High-risk group health record management system and method

Abstract

The invention discloses a high-risk group health archive management system and method, and the system carries out the semantic understanding and analysis of a health archive of a to-be-evaluated user through an advanced data processing technology and a deep learning algorithm, carries out the multi-mode interaction coupling of information in the health archive, so as to form comprehensive user feature representation. Therefore, automatic filing is realized by utilizing the classifier, whether the user belongs to a high-risk group or not is intelligently judged, and powerful technical support is provided for public health management.

Description

Technical Field

[0001] The present application relates to the field of intelligent management, and more specifically, to a health record management system and method for high-risk groups. Background Art

[0002] With the development of society and the aging of the population, the incidence of chronic diseases and non-communicable diseases continues to rise, posing a huge challenge to the public health system. In order to effectively prevent and manage these health problems, early identification and intervention of high-risk groups are particularly important. High-risk groups refer to individuals who have a higher risk of disease due to factors such as genetics, environment, lifestyle or medical history. Establishing health records for high-risk groups can help medical institutions effectively track individuals who face higher health risks.

[0003] However, traditional methods for identifying high-risk groups often rely on limited medical resources and the judgment of professionals, which is not only inefficient but also subjective and inaccurate. Therefore, an optimized health record management system and method for high-risk groups is expected. Summary of the invention

[0004] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides a high-risk population health record management system and method, which uses advanced data processing technology and deep learning algorithms to semantically understand and analyze the health records of the user to be evaluated, and multi-modally interactively couples the information in the health records to form a comprehensive user feature representation, thereby using a classifier to achieve automatic archiving, and intelligently judge whether the user belongs to the high-risk group, providing strong technical support for public health management.

[0005] According to one aspect of the present application, a high-risk population health record management system is provided, which includes: A health record acquisition module is used to obtain the health record of the user to be evaluated; A multimodal information extraction module, used to extract basic user information, user living habit information and user medical history data from the health record of the user to be evaluated; A semantic coding module, used for semantically coding the user basic information, the user living habit information and the user medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; A multimodal information interaction fusion module, used for interactively fusing the user basic information semantic coding vector, the user living habit semantic coding vector and the user medical history data semantic coding vector based on an equilibrium threshold and a coupling interaction weight to obtain a user multimodal information interaction coupling representation vector; The archiving result generating module is used to determine the archiving result based on the user multimodal information interaction coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

[0006] According to another aspect of the present application, a method for managing health records of high-risk groups is provided, comprising: Obtain the health records of the user to be evaluated; Extracting basic information of the user, information on the user's living habits and medical history data from the health record of the user to be evaluated; Performing semantic coding on the user basic information, the user living habit information and the user medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; Perform interactive fusion based on the equalization threshold and coupling interaction weight on the semantic coding vector of the user's basic information, the semantic coding vector of the user's living habits, and the semantic coding vector of the user's medical history data to obtain a user multimodal information interactive coupling representation vector; The archiving result is determined based on the user multimodal information interactive coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

[0007] Compared with the existing technology, the present application provides a health record management system and method for high-risk groups. It uses advanced data processing technology and deep learning algorithms to semantically understand and analyze the health records of users to be evaluated, and multimodally interactively couples the information in the health records to form a comprehensive user feature representation, thereby using classifiers to achieve automatic archiving and intelligently determine whether users belong to high-risk groups, providing strong technical support for public health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 is a block diagram of a high-risk population health record management system according to an embodiment of the present application; Figure 2 This is a system architecture diagram of a high-risk population health record management system according to an embodiment of the present application; Figure 3 A block diagram of a multimodal information interaction and fusion module in a high-risk population health record management system according to an embodiment of the present application; Figure 4 is a block diagram of an adaptive fusion unit in a high-risk population health record management system according to an embodiment of the present application; Figure 5 The present invention is a flowchart of a method for managing health records of high-risk groups according to an embodiment of the present application. DETAILED DESCRIPTION

[0010] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0011] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0012] 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 can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0013] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments 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 may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0014] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0015] Traditional methods for identifying high-risk groups often rely on limited medical resources and the judgment of professionals, which is not only inefficient but also subjective and inaccurate. Therefore, an optimized health record management system and method for high-risk groups is expected.

