Cross-modal data generation type driven chronic disease management method and platform

Abstract

The invention discloses a cross-modal data generation driven chronic disease management method and platform, and relates to the technical field related to medical data management, and the method comprises the steps: accessing a cross-modal database; acquiring a multi-party information receiving end, and extracting a feature vector representing a data granularity demand; outputting a clustering result, configuring the number of a plurality of transmission channels, and establishing connection between the plurality of transmission channels and a multi-party information receiving end; and setting a plurality of data granularities of the plurality of transmission channels, performing template processing, performing template processing on the chronic disease resource data transmitted by the plurality of transmission channels according to a plurality of granularity templates, and correspondingly transmitting the chronic disease resource data to the multi-party information receiving end. The technical problems that in the prior art, cross-modal data granularity and modal semantics are inconsistent, an appropriate granularity template is difficult to generate flexibly according to requirements, information transmission efficiency is low, and chronic disease management accuracy and real-time performance are poor are solved. The technical effect of improving the data transmission efficiency and the accuracy and real-time performance of chronic disease management is achieved.

Description

Technical Field

[0001] The present application relates to the technical field related to medical data management, and specifically to a chronic disease management method and platform driven by cross-modal data generation. Background Art

[0002] In chronic disease management, traditional medical and health management faces problems such as heterogeneous multimodal data sources, low data processing efficiency, and inaccurate information transmission, as chronic disease patients require long-term, multi-dimensional health data tracking and monitoring. Chronic disease management not only requires processing different types of data, such as medical images, laboratory test data, vital signs monitoring data, and patient self-reported data, but also has large differences in data distribution, format, and dimensions of each modality, leading to difficulties in information fusion and cross-modal analysis; it is also necessary to generate data templates of different granularities for different information needs in order to better support decision-making and early warning. The existing single data processing model cannot meet these diverse needs. Different chronic disease management scenarios have different requirements for information granularity, which makes it difficult to flexibly generate appropriate granularity templates according to needs, resulting in low efficiency in data transmission and display, and even affecting the accuracy of decision-making and chronic disease management.

[0003] Therefore, in the current relevant technologies, there are technical problems such as inconsistency between cross-modal data granularity and modal semantics, and difficulty in flexibly generating appropriate granularity templates according to needs, which leads to low information transmission efficiency and poor accuracy and real-time performance of chronic disease management. Summary of the invention

[0004] This application solves the technical problems in the prior art that the granularity of cross-modal data is inconsistent with the modal semantics, and it is difficult to flexibly generate appropriate granularity templates according to needs, resulting in low information transmission efficiency and poor accuracy and real-time performance of chronic disease management by providing a chronic disease management method and platform driven by cross-modal data generation. This achieves the technical effect of improving data transmission efficiency and the accuracy and real-time performance of chronic disease management.

[0005] The present application provides a chronic disease management method driven by cross-modal data generation, the method comprising: accessing a cross-modal database, the cross-modal database being stored in a cloud processor, the cloud processor comprising multiple transmission channels; obtaining a multi-party information receiving end, extracting a feature vector representing the data granularity requirements in the multi-party information receiving end; clustering the multi-party information receiving end according to the feature vector, outputting a clustering result, configuring the number of the multiple transmission channels according to the number of classes of the clustering result, and establishing a connection between the multiple transmission channels and the multi-party information receiving end; setting multiple data granularities of the multiple transmission channels, performing template processing respectively according to the multiple data granularities, outputting multiple granularity templates, downloading the multiple granularity templates to the multiple transmission channels, performing template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmitting them to the multi-party information receiving end accordingly.

[0006] In a possible implementation, the cross-modal data generatively driven chronic disease management method also performs the following processing: high-dimensional feature space mapping of the data of each modality in the cross-modal database, and outputting a multimodal high-dimensional feature data set corresponding to multiple modalities; generatively aligning the multimodal high-dimensional feature data set using a functional-adversarial network, and outputting an aligned multimodal feature data set; and using the aligned multimodal feature data set to perform modal data feature fusion to update the cross-modal database.

[0007] In a possible implementation, the cross-modal data generatively driven chronic disease management method also performs the following processing: wherein the functional-adversarial network includes a generator and a discriminator, the output end of the generator is connected to the input end of the discriminator, and the output end of the discriminator is connected to the input end of the generator; the generator and the discriminator are adversarially trained to obtain the modal data generated by the generator, input the generated modal data into the discriminator, calculate the difference distribution with the sample modal data, perform parameter learning with the goal of being less than a preset difference distribution, and obtain a trained functional-adversarial network.

[0008] In a possible implementation, the cross-modal data generation-driven chronic disease management method also performs the following processing: calculating the semantic difference between features in each modality of the aligned multimodal feature data set; storing in parallel the features whose semantic difference is greater than or equal to the preset semantic difference, and overwriting and storing the features whose semantic difference is less than the preset semantic difference.

[0009] In a possible implementation, the cross-modal data generation-driven chronic disease management method also performs the following processing: the cloud processor performs template processing on the cross-modal database to generate an initialization template, wherein the initialization template includes the number of initialization template partitions, the initialization information hierarchy structure, the initialization data update frequency, and the initialization data display accuracy; the template parameters of the initialization template are adjusted according to the multiple data granularities, and multiple granular templates are output, wherein the template parameters include the number of template partitions, the information hierarchy structure data, the update frequency, and the data display accuracy.

[0010] In a possible implementation, the cross-modal data generation-driven chronic disease management method also performs the following processing: the multi-party information receiving end performs a quality assessment on the received templated chronic disease resource data, and obtains a plurality of quality assessment indicators corresponding to the multi-party information receiving end; the plurality of quality assessment indicators are fed back to the cloud processor, and the template parameters of the plurality of granular templates are updated according to the cloud processor until the preset quality assessment indicators are met, and the updated plurality of granular templates are output.

