Urine KIM-1 level intelligent detection method and system and storage medium

Abstract

The invention discloses a urine KIM-1 level intelligent detection method and system and a storage medium, and relates to the related field of biomarker detection.The method comprises the steps that urine KIM-1 data is detected through a urine sample; connecting medical record management information, and identifying accompanying disease information of the current user; inputting the accompanying disease information into a correlation analysis model to obtain a first correlation index of the accompanying disease information; if the first association index is greater than the preset association index, acquiring an accompanying KIM-1 level feature corresponding to the accompanying disease information; according to the accompanying KIM-1 level characteristics and the urine KIM-1 data, renal injury risk prediction is carried out, a first prediction risk index is output, and the medical terminal carries out reminding according to the first prediction risk index. The technical problem of insufficient accuracy and comprehensiveness of kidney health assessment only based on the urine KIM-1 level in the prior art is solved, and the technical effect of improving the accuracy and comprehensiveness of an analysis result is achieved.

Description

Technical Field

[0001] The present application relates to the field of biomarker detection, and in particular to an intelligent detection method, system and storage medium for urine KIM-1 level. Background Art

[0002] In the field of biomarker detection and analysis, the levels of specific molecules in urine can often reflect the physiological state of the human body and potential health risks. Among them, Kidney Injury Molecule-1 (KIM-1) is an important biomarker in urine. Its level changes are closely related to the state of kidney function and are widely used in the assessment of kidney health. At present, some scientific research institutions and medical institutions have begun to use specific technical means to conduct separate tests on the level of KIM-1 in urine in order to obtain a preliminary assessment of kidney health. However, considering only the test results of urine KIM-1 levels as a single indicator will lead to a certain one-sidedness in the analysis results.

[0003] Among the current related technologies, kidney health assessment based solely on urine KIM-1 levels has technical problems such as lack of accuracy and comprehensiveness. Summary of the invention

[0004] The present application provides a method, system and storage medium for intelligent detection of urine KIM-1 level, detects urine KIM-1 data through the urine sample of the current user, connects the medical record management information, identifies the concomitant disease information of the current user, inputs the concomitant disease information into the correlation analysis model, obtains the relationship index (first correlation index) between the concomitant disease information and the urine KIM-1 level, and when the first correlation index is greater than the preset correlation index, further obtains the KIM-1 level characteristics corresponding to the concomitant disease information, and performs a comprehensive analysis in combination with the urine KIM-1 data, outputs a predicted risk index that can reflect the current user's kidney health status, and the medical terminal uses the predicted risk index to give a reminder and other technical means to achieve the technical effect of improving the accuracy and comprehensiveness of the analysis results.

[0005] The present application provides a method for intelligent detection of urine KIM-1 level, comprising: detecting urine KIM-1 data through a urine sample of a current user; connecting medical record management information to identify concomitant disease information of the current user; inputting the concomitant disease information into a correlation analysis model, and obtaining a first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level; if the first correlation index is greater than a preset correlation index, obtaining a concomitant KIM-1 level feature corresponding to the concomitant disease information; predicting the risk of renal injury according to the concomitant KIM-1 level feature and the urine KIM-1 data, outputting a first predicted risk index, and the medical terminal giving a reminder according to the first predicted risk index.

[0006] In a possible implementation, the risk of renal injury is predicted based on the accompanying KIM-1 level feature and the urine KIM-1 data, and the following processing is performed: historical urine KIM-1 sample data of the current user is obtained; the historical urine KIM-1 sample data and the accompanying KIM-1 level feature are used for feature combination to construct an input feature vector; model learning is performed on a multi-layer neural network based on the input feature vector to obtain a renal injury risk prediction model trained to convergence; risk prediction is performed on the accompanying KIM-1 level feature and the urine KIM-1 data based on the renal injury risk prediction model, and a first predicted risk indicator is output.

[0007] In a possible implementation, risk prediction is performed on the accompanying KIM-1 level characteristics and the urine KIM-1 data according to the renal injury risk prediction model, and the following processing is performed: the renal injury risk prediction model obtains a predicted urine KIM-1 curve based on the accompanying KIM-1 level characteristics; detects deviation data between the predicted urine KIM-1 curve and the urine KIM-1 data, and outputs a first predicted risk indicator based on the variance of the deviation data.