[0016] In the technical solution of the present application, a health record management system for high-risk groups is proposed. Figure 1 The block diagram of the high-risk population health record management system according to an embodiment of the present application. Figure 2: is a system architecture diagram of a high-risk population health record management system according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the health record management system for high-risk groups includes: a health record acquisition module 310, which is used to acquire the health record of the user to be evaluated; a multimodal information extraction module 320, which is used to extract user basic information, user living habit information and user medical history data from the health record of the user to be evaluated; a semantic coding module 330, which is used to semantically encode the user basic information, the user living habit information and the user medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; a multimodal information interactive fusion module 340, which is used to interactively fuse the user basic information semantic coding vector, the user living habit semantic coding vector and the user medical history data semantic coding vector based on an equalization threshold and a coupling interactive weight to obtain a user multimodal information interactive coupling representation vector; an archiving result generation module 350, which is used to determine the archiving result based on the user multimodal information interactive coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

[0017] In particular, the health record acquisition module 310 and the multimodal information extraction module 320 are used to obtain the health record of the user to be evaluated; and extract the user's basic information, user's living habit information and user's medical history data from the health record of the user to be evaluated. Here, in actual application scenarios, the health record usually contains comprehensive medical information of the user to be evaluated. For example, the basic information of the user includes age, gender, occupation, etc., which are closely related to the individual's health status and disease risk; the user's living habit information such as eating, exercising, sleeping, smoking and drinking habits, these habits directly affect the individual's health status, and unhealthy living habits may increase the risk of certain diseases; and the user's medical history data includes past medical history, family history, chronic disease conditions, surgical history, etc., which are helpful to understand the individual's health status and disease development trend. Although the health record contains comprehensive medical information of the user to be evaluated, the identification of high-risk groups still relies on professional medical personnel to judge and analyze, which may lead to inefficiency. Therefore, in the technical solution of this application, it is expected to use intelligent algorithms to analyze the above data to automatically determine whether the user to be evaluated belongs to the high-risk group, and provide support for the management of health records for high-risk groups.

[0018] In particular, the semantic encoding module 330 is used to semantically encode the user basic information, the user living habit information and the user medical history data to obtain the user basic information semantic encoding vector, the user living habit semantic encoding vector and the user medical history data semantic encoding vector. In the technical solution of the present application, the user basic information, the user living habit information and the user medical history data are semantically encoded to obtain the user basic information semantic encoding vector, the user living habit semantic encoding vector and the user medical history data semantic encoding vector. That is, the key semantic meanings are extracted from the user basic information, the user living habit information and the user medical history data through the processing method of semantic encoding, and the text data that is difficult to process directly is converted into a structured vector representation, which is helpful for the subsequent model to read and analyze. Specifically, in a specific example of the present application, a semantic encoder including a Transformer layer is used to semantically encode the user basic information, the user living habit information and the user medical history data to obtain the user basic information semantic encoding vector, the user living habit semantic encoding vector and the user medical history data semantic encoding vector. It is worth mentioning that the main feature of Transformer is based on the attention mechanism, which avoids some limitations of traditional recurrent neural networks (RNNs) and convolutional neural networks (CNNs), such as long-distance dependencies and parallel computing capabilities.

[0019] In particular, the multimodal information interaction fusion module 340 is used to interactively fuse the user basic information semantic coding vector, the user life habit semantic coding vector and the user medical history data semantic coding vector based on the equalization threshold and the coupling interaction weight to obtain the user multimodal information interaction coupling representation vector. In particular, in a specific example of the present application, if Figure 3 As shown, the multimodal information interaction fusion module 340 includes: an adaptive fusion unit 341, which is used to input the user basic information semantic coding vector and the user medical history data semantic coding vector into the equalization threshold feature vector adaptive fusion module to obtain the user basic information-user medical history semantic fusion representation vector; a nonlinear coupling unit 342, which is used to input the user basic information-user medical history semantic fusion representation vector and the user life habit semantic coding vector into the nonlinear coupling interaction module to obtain the user multimodal information interaction coupling representation vector.

