Early diagnosis data management system for patient with schistosome hepatic fibrosis

Abstract

The invention relates to the technical field of medical information systems, and particularly discloses an early diagnosis data management system for a patient with schistosome hepatic fibrosis, which is used for acquiring blood biomarker data of the patient, and comparing and analyzing real-time blood biomarker data and historical blood biomarker data to obtain early diagnosis data of the patient with the schistosome hepatic fibrosis. Classifying the blood biomarker data into reliable data and unreliable data using a logistic regression model; calculating a variability deviation index and a logic deviation index of the blood biomarker data through a dynamic time warping and auto-encoder technology, and evaluating time stability and internal logic consistency of the blood biomarker data; performing normalization processing on the variability deviation index and the logic deviation index, and calculating an accuracy coefficient of the data; dividing the data into accurate data and inaccurate data according to the accuracy coefficient, and transmitting the accurate data to a patient management system.

Description

Technical Field

[0001] The present invention relates to the technical field of medical information systems, and in particular to an early diagnosis data management system for patients with schistosomiasis liver fibrosis. Background Art

[0002] Schistosomal fibrosis is a chronic liver disease caused by schistosoma japonicum infection, characterized by the gradual fibrosis of liver tissue, which may lead to cirrhosis and even liver failure. As the disease progresses slowly and the early symptoms are not obvious, early diagnosis and treatment are crucial. However, in actual clinical practice, the diagnosis and treatment of schistosomal fibrosis face many challenges;

[0003] In recent years, with the development of information technology and big data analysis technology, the medical field has also begun to introduce various advanced data processing technologies to improve the effectiveness of diagnosis and treatment; for example, artificial intelligence technologies such as machine learning and deep learning have been applied to medical image analysis, disease prediction, etc.; however, the application of these technologies in the field of schistosomal liver fibrosis is not yet widespread, especially in the areas of real-time monitoring and dynamic evaluation.

[0004] In the management of patients with schistosomiasis liver fibrosis, timely diagnosis and treatment of patients can be achieved through real-time monitoring of patients' blood biomarker data. However, if there is data distortion in the process of monitoring patients' blood biomarker data, which makes the blood biomarker data inconsistent with the actual situation, it may lead to doctors' judgment of the treatment effect and make it impossible for doctors to adjust the treatment plan in time, thus affecting the patient's recovery process; patient management plans based on inaccurate data may lead to unnecessary medical interventions or delay necessary treatments. Summary of the invention

[0005] The purpose of the present invention is to provide an early diagnosis data management system for patients with schistosomiasis liver fibrosis to solve the above-mentioned problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The early diagnosis data management system for patients with schistosomiasis liver fibrosis includes: a data acquisition module, which is used to obtain blood biomarker data of patients, wherein the blood biomarker data includes: complete blood cell count, coagulation function index, immunological index and biochemical index;

[0008] A data verification module, which is used to analyze the credibility of the patient's blood biomarker data, compare the patient's real-time blood biomarker data with the corresponding historical blood biomarker data, and classify the patient's blood biomarker data into reliable data and unreliable data according to the comparison results;

[0009] A data change analysis module, wherein the data change analysis module is used to analyze the change of the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the variability deviation index of the blood biomarker data, and evaluate the variability of the patient's blood biomarker data;

[0010] A data correlation analysis module, which is used to analyze the correlation between indicators in the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the logical deviation index of the blood biomarker data, and evaluate the logic of the patient's blood biomarker data;

[0011] A comprehensive evaluation module, which is used to comprehensively analyze the variability deviation index and the logical deviation index of the patient's blood biomarker data to evaluate the accuracy of the blood biomarker data;

[0012] A data division and application module, which divides the patient's blood biomarker data into accurate data and inaccurate data according to the evaluation results, sends the accurate data to the patient management system, and provides a diagnosis and treatment plan based on the patient management system.

[0013] As a further solution of the present invention: the credibility of the patient's blood biomarker data is analyzed, the patient's real-time blood biomarker data is compared with the corresponding historical blood biomarker data, and according to the comparison result, the patient's blood biomarker data is divided into reliable data and unreliable data, specifically including:

[0014] performing feature extraction on the real-time blood biomarker data and the corresponding historical blood biomarker data;

[0015] A classification model was constructed using a logistic regression model to classify blood biomarker data into reliable data and unreliable data;

[0016] Wherein, the feature extraction includes: mean difference feature Δμ i , standard deviation difference characteristic Δσ i and change ratio characteristic R i ;

[0017] Construct the feature vector x = [|Δμ i |,|Δσ i |,|R i |];