[0008] In a possible implementation, the concomitant disease information is input into a correlation analysis model, and according to the correlation analysis model, a first correlation index of the concomitant disease information is obtained, and the following processing is performed: a training data group is obtained, wherein the training data group includes multiple disease type samples, the severity of each disease type, and urine KIM-1 data samples collected corresponding to each disease type and the severity of each disease type; a regression model is trained according to the training data group and identification information identifying the degree of fluctuation of the urine KIM-1 data samples, and when the accuracy of the regression model is greater than a preset accuracy, a correlation analysis model is obtained; the concomitant disease information is analyzed according to the correlation analysis model to obtain a first correlation index.

[0009] In a possible implementation, after obtaining the correlation analysis model, the following processing is also performed: obtaining multiple correlation indicators corresponding to the multiple disease type samples respectively; screening N disease type samples that are greater than the preset correlation indicators; constructing a disease type matching library based on the N disease type samples; matching the concomitant disease information of the current user based on the disease type matching library, and obtaining a matching return result.

[0010] In a possible implementation, after obtaining the matching return result, the following processing is also performed: if the match is successful, the accompanying KIM-1 level characteristics corresponding to the accompanying disease information are obtained, and the risk of renal injury is predicted based on the accompanying KIM-1 level characteristics and the urine KIM-1 data, and a first predicted risk index is output, wherein the accompanying KIM-1 level characteristics are the fluctuation characteristics of the urine KIM-1 data; if the match is unsuccessful, the risk of renal injury is predicted based on the urine KIM-1 data, and the first predicted risk index is output.

[0011] In a possible implementation, urine KIM-1 data is detected by using a urine sample of the current user, and the following processing is performed: the urine sample is dropped onto a fluorescence detection chip, and the urine sample on the fluorescence detection chip is irradiated by an irradiation unit of a user mobile terminal to collect fluorescence intensity change data; the fluorescence intensity change data is converted to obtain urine KIM-1 data, and the user mobile terminal uploads the urine KIM-1 data to the medical terminal.

[0012] The present application also provides a urine KIM-1 level intelligent detection system, including: a urine detection module, the urine detection module is used to detect urine KIM-1 data through a urine sample of a current user; a concomitant disease information identification module, the concomitant disease information identification module is used to connect medical record management information and identify the concomitant disease information of the current user; a correlation analysis module, the correlation analysis module is used to input the concomitant disease information into a correlation analysis model, and obtain a first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level; a concomitant KIM-1 level feature acquisition module, the concomitant KIM-1 level feature acquisition module is used to obtain the concomitant KIM-1 level feature corresponding to the concomitant disease information if the first correlation index is greater than a preset correlation index; a renal injury risk prediction module, the renal injury risk prediction module is used to predict the risk of renal injury according to the concomitant KIM-1 level feature and the urine KIM-1 data, and output a first predicted risk index, and the medical terminal reminds according to the first predicted risk index.

[0013] The present application also provides a computer-readable storage medium, including: a computer program stored thereon, which implements a method for intelligent detection of urine KIM-1 levels when executed by a processor.

[0014] The present application proposes a method, system and storage medium for intelligent detection of urine KIM-1 levels. First, urine KIM-1 data is detected through a urine sample of the current user. Then, the medical record management information is connected to identify the concomitant disease information of the current user. Then, the concomitant disease information is input into a correlation analysis model. According to the correlation analysis model, a first correlation index of the concomitant disease information is obtained, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level. If the first correlation index is greater than the preset correlation index, the concomitant KIM-1 level feature corresponding to the concomitant disease information is obtained. Finally, the risk of renal injury is predicted based on the concomitant KIM-1 level feature and the urine KIM-1 data, and the first predicted risk index is output. The medical terminal gives a reminder based on the first predicted risk index, thereby achieving the technical effect of improving the accuracy and comprehensiveness of the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0016] Figure 1 A schematic diagram of a flow chart of an intelligent method for detecting KIM-1 levels in urine provided in an embodiment of the present application.