[0020] Specifically, the adaptive fusion unit 341 is used to input the user basic information semantic coding vector and the user medical history data semantic coding vector into the equalization threshold feature vector adaptive fusion module to obtain the user basic information-user medical history semantic fusion representation vector. It should be understood that since the user's basic information and medical history are key factors in assessing individual health status, and there may be complex correlations and complementary information between the two. For example, some genetic tendencies may not be directly reflected in the medical history, but the family history in the basic information can provide clues in this regard. However, traditional fusion methods such as adding by position may ignore the importance between basic information and medical history information, resulting in the loss of important information in the fused feature representation. Therefore, in the technical solution of the present application, it is expected that the user basic information semantic coding vector and the user medical history data semantic coding vector are input into the equalization threshold feature vector adaptive fusion module to obtain the user basic information-user medical history semantic fusion representation vector. In particular, in a specific example of the present application, such as Figure 4 As shown, the adaptive fusion unit 341 includes: an equilibrium threshold dynamic fusion weight determination subunit 3411, which is used to determine the first equilibrium threshold dynamic fusion weight, the second equilibrium threshold dynamic fusion weight and the third equilibrium threshold dynamic fusion weight based on the user basic information semantic coding vector and the user medical history data semantic coding vector; an adaptive dynamic integration subunit 3412, which is used to adaptively and dynamically integrate the user basic information semantic coding vector and the user medical history data semantic coding vector with the first equilibrium threshold dynamic fusion weight, the second equilibrium threshold dynamic fusion weight and the third equilibrium threshold dynamic fusion weight to obtain the user basic information semantic dynamic interactive fusion item and the user medical history semantic dynamic interactive fusion item; an element-by-element interactive fusion subunit 3413, which is used to perform element-by-element fusion on the user basic information semantic dynamic interactive fusion item and the user medical history semantic dynamic interactive fusion item to obtain the user basic information-user medical history semantic fusion representation vector.

[0021] More specifically, the equalization threshold dynamic fusion weight determination subunit 3411 is used to determine the first equalization threshold dynamic fusion weight, the second equalization threshold dynamic fusion weight and the third equalization threshold dynamic fusion weight based on the user basic information semantic coding vector and the user medical history data semantic coding vector. Here, the establishment of the first equalization threshold dynamic fusion weight, the second equalization threshold dynamic fusion weight and the third equalization threshold dynamic fusion weight is allocated according to the importance of the data itself and the predictive ability of the features. In this way, the equalization threshold dynamic fusion weight can be dynamically adjusted, which also means that the model can adaptively respond to changes in data to adjust the feature contribution from the user basic information semantic coding vector and the user medical history data semantic coding vector.

[0022] In a specific example of the present application, a first balancing threshold dynamic fusion weight, a second balancing threshold dynamic fusion weight, and a third balancing threshold dynamic fusion weight are determined based on the user basic information semantic coding vector and the user medical history data semantic coding vector, including: cascading the user basic information semantic coding vector and the user medical history data semantic coding vector to obtain a user basic information-user medical history semantic cascade vector; calculating the position-by-position element-by-element addition of the user basic information semantic coding vector and the user medical history data semantic coding vector to obtain a user basic information-user medical history semantic element-by-element addition vector; calculating the position-by-element multiplication of the user basic information semantic coding vector and the user medical history data semantic coding vector To obtain the element-by-element multiplication vector of user basic information - user medical history semantics; multiply the first transformation vector by the user basic information - user medical history semantics cascade vector and add the result with the first bias parameter to obtain the cascade coefficient, multiply the second transformation vector by the user basic information - user medical history semantics element-by-element addition vector and add the result with the second bias parameter to obtain the addition coefficient, multiply the third transformation vector by the user basic information - user medical history semantics element-by-element multiplication vector and add the result with the third bias parameter to obtain the multiplication coefficient; respectively pass the cascade coefficient, addition coefficient and multiplication coefficient through the sigmoid activation function to obtain the first equalization threshold dynamic fusion weight, the second equalization threshold dynamic fusion weight and the third equalization threshold dynamic fusion weight.