[0018] Construct the weight vector w = [w 1 , w 2 , w 3 ];

[0019] The formula for calculating the classification model is:

[0020] Where P(x) represents the probability that the blood biomarker data x is reliable data, and w 1 、w 2 and w 3 are preset weight coefficients, and w 1 、w 2 and w 3 are all greater than 0, b is a preset constant, x is a feature vector, including mean difference, standard deviation difference and change ratio, and w is a weight vector, including w 1 、w 2 and w 3 , w T represents the transpose of the weight vector, that is, converting the column vector into a row vector;

[0021] Determine whether the probability P(x) of reliable data is greater than or equal to a preset threshold. If so, it is recorded as reliable data, otherwise it is recorded as unreliable data.

[0022] As a further solution of the present invention: the mean difference feature Δμ i , standard deviation difference characteristic Δσ i and change ratio characteristic R i The acquisition process is:

[0023] Extract the mean difference features of real-time blood biomarker data and historical blood biomarker data:

[0024] The calculation formula of the mean difference is: Δμ i =μ inew -μ ihist ;

[0025] In the formula, μ inew represents the mean value of the ith indicator in real-time blood biomarker data, μ ihist represents the mean value of the i-th indicator in the historical blood biomarker data, Δμ i represents the mean difference of the i-th indicator, i represents the i-th indicator;

[0026] Extract the standard deviation difference features of real-time blood biomarker data and historical blood biomarker data:

[0027] The calculation formula of the standard deviation difference is: Δσ i =σ inew -σ ihist ;

[0028] In the formula, σ inewrepresents the standard deviation of the ith indicator in real-time blood biomarker data, σ ihist represents the standard deviation of the i-th indicator in the historical blood biomarker data, Δσ i represents the standard deviation difference of the i-th indicator;

[0029] Extracting change ratio features of real-time blood biomarker data and historical blood biomarker data;

[0030] Wherein, the calculation formula of the change ratio is:

[0031] In the formula, R i It represents the change ratio of the i-th indicator;

[0032] The indicators include complete blood cell count, coagulation function index, immunological index and biochemical index in the blood biomarker data.

[0033] As a further solution of the present invention: when the patient's blood biomarker data is reliable data, the change of the patient's blood biomarker data is analyzed, the variability deviation index of the blood biomarker data is calculated, and the variability of the patient's blood biomarker data is evaluated;

[0034] The process of obtaining the variability deviation index of the blood biomarker data is as follows:

[0035] The patient's blood biomarker data is constructed as a time series, assuming y = [y 1 ,y 2 ,......,y n ] represents the observed value at time point t;

[0036] Treat multiple blood biomarker data measured on the same patient as different time series and calculate the DTW distances between them. Calculate the variability deviation index of the blood biomarker data by performing statistical analysis on these distances.

[0037] The calculation formula of the variability deviation index of the blood biomarker data is:

[0038]

[0039] In the formula, Z nt represents the variability deviation index of blood biomarker data, and C[s][j] represents the DTW distance between sequence X and sequence Y.

[0040] As a further solution of the present invention: the DTW distance calculation process is:

[0041] Define the cost matrix, calculate the Euclidean distance between each pair of points, and construct the cost matrix D;

[0042] For two sequences X and Y, the element D[s][j] in the cost matrix D represents the distance between X[s] and Y[j];

[0043] Where s represents the number of observation points in sequence X, s is a positive integer greater than 0, j represents the number of observation points in sequence Y, j is a positive integer greater than 0;

[0044] Fill the cumulative distance matrix C using a dynamic programming algorithm;

[0045] The element C[s][j] of the cumulative distance matrix C represents the cumulative distance between the first s points of sequence X and the first j points of sequence Y;

[0046] The updating rule of the cumulative distance matrix is:

[0047] If s=0 or j=0, C[s][j]=D[s][j];

[0048] If s≠0 and j≠0, C[s][j]=D[s][j]+min(C[s-1][j],C[s][j-1],C[s-1][j-1]);

[0049] The DTW distance is the last element of the cumulative distance matrix, namely C[n][n].

