Serum specimen typing identification system based on patient centralized detection

Abstract

The invention relates to the technical field of serum specimen typing identification, in particular to a serum specimen typing identification system based on patient centralized detection. Comprising a data acquisition module, a data processing module, a data analysis module and a typing identification module, the data acquisition module is used for collecting and recording serum index data of a patient; according to the method, serum index data, including C-reactive protein, glucose, lactic dehydrogenase, TNF-alpha, IgG and IgM, of the patient are collected, an index matrix is constructed, and homogenization treatment is performed, so that a serum index distribution map of the patient is formed, the relationship between serum index values of the patient is conveniently and intuitively reflected, the synergistic reaction effect of each index on inflammation is reflected, and the sensitivity of the patient is improved. Meanwhile, a serum specimen of a new patient is subjected to typing, disease indexes are calculated, and medical staff can conveniently carry out targeted treatment on the patient.

Description

Technical Field

[0001] The invention relates to the technical field of serum specimen typing and identification, in particular to a serum specimen typing and identification system based on centralized patient testing. Background Art

[0002] With the continuous development of society and the evolution of lifestyle, the incidence of inflammation in the population is gradually increasing. Serum testing is a technology widely used in medical diagnosis and health assessment. Through serum testing, a lot of physiological and pathological information can be obtained. Serum testing can obtain indicators such as C-reactive protein, glucose, lactate dehydrogenase, TNF-α, IgG and IgM. Based on the analysis of these indicators, the inflammation status of the patient population can be revealed, which is convenient for medical personnel to carry out targeted treatment for patients. The serum sample analysis system in the prior art can generally only directly measure the index values ​​of serum, but cannot analyze and display the relationship between the serum index values ​​and the synergistic response of each index to inflammation. It is also impossible to type the serum samples of new patients and classify them into existing patient groups for targeted treatment, and it is also impossible to calculate the disease index. In summary, how to design a serum specimen typing and identification system that can reflect the relationship between serum index values, reflect the synergistic response of various indicators to inflammation, type serum specimens of new patients, and calculate disease indicators has become a problem to be solved. Summary of the invention

[0003] The purpose of the present invention is to provide a serum specimen typing and identification system based on centralized patient testing to solve the problems existing in the prior art.

[0004] In order to achieve the above objectives, the following technical solutions are provided: A serum specimen typing and identification system based on centralized patient testing, comprising: a data acquisition module, a data processing module, a data analysis module and a typing and identification module; The data acquisition module collects and records the patient's serum index data; The serum index data include C-reactive protein, glucose, lactate dehydrogenase, TNF-α, IgG and IgM; The data acquisition module is provided with a spectrophotometer, which can measure absorbance to measure serum index data; Wherein, the measuring wavelength range of the spectrophotometer is 200-800nm; The data processing module is used to process the patient's serum index data to obtain homogenized serum index data; The data analysis module is used to perform data cleaning and feature extraction on the homogenized serum index data, and generate corresponding serum index distribution maps by clustering based on the processing results; The typing and identification module is used to perform typing and identification of the patient's health status; The data acquisition module is also provided with a data storage unit and a detection unit. The process of the data acquisition module collecting and recording the patient's serum index data includes: Input the serum index data list, set the collection cycle and equipment parameters, initialize the data storage unit, the detection unit is used to detect the patient's serum index data according to the serum index data list based on the absorbance measurement function of the spectrophotometer, and store the detection results in the data storage unit to obtain the patient's serum index data; The process of the data processing module processing the patient's serum index data includes: Input the patient's serum index data, store the serum index data in matrix form, and construct and obtain the index matrix X=[{x}_{ij}] ,in represents the element in the i-th row and j-th column of the matrix X, whose value is the j-th serum index value of the i-th sample, i=1, 2, ..., n, n>0 and n is an integer, j=1, 2, ..., 6; Calculate the mean and standard deviation of each column in the indicator matrix column by column to obtain the mean and standard deviation of each serum indicator data. Subtract the mean of the serum indicator data corresponding to each column from the serum indicator data in the indicator matrix in turn to obtain the mean-averaged serum indicator data. Then divide each column of the mean-averaged serum indicator data by the standard deviation of the serum indicator data corresponding to the column to obtain the normalized indicator data.