[0011] In a possible implementation, the cross-modal data generation-driven chronic disease management method also performs the following processing: using the multi-party information receiving end as an edge node and the cloud processor as a central node to build a resource topology management platform; based on the resource topology management platform, the granularity templates of the multiple transmission channels are updated and synchronized.

[0012] In a possible implementation, the cross-modal data generation-driven chronic disease management method also performs the following processing: wherein the feature vector of each information receiving end includes data precision requirements, data interaction complexity, data modality preference, and granularity level requirements; the similarity between the feature vectors of every two information receiving ends among the multiple information receiving ends is calculated to obtain a similarity set; the multiple information receiving ends are clustered using the similarity set, and the clustering results are output.

[0013] The present application also provides a chronic disease management platform driven by cross-modal data generation, including: a cross-modal database access unit, used to access a cross-modal database, the cross-modal database is stored in a cloud processor, and the cloud processor includes multiple transmission channels; a feature vector extraction unit, used to obtain a multi-party information receiving end, and extract a feature vector representing the data granularity requirements in the multi-party information receiving end; a transmission channel configuration unit, used to cluster the multi-party information receiving end according to the feature vector, output the clustering result, configure the number of the multiple transmission channels according to the number of classes of the clustering result, and establish a connection between the multiple transmission channels and the multi-party information receiving end; a granularity template output unit, used to set multiple data granularities of the multiple transmission channels, perform template processing according to the multiple data granularities, output multiple granularity templates, download the multiple granularity templates to the multiple transmission channels, and perform template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmit them to the multi-party information receiving end accordingly.

[0014] It is intended to access the cross-modal database through the cross-modal data generation-driven chronic disease management method and platform proposed in this application; obtain multiple information receiving terminals and extract feature vectors that represent data granularity requirements; cluster the multiple information receiving terminals, output clustering results, configure the number of multiple transmission channels, and establish connections between multiple transmission channels and multiple information receiving terminals; set multiple data granularities for multiple transmission channels, perform template processing, and perform template processing on the chronic disease resource data transmitted by multiple transmission channels according to multiple granularity templates, and transmit them to the multiple information receiving terminals accordingly. It solves the technical problems existing in the prior art that the cross-modal data granularity is inconsistent with the modal semantics, and it is difficult to flexibly generate appropriate granularity templates according to needs, resulting in low information transmission efficiency and poor accuracy and real-time performance of chronic disease management, and achieves the technical effect of improving data transmission efficiency and the accuracy and real-time performance of chronic disease management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure are briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the platform according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed precisely 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 flow chart of a cross-modal data generation-driven chronic disease management method provided in an embodiment of the present application.

[0017] Figure 2A schematic diagram of the structure of a chronic disease management platform driven by cross-modal data generation provided in an embodiment of the present application.

[0018] Explanation of the reference numerals: cross-modal database access unit 10 , feature vector extraction unit 20 , transmission channel configuration unit 30 , granular template output unit 40 . 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 the objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, platform, 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 units 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 chronic disease management method driven by cross-modal data generation, such as Figure 1 As shown, the method includes: Step S100: accessing a cross-modal database, wherein the cross-modal database is stored in a cloud processor, and the cloud processor includes a plurality of transmission channels.

[0023] Preferably, the cross-modal database is used to store different types of modal data, wherein modal data refers to data from different sources and forms, including but not limited to medical imaging data, electronic health records (EHR), laboratory test results, sensor data, patient self-report data, etc. The cross-modal database can uniformly manage data from different modalities (data of different formats and characteristics) and support their integration and analysis; the cross-modal database is stored in a cloud processor, which means that all modal data are centrally stored on a cloud server, providing efficient computing resources and storage space for data access, processing and analysis. The cloud processor has strong computing power, can support large-scale data processing, and can realize real-time analysis and response of cross-modal data; and the cloud processor includes multiple different paths or channels for data transmission, each channel is used for the transmission of data of different granularity or different types, and different transmission channels can be optimized for different data requirements (such as different user terminals or different information display levels). Each transmission channel can process different data granularity, or adjust the format, type or processing capability of data transmission according to user needs, which helps to flexibly transmit and process different data streams according to specific circumstances, thereby improving the platform response capability and information transmission efficiency, thereby achieving more refined and efficient cross-modal data management.

[0024] Furthermore, step S100 also includes step S110, performing high-dimensional feature space mapping on the data of each modality in the cross-modal database, and outputting a multimodal high-dimensional feature data set corresponding to multiple modalities; step S120, using a functional-adversarial network to generatively align the multimodal high-dimensional feature data set, and outputting the aligned multimodal feature data set; step S130, using the aligned multimodal feature data set to perform modal data feature fusion and update the cross-modal database.

[0025] Preferably, medical data of multiple different modalities stored in the database (such as medical imaging data, clinical examination data, patient self-reported symptoms, sensor monitoring data, etc.) are mapped into a high-dimensional feature space, that is, these data of different modalities are converted into representations in the high-dimensional feature space. The data of each modality may have different structures and dimensions. For example, imaging data is a two-dimensional or three-dimensional matrix, text data may be a sequence, sensor data is a time series, etc. By mapping to a high-dimensional feature space, data from different sources (such as images, time series, physiological signals, etc.) can share the same space, and then output the corresponding multimodal high-dimensional feature data set.