[0050] As a further solution of the present invention: when the patient's blood biomarker data is reliable data, the interrelationships of the indicators in the patient's blood biomarker data are analyzed, the logical deviation index of the blood biomarker data is calculated, and the logic of the patient's blood biomarker data is evaluated;

[0051] The process of obtaining the logical deviation index of the blood biomarker data is as follows:

[0052] The patient's blood biomarker data is organized into a data matrix E, where each row represents a blood biomarker data, each column represents an indicator, and the dimension of the data matrix E is set to m*q;

[0053] Among them, m represents the number of data, m is a positive integer greater than 0, and q represents the number of indicators;

[0054] The goal of training the autoencoder is to minimize the error between the input blood biomarker data and the reconstructed blood biomarker data, using mean squared error as the loss function and calculating the logistic deviation index;

[0055] The calculation formula of the logic deviation index is:

[0056]

[0057] In the formula, Z q represents the normalized value of the qth blood biomarker data, Z yq represents the reconstructed data of the qth blood biomarker data, ‖Z q -Z yq ‖ 2 represents the reconstruction error of the qth blood biomarker data, L nt Represents the logistic deviation index for blood biomarker data.

[0058] As a further solution of the present invention: the process of training the autoencoder is:

[0059] Build an autoencoder model, which consists of an encoder and a decoder;

[0060] The encoder maps high-dimensional input data to a low-dimensional representation;

[0061] The calculation formula for mapping the high-dimensional input data to low-dimensional is:

[0062] h=f(W e Z+b e );

[0063] In the formula, h represents the mapping of high-dimensional input data to low-dimensional representation, f is the activation function, and W e is the weight matrix of the encoder, b e is the bias vector of the encoder, and Z is the normalized value of the data;

[0064] The decoder maps the low-dimensional representation back to a high-dimensional space;

[0065] The calculation formula for mapping the low-dimensional representation back to the high-dimensional space is:

[0066] Z Y =g(W d h+b d );

[0067] In the formula, Z Y Represents mapping the low-dimensional representation back to the high-dimensional space representation, W d is the weight matrix of the decoder, b d is the bias vector of the decoder, and g is the activation function, which is the same as the activation function of the encoder.

[0068] As a further solution of the present invention: the variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data are comprehensively analyzed to calculate the accuracy coefficient of the patient's blood biomarker data and evaluate the accuracy of the blood biomarker data;

[0069] The accuracy coefficient acquisition process of the patient's blood biomarker data is as follows:

[0070] The variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data are normalized, and the accuracy coefficient of the patient data is calculated by the variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data after the normalization.

[0071] As a further solution of the present invention: the patient's blood biomarker data is divided into accurate data and inaccurate data according to the accuracy evaluation result of the blood biomarker data, specifically including:

[0072] It is determined whether the accuracy coefficient of the blood biomarker data is greater than or equal to a preset accuracy coefficient threshold of the blood biomarker data. If so, it is recorded as accurate data; if not, it is recorded as inaccurate data.

[0073] As a further solution of the present invention: the comprehensive evaluation module at least includes:

[0074] a blood biomarker data variability deviation index calculation unit, the blood biomarker data variability deviation index calculation unit is used to calculate the variability deviation index of the blood biomarker data;

[0075] A logic deviation index calculation unit for blood biomarker data is used to calculate the logic deviation index of the blood biomarker data.

[0076] Beneficial effects of the present invention:

[0077] (1) The present invention introduces advanced data analysis techniques, such as logistic regression model, dynamic time warping and autoencoder, to analyze data at multiple levels and angles, thus greatly improving the accuracy of diagnosis;

[0078] Specifically, the data verification module calculates mean difference, standard deviation difference and change ratio characteristics by comparing and analyzing the real-time blood biomarker data with the historical blood biomarker data, and uses the logistic regression model to classify the data into reliable data and unreliable data; this step effectively excludes abnormal data caused by operational errors or equipment failures, and ensures the basic data quality for subsequent analysis; in addition, by calculating the variability deviation index and logical deviation index of blood biomarker data, the temporal stability of the data and its internal logical consistency are further verified, thereby providing doctors with a more reliable data basis and making the diagnosis results more accurate;

[0079] (2) The data acquisition module can efficiently obtain the patient's blood biomarker data and standardize it to ensure the consistency and availability of the data; the data change analysis module and the data correlation analysis module can automatically identify abnormal changes and internal correlations in the data through dynamic time warping and autoencoder technology, providing doctors with more intuitive and easy-to-understand data reports; this intelligent data processing method not only saves a lot of human resources, but also avoids omissions and errors that may occur when manually processing blood biomarker data; the comprehensive evaluation module calculates the accuracy coefficient of blood biomarker data, comprehensively analyzes the variability deviation index and the logical deviation index, and provides doctors with quantitative indicators of data accuracy; this means that doctors no longer need to rely on subjective judgment to evaluate the reliability of blood biomarker data, but can rely on the accuracy coefficient automatically generated by the system to make decisions. This decision-making process based on objective data not only improves work efficiency, but also provides a solid scientific basis for medical decision-making; the data division and application module can automatically divide the data into accurate data and inaccurate data according to the evaluation results, and transmit the accurate data to the patient management system; this process not only simplifies the data management process, but also provides doctors with real-time updates of patient health status, allowing doctors to review the latest patient data at any time and adjust treatment plans in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] The present invention will be further described below in conjunction with the accompanying drawings.