[0005] Preferably, the data cleaning process includes: Sort the normalized index data of each column in order of numerical value, obtain the value of each column closest to the 30% position as the first eigenvalue, and the value of the position closest to the 70% position as the second eigenvalue, subtract the two to obtain the characteristic difference, set the abnormal coefficient, calculate the product of the characteristic difference and the abnormal coefficient, take the sum of the product of the characteristic difference and the abnormal coefficient and the first eigenvalue as the upper limit, take the difference between the second eigenvalue and the product of the characteristic difference and the abnormal coefficient as the lower limit, and judge whether there is data with values ​​outside the upper limit and the lower limit in each row of normalized index data; If so, remove the homogenized indicator data of this bank; and obtain the cleaned indicator data.

[0006] Preferably, the feature extraction process includes: Get the matrix A=[{a}_{mj}] of the cleaned indicator data ,in Represents the element in the mth row and jth column of matrix A, whose value is the jth serum index value of the mth sample, m = 1, 2, ..., n, n > 0 and n is an integer, m ≤ i; j = 1, 2, ..., 6; construct the covariance matrix B of matrix A, perform eigendecomposition on matrix B, and obtain the eigenvalues ​​λ1, λ2, ..., λ j and the corresponding eigenvectors v1, v2, ..., v j , sort the eigenvalues ​​from large to small, set the feature threshold, take the first z largest eigenvalues ​​whose cumulative variance contribution rate is greater than the feature threshold and their corresponding eigenvectors, and construct the dimension reduction projection matrix C=[{v}_{1},{v}_{2},......,{v}_{z}] , where v z Represents the elements of the zth column in matrix C. The matrix A is linearly transformed using the dimension reduction projection matrix C to obtain the matrix D=[{d}_{mz}] ,in Represents the element in the mth row and zth column of matrix D.

[0007] Preferably, the generation process of the serum index distribution map includes: The corresponding indicator data in the input matrix D are obtained, the probability density function of the potential distribution corresponding to the indicator data is obtained, and it is compared with the preset distribution model to obtain the multidimensional feature data corresponding to each sample in the matrix D and the distribution membership corresponding to each sample in the corresponding feature data, and the corresponding Gaussian mixture distribution map is constructed based on the distribution membership to obtain the serum indicator distribution map.

[0008] Preferably, the process of the typing identification module performing typing identification on the patient's health status includes: Group patient samples, construct multiple patient groups, calculate the mean and standard deviation of each serum index in multiple patient groups, input serum index data of new patients, obtain serum index weights, set distance deviation adjustment factors, calculate weighted distance metrics from serum index data of new patients to known patient groups, and assign new patients to the closest patient group according to the minimum value of weighted distance metrics; Read the serum reference distribution data of the healthy population, obtain the mean and standard deviation of the j-th serum index of the healthy population, calculate the standardized deviation of each serum index of the patient relative to the index distribution of the healthy population, integrate the single deviation value using the distribution deviation term formula, accumulate the health deviation contribution of each serum index of the patient, obtain the distribution deviation term, calculate the correlation coefficient matrix between the serum indicators of the patient, calculate the synergistic deviation integral for the significantly correlated indicator pairs, and perform weighted accumulation to obtain the synergistic abnormal term, substitute the distribution deviation term and the synergistic abnormal term into the disease index calculation formula, and calculate the comprehensive disease index.

[0009] Preferably, the calculation formula of the weighted distance metric is: Formula 1 Where: For a new patient P to a known patient group G r The weighted distance metric, X p,j is the jth serum index value of patient P; μ Gr,j , σ Gr,j For group G r The mean and standard deviation of the jth serum index; w DC,j is the weight of the jth serum index in the weighted distance metric calculation, indicating its importance to the classification; γ is the distance deviation adjustment factor.

[0010] Preferably, the disease index calculation formula is: Formula 2 Where: S H is the patient's disease index; X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; w SH,j is the weight of the jth serum index in the calculation of disease index; it reflects the degree of influence of this index on health status; δ is the deviation index, which is used to control the nonlinear influence of index deviation on probability; ρ jk is the correlation coefficient between the jth and kth serum indicators, quantifying their synergistic effects; α and β are global adjustment coefficients, respectively controlling the weights of the deviation of a single indicator and the synergistic abnormality on the disease indicator; h jk is the degree of joint deviation, and its calculation formula is as follows: Formula 3 Where: X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; X p,k is the kth serum index value of patient P; μ H,k , σ H,k is the mean and standard deviation of the kth serum index of healthy people.

