when a sample is marked as a boundary sample, the diagnostic support output module 160 prioritizes the modified risk score R. j Risk stratification is performed. This avoids the system directly using initial identification results to output conclusions for boundary samples.

[0080] In one implementation, the diagnostic support output module 160 can classify the risk level of candidate diseases into low risk, borderline risk, medium risk, and high risk. Low risk indicates that the current antibody combination pattern, clinical support information, and historical trends do not show significant support for the corresponding candidate disease; borderline risk indicates the presence of weak positives, single positives, gray zone results, or mild clinical support, but not enough to form a clear disease indication; medium risk indicates that multiple antibody items, clinical support characteristics, or historical trends show some consistency with the corresponding candidate disease; high risk indicates that there is high consistency among the core antibodies, auxiliary antibodies, clinical support characteristics, and historical trends related to the candidate disease, or that the revised risk score is significantly higher than other candidate disease categories. These risk levels are used to indicate the level of attention required by medical personnel, rather than replacing clinical diagnostic conclusions.

[0081] The diagnostic support output module 160 is also used to generate result interpretation information. This interpretation information includes at least the candidate disease category, risk level, major positive antibody item, gray zone or weakly positive antibody item, supporting clinical features, historical trends, whether it belongs to the boundary sample, and the basis for boundary sample correction. For example, when the risk level of a candidate disease is upgraded, the output may include statements such as "high consistency of related antibody combinations," "related antibodies show an increasing trend compared to the individual's historical baseline," or "clinical symptoms or complement indicators support the disease." When the risk level of a candidate disease is downgraded, the output may include statements such as "only a single weak positive exists," "lack of support from same group antibodies," "cross-disease antibody pointing conflict," "insufficient test reliability," or "insufficient support from clinical features." In this way, medical personnel can not only see the risk level but also understand the main basis for the system to form that risk level.

[0082] Furthermore, the diagnostic support output module 160 can output further examination suggestions based on risk level and borderline sample type. For high-risk samples, suggestions can be given such as referring to the rheumatology and immunology department for further evaluation, supplementing specific antibody testing, and making a comprehensive judgment based on imaging or organ function tests. For medium-risk samples, suggestions can be given such as repeating relevant antibody tests, supplementing complement, inflammatory markers, urinalysis, or testing for related organ involvement indicators. For borderline risk samples, suggestions can be given such as repeating the same antibody test after a preset follow-up period, supplementing with the same group of antibodies, verifying sample status, or resampling and testing. For low-risk samples, suggestions can be given such as that no obvious antibody combination supporting the corresponding candidate disease has been found, but follow-up can be decided based on clinical manifestations.

[0083] The output of the diagnostic support output module 160 can include test reports, graphical interfaces, structured data tables, or data packages that interface with a hospital information system. The test report may include basic information about the examinee, sample information, test results for each antibody item, risk level of candidate diseases, boundary sample markings, result interpretation, and recommended examinations. The graphical interface can display the risk levels and related evidence for different candidate diseases in the form of lists, radar charts, risk bars, color-coded indicators, or time trend graphs. The structured data table may include candidate disease number, candidate disease name, risk score, risk level, major contributing antibody, clinical supporting information, correction status, and recommended re-examination items for subsequent storage, retrieval, or follow-up comparison. (See attached...) Figure 4 The image shows an exemplary user interface of the support system.

[0084] Example 3: This example should be understood as including at least all the features of any of the foregoing examples, and further improving upon them; For example, as shown in the appendix Figure 4 The following diagram illustrates the implementation of the computer system 500 used in the support system 10; the computer system 500 can be applied to the data storage, calculation, and result output processes of each working module in the identification and judgment system.