[0081] Figure 1 This is a flowchart of the early diagnosis data management system for patients with schistosomiasis liver fibrosis of the present invention;

[0082] Figure 2 It is a flowchart of the specific steps of the present invention. DETAILED DESCRIPTION

[0083] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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.

[0084] Embodiment 1:

[0085] See also Figure 1 or Figure 2 As shown, the present invention is an early diagnosis data management system for patients with schistosomiasis liver fibrosis, comprising:

[0086] A data acquisition module, wherein the data acquisition module is used to obtain the patient's blood biomarker data, wherein the blood biomarker data includes: complete blood cell count, coagulation function index, immunological index and biochemical index;

[0087] A data verification module, which is used to analyze the credibility of the patient's blood biomarker data, compare the patient's real-time blood biomarker data with the corresponding historical blood biomarker data, and classify the patient's blood biomarker data into reliable data and unreliable data according to the comparison results;

[0088] A data change analysis module, wherein the data change analysis module is used to analyze the change of the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the variability deviation index of the blood biomarker data, and evaluate the variability of the patient's blood biomarker data;

[0089] A data correlation analysis module, which is used to analyze the correlation between indicators in the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the logical deviation index of the blood biomarker data, and evaluate the logic of the patient's blood biomarker data;

[0090] A comprehensive evaluation module, which is used to comprehensively analyze the variability deviation index and the logical deviation index of the patient's blood biomarker data to evaluate the accuracy of the blood biomarker data;

[0091] A data division and application module, which divides the patient's blood biomarker data into accurate data and inaccurate data according to the evaluation results, sends the accurate data to the patient management system, and provides a diagnosis and treatment plan based on the patient management system.

[0092] Embodiment 2:

[0093] On the basis of Example 1, the patient's blood biomarker data is obtained, and the real-time blood biomarker data of different patients is compared with the corresponding historical blood biomarker data, and then the blood biomarker data is divided into reliable data and unreliable data, specifically including:

[0094] Preprocess the real-time blood biomarker data collected from patients, including cleaning invalid data and formatting data to ensure data consistency and availability;

[0095] Pre-processed blood biomarker data included complete blood count, coagulation function indexes, immunological indexes, and biochemical indexes;

[0096] extracting features from the real-time blood biomarker data and the corresponding historical blood biomarker data;

[0097] The extracted features include:

[0098] Extract the mean difference features of real-time blood biomarker data and historical blood biomarker data:

[0099] The calculation formula of the mean difference is: Δμ i =μ inew -μ ihist ;

[0100] In the formula, μ inew represents the mean value of the ith indicator in real-time blood biomarker data, μ ihist represents the mean value of the i-th indicator in the historical blood biomarker data, Δμ i represents the mean difference of the i-th indicator, i represents the i-th indicator;

[0101] Extract the standard deviation difference features of real-time blood biomarker data and historical blood biomarker data:

[0102] The calculation formula of the standard deviation difference is: Δσ i =σ inew -σ ihist ;

[0103] In the formula, σ inew represents the standard deviation of the ith indicator in real-time blood biomarker data, σ ihist represents the standard deviation of the i-th indicator in the historical blood biomarker data, Δσ i represents the standard deviation difference of the i-th indicator;

[0104] Extracting change ratio features of real-time blood biomarker data and historical blood biomarker data;

[0105] Wherein, the calculation formula of the change ratio is:

[0106] In the formula, R i It represents the change ratio of the i-th indicator;

[0107] A classification model was constructed using a logistic regression model to classify blood biomarker data into reliable data and unreliable data;

[0108] The training dataset consists of historical blood biomarker data and their labels. The labels can be annotated based on the doctor’s experience or known reliable data, including:

[0109] Construct the feature vector x = [|Δμi |,|Δσ i |,|R i |];

[0110] Construct the weight vector w = [w 1 , w 2 , w 3 ];

[0111] The calculation formula of the classification model is:

[0112] Where P(x) represents the probability that the blood biomarker data x is reliable data, and w 1 、w 2 and w 3 are preset weight coefficients, and w 1 、w 2 and w 3 are all greater than 0, b is a preset constant with a value of 2, x is a feature vector, including mean difference, standard deviation difference, and change ratio, and w is a weight vector, including w 1 、w 2 and w 3 , w T represents the transpose of the weight vector, that is, converting the column vector into a row vector;