[0023] More specifically, the adaptive dynamic integration subunit 3412 is used to adaptively and dynamically integrate the user basic information semantic coding vector and the user medical history data semantic coding vector with the first balanced threshold dynamic fusion weight, the second balanced threshold dynamic fusion weight and the third balanced threshold dynamic fusion weight to obtain the user basic information semantic dynamic interaction fusion item and the user medical history semantic dynamic interaction fusion item. In a specific example of the present application, first, the average value of the first balanced threshold dynamic fusion weight, the second balanced threshold dynamic fusion weight and the third balanced threshold dynamic fusion weight is used as the user medical history semantic dynamic interaction fusion coefficient; then, the value obtained by subtracting the user medical history semantic dynamic interaction fusion coefficient is used as the user basic information semantic dynamic interaction fusion coefficient; then, the point multiplication of the user basic information semantic dynamic interaction fusion coefficient and the user basic information semantic coding vector is calculated to obtain the user basic information semantic dynamic interaction fusion item; and the point multiplication of the user medical history semantic dynamic interaction fusion coefficient and the user medical history data semantic coding vector is calculated to obtain the user medical history semantic dynamic interaction fusion item.

[0024] More specifically, the element-by-element interactive fusion subunit 3413 is used to perform element-by-element fusion on the user basic information semantic dynamic interactive fusion item and the user medical history semantic dynamic interactive fusion item to obtain the user basic information-user medical history semantic fusion representation vector. That is, the use of the first balanced threshold dynamic fusion weight, the second balanced threshold dynamic fusion weight, and the third balanced threshold dynamic fusion weight can adaptively and dynamically integrate the basic information semantic feature distribution and the medical history semantic feature distribution, and ensure that fine-grained information is properly considered by element-by-element fusion, rather than simply processing the vector as a whole.

[0025] In summary, in the above embodiment, the user basic information semantic coding vector and the user medical history data semantic coding vector are input into the balanced threshold feature vector adaptive fusion module to obtain the user basic information-user medical history semantic fusion representation vector, including: using the balanced threshold feature vector adaptive fusion module to adaptively fuse the user basic information semantic coding vector and the user medical history data semantic coding vector according to the following formula to obtain the user basic information-user medical history semantic fusion representation vector; wherein the formula is: ;in, and are the semantic coding vector of the user's basic information and the semantic coding vector of the user's medical history data, respectively. Indicates cascade, represents vector multiplication, It means adding elements by position. represents element-by-element multiplication by position, , and are the first transformation vector, the second transformation vector and the third transformation vector respectively, , and are the first bias parameter, the second bias parameter and the third bias parameter respectively, represents the sigmoid activation function, , and They are the first balance threshold dynamic fusion weight, the second balance threshold dynamic fusion weight and the third balance threshold dynamic fusion weight, , It is the semantic fusion representation vector of user basic information and user medical history.

[0026] It is worth mentioning that in other specific examples of the present application, the user basic information semantic coding vector and the user medical history data semantic coding vector can also be input into the balanced threshold feature vector adaptive fusion module in other ways to obtain the user basic information-user medical history semantic fusion representation vector, for example: input the user basic information semantic coding vector and the user medical history data semantic coding vector; define the balanced threshold feature vector, which is used to control the fusion weight between the basic information and the medical history data; use the adaptive fusion module to fuse the basic information semantic coding vector and the medical history data semantic coding vector; wherein the adaptive fusion module may include a neural network layer, an attention mechanism or other model components to learn how to effectively fuse the two types of information; according to the settings in the balanced threshold feature vector, the module will automatically adjust the fusion ratio of the two types of information; in the adaptive fusion module, the semantic coding vectors of the basic information and the medical history data are fused to obtain the user basic information-user medical history semantic fusion representation vector.

[0027] Specifically, the nonlinear coupling unit 342 is used to input the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector into the nonlinear coupling interaction module to obtain the user multimodal information interactive coupling representation vector. In particular, the nonlinear coupling interaction module introduces the operation of quadratic terms in the data processing process to achieve nonlinear optimization of data distribution, so as to better fit the complex interaction between the user basic information-user medical history semantic fusion features and the user life habit semantic features. At the same time, the coupling interaction weight is applied to adjust the relative importance of different features in the nonlinear transformation interaction to balance the scale of different feature values ​​and realize dynamic adjustment and adaptive change of feature distribution. In this way, the user multimodal information interactive coupling representation vector can fully understand the complex semantic interaction between the user basic information, user medical history information and life habit information, and form the user's comprehensive health characteristics.