[0026] Preferably, a functional-adversarial network is used to generatively align multimodal high-dimensional feature data sets, that is, to align high-dimensional feature data of multiple modalities so that the data of these modalities can not only match in structure but also keep consistent in semantics and statistical features. Through the functional-adversarial network, virtual data with the same semantics as the original modal data can be generated to ensure alignment between different modalities. The functional-adversarial network is a method based on a generative adversarial network. By training two neural networks (a generator and a discriminator), data is generated in the generator, and the discriminator evaluates the difference between the generated data and the real data. The functional-adversarial network It focuses on data alignment at the functional level, and can solve the problem of differences between modalities by generating data that is consistent with the distribution of the original data; then, the features in the aligned multimodal data are fused, that is, the data of different modalities are weighted averaged and spliced ​​in the same space, so that the fused data can retain the unique information of different modal data at the same time, reduce redundant information and find the optimal combination between multiple modalities; finally, the high-dimensional feature data set obtained after generative alignment and feature fusion is inserted into the original cross-modal database as a new data record, thereby updating the content of the modal database and improving the accuracy and consistency of cross-modal data analysis.

[0027] Furthermore, step S120 also includes step S121, wherein the functional-adversarial network includes a generator and a discriminator, the output end of the generator is connected to the input end of the discriminator, and the output end of the discriminator is connected to the input end of the generator; step S122, performing adversarial training on the generator and the discriminator, obtaining the modal data generated by the generator, inputting the generated modal data into the discriminator, calculating the difference distribution with the sample modal data, performing parameter learning with the goal of being less than a preset difference distribution, and obtaining a trained functional-adversarial network.

[0028] Preferably, the functional-adversarial network (F-GAN) is an extension of the traditional generative adversarial network (GAN), which usually includes two core parts, a generator and a discriminator. The generator generates fake data, and the discriminator judges the authenticity of the data. The two parts compete with each other and gradually optimize the ability of the generator, so that the generated data is closer and closer to the real data. In the functional-adversarial network (F-GAN), the design goals of the generator and the discriminator are used to solve the problem of cross-modal data generation and alignment. The generator not only generates "fake" data, but also needs to generate outputs that match the target modal data in a specific functional space, while the discriminator needs to evaluate the difference between the generated data and the real data according to the goal of "functional alignment"; the output of the generator is connected to the input of the discriminator. The modal data generated by the generator (for example, synthetic medical images or text data) will be input into the discriminator, whose task is to determine whether the data matches the real data; the output of the discriminator (that is, the discriminator's score or difference measurement result) will be returned to the generator to guide the optimization of the generator. The generator adjusts its parameters based on the output of the discriminator so that the generated data is closer to the distribution of the target modal data.

[0029] Preferably, the generator and the discriminator are subjected to adversarial training, that is, through the game between the generator and the discriminator, the generator continuously generates data that is more in line with the expected goal, while the discriminator continuously improves its recognition ability. During the training process, the generator will gradually adjust its generation strategy according to the feedback of the discriminator to minimize the difference between the generated data and the real data. The discriminator learns how to make effective judgments in the functional space by identifying the difference between the generated data and the real data. Specifically, at the beginning of the training, the generator receives input (such as random noise, partial information of the target modality, etc.) and generates modal data (medical images or other types of health data). The discriminator receives the modal data output by the generator and the real modal data (from the original sample or the cross-modal database), calculates the difference distribution between the two, and minimizes the difference between the two. Parameter learning is performed with different distribution as the goal, for example, using some distance metric or discriminant criterion to measure similarity or difference, and then adjusting the generator parameters through optimization algorithms (such as gradient descent) to make the generated data as close as possible to the real data in the difference distribution. The goal of the discriminator is to more accurately evaluate the difference between the generated data and the real data. During the whole process, the generator and the discriminator are constantly competing and reaching a balance through adversarial training. The data generated by the generator is highly consistent with the real data in the functional space, and ensures that the generated data can meet certain functional requirements, such as semantic consistency or alignment goals of cross-modal data; finally, the generator and the discriminator together constitute a functional-adversarial network. The generator can generate modal data that meets the functional goals based on the input data, and the discriminator can evaluate the difference between these generated data and the real data.

[0030] Furthermore, step S130 also includes step S131, calculating the semantic difference between features in each modality of the aligned multimodal feature data set; step S132, storing in parallel the features whose semantic difference is greater than or equal to the preset semantic difference, and overwriting and storing the features whose semantic difference is less than the preset semantic difference.

[0031] Preferably, the semantic difference between the features in each modality of the aligned multimodal feature dataset is calculated, wherein the semantic difference is a measure of the similarity or difference between features of different modalities (or different features in the same modality) at the semantic level. Specifically, the semantic difference is measured by calculating a certain distance between features (such as Euclidean distance, cosine similarity, etc.), which reflects the similarity of features in the semantic space. For example, a similarity metric (such as cosine similarity) or a distance metric (such as L2 norm) is used to express that a smaller semantic difference means that the two features are semantically more similar, and a larger semantic difference means that the two features are semantically different, indicating that they may belong to different categories or come from different modalities in representation.

[0032] Preferably, the preset semantic difference refers to a preset semantic difference setting threshold to determine which features are stored in parallel and which features are stored in an overlay manner. Specifically, two features with a semantic difference greater than or equal to the preset semantic difference are stored in parallel, that is, the two features will retain their respective storage locations to ensure that they can be used separately in subsequent processing. For example, medical images and genetic data may have large semantic differences in the feature space and should be stored in parallel; if the semantic difference between two features is less than the preset threshold, they will be stored in an overlay manner, that is, these features will be stored in the same location, overwriting the original feature representation, indicating that they are highly consistent in semantics. For example, feature data generated by two sensors (such as a blood glucose meter and a blood pressure monitoring device) can be stored in an overlay manner if their differences in the semantic space are small; thereby improving the storage efficiency of cross-modal data and optimizing the entire data processing flow of chronic disease management.