[0085] For example, computer system 500 includes bus 502 or other communication mechanism for transmitting information, and one or more processors 504 coupled to bus 502 for processing information; processor 504 may be, for example, one or more general-purpose microprocessors. Computer system 500 also includes main memory 506, such as random access memory (RAM), cache and / or other dynamic storage devices, coupled to bus 502 for storing information and instructions to be executed by processor 504; main memory 506 can also be used to store temporary variables or other intermediate information during the execution of instructions executed by processor 504; when these instructions are stored in storage media accessible to processor 504, they present computer system 500 as a dedicated machine customized to perform the operations specified in the instructions; The computer system 500 may also include a read-only memory (ROM) 508 or other static storage device coupled to the bus 502 for storing static information and instructions of the processor 504; among which, storage devices 510 such as disks, optical disks or USB drives (flash drives) will be coupled to the bus 502 for storing information and instructions. Furthermore, the bus 502 may also include a display 512 for displaying various information, data, media, etc., and an input device 514 for allowing users of the computer system 500 to control, manipulate, and / or interact with the computer system 500. A preferred method of interacting with the management system may be through a cursor control device 516, such as a computer mouse or a similar control / navigation mechanism; Furthermore, the computer system 500 may also include a network device 518 coupled to the bus 502; wherein the network device 518 may include components such as wired network cards, wireless network cards, switching chips, routers, switches, etc. Generally speaking, the terms “engine,” “component,” “system,” and “database” used in this article can refer to the logic embodied in hardware or firmware, or to a set of software instructions that may have entries and exit points, written in programming languages ​​such as Java, C, or C++; software components can be compiled and linked into executable programs, installed in dynamic link libraries, or written in interpreted programming languages ​​such as BASIC, Perl, or Python; it should be understood that software components can be called from other components or from themselves, and / or can be called in response to detected events or interrupts; Software components configured to execute on a computing device may be provided on computer-readable media, such as optical discs, digital video discs, flash drives, magnetic disks, or any other tangible media, or as digital downloads (and may be initially stored) in compressed or installable formats that require installation, decompression, or decryption prior to execution; such software code may be stored, in part or in whole, on a memory device executing the computing device; software instructions may be embedded in firmware, such as EPROM; it should also be understood that hardware components may consist of connected logic units (e.g., gates and flip-flops), and / or may consist of programmable units (e.g., programmable gate arrays or processors); Computer system 500 includes technologies described herein that can be implemented using custom hardwired logic, one or more ASICs or FPGAs, firmware and / or program logic, which, when combined with the computer system, enables computer system 500 to become a dedicated computing device. According to one or more embodiments, the techniques described herein are executed by computer system 500 in response to processor 504 executing one or more sequences of one or more instructions contained in main memory 506; such instructions may be read into main memory 506 from another storage medium such as storage device 510; execution of the sequence of instructions contained in main memory 506 causes processor 504 to perform the processing steps described herein; in alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions. As used herein, the term "non-transitory medium" and similar terms refer to any medium that stores data and / or instructions that enable a machine to operate in a particular manner; such non-transitory medium may include non-volatile medium and / or volatile medium; non-volatile medium includes, for example, optical discs or magnetic disks, such as storage device 510; volatile medium includes dynamic memory, such as main memory 506. Common forms of non-transitory media include, for example, floppy disks, hard disks, solid-state drives, magnetic tapes or any other magnetic data storage media, CD-ROMs, any other optical data storage media, any physical media with a hole pattern, RAM, PROM and EPROM, FLASH-EPROM, NVRAM, any other memory chips or cartridges and their network versions. Non-transient media are different from transmission media, but can be used in conjunction with transmission media; transmission media participate in information transmission between non-transient media; for example, transmission media include coaxial cables, copper wires and optical fibers, including the wires that constitute bus 502; transmission media can also take the form of sound waves or light waves, such as radio waves and infrared data communication.

[0086] While the invention has been described above with reference to various embodiments, it should be understood that many changes and modifications can be made without departing from the scope of the invention. That is, the methods, systems, and devices discussed above are examples. Various configurations can be appropriately omitted, substituted, or added to various processes or components. For example, in alternative configurations, methods can be performed in a different order than those described, and / or various components can be added, omitted, and / or combined. Moreover, features described with respect to certain configurations can be combined in various other configurations, such as different aspects and elements of the configuration can be combined in a similar manner. Furthermore, the elements therein can be updated as the technology develops; that is, many elements are examples and do not limit the scope of this disclosure or the claims.

[0087] Specific details are provided in the specification to offer a thorough understanding of exemplary configurations, including implementations. However, configurations can be practiced without these specific details; for example, well-known circuits, processes, algorithms, structures, and techniques have been shown without unnecessary detail to avoid obscuring the configuration. This description provides only exemplary configurations and does not limit the scope, applicability, or configuration of the claims. Rather, the foregoing description of the configurations will provide those skilled in the art with an enabling description for implementing the described techniques. Various changes can be made to the function and arrangement of the elements without departing from the spirit or scope of this disclosure.

[0088] In summary, the above detailed description is intended to be illustrative rather than restrictive, and it should be understood that these embodiments are for illustrative purposes only and not for limiting the scope of protection of the invention. After reading the description of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent changes and modifications also fall within the scope defined by the claims of this invention.