[0113] If the probability of reliable data P(x) is greater than 0.5, the corresponding data is considered reliable and recorded as reliable data;

[0114] If the probability P(x) of reliable data is less than or equal to 0.5, the corresponding data is considered unreliable and recorded as unreliable data;

[0115] It should be noted that the probability of reliable data P(x) is a value between 0 and 1, which indicates the probability that data x is reliable data. The higher the probability value of the reliable data probability P(x), the more reliable the data is. Otherwise, it means that there may be certain problems with the data. By setting the threshold value of 0.5, it is possible to distinguish whether the data is reliable, thereby providing doctors with a reliable basis for treatment.

[0116] Embodiment 3:

[0117] On the basis of Example 2, when the patient's blood biomarker data is reliable data, the change of the patient's blood biomarker data is analyzed, the variability deviation index of the blood biomarker data is calculated, and the variability of the patient's blood biomarker data is evaluated, specifically including:

[0118] The patient's blood biomarker data is constructed as a time series, assuming y = [y 1 ,y 2 ,......,y n] represents the observed value at time point t (such as the measured value of a blood biomarker indicator);

[0119] Treat multiple blood biomarker data measured on the same patient as different time series and calculate the DTW distances between them. Calculate the variability deviation index of the blood biomarker data by performing statistical analysis on these distances.

[0120] It should be noted that the DTW distance is a nonlinear distance metric that allows time series to be stretched and compressed on the time axis to find the best matching path between two sequences.

[0121] The DTW distance calculation process is:

[0122] Define the cost matrix, calculate the Euclidean distance between each pair of points, and construct the cost matrix D;

[0123] For two sequences X and Y, the element D[s][j] in the cost matrix D represents the distance between X[s] and Y[j];

[0124] Where s represents the number of observation points in sequence X, s is a positive integer greater than 0, j represents the number of observation points in sequence Y, j is a positive integer greater than 0;

[0125] Fill the cumulative distance matrix C using a dynamic programming algorithm;

[0126] The element C[s][j] of the cumulative distance matrix C represents the cumulative distance between the first s points of sequence X and the first j points of sequence Y;

[0127] The updating rule of the cumulative distance matrix is:

[0128] If s=0 or j=0, C[s][j]=D[s][j];

[0129] If s≠0 and j≠0, C[s][j]=D[s][j]+min(C[s-1][j],C[s][j-1],C[s-1][j-1]);

[0130] The DTW distance is the last element of the cumulative distance matrix, namely C[n][n];

[0131] The calculation formula of the variability deviation index is:

[0132]

[0133] In the formula, Z nt represents the variability deviation index of blood biomarker data, C[s][j] represents the DTW distance between sequence X and sequence Y;

[0134] It should be noted that the DTW distance calculation method is used to calculate the variability deviation index of the data, thereby providing doctors with valuable information about the stability of the data. The larger the variability deviation index of the data, the more unstable the data.

[0135] Embodiment 4:

[0136] On the basis of Example 3, when the patient's blood biomarker data is reliable data, the relationship between the indicators in the patient's blood biomarker data is analyzed, the logical deviation index of the blood biomarker data is calculated, and the logic of the patient's blood biomarker data is evaluated, specifically:

[0137] The patient's blood biomarker data is organized into a data matrix E, where each row represents a blood biomarker data, and each column represents an indicator. The dimension of the data matrix E is set to m*q, where m represents the number of data, m is a positive integer greater than 0, and q represents the number of indicators;

[0138] Build an autoencoder model, which consists of an encoder and a decoder;

[0139] The encoder maps the high-dimensional input blood biomarker data to a low-dimensional representation (hidden layer);

[0140] The calculation formula for mapping the high-dimensional input blood biomarker data to low-dimensional data is:

[0141] h=f(W e Z+b e );

[0142] Where h represents the mapping of high-dimensional input blood biomarker data to low-dimensional representation, f is the activation function, and W e is the weight matrix of the encoder, b e is the bias vector of the encoder, and Z is the normalized value of the data;

[0143] The decoder maps the low-dimensional representation back to a high-dimensional space;

[0144] The calculation formula for mapping the low-dimensional representation back to the high-dimensional space is:

[0145] Z Y =g(W d h+b d );