[0017] Figure 2 This is a schematic diagram of the structure of a urine KIM-1 level intelligent detection system provided in an embodiment of the present application.

[0018] Explanation of the accompanying reference numerals: urine detection module 10 , concomitant disease information identification module 20 , correlation analysis module 30 , concomitant KIM-1 level feature acquisition module 40 , renal injury risk prediction module 50 . DETAILED DESCRIPTION

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0020] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0021] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0022] The present application embodiment provides a method for intelligent detection of urine KIM-1 level, such as Figure 1 As shown, the method includes: step S100, detecting urine KIM-1 data through the urine sample of the current user. Specifically, the urine sample of the current user is collected, and the KIM-1 molecules in the urine sample are detected using a specific detection technology (such as fluorescence detection, biochemical analysis, etc.), so as to obtain the concentration or activity data of KIM-1 in the urine. Among them, KIM-1 is a molecule that is upregulated when the kidney is damaged, and its level reflects the health of the kidney.

[0023] In a possible implementation, urine KIM-1 data is detected by a urine sample of the current user, and step S100 further includes step S110, wherein the urine sample is dropped onto a fluorescence detection chip, and the urine sample on the fluorescence detection chip is irradiated based on an irradiation unit of a user mobile terminal to collect fluorescence intensity change data. Specifically, a urine sample is collected from the current user, and an appropriate amount of the urine sample is dropped into a designated area on the fluorescence detection chip according to the instructions, wherein the fluorescence detection chip is a detection carrier, and a fluorescent substance that can bind to a specific component (KIM-1) in urine is coated on the surface. When the urine sample contacts the fluorescent substance on the chip, a specific chemical reaction occurs, resulting in a change in fluorescence intensity. When the fluorescent substance on the fluorescence detection chip reacts with the urine sample, the fluorescence intensity changes. These changes are captured by an irradiation unit (a part of the user's mobile terminal used to emit light to illuminate the fluorescence detection chip) of a user's mobile terminal (a portable electronic device such as a mobile phone or a tablet computer with specific functions held by the user) and converted into fluorescence intensity change data. Step S120, converting the fluorescence intensity change data to obtain urine KIM-1 data, and the user mobile terminal uploads the urine KIM-1 data to the medical terminal. Specifically, based on the reaction relationship between the known fluorescent substance and KIM-1 and the standard curve, the collected fluorescence intensity change data is converted to obtain the concentration or level of KIM-1 in the urine. Finally, the user mobile terminal uploads the urine KIM-1 data to the medical terminal for subsequent analysis and prediction. This implementation method is based on the fluorescence detection chip and the user mobile terminal to collect and obtain urine KIM-1 data. Users can collect urine samples at home or other convenient places, and test them through the user mobile terminal without going to a medical institution, thereby improving the convenience and accessibility of the test.

[0024] Step S200, connect the medical record management information and identify the concomitant disease information of the current user. Specifically, through the medical record management information system, the medical record management information system records the user's health status, medical history, diagnosis, treatment and other information, and obtains the current user's medical record information from the medical record management information system. Then, extract the user's concomitant disease information from the medical record information, that is, the disease information that the user currently or has suffered from, which may affect kidney health, such as hypertension, diabetes, etc.

[0025] Step S300, input the concomitant disease information into the correlation analysis model, and obtain the first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level. Specifically, the acquired concomitant disease information is input into a pre-trained correlation analysis model, and the correlation analysis model is a machine learning model used to analyze the degree of correlation between different variables. The model calculates the degree of correlation between the current user's concomitant disease and the urine KIM-1 level by analyzing the relationship between a large number of diseases and the urine KIM-1 level, that is, the first correlation index. The first correlation index is a quantitative value, which indicates the degree of correlation between the concomitant disease and the urine KIM-1 level.