Claims

1. A system for detecting and clinically diagnosing antibodies in autoimmune diseases, characterized in that, The support system includes: a multi-channel antibody detection module, an information management module, a feature vector construction module, an antibody combination pattern recognition module, a boundary sample correction module, and a diagnostic support output module; The multi-channel antibody detection module is used to perform parallel detection of multiple antibody items on the body fluid sample of the subject and obtain the detection results of each antibody item. The information management module is used to acquire the examinee's clinical correlation information, historical test records, sample correlation information, and testing process information, and to perform basic quality screening on the test samples based on the sample correlation information and testing process information; The feature vector construction module is used to construct a multidimensional diagnostic feature vector based on the detection results, clinical correlation information, and historical test records; the multidimensional diagnostic feature vector includes an antibody strength sub-vector, an antibody combination feature vector, a clinical support sub-vector, and a historical trend sub-vector; The antibody combination pattern recognition module is used to identify the degree of matching between the subject sample and different autoimmune disease-related antibody combination patterns based on the multidimensional diagnostic feature vector, generate candidate disease categories and their initial risk scores, and determine whether the subject sample is a boundary sample based on preset boundary sample determination conditions. The boundary sample correction module is used to correct the initial risk score of a boundary sample when the detected sample is identified as a boundary sample, based on the antibody combination consistency parameter, disease pointing conflict parameter, historical offset characteristics, and detection confidence set of the boundary sample, to obtain a corrected risk score. The diagnostic support output module is used to output diagnostic support results based on the initial risk score when the test sample is not identified as a borderline sample; and to output diagnostic support results based on the revised risk score when the test sample is identified as a borderline sample; wherein the diagnostic support results include the suspected disease category, risk level, and further examination recommendations.

2. The support system as described in claim 1, characterized in that, The information management module further includes performing the following operations: when the test sample does not meet the minimum usability requirements, preventing the test sample from entering the subsequent antibody combination pattern recognition process; when the test sample meets the minimum usability requirements, adding a primary confidence mark to the antibody test result of the corresponding test sample.

3. The support system as described in claim 1, characterized in that, The feature vector construction module, when constructing multidimensional diagnostic feature vectors, includes the following steps: Calculate the standardized antibody strength value x of the i-th antibody item based on the original detection value A of the i-th antibody item. i And composed of multiple standardized antibody strength values ​​x i This constitutes the antibody strength subvector XA; Based on the set of antibody items related to the j-th candidate disease and their corresponding standardized antibody strength values, an antibody combination feature vector XC is formed. The clinical association information is converted into a clinical support subvector XL; Based on the offset of the current detection value of the same detection item relative to the individual's historical baseline, a historical trend sub-vector XH is formed; The antibody strength sub-vector, antibody combination feature vector, clinical support sub-vector, and historical trend sub-vector are concatenated or fused to obtain the multidimensional diagnostic feature vector.

4. The support system as described in claim 1, characterized in that, The feature vector construction module is further configured to: generate a detection confidence set Q={q1,q2,…,q...} based on one or more data information output by the information management module, including the initial confidence marker, sample status, detection batch information, quality control signal, repeated detection difference, and channel background signal. i ,…,q m }, q i denoted as the confidence coefficient of the detection result of the i-th antibody item, and m represents the number of antibody items included in the detection.

5. The support system as described in claim 1, characterized in that, The boundary sample correction module is configured as follows: The system receives the candidate disease category, initial risk score, antibody combination consistency parameter, disease pointing conflict parameter, standardized antibody strength value, historical offset features, and detection confidence set corresponding to the sample determined as a boundary sample by the antibody combination pattern recognition module. A boundary correction factor is generated based on the relevant antibody set corresponding to the j-th candidate disease category, the antibody combination consistency parameter, the disease pointing conflict parameter, the historical offset features, and the detection confidence set. The initial risk score for the j-th candidate disease category is corrected based on the boundary correction factor to obtain the corrected risk score; Wherein, when the antibody combination consistency parameter of the j-th candidate disease category reaches a preset threshold and the relevant antibody item shows a credible upward trend relative to the individual's historical baseline, the boundary correction factor is used to adjust or maintain the risk score of the corresponding candidate disease category. When the disease-pointing conflict parameter reaches a preset threshold, or when the relevant antibody item is within the boundary range and the detection confidence is lower than the preset threshold, the boundary correction factor is used to lower the risk score of the corresponding candidate disease category.

6. The support system as described in claim 1, characterized in that, The boundary sample determination conditions include at least one of the following: At least one antibody item associated with a candidate autoimmune disease has a standardized antibody strength value that is between the upper limit of negative and the threshold for positive judgment, or is in the vicinity of weak positive. The presence of only a single positive antibody, with insufficient auxiliary antibody features or clinical supporting features associated with the same candidate autoimmune disease; Two or more candidate autoimmune disease categories have similar initial risk scores, or the disease-pointing conflict parameter reaches a preset conflict threshold. The antibody test result was positive or weakly positive, but the clinical symptoms, inflammatory markers, complement markers, or organ involvement indicators did not provide corresponding support.

Images