[0146] In the formula, Z Y Represents mapping the low-dimensional representation back to the high-dimensional space representation, W d is the weight matrix of the decoder, b dis the bias vector of the decoder, g is the activation function, which means it is the same as the activation function of the encoder;

[0147] The goal of training the autoencoder is to minimize the error between the input blood biomarker data and the reconstructed blood biomarker data, using mean squared error as the loss function and calculating the logistic deviation index;

[0148] The calculation formula of the logic deviation index is:

[0149]

[0150] In the formula, Z q represents the normalized value of the qth blood biomarker data, Z yq represents the reconstructed blood biomarker data of the qth blood biomarker data, ‖Z q -Z yq ‖ 2 represents the reconstruction error of the qth blood biomarker data, L nt Represents the logistic deviation index for blood biomarker data;

[0151] It should be noted that the autoencoder is used to evaluate the logic of the patient's blood biomarker data and calculate the logic deviation index. The smaller the logic deviation index, the more consistent the distribution of the blood biomarker data after reconstruction and the better the logic. This method can not only effectively evaluate the consistency of the data, but also capture the nonlinear relationship in the data.

[0152] Embodiment 5:

[0153] On the basis of Example 4, the variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data are comprehensively analyzed to evaluate the accuracy of the blood biomarker data, specifically including:

[0154] Normalizing the variability deviation index of the patient's blood biomarker data and the logistic deviation index of the patient's blood biomarker data, and calculating the accuracy coefficient of the patient's blood biomarker data by using the normalized variability deviation index of the patient's blood biomarker data and the logistic deviation index of the patient's blood biomarker data;

[0155] The calculation formula of the accuracy coefficient is:

[0156]

[0157] In the formula, Q nt is the accuracy coefficient of blood biomarker data, Z nt is the variability bias index of blood biomarker data, Lnt is the logistic deviation index for blood biomarker data, a 1 and a 2 is the preset proportional coefficient, and a 1 and a 2 All are greater than 0;

[0158] It should be noted that: from the calculation formula, it can be seen that the variability deviation index of blood biomarker data is inversely proportional to the accuracy coefficient of blood biomarker data, and the logical deviation index of blood biomarker data is inversely proportional to the accuracy coefficient of blood biomarker data; and when the accuracy coefficient Q of blood biomarker data is nt The larger it is, the higher the accuracy of the blood biomarker data.

[0159] Based on the evaluation results, the patient's blood biomarker data is divided into accurate data and inaccurate data, and the accurate data is sent to the patient management system, including:

[0160] The accuracy coefficient Q of blood biomarker data nt Compare with the accuracy coefficient threshold of the preset data;

[0161] If the accuracy coefficient Q of blood biomarker data is nt If the value is greater than or equal to the accuracy coefficient threshold of the preset blood biomarker data, it means that the corresponding blood biomarker data is accurate and is recorded as accurate data;

[0162] If the accuracy coefficient Q of blood biomarker data is nt If the value is less than the preset accuracy coefficient threshold of the blood biomarker data, it means that the corresponding blood biomarker data is inaccurate and is recorded as inaccurate data;

[0163] Obtain all accurate data and transfer all accurate data to the patient management system so that subsequent doctors can give treatment plans based on accurate data;

[0164] It should be noted that: transferring all accurate data to the patient management system includes: if the data is stored in a database, the blood biomarker data can be directly imported into the blood biomarker database of the patient management system through the blood biomarker database connection; if the system supports file upload, the data can be saved as a CSV file and imported through the upload function provided by the system; if the patient management system provides an API interface, a script or program can be written to upload the data directly through the API.

[0165] The working principle of the present invention is as follows: the data acquisition module is responsible for acquiring the patient's blood biomarker data, covering various information such as complete blood cell count, coagulation function index, immunological index and biochemical index; the data verification module performs credibility analysis on these blood biomarker data, calculates the mean difference, standard deviation difference and change ratio characteristics by comparing with the historical blood biomarker data, and uses the logistic regression model to classify the blood biomarker data into reliable data and unreliable data;

[0166] When the blood biomarker data are confirmed to be reliable, the data change analysis module constructs these data into time series and uses the dynamic time warping method to calculate the distance between different time series, thereby evaluating the variability deviation index of the blood biomarker data; the data correlation analysis module calculates the error between the input blood biomarker data and the reconstructed blood biomarker data by constructing an autoencoder model to obtain the logical deviation index of the blood biomarker data; the comprehensive evaluation module normalizes the variability deviation index and the logical deviation index, and calculates the accuracy coefficient of the blood biomarker data; the blood biomarker data division and application module divides the blood biomarker data into accurate data and inaccurate data based on the comparison results of the accuracy coefficient with the preset threshold, and transmits the accurate data to the patient management system, so that doctors can formulate personalized treatment plans based on the accurate data.