[0026] In a possible implementation, the concomitant disease information is input into a correlation analysis model, and a first correlation index of the concomitant disease information is obtained according to the correlation analysis model. Step S300 further includes step S310, obtaining a training data set, wherein the training data set includes a plurality of disease type samples, the severity of each disease type, and urine KIM-1 data samples corresponding to each disease type and the severity of each disease type. Specifically, the training data set is a set containing a plurality of data points, which is used to train a machine learning model. The training data set includes a plurality of disease type samples (different diseases, such as diabetes, hypertension, etc.), the severity of each disease type (such as mild, moderate, severe), and urine KIM-1 data samples corresponding to each disease type and the severity of each disease type (KIM-1 level data detected from the user's urine under different disease types and severities). Step S320, according to the training data set and the identification information identifying the fluctuation degree of the urine KIM-1 data sample, a regression model is trained, and when the accuracy of the regression model is greater than the preset accuracy, a correlation analysis model is obtained. Specifically, the regression model is a machine learning model for predicting the relationship between the urine KIM-1 level (a continuous value) and the type and severity of the disease. The identification information for identifying the fluctuation degree of the urine KIM-1 data sample is used to measure the fluctuation degree of the data, which can be the standard deviation, variance or other statistics of the urine KIM-1 data, and is used to enable the model to better learn the pattern in the data. The accuracy of the regression model is an indicator for measuring the consistency between the model prediction result and the actual result. When the accuracy of the model is higher than the preset accuracy, it is considered that the model has been fully trained and can be used for subsequent analysis and prediction. Step S330, analyze the concomitant disease information according to the correlation analysis model to obtain a first correlation index. Specifically, the concomitant disease information is input into the correlation analysis model, and the correlation analysis model analyzes the concomitant disease information of the current user, and outputs a first correlation index, which is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level. This implementation method uses the powerful ability of machine learning to predict the changes in the user's urine KIM-1 level through a data-driven method, thereby improving the accuracy and efficiency of the prediction.

[0027] In a possible implementation, after obtaining the correlation analysis model, step S300 further includes step S340, obtaining multiple correlation indicators corresponding to the multiple disease type samples. Specifically, according to the correlation analysis model obtained by training, each disease type sample (including different disease types and their corresponding severity) is analyzed, and the correlation indicators between each of them and the urine KIM-1 level are calculated. These correlation indicators reflect the degree of influence of different disease types and their severity on the urine KIM-1 level. Step S350, screening N disease type samples greater than the preset correlation index. Specifically, according to the preset correlation index threshold (the threshold for judging whether the correlation between the concomitant disease and the urine KIM-1 level is significant), those disease type samples with higher correlation indicators are screened. These samples are considered to be disease types that are closely related to the urine KIM-1 level, so they need to be paid special attention in subsequent analysis. Wherein, N is the number of disease type samples screened.

[0028] Step S360, construct a disease type matching library based on the N disease type samples. Specifically, the screened N disease type samples and their related information (such as disease name, severity, associated indicators, etc.) are organized into a database, namely the disease type matching library, which is used to subsequently match the concomitant disease information of the current user. Step S370, match the concomitant disease information of the current user according to the disease type matching library, and obtain a matching return result. Specifically, the concomitant disease information of the current user is matched with the disease type in the disease type matching library to find out whether there is a disease type sample similar to the disease information of the current user.

[0029] Step S380, if the match is successful, obtain the accompanying KIM-1 level feature corresponding to the accompanying disease information, predict the risk of renal injury according to the accompanying KIM-1 level feature and the urine KIM-1 data, and output a first predicted risk index, wherein the accompanying KIM-1 level feature is the fluctuation feature of the urine KIM-1 data. Specifically, if the match is successful, that is, disease type samples similar to the disease information of the current user are found, then the accompanying KIM-1 level features corresponding to these samples (fluctuation features of urine KIM-1 data, etc.) are obtained, and the risk of renal injury is predicted in combination with the urine KIM-1 data of the current user, and the first predicted risk index is output. Step S390, if the match is unsuccessful, predict the risk of renal injury according to the urine KIM-1 data, and output the first predicted risk index. Specifically, if the match is unsuccessful, that is, no disease type samples similar to the disease information of the current user are found, then the risk of renal injury is predicted only according to the urine KIM-1 data of the current user, and the first predicted risk index is output. This implementation method can quickly identify disease type samples similar to the current user's disease information by building a disease type matching library, and obtain the corresponding KIM-1 level characteristics of these samples. In this way, when predicting the risk of kidney injury, various factors can be considered more comprehensively to improve the accuracy and reliability of the prediction. At the same time, for users who have not been successfully matched, basic kidney injury risk prediction can also be performed based on urine KIM-1 data to ensure that each user can get a corresponding risk assessment.