[0011] The beneficial effects of the present invention are: 1. The present invention collects serum index data of patients, including C-reactive protein, glucose, lactate dehydrogenase, TNF-α, IgG and IgM, constructs an index matrix, performs homogenization processing, and forms a distribution map of serum indexes of patients, which is convenient for intuitively reflecting the relationship between serum index values ​​of patients, reflecting the synergistic response of various indicators to inflammation, and displaying the inflammation situation of the patient population. At the same time, serum specimens of new patients are typed, and disease indicators are calculated, which is convenient for medical personnel to carry out targeted treatment of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0013] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions of the schematic diagrams in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0014] like Figure 1 As shown, a serum specimen typing and identification system based on centralized patient testing includes: a data acquisition module, a data processing module, a data analysis module and a typing and identification module; The data collection module collects and records the patient's serum index data; Serum index data included C-reactive protein, glucose, lactate dehydrogenase, TNF-α, IgG, and IgM; The data acquisition module is equipped with a spectrophotometer, which can measure absorbance to measure serum index data; The measurement wavelength range of the spectrophotometer is 200-800nm; The data processing module is used to process the patient's serum index data to obtain homogenized serum index data; The data analysis module is used to perform data cleaning and feature extraction on the homogenized serum index data, and generate the corresponding serum index distribution map through clustering based on the processing results; The typing and identification module is used to identify the patient's health status; The data acquisition module is also provided with a data storage unit and a detection unit. The process of the data acquisition module collecting and recording the patient's serum index data includes: Input the serum index data list, set the collection cycle and equipment parameters, initialize the data storage unit, and the detection unit is used to detect the patient's serum index data according to the serum index data list based on the absorbance measurement function of the spectrophotometer, and store the detection results in the data storage unit to obtain the patient's serum index data; The process of data processing module processing the patient's serum index data includes: Input the patient's serum index data, store the serum index data in matrix form, and construct and obtain the index matrix X=[{x}_{ij}] ,in represents the element in the i-th row and j-th column of the matrix X, whose value is the j-th serum index value of the i-th sample, i=1, 2, ..., n, n>0 and n is an integer, j=1, 2, ..., 6; Calculate the mean and standard deviation of each column in the indicator matrix column by column to obtain the mean and standard deviation of each serum indicator data, subtract the mean of the serum indicator data corresponding to the column from the serum indicator data in each column of the indicator matrix in turn to obtain the mean-free serum indicator data, and then divide each column of the mean-free serum indicator data by the standard deviation of the serum indicator data corresponding to the column to obtain the normalized indicator data;

[0015] In some embodiments, the data cleaning process includes: Sort the normalized index data of each column in order of numerical value, obtain the value of each column closest to the 30% position as the first eigenvalue, and the value of the position closest to the 70% position as the second eigenvalue, subtract the two to obtain the characteristic difference, set the abnormal coefficient, calculate the product of the characteristic difference and the abnormal coefficient, take the sum of the product of the characteristic difference and the abnormal coefficient and the first eigenvalue as the upper limit, take the difference between the second eigenvalue and the product of the characteristic difference and the abnormal coefficient as the lower limit, and judge whether there is data with values ​​outside the upper limit and the lower limit in each row of normalized index data; If yes, remove the homogenized index data of this bank; obtain the cleaned index data; In this embodiment, the abnormal coefficient can be set to 1.5. For example, if the first eigenvalue of a certain indicator in the data is 1.5 and the second eigenvalue is 0.5, then the characteristic difference is 1.0, and the product of the characteristic difference and the abnormal coefficient is 1.5. Therefore, the upper limit is the sum of the first eigenvalue and the second eigenvalue, which is 3.0, and the lower limit is the difference between the second eigenvalue and the first eigenvalue, which is -1.0. Any value outside the range of (-1.0, 3.0) is regarded as an abnormal value, which is used to clean and remove the serum indicator data of the concentrated test patient group that is not representative;