[0167] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0168] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0169] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0170] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0171] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An early diagnosis data management system for patients with schistosomiasis liver fibrosis, characterized in that: include: A data acquisition module, wherein the data acquisition module is used to obtain the patient's blood biomarker data, wherein the blood biomarker data includes: complete blood cell count, coagulation function index, immunological index and biochemical index; A data verification module, which is used to analyze the credibility of the patient's blood biomarker data, compare the patient's real-time blood biomarker data with the corresponding historical blood biomarker data, and classify the patient's blood biomarker data into reliable data and unreliable data according to the comparison results; A data change analysis module, wherein the data change analysis module is used to analyze the change of the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the variability deviation index of the blood biomarker data, and evaluate the variability of the patient's blood biomarker data; A data correlation analysis module, which is used to analyze the correlation between indicators in the patient's blood biomarker data when the patient's blood biomarker data is reliable data, calculate the logical deviation index of the blood biomarker data, and evaluate the logic of the patient's blood biomarker data; A comprehensive evaluation module, which is used to comprehensively analyze the variability deviation index and the logical deviation index of the patient's blood biomarker data to evaluate the accuracy of the blood biomarker data; A data division and application module, which divides the patient's blood biomarker data into accurate data and inaccurate data according to the evaluation results, sends the accurate data to the patient management system, and provides a diagnosis and treatment plan based on the patient management system.

2. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 1, characterized in that: The credibility of the patient's blood biomarker data is analyzed, and the patient's real-time blood biomarker data is compared with the corresponding historical blood biomarker data, and the patient's blood biomarker data is divided into reliable data and unreliable data according to the comparison result, specifically including: performing feature extraction on the real-time blood biomarker data and the corresponding historical blood biomarker data; A classification model was constructed using a logistic regression model to classify blood biomarker data into reliable data and unreliable data; Wherein, the feature extraction includes: mean difference feature Δμ i , standard deviation difference characteristic Δσ i and change ratio characteristic R i ; Construct the feature vector x = [|Δμ i |,|Δσ i |,|R i |]; Construct a weight vector w = [w1, w2, w3]; The calculation formula of the classification model is: Where P(x) represents the probability that the blood biomarker data x is reliable data, w1, w2 and w3 are preset weight coefficients, and w1, w2 and w3 are all greater than 0, b is a preset constant, x is a feature vector, including mean difference, standard deviation difference and change ratio, w is a weight vector, including w1, w2 and w3, w T represents the transpose of the weight vector, that is, converting the column vector into a row vector; Determine whether the probability P(x) of reliable data is greater than or equal to a preset threshold. If so, it is recorded as reliable data, otherwise it is recorded as unreliable data.

3. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 2, characterized in that: The mean difference characteristic Δμ i , standard deviation difference characteristic Δσ i and change ratio characteristic R i The acquisition process is: Extract the mean difference features of real-time blood biomarker data and historical blood biomarker data: The calculation formula of the mean difference is: Δμ i =μ inew -μ ihist ; In the formula, μ inew represents the mean value of the ith indicator in real-time blood biomarker data, μ ihist represents the mean value of the i-th indicator in the historical blood biomarker data, Δμ i represents the mean difference of the i-th indicator, i represents the i-th indicator; Extract the standard deviation difference features of real-time blood biomarker data and historical blood biomarker data: The calculation formula of the standard deviation difference is: Δσ i =σ inew -σ ihist ; In the formula, σ inew represents the standard deviation of the ith indicator in real-time blood biomarker data, σ ihist represents the standard deviation of the i-th indicator in the historical blood biomarker data, Δσ i represents the standard deviation difference of the i-th indicator; Extracting change ratio features of real-time blood biomarker data and historical blood biomarker data; Wherein, the calculation formula of the change ratio is: In the formula, R i It represents the change ratio of the i-th indicator; The indicators include complete blood cell count, coagulation function index, immunological index and biochemical index in the blood biomarker data.

4. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 1, characterized in that: When the patient's blood biomarker data is reliable data, analyzing the change of the patient's blood biomarker data, calculating the variability deviation index of the blood biomarker data, and evaluating the variability of the patient's blood biomarker data; The process of obtaining the variability deviation index of the blood biomarker data is as follows: The patient's blood biomarker data is constructed as a time series, assuming y = [y1, y2, ..., y n ] represents the observed value at time point t; Treat multiple blood biomarker data measured on the same patient as different time series and calculate the DTW distances between them. Calculate the variability deviation index of the blood biomarker data by performing statistical analysis on these distances. The calculation formula of the variability deviation index of the blood biomarker data is: In the formula, Z nt represents the variability deviation index of blood biomarker data, C[s][j] represents the DTW distance between sequence X and sequence Y, n represents the number of DTW distances, and n is a positive integer greater than 0.

5. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 4, characterized in that: The DTW distance calculation process is: Define the cost matrix, calculate the Euclidean distance between each pair of points, and construct the cost matrix D; For two sequences X and Y, the element D[s][j] in the cost matrix D represents the distance between X[s] and Y[j]; Where s represents the number of observation points in sequence X, s is a positive integer greater than 0, j represents the number of observation points in sequence Y, j is a positive integer greater than 0; Fill the cumulative distance matrix C using a dynamic programming algorithm; The element C[s][j] of the cumulative distance matrix C represents the cumulative distance between the first s points of sequence X and the first j points of sequence Y; The updating rule of the cumulative distance matrix is: If s=0 or j=0, C[s][j]=D[s][j]; If s≠0 and j≠0, C[s][j]=D[s][j]+min(C[s-1][j],C[s][j-1],C[s-1][j-1]); The DTW distance is the last element of the cumulative distance matrix, namely C[n][n].

6. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 1, characterized in that: When the patient's blood biomarker data is reliable data, the interrelationships of the indicators in the patient's blood biomarker data are analyzed, the logical deviation index of the blood biomarker data is calculated, and the logic of the patient's blood biomarker data is evaluated; The process of obtaining the logical deviation index of the blood biomarker data is as follows: The patient's blood biomarker data is organized into a data matrix E, where each row represents a blood biomarker data, each column represents an indicator, and the dimension of the data matrix E is set to m*q; Among them, m represents the number of data, m is a positive integer greater than 0, and q represents the number of indicators; The goal of training the autoencoder is to minimize the error between the input blood biomarker data and the reconstructed blood biomarker data, using mean squared error as the loss function and calculating the logistic deviation index; The calculation formula of the logic deviation index is: In the formula, Z q represents the normalized value of the qth blood biomarker data, Z yq represents the reconstructed data of the qth blood biomarker data, ‖Z q -Z yq ‖ 2 represents the reconstruction error of the qth blood biomarker data, L nt Represents the logistic deviation index for blood biomarker data.

7. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 6, characterized in that: The process of training the autoencoder is: Build an autoencoder model, which consists of an encoder and a decoder; The encoder maps high-dimensional input data to a low-dimensional representation; The calculation formula for mapping the high-dimensional input data to low-dimensional is: h=f(W e Z+b e ); In the formula, h represents the mapping of high-dimensional input data to low-dimensional representation, f is the activation function, and W e is the weight matrix of the encoder, b e is the bias vector of the encoder, and Z is the normalized value of the data; The decoder maps the low-dimensional representation back to a high-dimensional space; The calculation formula for mapping the low-dimensional representation back to the high-dimensional space is: Z Y =g(W d h+b d ); In the formula, Z Y Represents mapping the low-dimensional representation back to the high-dimensional space representation, W d is the weight matrix of the decoder, b d is the bias vector of the decoder, and g is the activation function, which is the same as the activation function of the encoder.

8. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 1, characterized in that: The variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data are comprehensively analyzed to calculate the accuracy coefficient of the patient's blood biomarker data and evaluate the accuracy of the blood biomarker data; The accuracy coefficient acquisition process of the patient's blood biomarker data is as follows: The variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data are normalized, and the accuracy coefficient of the patient data is calculated by the variability deviation index of the patient's blood biomarker data and the logical deviation index of the patient's blood biomarker data after the normalization.

9. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 8, characterized in that: The method of dividing the patient's blood biomarker data into accurate data and inaccurate data according to the accuracy evaluation result of the blood biomarker data specifically includes: It is determined whether the accuracy coefficient of the blood biomarker data is greater than or equal to a preset accuracy coefficient threshold of the blood biomarker data. If so, it is recorded as accurate data; if not, it is recorded as inaccurate data.

10. The early diagnosis data management system for patients with schistosomiasis-induced liver fibrosis according to claim 1, characterized in that: The comprehensive evaluation module at least includes: a blood biomarker data variability deviation index calculation unit, the blood biomarker data variability deviation index calculation unit is used to calculate the variability deviation index of the blood biomarker data; A logic deviation index calculation unit for blood biomarker data is used to calculate the logic deviation index of the blood biomarker data.

Images