[0030] Step S400, if the first correlation index is greater than the preset correlation index, the accompanying KIM-1 level characteristics corresponding to the accompanying disease information are obtained. Specifically, it is determined whether the first correlation index is greater than the preset correlation index. If it is greater, the urine KIM-1 level characteristics corresponding to the accompanying disease information of the current user are further obtained, including the normal range of KIM-1 levels under specific diseases, abnormal change patterns, etc.

[0031] Step S500, predict the risk of renal injury based on the accompanying KIM-1 level characteristics and the urine KIM-1 data, output a first predicted risk index, and the medical terminal reminds according to the first predicted risk index. Specifically, the renal injury risk prediction model is used to predict the risk of renal injury in combination with the accompanying KIM-1 level characteristics and the urine KIM-1 data. The prediction result is output in the form of a first predicted risk index, which reflects the user's current risk level of renal injury. The first predicted risk index can be a probability value (such as the probability of renal injury), a score (such as a risk score) or a classification label (such as low risk, medium risk, high risk). The medical terminal receives the first predicted risk index and sends a reminder to the user or medical staff according to preset rules. The embodiment of the present application detects urine KIM-1 data through the urine sample of the current user, connects the medical record management information, identifies the concomitant disease information of the current user, inputs the concomitant disease information into the correlation analysis model, obtains the relationship index (first correlation index) between the concomitant disease information and the urine KIM-1 level, and when the first correlation index is greater than the preset correlation index, further obtains the KIM-1 level characteristics corresponding to the concomitant disease information, and performs a comprehensive analysis in combination with the urine KIM-1 data, outputs a predicted risk index that can reflect the current user's kidney health status, and the medical terminal uses technical means such as giving reminders based on the predicted risk index to achieve the technical effect of improving the accuracy and comprehensiveness of the analysis results.

[0032] In a possible implementation, the risk of renal injury is predicted based on the accompanying KIM-1 level feature and the urine KIM-1 data, and step S500 further includes step S510, obtaining the historical urine KIM-1 sample data of the current user. Specifically, the urine KIM-1 test data previously submitted by the current user is retrieved from the medical record management information system. These data are organized into a time series or a list sorted by test time. These data provide a historical record of the user's kidney health status. Step S520, the historical urine KIM-1 sample data and the accompanying KIM-1 level feature are combined to construct an input feature vector. Specifically, the historical urine KIM-1 sample data is merged with the accompanying KIM-1 level feature, and these features are normalized or standardized to ensure that they have similar weights in model training. Construct an input feature vector containing all relevant features for subsequent model training.

[0033] Step S530, model learning is performed on the multi-layer neural network according to the input feature vector to obtain a renal injury risk prediction model trained to convergence. Specifically, a multi-layer neural network architecture (such as a deep neural network, a convolutional neural network, etc.) is selected, which can handle complex feature relationships in the input feature vector. The input feature vector is used as training data to train the multi-layer neural network. The prediction error is minimized by iterative optimization (such as gradient descent) until the model converges (that is, the prediction performance is no longer significantly improved). The trained model is saved as a renal injury risk prediction model. Step S540, risk prediction is performed on the accompanying KIM-1 level feature and the urine KIM-1 data according to the renal injury risk prediction model, and a first predicted risk index is output. Specifically, the accompanying KIM-1 level feature and the urine KIM-1 data of the current user are input into the trained renal injury risk prediction model, and the model calculates the renal injury risk according to the input feature vector, and outputs a quantitative risk index as the first predicted risk index. This implementation method combines historical urine KIM-1 data and concomitant disease information, and learns complex feature relationships through a multi-layer neural network. The historical urine KIM-1 data provides time series information on the user's kidney health status, which helps capture the changing trend of kidney function over time and improves the accuracy of predictions.