[0028] In a specific example of the present application, the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector are input into a nonlinear coupling interaction module to obtain the user multimodal information interactive coupling representation vector, including: calculating the position-by-position response of the user basic information-user medical history semantic fusion representation vector relative to the user life habit semantic encoding vector to obtain a user multimodal information associated response interaction feature vector; and performing nonlinear coupling optimization on the user multimodal information associated response interaction feature vector to obtain the user multimodal information interactive coupling representation vector.

[0029] More specifically, in an embodiment of the present application, the specific implementation process of calculating the position-by-position response of the user basic information-user medical history semantic fusion representation vector relative to the user life habit semantic encoding vector to obtain the user multimodal information association response interaction feature vector is: dividing the i-th eigenvalue of the user basic information-user medical history semantic fusion representation vector by the i-th eigenvalue of the user life habit semantic encoding vector to obtain the i-th eigenvalue of the user multimodal information association response interaction feature vector. The specific implementation process of performing nonlinear coupling optimization on the user multimodal information association response interaction feature vector to obtain the user multimodal information interaction coupling representation vector is: determining a first coupling interaction weight, a second coupling interaction weight, a third coupling interaction weight and a fourth coupling interaction weight; calculating the i-th eigenvalue of the user multimodal information association response interaction feature vector. The first coupling interaction quadratic term of the eigenvalue is calculated, the quadratic coefficient of the first coupling interaction quadratic term is the first coupling interaction weight, the first coefficient of the first coupling interaction quadratic term is the second coupling interaction weight, and the constant coefficient of the first coupling interaction quadratic term is zero; calculate the first coupling interaction quadratic term in the user multimodal information association response interaction eigenvector The second coupling interaction quadratic term of the eigenvalue, the quadratic coefficient of the second coupling interaction quadratic term is the third coupling interaction weight, the first coefficient of the second coupling interaction quadratic term is the fourth coupling interaction weight, and the constant coefficient of the second coupling interaction quadratic term is one; The first coupling interaction quadratic term of the eigenvalue is divided by the The second coupling interaction quadratic term of the eigenvalue is used to obtain the first feature values.

[0030] In summary, in the above embodiment, the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector are input into a nonlinear coupling interaction module to obtain the user multimodal information interactive coupling representation vector, including: using the nonlinear coupling interaction module to perform nonlinear coupling interaction on the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector according to the following formula to obtain the user multimodal information interactive coupling representation vector; wherein the formula is: ;in, and They are the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector. eigenvalues, , , and are the first coupling interaction weight, the second coupling interaction weight, the third coupling interaction weight and the fourth coupling interaction weight, is the first in the user multimodal information interaction coupling representation vector feature values.

[0031] It is worth mentioning that in other specific examples of the present application, the user basic information semantic coding vector, the user life habit semantic coding vector and the user medical history data semantic coding vector can also be interactively fused based on the balanced threshold and the coupling interaction weight to obtain the user multimodal information interactive coupling representation vector in other ways, for example: input the user basic information semantic coding vector, the user life habit semantic coding vector and the user medical history data semantic coding vector; define the coupling interaction weights for interactive fusion, which can be learned parameters used to control the importance between different information sources; set the balanced threshold, which can be a threshold used to control the intensity of information interaction. When the similarity between the two information sources exceeds the threshold, interactive fusion will be performed; interactive fusion is performed on the user basic information semantic coding vector, the user life habit semantic coding vector and the user medical history data semantic coding vector; for each pair of vectors, the similarity or correlation between them is calculated, and interactive fusion is performed according to the coupling interaction weight and the balanced threshold; the interactively fused vectors are integrated to obtain the user multimodal information interactive coupling representation vector.