[0016] In some embodiments, the feature extraction process includes: Get the matrix A=[{a}_{mj}] of the cleaned indicator data ,in Represents the element in the mth row and jth column of matrix A, whose value is the jth serum index value of the mth sample, m = 1, 2, ..., n, n > 0 and n is an integer, m ≤ i; j = 1, 2, ..., 6; construct the covariance matrix B of matrix A, perform eigendecomposition on matrix B, and obtain the eigenvalues ​​λ1, λ2, ..., λ j and the corresponding eigenvectors v1, v2, ..., v j , sort the eigenvalues ​​from large to small, set the feature threshold, take the first z largest eigenvalues ​​whose cumulative variance contribution rate is greater than the feature threshold and their corresponding eigenvectors, and construct the dimension reduction projection matrix C=[{v}_{1},{v}_{2},......,{v}_{z}] , where v z Represents the elements of the zth column in matrix C. The matrix A is linearly transformed using the dimension reduction projection matrix C to obtain the matrix D=[{d}_{mz}] ,in Represents the element in the mth row and zth column of the matrix D;

[0017] In this embodiment, a covariance matrix B is constructed to capture the linear correlation between features. The eigenvalues ​​and eigenvectors of the covariance matrix reflect the main direction and variance of the data distribution. The covariance matrix B is eigen-decomposed, and the obtained eigenvalues ​​reflect the variance of the data in the direction of the eigenvector, and the eigenvector represents the main change direction of the data. The eigenvalues ​​are sorted from large to small, and the first z largest eigenvalues ​​and their corresponding eigenvectors are retained to construct a dimensionality reduction projection matrix C. The selection of z is calculated by the cumulative variance contribution rate:

[0018] In this embodiment, the feature threshold is set to 90% to retain most of the variance information; The original data is linearly transformed using the dimension reduction projection matrix C to obtain the transformed matrix D=[{d}_{mz}] ; The matrix D has dimensions m×z;

[0019] In some embodiments, the process of generating a serum index distribution map includes: Input the corresponding indicator data in the matrix D, obtain the probability density function of the potential distribution corresponding to the indicator data, compare it with the preset distribution model, obtain the multidimensional feature data corresponding to each sample in the matrix D and the distribution membership corresponding to each sample in the corresponding feature data, and construct the corresponding Gaussian mixture distribution map based on the distribution membership to obtain the serum indicator distribution map;

[0020] In this embodiment, it is assumed that the data comes from a mixture of multiple Gaussian distributions, each Gaussian distribution is described by a mean, covariance and weight, and the distribution membership of each sample is obtained by calculating the posterior probability that the sample belongs to each Gaussian distribution, allowing the sample to belong to multiple distributions at the same time. For the data set to be analyzed, it is assumed that it comes from a mixture model of K Gaussian distributions, each distribution is represented by a mean vector, a covariance matrix and a weight, and the probability of each sample is calculated using a Gaussian probability density function. According to the initial parameters (mean, covariance, weight), the probability density value of each sample under all K Gaussian distributions is calculated; the degree of membership of each sample to each Gaussian distribution is calculated, and a membership matrix is ​​established. According to the membership matrix, the samples are distributed to K Gaussian distributions to generate clustering results, and different colors and shapes are used to represent different Gaussian distributions to show the clustering characteristics and probability density distribution of the data;

[0021] In some embodiments, the process of the typing identification module performing health status typing identification on the patient includes: Group patient samples, construct multiple patient groups, calculate the mean and standard deviation of each serum index in multiple patient groups, input serum index data of new patients, obtain serum index weights, set distance deviation adjustment factors, calculate weighted distance metrics from serum index data of new patients to known patient groups, and assign new patients to the closest patient group according to the minimum value of weighted distance metrics; Read the serum reference distribution data of healthy people, obtain the mean and standard deviation of the jth serum index of healthy people, calculate the standardized deviation of each serum index of patients relative to the index distribution of healthy people, integrate the single deviation value using the distribution deviation term formula, accumulate the health deviation contribution of each serum index of patients, obtain the distribution deviation term, calculate the correlation coefficient matrix between the serum indicators of patients, calculate the synergistic deviation integral for the significantly correlated index pairs, and perform weighted accumulation to obtain the synergistic abnormal term, substitute the distribution deviation term and the synergistic abnormal term into the disease index calculation formula, and calculate the comprehensive disease index;

[0022] In some embodiments, the weighted distance metric is calculated as: Formula 1 Where: For a new patient P to a known patient group G r The weighted distance metric, X p,j is the jth serum index value of patient P; μ Gr,j , σ Gr,j For group G r The mean and standard deviation of the jth serum index; w DC,j is the weight of the jth serum index in the weighted distance metric calculation, indicating its importance to the classification; γ is the distance deviation adjustment factor;