[0034] In a possible implementation, the risk prediction of the accompanying KIM-1 level feature and the urine KIM-1 data is performed according to the renal injury risk prediction model, and step S540 further includes step S541, wherein the renal injury risk prediction model obtains a predicted urine KIM-1 curve according to the accompanying KIM-1 level feature. Specifically, the renal injury risk prediction model first receives the accompanying KIM-1 level feature as input, and these features include the KIM-1 level change trend, normal range or other indicators related to the KIM-1 level obtained based on the user's accompanying disease information. The model uses these features and combines the complex relationships learned internally to generate a predicted urine KIM-1 curve. This curve represents the possible urine KIM-1 level change trend of the user in the future period of time under the given accompanying KIM-1 level feature. The predicted urine KIM-1 curve is presented in the form of time series data, including a series of time points and corresponding KIM-1 level prediction values. Step S542, detect the deviation data between the predicted urine KIM-1 curve and the urine KIM-1 data, and output the first predicted risk index according to the variance of the deviation data. Specifically, the actual urine KIM-1 data of the current user is obtained, the actual urine KIM-1 data is compared with the predicted urine KIM-1 curve, and the deviation data between them is calculated by calculating the difference between the actual value and the predicted value at each time point. The variance of these deviation data is calculated to quantify the degree of discreteness between the predicted value and the actual value. The larger the variance, the greater the difference between the predicted value and the actual value, which means a higher risk of renal injury. According to the calculated variance, a quantitative first predicted risk index is output, which indicates the level of renal injury risk currently faced by the user. This implementation method generates a predicted urine KIM-1 curve based on the accompanying KIM-1 level characteristics, providing a basis for evaluating the changing trend of the user's future urine KIM-1 level. By comparing the actual urine KIM-1 data with the predicted curve, the difference between the predicted value and the actual value is quantified. The variance of the deviation data is used as a quantitative indicator to intuitively represent the user's current kidney injury risk level, thereby improving the accuracy of kidney injury risk assessment.

[0035] In the above, refer to Figure 1 A method for intelligent detection of urine KIM-1 level according to an embodiment of the present invention is described in detail. Figure 2 A urine KIM-1 level intelligent detection system according to an embodiment of the present invention is described.

[0036] According to an embodiment of the present invention, a urine KIM-1 level intelligent detection system is used to solve the technical problem of insufficient accuracy and comprehensiveness of kidney health assessment based only on urine KIM-1 levels, thereby achieving the technical effect of improving the accuracy and comprehensiveness of analysis results. A urine KIM-1 level intelligent detection system includes: a urine detection module 10, a concomitant disease information identification module 20, a correlation analysis module 30, a concomitant KIM-1 level feature acquisition module 40, and a kidney injury risk prediction module 50.

[0037] The urine detection module 10 is used to detect urine KIM-1 data through the urine sample of the current user; the concomitant disease information identification module 20 is used to connect the medical record management information and identify the concomitant disease information of the current user; the correlation analysis module 30 is used to input the concomitant disease information into the correlation analysis model, and obtain the first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level; the concomitant KIM-1 level feature acquisition module 40 is used to obtain the concomitant KIM-1 level feature corresponding to the concomitant disease information if the first correlation index is greater than the preset correlation index; the renal injury risk prediction module 50 is used to predict the risk of renal injury based on the concomitant KIM-1 level feature and the urine KIM-1 data, and output the first predicted risk index, and the medical terminal gives a reminder based on the first predicted risk index.

[0038] The specific configuration of the renal injury risk prediction module 50 will be described in detail below. As described above, the renal injury risk prediction is performed based on the accompanying KIM-1 level feature and the urine KIM-1 data, and the renal injury risk prediction module 50 may further include: a historical urine KIM-1 sample data acquisition unit for acquiring the historical urine KIM-1 sample data of the current user; a feature combination unit for combining the historical urine KIM-1 sample data and the accompanying KIM-1 level feature to construct an input feature vector; a multi-layer neural network learning unit for model learning of the multi-layer neural network according to the input feature vector to obtain a renal injury risk prediction model trained to convergence; a risk prediction unit for performing risk prediction on the accompanying KIM-1 level feature and the urine KIM-1 data according to the renal injury risk prediction model, and outputting a first predicted risk index.