[0032] In particular, the archiving result generating module 350 is used to determine the archiving result based on the user multimodal information interactive coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group. That is, in the technical solution of the present application, the user multimodal information interactive coupling representation vector is input into the classifier-based automatic archiving device to obtain the archiving result, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

[0033] Preferably, inputting the user multimodal information interactive coupling representation vector into a classifier-based automatic archiver to obtain an archive result comprises: Multiplying the user multimodal information interaction coupling representation vector by the length of the user multimodal information interaction coupling representation vector and the square root of the length of the user multimodal information interaction coupling representation vector respectively to obtain a user multimodal information interaction coupling full-width representation vector and a user multimodal information interaction coupling half-width representation vector; Performing point subtraction on the user multimodal information interaction coupling full-scale representation vector and a norm of the user multimodal information interaction coupling representation vector, and calculating the square root of the absolute value of each position of the point subtraction result vector to obtain the user multimodal information interaction coupling full-scale semantic change vector; Performing point subtraction on the user multimodal information interaction coupling half-width representation vector and the bi-norm of the user multimodal information interaction coupling representation vector, and calculating the square root of the absolute value of each position of the point subtraction result vector to obtain the user multimodal information interaction coupling half-width semantic change vector; Respectively calculating the logarithmic value with base 2 of each eigenvalue of the user multimodal information interaction coupling full-width semantic change vector and the user multimodal information interaction coupling half-width semantic change vector to obtain the user multimodal information interaction coupling full-width semantic change information vector and the user multimodal information interaction coupling half-width semantic change information vector; Calculating the weighted sum of the user multimodal information interaction coupling full-width semantic change information vector and the user multimodal information interaction coupling half-width semantic change information vector using the balance hyperparameter as a weight to obtain an optimized user multimodal information interaction coupling representation vector; The optimized user multimodal information interactive coupling representation vector is input into a classifier-based automatic archiver to obtain an archiving result.

[0034] Here, for the user basic information semantic coding vector, the user life habit semantic coding vector and the user medical history data semantic coding vector, due to the source semantic differences of the user basic information, the user life habit information and the user medical history data, after passing through the balanced threshold feature vector adaptive fusion module and the nonlinear coupling interaction module, due to the balanced threshold adaptive fusion weight difference and the nonlinear coupling interaction weight difference caused by the feature semantic differences, the user multimodal information interactive coupling representation vector will also have the problem of insufficient aggregation of the fusion-interaction semantic feature distribution, thereby affecting its classification convergence efficiency through the classifier, that is, affecting the efficiency of classification training and the accuracy of the classification results.

[0035] Based on this, the applicant of the present application takes the user multimodal information interaction coupling representation vector as a feature set, and in the case of the change semantic representation in units of the feature value of the user multimodal information interaction coupling representation vector, dynamically aggregates the semantic set composed of different change semantics of the user multimodal information interaction coupling representation vector as a whole without ignoring the individual semantic change information. Thus, the collective expression of the individual features of the user multimodal information interaction coupling representation vector and the aggregation scale of the user multimodal information interaction coupling representation vector is used as the feature full amplitude and half amplitude, and the low-rank negative correlation of different dimensions of the overall semantics of the feature set of the user multimodal information interaction coupling representation vector is used as the phase and scaled to dynamically adjust the change relationship of the semantic content of the user multimodal information interaction coupling representation vector. In this way, the overall semantic information expression aggregation of the feature set of the user multimodal information interaction coupling representation vector is improved, thereby improving the classification convergence efficiency of the user multimodal information interaction coupling representation vector when it is classified by a classifier-based automatic archiver.

[0036] As described above, the high-risk population health record management system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a high-risk population health record management algorithm. In one possible implementation, the high-risk population health record management system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the high-risk population health record management system 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the high-risk population health record management system 300 can also be one of the many hardware modules of the wireless terminal.

[0037] Alternatively, in another example, the high-risk population health record management system 300 and the wireless terminal may also be separate devices, and the high-risk population health record management system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0038] Furthermore, a method for managing health records of high-risk groups is also provided.