[0033] Step S200: obtaining a multi-party information receiving end, and extracting a feature vector representing the data granularity requirement in the multi-party information receiving end.

[0034] Preferably, the multi-party information receiving end includes multiple terminal devices that need to receive processed data, which can be different types of data receiving devices or systems, such as hospital management systems (used to display patients' health data, diagnostic results, etc.), patient-side devices (patients' mobile phone applications), health monitoring devices (such as blood glucose meters, heart rate monitors, etc.), family doctors, mobile devices, etc. These different receiving ends (or terminals) have different data requirements, some require low-granularity information (such as trend charts), and some require high-granularity detailed data (such as detailed blood glucose change curves, image analysis results, etc.); then extract the feature vector representing the data granularity requirements from the multi-party information receiving end, wherein the feature vector is used to capture certain features of the information and represent the granularity requirements of each receiving end. Specifically, the granularity requirements refer to the requirements of different receiving ends for the level of data details. For example, the patient-side device may only need to provide brief information on health trends, while the hospital management system may require detailed diagnostic data.

[0035] Preferably, a feature vector is extracted from each receiving end to represent the data granularity requirements of the receiving end. By extracting the feature vector of the receiving end, the data granularity requirements of each receiving end are quantified. The information contained in the feature vector may include the data granularity requirements (for example, whether detailed data at each time point is required, or only the trend after aggregation is required), the data format (for example, whether charts, tables, or raw data are required), the complexity of the data (for example, whether deeply analyzed data is required, or only surface data), and the modal preference of the data. According to the specific needs of each receiving end, data of different granularities are generated and provided, so as to achieve more personalized and accurate health data transmission and display.

[0036] Step S300, clustering the multi-party information receiving terminals according to the feature vectors, outputting clustering results, configuring the number of the multiple transmission channels according to the number of classes of the clustering results, and establishing connections between the multiple transmission channels and the multi-party information receiving terminals.

[0037] Preferably, the receivers are grouped according to the feature vectors by a clustering algorithm (such as K-means, hierarchical clustering, DBSCAN, etc.), that is, according to the similarity of the feature vectors of the receivers, the receivers with similar requirements are classified into the same category. Specifically, K-means is used to cluster the multi-party information receivers. First, the number of clusters is determined. For example, the receivers are divided into two categories (including receivers with low-granularity data requirements and receivers with high-granularity data requirements). Then, K centroids are initialized to represent the center of each class. In the first round of iteration, K-means will randomly select K data points as the initial centroids, and then assign each receiver to the closest centroid. Usually The Euclidean distance is used to calculate the distance between each receiver and the centroid, and the receiver is assigned to the nearest centroid. All receivers are assigned to a cluster, and K-means recalculates the new centroid of each cluster (i.e., the average position of all data points assigned to this cluster). The receivers are then assigned to the nearest centroid and updated repeatedly until the centroid position no longer changes or changes very little. The clustering results are finally output, for example, receivers that require high-granularity data (such as hospital management systems) and receivers that require low-granularity data (such as patient mobile phone applications). Each cluster represents a class of receivers with similar needs, and different transmission channels are configured based on these clustering results.

[0038] Preferably, a corresponding number of transmission channels are configured according to the number of classes of the clustering results, and each clustering class corresponds to a transmission channel, that is, each receiving end with similar data requirements will receive data through an independent transmission channel, wherein the transmission channel is used to transmit data of different granularities and formats according to the requirements of each class, and then connect with multiple information receiving ends based on the number of transmission channels. Specifically, these receiving ends are connected to corresponding transmission channels according to the class to which each receiving end belongs. Each receiving end will receive template data of corresponding granularity through the transmission channel to which it is connected according to its own data granularity requirements, ensuring that each receiving end receives the correct granularity data that meets its requirements, and providing customized and efficient data transmission for each type of receiving end, thereby effectively managing the needs of different receiving ends and ensuring the efficiency and accuracy of data transmission.

[0039] Furthermore, step S300 also includes step S310, wherein the feature vector of each information receiving end includes data precision requirements, data interaction complexity, data modality preference and granularity level requirements; step S320, calculating the similarity between the feature vectors of every two information receiving ends among the multiple information receiving ends, and obtaining a similarity set; step S330, using the similarity set to cluster the multiple information receiving ends, and outputting the clustering results.

[0040] Preferably, the feature vector of each information receiving end includes the following four dimensions, namely, data precision requirement, data interaction complexity, data modality preference and granularity level requirement. Specifically, data precision requirement refers to the receiving end's requirement for the detail level of data. For example, some receiving ends require fine-grained real-time data, while other receiving ends only require rough trend information; data interaction complexity refers to the receiving end's requirement for the complexity of data interaction. For example, some receiving ends require complex interactive operations with data (such as visualization and in-depth analysis), while other receiving ends may only require basic data display; data modality preference refers to the receiving end's preference for different data modalities, such as whether to prefer processing image data, sensor data, text data, etc.; granularity level requirement refers to the receiving end's hierarchical requirement for data granularity, such as whether raw data (low granularity) or aggregated data (high granularity) is required.