[0023] In this embodiment, w DC,j It can take values ​​such as 0.4, 0.5, 0.6, etc., or it can uniformly divide the weights according to j indicators. γ usually takes a value greater than 1, such as 1.5, to amplify the impact of large deviations on distance and enhance the sensitivity of classification to patients;

[0024] In some embodiments, the disease index calculation formula is: Formula 2 Where: S H is the patient's disease index; X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; w SH,j is the weight of the jth serum index in the calculation of disease index; it reflects the degree of influence of this index on health status; δ is the deviation index, which is used to control the nonlinear influence of index deviation on probability; ρ jk is the correlation coefficient between the jth and kth serum indicators, quantifying their synergistic effects; α and β are global adjustment coefficients, respectively controlling the weights of the deviation of a single indicator and the synergistic abnormality on the disease indicator; h jk is the degree of joint deviation, and its calculation formula is as follows: Formula 3 Where: X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; X p,k is the kth serum index value of patient P; μ H,k , σ H,k is the mean and standard deviation of the kth serum index of healthy people; In this embodiment, the disease index is used to reveal the inflammation status in the patient's body. SH,j The value can be 0.3, 0.4, 0.5, etc. The secondary indicator w SH,j The value can be 0.1, 0.2, etc. The δ deviation index can be 2 in general scenarios and 3 in highly sensitive abnormal scenarios. jk The correlation coefficient generally ranges from [-1, 1]. The lower the correlation, the closer the value is to 0. The global adjustment coefficient α generally ranges from [0.5, 5]. The higher the value, the greater the impact of single item deviation. The global adjustment coefficient β generally ranges from [0.1, 2]. The higher the value, the greater the impact of collaborative anomalies between indicators.

Claims

1. A serum specimen typing and identification system based on centralized patient testing, characterized in that: include: Data collection module, data processing module, data analysis module and typing identification module; The data acquisition module collects and records the patient's serum index data; The serum index data include C-reactive protein, glucose, lactate dehydrogenase, TNF-α, IgG and IgM; The data acquisition module is provided with a spectrophotometer, which can measure absorbance to measure serum index data; Wherein, the measuring wavelength range of the spectrophotometer is 200-800nm; The data processing module is used to process the patient's serum index data to obtain homogenized serum index data; The data analysis module is used to perform data cleaning and feature extraction on the homogenized serum index data, and generate corresponding serum index distribution maps by clustering based on the processing results; The typing and identification module is used to perform typing and identification of the patient's health status; The data acquisition module is also provided with a data storage unit and a detection unit. The process of the data acquisition module collecting and recording the patient's serum index data includes: Input the serum index data list, set the collection cycle and equipment parameters, initialize the data storage unit, the detection unit is used to detect the patient's serum index data according to the serum index data list based on the absorbance measurement function of the spectrophotometer, and store the detection results in the data storage unit to obtain the patient's serum index data; The process of the data processing module processing the patient's serum index data includes: Input the patient's serum index data, store the serum index data in matrix form, and construct and obtain the index matrix X=[{x}_{ij}] ,in represents the element in the i-th row and j-th column of the matrix X, whose value is the j-th serum index value of the i-th sample, i=1, 2, ..., n, n>0 and n is an integer, j=1, 2, ..., 6; Calculate the mean and standard deviation of each column in the indicator matrix column by column to obtain the mean and standard deviation of each serum indicator data. Subtract the mean of the serum indicator data corresponding to each column from the serum indicator data in the indicator matrix in turn to obtain the mean-averaged serum indicator data. Then divide each column of the mean-averaged serum indicator data by the standard deviation of the serum indicator data corresponding to the column to obtain the normalized indicator data.

2. A serum specimen typing and identification system based on centralized patient testing according to claim 1, characterized in that: The data cleaning process includes: Sort the normalized index data of each column in order of numerical value, obtain the value of each column closest to the 30% position as the first eigenvalue, and the value of the position closest to the 70% position as the second eigenvalue, subtract the two to obtain the characteristic difference, set the abnormal coefficient, calculate the product of the characteristic difference and the abnormal coefficient, take the sum of the product of the characteristic difference and the abnormal coefficient and the first eigenvalue as the upper limit, take the difference between the second eigenvalue and the product of the characteristic difference and the abnormal coefficient as the lower limit, and judge whether there is data with values ​​outside the upper limit and the lower limit in each row of normalized index data; If so, remove the homogenized indicator data of this bank; and obtain the cleaned indicator data.