[0039] Figure 5 FIG. 1 is a flow chart of a method for managing health records of high-risk groups according to an embodiment of the present application. Figure 5As shown, according to the embodiment of the present application, the health record management method for high-risk groups includes the following steps: S1, obtaining the health record of the user to be evaluated; S2, extracting the user's basic information, the user's living habit information and the user's medical history data from the health record of the user to be evaluated; S3, semantically encoding the user's basic information, the user's living habit information and the user's medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; S4, interactively fusing the user's basic information semantic coding vector, the user's living habit semantic coding vector and the user's medical history data semantic coding vector based on an equalization threshold and a coupling interaction weight to obtain a user multimodal information interactive coupling representation vector; S5, determining the archiving result based on the user multimodal information interactive coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

[0040] In summary, according to the embodiment of the present application, the health record management method for high-risk groups is explained, which uses advanced data processing technology and deep learning algorithms to semantically understand and analyze the health records of the users to be evaluated, and performs multimodal interactive coupling of the information in the health records to form a comprehensive user feature representation, thereby using classifiers to achieve automatic archiving, and to intelligently determine whether the user belongs to a high-risk group, providing strong technical support for public health management.

[0041] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A health record management system for high-risk groups, characterized in that: include: A health record acquisition module is used to obtain the health record of the user to be evaluated; A multimodal information extraction module, used to extract basic user information, user living habit information and user medical history data from the health record of the user to be evaluated; A semantic coding module, used for semantically coding the user basic information, the user living habit information and the user medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; A multimodal information interaction fusion module, used for interactively fusing the user basic information semantic coding vector, the user living habit semantic coding vector and the user medical history data semantic coding vector based on an equilibrium threshold and a coupling interaction weight to obtain a user multimodal information interaction coupling representation vector; The archiving result generating module is used to determine the archiving result based on the user multimodal information interaction coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

2. The high-risk population health record management system according to claim 1 is characterized in that: The semantic encoding module is used to: A semantic encoder including a Transformer layer is used to semantically encode the user basic information, the user living habit information and the user medical history data to obtain a semantic encoding vector of the user basic information, a semantic encoding vector of the user living habit and a semantic encoding vector of the user medical history data.

3. The high-risk population health record management system according to claim 2 is characterized in that: The multimodal information interaction fusion module includes: An adaptive fusion unit, used for inputting the user basic information semantic coding vector and the user medical history data semantic coding vector into an equalization threshold feature vector adaptive fusion module to obtain a user basic information-user medical history semantic fusion representation vector; The nonlinear coupling unit is used to input the user basic information-user medical history semantic fusion representation vector and the user life habit semantic encoding vector into a nonlinear coupling interaction module to obtain the user multimodal information interactive coupling representation vector.

4. The high-risk population health record management system according to claim 3 is characterized in that: The adaptive fusion unit comprises: A subunit for determining a dynamic fusion weight of a balanced threshold, configured to determine a first dynamic fusion weight of a balanced threshold, a second dynamic fusion weight of a balanced threshold, and a third dynamic fusion weight of a balanced threshold based on the semantic coding vector of the user's basic information and the semantic coding vector of the user's medical history data; An adaptive dynamic integration subunit, used for adaptively and dynamically integrating the user basic information semantic coding vector and the user medical history data semantic coding vector with the first equalization threshold dynamic fusion weight, the second equalization threshold dynamic fusion weight and the third equalization threshold dynamic fusion weight to obtain a user basic information semantic dynamic interaction fusion item and a user medical history semantic dynamic interaction fusion item; The element-by-element interactive fusion subunit is used to perform element-by-element fusion on the user basic information semantic dynamic interactive fusion item and the user medical history semantic dynamic interactive fusion item to obtain the user basic information-user medical history semantic fusion representation vector.