[0041] Preferably, the similarity between the feature vectors of each information receiving terminal is calculated to measure the similarity of the receiving terminals in terms of data requirements. For example, the cosine similarity is used to calculate the similarity of two feature vectors. The closer the value is to 1, the more similar it is, and the value is 0, which means it is completely dissimilar. Or the Euclidean distance is used to calculate the distance between two feature vectors in a multidimensional space. The smaller the distance, the higher the similarity between the receiving terminals. Then, the similarity values ​​of multiple receiving terminals are obtained to form a similarity set. Then, the receiving terminals are divided into K classes based on the similarity. K is a preset number of clusters. K-means clustering divides receiving terminals with high similarity into the same class, that is, the feature vectors of each receiving terminal will be assigned to a cluster according to their similarity. Receiving terminals with high similarity will be divided into the same class, so that the requirements of these receiving terminals (such as data refinement, interaction complexity, etc.) are processed similarly. Finally, the clustering results are output, including the class to which each receiving terminal belongs and the representative features of each class, such as the center (i.e., centroid) of the class, so as to provide customized data processing and transmission solutions for different receiving terminals.

[0042] Step S400, set multiple data granularities of the multiple transmission channels, perform template processing according to the multiple data granularities respectively, output multiple granularity templates, download the multiple granularity templates to the multiple transmission channels, perform template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmit them to the multi-party information receiving end accordingly.

[0043] Preferably, multiple data granularities of multiple transmission channels are set, that is, corresponding data granularity (including data detail level or processing level) is set for each transmission channel. Different receiving ends may have different requirements for data granularity. Data granularity can be divided into multiple levels: low granularity, such as health trend charts or summary information, suitable for display to patients or users; medium granularity, such as partial detailed information or periodic analysis data of health data, suitable for use by family doctors, nurses, etc.; high granularity, such as detailed physiological parameter time series data, imaging data, etc., suitable for in-depth analysis by hospital management systems; then template processing is performed according to multiple data granularities. That is, according to different data granularity requirements, a corresponding data template is generated for each granularity, and then multiple granularity templates are output. Each template represents a data format at a different level of detail (that is, each transmission channel will correspond to a granularity template), which may include low-granularity templates (used to display summary data, overview diagrams, simplified trends, etc.), medium-granularity templates (displaying periodic health data, partially detailed analysis charts), and high-granularity templates (displaying detailed time series data, complete health records, medical images, etc.), and then the data of different granularities are converted into a standard format suitable for display or processing to ensure that the data display format and detail level meet the needs of the receiving end.

[0044] Preferably, a corresponding granularity template is downloaded for each transmission channel to ensure that the data is transmitted in a predefined format. For example, a low-granularity template is downloaded to transmission channel 1 for use at a receiving end that requires low-granularity data (such as a patient application); a medium-granularity template is downloaded to transmission channel 2 for use at a receiving end that requires medium-granularity data (such as a family doctor); and a high-granularity template is downloaded to transmission channel 3 for use at a receiving end that requires high-granularity data (such as a hospital management system); ensuring that each transmission channel can transmit appropriate templated data according to the needs of the receiving end.

[0045] Preferably, when the platform transmits chronic disease resource data through a transmission channel, template processing will format the original chronic disease resource data to meet the needs and standards of the receiving end, that is, the original chronic disease data (such as the patient's physiological data, medical images, health monitoring data, etc.) will be organized and displayed according to the definition of the template, and finally transmitted to multiple information receiving ends through the corresponding transmission channels, for example, patient application devices (receiving low-granularity data, viewing health trend charts, etc.), family doctors (receiving medium-granularity data, such as analysis charts, cycle data, etc.), hospital management systems (receiving high-granularity data, such as detailed physiological data, medical images, patients' complete health records, etc.), ensuring that the data received by different receiving ends meets the granularity and format of their requirements, thereby improving the data transmission and display effects in chronic disease resource management and meeting the needs of different users.

[0046] Furthermore, step S400 also includes step S410, wherein the cloud-based processor performs template processing on the cross-modal database to generate an initialization template, wherein the initialization template includes the number of initialization template partitions, the initialization information hierarchy structure, the initialization data update frequency, and the initialization data display accuracy; and step S420, wherein the template parameters of the initialization template are adjusted according to the multiple data granularities, and multiple granularity templates are output, wherein the template parameters include the number of template partitions, the information hierarchy structure data, the update frequency, and the data display accuracy.

[0047] Preferably, the data in the cross-modal database (such as medical images, electronic health records, genomic data, sensor data, etc.) is templated through a cloud processor to generate an initialization template, that is, converted into a standardized format and structure to ensure that data of different types and formats can be uniformly displayed to different users (such as patients, doctors, researchers, etc.), and to resolve differences in format, granularity, level of detail, etc. of cross-modal data. Specifically, the number of initialization template partitions, the initialization information hierarchy, the initialization data update frequency, and the initialization data display accuracy are defined, wherein the number of initialization template partitions determines the partition and display area of ​​the data in the template, each partition represents a different type of data display area, reflecting the level and complexity of the data display, the more template partitions, the more detailed the data display, for example, there is one partition for displaying the patient's basic information (such as name, age, gender, etc.), and another partition for displaying health status (such as blood sugar level, heart rate, body temperature, etc.).

[0048] Preferably, the initialization information hierarchy defines the hierarchical structure of the data in the template, that is, the order in which the data is displayed from overview to detail, for example, the top level displays high-level health overview data (overall health status, trends, etc.); the secondary level displays more detailed data (such as the results of the most recent examination, the patient's various physiological indicators, etc.); the bottom level displays the most detailed data (such as daily health monitoring data, laboratory test data, etc.). The initialization data update frequency defines the frequency of data updates. Different types of data will have different update frequency requirements. For example, for real-time monitoring data (such as patients' blood sugar levels, heart rates, etc.), frequent updates are required, which may be updated every minute or every hour; for static data (such as diagnostic reports, historical medical records, etc.), the update frequency is lower, which may be updated weekly or monthly to ensure the timeliness of the data, and set reasonable update times according to different data types. The initialization data display accuracy defines the accuracy level of data display, that is, the degree of detail of the displayed data. For example, for receivers that require detailed data display (such as doctors or researchers), complete experimental results, time series data, etc. may be displayed; for receivers that require simplified display (such as patients or health monitoring equipment), only trend charts or summary data may be displayed; data display accuracy can be flexibly adjusted according to the needs of the receiver, so that the data display will not be too complicated.