[0039] Among them, the risk prediction of the accompanying KIM-1 level characteristics and the urine KIM-1 data is performed according to the renal injury risk prediction model, and the risk prediction unit may further include: a predicted urine KIM-1 curve acquisition subunit is used for the renal injury risk prediction model to obtain a predicted urine KIM-1 curve according to the accompanying KIM-1 level characteristics; a deviation data detection subunit is used to detect the deviation data between the predicted urine KIM-1 curve and the urine KIM-1 data, and output a first prediction risk indicator according to the variance of the deviation data.

[0040] The specific configuration of the correlation analysis module 30 will be described in detail below. As described above, the concomitant disease information is input into the correlation analysis model, and the first correlation index of the concomitant disease information is obtained according to the correlation analysis model. The correlation analysis module 30 may further include: a training data group acquisition unit for acquiring a training data group, wherein the training data group includes a plurality of disease type samples, the severity of each disease type, and urine KIM-1 data samples collected corresponding to each disease type and the severity of each disease type; a regression model training unit for performing regression model training according to the training data group and the identification information identifying the fluctuation degree of the urine KIM-1 data sample, and when the accuracy of the regression model is greater than the preset accuracy, the correlation analysis model is obtained; and a first correlation index acquisition unit for analyzing the concomitant disease information according to the correlation analysis model to obtain the first correlation index.

[0041] Among them, after obtaining the correlation analysis model, the correlation analysis module 30 may further include: a plurality of correlation indicator acquisition units for obtaining a plurality of correlation indicators corresponding to the plurality of disease type samples respectively; a disease type sample screening unit for screening N disease type samples greater than the preset correlation indicators; a disease type matching library construction unit for constructing a disease type matching library based on the N disease type samples; a matching unit for matching the concomitant disease information of the current user based on the disease type matching library to obtain a matching return result.

[0042] Among them, after obtaining the matching return result, the correlation analysis module 30 may further include: a renal injury risk prediction unit is used to obtain the accompanying KIM-1 level characteristics corresponding to the accompanying disease information if the matching is successful, and predict the risk of renal injury according to the accompanying KIM-1 level characteristics and the urine KIM-1 data, and output a first predicted risk index, wherein the accompanying KIM-1 level characteristics are the fluctuation characteristics of the urine KIM-1 data; if the matching is unsuccessful, predict the risk of renal injury according to the urine KIM-1 data, and output the first predicted risk index.

[0043] The specific configuration of the urine detection module 10 will be described in detail below. As described above, the urine KIM-1 data is detected by the urine sample of the current user, and the urine detection module 10 may further include: a fluorescence intensity change data acquisition unit for dropping the urine sample onto the fluorescence detection chip, and irradiating the urine sample on the fluorescence detection chip based on the irradiation unit of the user's mobile terminal to collect fluorescence intensity change data; a urine KIM-1 data uploading unit for converting the fluorescence intensity change data to obtain urine KIM-1 data, and the user's mobile terminal uploads the urine KIM-1 data to the medical terminal.

[0044] The urine KIM-1 level intelligent detection system provided by the embodiment of the present invention can execute the urine KIM-1 level intelligent detection method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0045] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0046] Based on the foregoing embodiments, the embodiments of the present application further provide a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, a method for intelligent detection of urine KIM-1 levels as described in any of the previous embodiments can be implemented.

[0047] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for intelligent detection of urine KIM-1 level, characterized in that: The method comprises: Detect urine KIM-1 data through the urine sample of the current user; Connect to medical record management information and identify the current user's concomitant disease information; Inputting the concomitant disease information into a correlation analysis model, and obtaining a first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level; If the first correlation index is greater than a preset correlation index, obtaining a concomitant KIM-1 level feature corresponding to the concomitant disease information; The risk of renal injury is predicted based on the accompanying KIM-1 level characteristics and the urine KIM-1 data, and a first predicted risk indicator is output, and the medical terminal issues a reminder based on the first predicted risk indicator.

2. The method for intelligent detection of urine KIM-1 level according to claim 1, characterized in that: Predicting the risk of renal injury based on the accompanying KIM-1 level characteristics and the urine KIM-1 data, the method comprising: Get the historical urine KIM-1 sample data of the current user; Combining the historical urine KIM-1 sample data with the accompanying KIM-1 level features to construct an input feature vector; Performing model learning on the multi-layer neural network according to the input feature vector to obtain a renal injury risk prediction model trained to convergence; The accompanying KIM-1 level characteristics and the urine KIM-1 data are subjected to risk prediction according to the renal injury risk prediction model, and a first predicted risk index is output.