5. The high-risk population health record management system according to claim 4 is characterized in that: The equalization threshold dynamic fusion weight determination subunit is used to: Cascading the user basic information semantic encoding vector and the user medical history data semantic encoding vector to obtain a user basic information-user medical history semantic cascade vector; Calculate the position-by-position addition of the semantic coding vector of the user basic information and the semantic coding vector of the user medical history data to obtain the user basic information-user medical history semantic element-by-element addition vector; Calculate the position-by-position element-by-element multiplication of the user basic information semantic encoding vector and the user medical history data semantic encoding vector to obtain a user basic information-user medical history semantic element-by-element multiplication vector; The first transformation vector is multiplied by the user basic information-user medical history semantic cascade vector and then added to the first bias parameter to obtain a cascade coefficient, the second transformation vector is multiplied by the user basic information-user medical history semantic element-by-element addition vector and then added to the second bias parameter to obtain an addition coefficient, and the third transformation vector is multiplied by the user basic information-user medical history semantic element-by-element multiplication vector and then added to the third bias parameter to obtain a multiplication coefficient; The cascade coefficient, the addition coefficient and the multiplication coefficient are respectively passed through a sigmoid activation function to obtain the first equalization threshold dynamic fusion weight, the second equalization threshold dynamic fusion weight and the third equalization threshold dynamic fusion weight.

6. The high-risk population health record management system according to claim 5 is characterized in that: The adaptive dynamic integration subunit is used for: The average value of the first balanced threshold dynamic fusion weight, the second balanced threshold dynamic fusion weight and the third balanced threshold dynamic fusion weight is used as the user medical history semantic dynamic interaction fusion coefficient; The value obtained by subtracting the user medical history semantic dynamic interaction fusion coefficient from one is used as the user basic information semantic dynamic interaction fusion coefficient; Calculate the point multiplication of the user basic information semantic dynamic interaction fusion coefficient and the user basic information semantic encoding vector to obtain the user basic information semantic dynamic interaction fusion item; The user medical history semantic dynamic interaction fusion coefficient is calculated and multiplied by the point of the user medical history data semantic encoding vector to obtain the user medical history semantic dynamic interaction fusion item.

7. The high-risk population health record management system according to claim 6 is characterized in that: The nonlinear coupling unit comprises: A position-by-position association response interaction subunit, used to calculate the position-by-position response of the user basic information-user medical history semantic fusion representation vector relative to the user life habit semantic encoding vector to obtain a user multimodal information association response interaction feature vector; The multimodal information interaction coupling subunit is used to perform nonlinear coupling optimization on the user multimodal information associated response interaction feature vector to obtain the user multimodal information interaction coupling representation vector.

8. The high-risk population health record management system according to claim 7 is characterized in that: The multimodal information interactive coupling subunit is used to: determining a first coupling interaction weight, a second coupling interaction weight, a third coupling interaction weight, and a fourth coupling interaction weight; Calculate the user's multimodal information association response interaction feature vector a first coupling interaction quadratic term of a eigenvalue, the quadratic coefficient of the first coupling interaction quadratic term is the first coupling interaction weight, the first coefficient of the first coupling interaction quadratic term is the second coupling interaction weight, and the constant coefficient of the first coupling interaction quadratic term is zero; Calculate the user's multimodal information association response interaction feature vector a second coupling interaction quadratic term of a eigenvalue, the quadratic coefficient of the second coupling interaction quadratic term is the third coupling interaction weight, the first coefficient of the second coupling interaction quadratic term is the fourth coupling interaction weight, and the constant coefficient of the second coupling interaction quadratic term is one; The said The first coupling interaction quadratic term of the eigenvalue is divided by the The second coupling interaction quadratic term of the eigenvalue is used to obtain the first feature values.

9. The high-risk population health record management system according to claim 8, characterized in that: The archiving result generating module is used for: The user multimodal information interactive coupling representation vector is input into a classifier-based automatic archiver to obtain the archiving result.

10. A method for managing health records of high-risk groups, characterized in that: include: Obtain the health records of the user to be evaluated; Extracting basic information of the user, information on the user's living habits and medical history data from the health record of the user to be evaluated; Performing semantic coding on the user basic information, the user living habit information and the user medical history data to obtain a user basic information semantic coding vector, a user living habit semantic coding vector and a user medical history data semantic coding vector; Perform interactive fusion based on the equalization threshold and coupling interaction weight on the semantic coding vector of the user's basic information, the semantic coding vector of the user's living habits, and the semantic coding vector of the user's medical history data to obtain a user multimodal information interactive coupling representation vector; The archiving result is determined based on the user multimodal information interactive coupling representation vector, and the archiving result is used to indicate whether the user to be evaluated is a high-risk group.

Images