[0049] Preferably, different receiving ends have different requirements for data granularity, and then the various parameters of the template are adjusted according to the different data granularity requirements. Specifically, if the receiving end needs to display more data levels (such as detailed health analysis), the number of partitions of the template may be increased to divide the data into more display areas; for receiving ends that require more detailed data (such as hospital systems), the hierarchical structure will be more complex and in-depth, and for receiving ends with lower granularity requirements (such as patient applications), the hierarchical structure may be simplified; for receiving ends with high-granularity data (such as real-time monitoring systems), the data update frequency is higher, while low-granularity data (such as health trend charts) are updated less frequently; a receiving end with high-granularity data may need to display precise time series data or detailed laboratory results, and its display accuracy will be higher, while a receiving end with low-granularity data may only need rough trends or overview data, and the display accuracy is set to be lower; finally, multiple granularity templates are generated, each template corresponds to a different level of requirements, and data of different granularities will be displayed according to the specific needs of the receiving end, thereby ensuring that the data in the cross-modal database can provide the most suitable display format according to the needs of different receiving ends, while optimizing the efficiency of data processing and transmission.

[0050] Furthermore, step S420 also includes step S421, wherein the multi-party information receiving end performs quality assessment on the received templated chronic disease resource data to obtain multiple quality assessment indicators corresponding to the multi-party information receiving end; step S422, feeds back the multiple quality assessment indicators to the cloud processor, updates the template parameters of the multiple granular templates according to the cloud processor until the preset quality assessment indicators are met, and outputs the updated multiple granular templates.

[0051] Preferably, multiple information receiving terminals are used to conduct quality assessment on templated chronic disease resource data, that is, to check all aspects of the data to ensure its accuracy, completeness, timeliness and availability, which may include assessing whether the data is true and correct, and whether it reflects the actual health status; whether the data is complete and whether key parts are missing; whether the data is consistent with other data sources and whether there are conflicts; whether the data is updated in a timely manner and whether it reflects the latest health status; whether the displayed data meets the preset display accuracy requirements (such as accurate to days, hours, minutes, etc.); each receiving terminal generates multiple quality assessment indicators for the quality assessment of the data, reflecting the quality status of the data in different dimensions. For example, for patient-side applications, more attention is paid to data accuracy and display accuracy, and for hospital management systems, more attention may be paid to data integrity, timeliness and consistency.

[0052] Preferably, each receiving end will feed back the evaluation results of the data quality (i.e., multiple quality evaluation indicators) to the cloud processor to help understand whether the current data quality meets the needs of different receiving ends and whether the template needs to be adjusted to improve the data quality, and then adjust the template parameters according to the received quality evaluation indicators to meet the quality requirements of the receiving end. Specifically, if the data display is not clear enough, the display area in the template may be increased, or the size of the partition may be adjusted; if the data hierarchy is not clear enough, or more detailed information is needed to be displayed, the cloud processor will adjust the hierarchical structure of the template to make the data stratification more reasonable; if the feedback shows that the timeliness of data updates is insufficient, the cloud processor may increase the frequency of data updates to ensure that the receiving end receives the latest data; if the data display accuracy is insufficient, the cloud processor can adjust the display accuracy of the template, such as providing more detailed timestamps, more accurate numerical displays, etc.

[0053] Preferably, the cloud processor continuously optimizes the template according to the feedback from the receiving end until the data quality meets the preset quality assessment standard, wherein the preset quality assessment index is the set ideal data quality standard, which may include data accuracy reaching a certain standard, data timeliness requirements, such as hourly updates, data integrity requirements, such as no missing data fields, and whether the display accuracy meets expectations, such as accurate to the hour or minute; when all template parameters are adjusted to meet the preset quality standards, the cloud processor will generate and output multiple updated granular templates, and redistribute them to the corresponding receiving ends to ensure that the data quality is optimal, while ensuring that different receiving ends can receive data displays that meet their quality requirements.

[0054] Furthermore, step S422 also includes step A, taking the multi-party information receiving end as an edge node and the cloud processor as a central node to build a resource topology management platform; step B, based on the resource topology management platform, performing update frequency synchronization processing on the granularity templates of the multiple transmission channels.

[0055] Preferably, multiple information receiving terminals (such as hospital management systems, patients' mobile applications, health monitoring devices, etc.) are used as edge nodes, responsible for receiving and displaying data according to their own needs (such as different data granularity or data format); the cloud processor is used as the central node, responsible for centralized management of data storage, processing, template generation, update and other tasks, including processing cross-modal data aggregation, templating, and adjusting the granularity, accuracy, and update frequency of data transmission according to the feedback from the receiving terminal; a resource topology management platform is constructed, responsible for monitoring and coordinating data flows, template management and updates between edge nodes (receiving terminals) and central nodes (cloud processors), specifically including node management, management of resources of each receiving terminal (edge ​​node) and cloud processor (central node), including storage capacity, computing power, network connection, etc.; data flow management, coordination of data transmission from the cloud to edge nodes, to ensure data accuracy, timeliness and security; template management, management of the generation, update and distribution of cross-modal data templates, to ensure that different receiving terminals can obtain data display templates that suit their needs), thereby more efficiently coordinating and managing data transmission, processing and updating.