3. The method for intelligent detection of urine KIM-1 level according to claim 2, characterized in that: The risk prediction of the accompanying KIM-1 level characteristics and the urine KIM-1 data is performed according to the renal injury risk prediction model, the method comprising: The renal injury risk prediction model obtains a predicted urine KIM-1 curve according to the accompanying KIM-1 level characteristics; The deviation data between the predicted urine KIM-1 curve and the urine KIM-1 data is detected, and a first predicted risk index is output according to the variance of the deviation data.

4. The method for intelligent detection of urine KIM-1 level according to claim 1, characterized in that: Inputting the concomitant disease information into a correlation analysis model, and obtaining a first correlation index of the concomitant disease information according to the correlation analysis model, the method comprising: Acquire a training data set, wherein the training data set includes a plurality of disease type samples, the severity of each disease type, and urine KIM-1 data samples collected corresponding to each disease type and the severity of each disease type; Performing regression model training according to the training data group and identification information identifying the degree of fluctuation of urine KIM-1 data samples, and obtaining a correlation analysis model when the accuracy of the regression model is greater than a preset accuracy; The concomitant disease information is analyzed according to the correlation analysis model to obtain a first correlation index.

5. The method for intelligent detection of urine KIM-1 level according to claim 4, characterized in that: After obtaining the correlation analysis model, the method further includes: Obtaining multiple correlation indicators corresponding to the multiple disease type samples respectively; Screening N disease type samples with values ​​greater than the preset correlation index; Constructing a disease type matching library according to the N disease type samples; The concomitant disease information of the current user is matched according to the disease type matching library to obtain a matching return result.

6. The method for intelligent detection of urine KIM-1 level according to claim 5, characterized in that: After obtaining the matching return results, the method includes: If the match is successful, obtaining the accompanying KIM-1 level feature corresponding to the accompanying disease information, predicting the risk of renal injury according to the accompanying KIM-1 level feature and the urine KIM-1 data, and outputting a first predicted risk index, wherein the accompanying KIM-1 level feature is a fluctuation feature of the urine KIM-1 data; If the match is unsuccessful, renal injury risk prediction is performed based on the urine KIM-1 data, and a first predicted risk index is output.

7. The method for intelligent detection of urine KIM-1 level according to claim 1, characterized in that: Detect urine KIM-1 data through the current user's urine sample, including: Dropping the urine sample onto the fluorescence detection chip, irradiating the urine sample on the fluorescence detection chip based on the irradiation unit of the user mobile terminal, and collecting fluorescence intensity change data; The fluorescence intensity change data is converted to obtain urine KIM-1 data, and the user mobile terminal uploads the urine KIM-1 data to the medical terminal.

8. An intelligent detection system for urine KIM-1 level, characterized in that: The system is used to implement the method for intelligent detection of urine KIM-1 level according to any one of claims 1 to 7, and the system comprises: A urine detection module, the urine detection module is used to detect urine KIM-1 data through a urine sample of the current user; A concomitant disease information identification module, which is used to connect to the medical record management information and identify the concomitant disease information of the current user; A correlation analysis module, the correlation analysis module is used to input the concomitant disease information into a correlation analysis model, and obtain a first correlation index of the concomitant disease information according to the correlation analysis model, wherein the first correlation index is used to characterize the relationship between the user's concomitant disease and the urine KIM-1 level; A concomitant KIM-1 level feature acquisition module, wherein the concomitant KIM-1 level feature acquisition module is used to acquire the concomitant KIM-1 level feature corresponding to the concomitant disease information if the first correlation index is greater than a preset correlation index; A renal injury risk prediction module is used to predict the risk of renal injury based on the accompanying KIM-1 level characteristics and the urine KIM-1 data, output a first predicted risk indicator, and the medical terminal issues a reminder based on the first predicted risk indicator.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, an intelligent detection method for urine KIM-1 level according to any one of claims 1 to 7 is implemented.

Images