3. A serum specimen typing and identification system based on centralized patient testing according to claim 2, characterized in that: The feature extraction process includes: Get the matrix A=[{a}_{mj}] of the cleaned indicator data ,in Represents the element in the mth row and jth column of matrix A, whose value is the jth serum index value of the mth sample, m = 1, 2, ..., n, n > 0 and n is an integer, m ≤ i; j = 1, 2, ..., 6; construct the covariance matrix B of matrix A, perform eigendecomposition on matrix B, and obtain the eigenvalues ​​λ1, λ2, ..., λ j and the corresponding eigenvectors v1, v2, ..., v j , sort the eigenvalues ​​from large to small, set the feature threshold, take the first z largest eigenvalues ​​whose cumulative variance contribution rate is greater than the feature threshold and their corresponding eigenvectors, and construct the dimension reduction projection matrix C=[{v}_{1},{v}_{2},......,{v}_{z}] , where v z Represents the elements of the zth column in matrix C. The matrix A is linearly transformed using the dimension reduction projection matrix C to obtain the matrix D=[{d}_{mz}] ,in Represents the element in the mth row and zth column of matrix D.

4. A serum specimen typing and identification system based on centralized patient testing according to claim 3, characterized in that: The generation process of the serum index distribution map includes: The corresponding indicator data in the input matrix D are obtained, the probability density function of the potential distribution corresponding to the indicator data is obtained, and it is compared with the preset distribution model to obtain the multidimensional feature data corresponding to each sample in the matrix D and the distribution membership corresponding to each sample in the corresponding feature data, and the corresponding Gaussian mixture distribution map is constructed based on the distribution membership to obtain the serum indicator distribution map.

5. A serum specimen typing and identification system based on centralized patient testing according to claim 4, characterized in that: The process of the typing identification module performing health status typing identification on the patient includes: Group patient samples, construct multiple patient groups, calculate the mean and standard deviation of each serum index in multiple patient groups, input serum index data of new patients, obtain serum index weights, set distance deviation adjustment factors, calculate weighted distance metrics from serum index data of new patients to known patient groups, and assign new patients to the closest patient group according to the minimum value of weighted distance metrics; Read the serum reference distribution data of the healthy population, obtain the mean and standard deviation of the j-th serum index of the healthy population, calculate the standardized deviation of each serum index of the patient relative to the index distribution of the healthy population, integrate the single deviation value using the distribution deviation term formula, accumulate the health deviation contribution of each serum index of the patient, obtain the distribution deviation term, calculate the correlation coefficient matrix between the serum indicators of the patient, calculate the synergistic deviation integral for the significantly correlated indicator pairs, and perform weighted accumulation to obtain the synergistic abnormal term, substitute the distribution deviation term and the synergistic abnormal term into the disease index calculation formula, and calculate the comprehensive disease index.

6. A serum specimen typing and identification system based on centralized patient testing according to claim 5, characterized in that: The calculation formula of the weighted distance metric is: Formula 1 Where: For a new patient P to a known patient group G r The weighted distance metric, X p,j is the jth serum index value of patient P; μ Gr,j , σ Gr,j For group G r The mean and standard deviation of the jth serum index; w DC,j is the weight of the jth serum index in the weighted distance metric calculation, indicating its importance to the classification; γ is the distance deviation adjustment factor.

7. A serum specimen typing and identification system based on centralized patient testing according to claim 6, characterized in that: The disease index calculation formula is: Formula 2 Where: S H is the patient's disease index; X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; w SH,j is the weight of the jth serum index in the calculation of disease index; Reflects the impact of the indicator on health status; δ is the deviation index, which is used to control the nonlinear effect of indicator deviation on probability; ρ jk is the correlation coefficient between the jth and kth serum indicators, quantifying their synergistic effects; α and β are global adjustment coefficients, respectively controlling the weights of the deviation of a single indicator and the synergistic abnormality on the disease indicator; h jk is the degree of joint deviation, and its calculation formula is as follows: Formula 3 Where: X p,j is the jth serum index value of patient P; μ H,j , σ H,j is the mean and standard deviation of the jth serum index of healthy people; X p,k is the kth serum index value of patient P; μ H,k , σ H,k is the mean and standard deviation of the kth serum index of healthy people.

Images