[0056] Preferably, different transmission channels are responsible for transmitting data templates of different granularities (such as low granularity, medium granularity, and high granularity). Each transmission channel will transmit the data template suitable for a specific receiving end to the target receiving end, and then synchronize the update frequency of the granularity templates of multiple transmission channels through the resource topology management platform. Specifically, due to different data granularities and different data requirements of the receiving end, the frequency of data updates may be different. For example, real-time monitoring data (such as blood sugar levels, heart rate, etc.) may need to be updated once a minute, while summary data (such as health trends) may only need to be updated daily or weekly. Through the coordination of the platform, templates of different transmission channels (low granularity, medium granularity, and high granularity) are updated at a consistent frequency to ensure that each receiving end can obtain the latest template data that meets its needs within an appropriate time, thereby improving the efficiency, timeliness and reliability of data transmission.

[0057] In the above, refer to Figure 1 The cross-modal data generation-driven chronic disease management method according to an embodiment of the present invention is described in detail. Figure 2 A chronic disease management platform driven by cross-modal data generation according to an embodiment of the present invention is described.

[0058] The cross-modal data generation-driven chronic disease management platform according to the embodiment of the present invention is used to solve the technical problems existing in the prior art, such as the inconsistency between the cross-modal data granularity and the modal semantics, the difficulty in flexibly generating appropriate granularity templates according to needs, resulting in low information transmission efficiency, poor accuracy and real-time performance of chronic disease management, and achieves the technical effect of improving data transmission efficiency and the accuracy and real-time performance of chronic disease management. Figure 2 As shown, the cross-modal data generation-driven chronic disease management platform includes: a cross-modal database access unit 10, a feature vector extraction unit 20, a transmission channel configuration unit 30, and a granular template output unit 40.

[0059] A cross-modal database access unit 10 is used to access a cross-modal database, which is stored in a cloud processor, and the cloud processor includes multiple transmission channels; a feature vector extraction unit 20 is used to obtain a multi-party information receiving end and extract a feature vector representing the data granularity requirements in the multi-party information receiving end; a transmission channel configuration unit 30 is used to cluster the multi-party information receiving end according to the feature vector, output the clustering result, configure the number of the multiple transmission channels according to the number of classes of the clustering result, and establish a connection between the multiple transmission channels and the multi-party information receiving end; a granularity template output unit 40 is used to set multiple data granularities of the multiple transmission channels, perform template processing according to the multiple data granularities, output multiple granularity templates, download the multiple granularity templates to the multiple transmission channels, and perform template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmit them to the multi-party information receiving end accordingly.

[0060] The specific configuration of the cross-modal database access unit 10 will be described in detail below. The cross-modal database access unit 10 further includes: mapping the data of each modality in the cross-modal database to a high-dimensional feature space, and outputting a multi-modal high-dimensional feature data set corresponding to multiple modalities; using a functional-adversarial network to generatively align the multi-modal high-dimensional feature data set, and outputting the aligned multi-modal feature data set; using the aligned multi-modal feature data set to perform modal data feature fusion, and update the cross-modal database.

[0061] The specific configuration of the cross-modal database access unit 10 will be described in detail below. The cross-modal database access unit 10 further includes: wherein the functional-adversarial network includes a generator and a discriminator, the output end of the generator is connected to the input end of the discriminator, and the output end of the discriminator is connected to the input end of the generator; the generator and the discriminator are adversarially trained, the modal data generated by the generator is obtained, the generated modal data is input into the discriminator, the difference distribution with the sample modal data is calculated, and parameter learning is performed with the goal of being less than the preset difference distribution, and the trained functional-adversarial network is obtained.

[0062] The specific configuration of the cross-modal database access unit 10 will be described in detail below. The cross-modal database access unit 10 further includes: calculating the semantic difference between the features in each modality of the aligned multimodal feature data set; storing the features whose semantic difference is greater than or equal to the preset semantic difference in parallel, and storing the features whose semantic difference is less than the preset semantic difference in overwriting.

[0063] The specific configuration of the granular template output unit 40 will be described in detail below. The granular template output unit 40 further includes: the cloud processor performs template processing on the cross-modal database to generate an initialization template, the initialization template includes the number of initialization template partitions, the initialization information hierarchy structure, the initialization data update frequency and the initialization data display accuracy; the template parameters of the initialization template are adjusted according to the multiple data granularities, and multiple granular templates are output, wherein the template parameters include the number of template partitions, the information hierarchy structure data, the update frequency and the data display accuracy.

[0064] The specific configuration of the granular template output unit 40 will be described in detail below. The granular template output unit 40 further includes: using the multi-party information receiving end as an edge node and the cloud processor as a central node to build a resource topology management platform; and performing update frequency synchronization processing on the granular templates of the multiple transmission channels based on the resource topology management platform.

[0065] The specific configuration of the granular template output unit 40 will be described in detail below. The granular template output unit 40 further includes: obtaining a warning matching strategy logic library according to the hierarchical warning mechanism; labeling and classifying the warning matching strategy logic library according to the battery fault label set to obtain a battery warning label strategy library; performing strategy matching analysis on the battery module fault prediction parameters based on the battery warning label strategy library to determine the battery fault warning strategy, wherein the battery fault warning strategy includes a warning signal type and a warning mode level.

[0066] The specific configuration of the transmission channel configuration unit 30 will be described in detail below. The transmission channel configuration unit 30 further includes: wherein the feature vector of each information receiving end includes data precision requirements, data interaction complexity, data modality preference, and granularity level requirements; calculating the similarity between the feature vectors of every two information receiving ends between the multiple information receiving ends, and obtaining a similarity set; clustering the multiple information receiving ends using the similarity set, and outputting the clustering result.

[0067] The cross-modal data generation-driven chronic disease management platform provided in the embodiments of the present invention can execute the cross-modal data generation-driven chronic disease management method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0068] Although the present application makes various references to certain modules in the platform 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 implemented; 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.

[0069] The above specific implementations do not constitute a limitation on the protection scope of this 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 this application should be included in the protection scope of this application.

Claims

1. A chronic disease management method driven by cross-modal data generation, characterized in that: The method comprises: Accessing a cross-modal database, wherein the cross-modal database is stored in a cloud processor, wherein the cloud processor includes a plurality of transmission channels; Acquire multiple information receiving terminals, and extract feature vectors representing data granularity requirements from the multiple information receiving terminals; Clustering the multi-party information receiving terminals according to the feature vectors, outputting clustering results, configuring the number of the multiple transmission channels according to the number of clustering results, and establishing connections between the multiple transmission channels and the multi-party information receiving terminals; Set multiple data granularities for the multiple transmission channels, perform template processing respectively according to the multiple data granularities, output multiple granularity templates, download the multiple granularity templates to the multiple transmission channels, perform template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmit them accordingly to the multiple information receiving ends.

2. The cross-modal data generation-driven chronic disease management method according to claim 1, characterized in that: After accessing the cross-modal database, the method includes: Performing high-dimensional feature space mapping on the data of each modality in the cross-modal database, and outputting a multi-modal high-dimensional feature data set corresponding to multiple modalities; Generatively aligning the multimodal high-dimensional feature dataset using a functional adversarial network, and outputting an aligned multimodal feature dataset; The aligned multimodal feature dataset is used to perform modal data feature fusion and update the cross-modal database.

3. The cross-modal data generation-driven chronic disease management method according to claim 2, characterized in that: Methods for Training Functional-Adversarial Networks include: The functional adversarial network includes a generator and a discriminator, the output end of the generator is connected to the input end of the discriminator, and the output end of the discriminator is connected to the input end of the generator; The generator and the discriminator are subjected to adversarial training, the modal data generated by the generator is obtained, the generated modal data is input into the discriminator, the difference distribution with the sample modal data is calculated, parameter learning is performed with the goal of being less than a preset difference distribution, and a trained functional-adversarial network is obtained.

4. The cross-modal data generation-driven chronic disease management method according to claim 3, characterized in that: The method of performing modal data feature fusion using the aligned multimodal feature data set to update the cross-modal database includes: Calculate the semantic difference between the features in each modality of the aligned multimodal feature dataset; Features whose semantic difference is greater than or equal to a preset semantic difference are stored in parallel, and features whose semantic difference is less than the preset semantic difference are stored in an overwritten manner.

5. The cross-modal data generation-driven chronic disease management method according to claim 1, characterized in that: Performing template processing respectively according to the multiple data granularities and outputting multiple granularity templates, the method includes: The cloud processor performs template processing on the cross-modal database to generate an initialization template, wherein the initialization template includes the number of initialization template partitions, the initialization information hierarchical structure, the initialization data update frequency, and the initialization data display accuracy; The template parameters of the initialization template are adjusted according to the multiple data granularities to output multiple granularity templates, wherein the template parameters include the number of template partitions, information hierarchical structure data, update frequency, and data display accuracy.

6. The cross-modal data generation-driven chronic disease management method according to claim 5, characterized in that: After outputting the multiple granularity templates, the method further includes: The multi-party information receiving end performs quality assessment on the received templated chronic disease resource data, and obtains a plurality of quality assessment indicators corresponding to the multi-party information receiving end; The multiple quality assessment indicators are fed back to the cloud processor, and the template parameters of the multiple granularity templates are updated according to the cloud processor until the preset quality assessment indicators are met, and the updated multiple granularity templates are output.

7. The cross-modal data generation-driven chronic disease management method according to claim 6, characterized in that: The multi-party information receiving end is used as an edge node, and the cloud processor is used as a central node to build a resource topology management platform; Based on the resource topology management platform, update frequency synchronization processing is performed on the granularity templates of the multiple transmission channels.

8. The cross-modal data generation-driven chronic disease management method according to claim 1, characterized in that: Clustering the multiple information receiving terminals according to the feature vectors and outputting the clustering results. include: Among them, the feature vector of each information receiving end includes data precision requirements, data interaction complexity, data modality preference, and granularity level requirements; Calculating the similarity between feature vectors of every two information receiving ends among the multiple information receiving ends to obtain a similarity set; The multiple information receiving terminals are clustered using the similarity set, and a clustering result is output.

9. A cross-modal data generation-driven chronic disease management platform, characterized by: The platform is used to implement the cross-modal data generation-driven chronic disease management method according to any one of claims 1 to 8, and the platform includes: A cross-modal database access unit, used to access a cross-modal database, wherein the cross-modal database is stored in a cloud processor, and the cloud processor includes a plurality of transmission channels; A feature vector extraction unit, used to obtain multiple information receiving terminals and extract feature vectors representing data granularity requirements from the multiple information receiving terminals; a transmission channel configuration unit, configured to cluster the multi-party information receiving terminals according to the feature vectors, output a clustering result, configure the number of the multiple transmission channels according to the number of classes of the clustering result, and establish connections between the multiple transmission channels and the multi-party information receiving terminals; A granularity template output unit is used to set multiple data granularities of the multiple transmission channels, perform template processing according to the multiple data granularities, output multiple granularity templates, download the multiple granularity templates to the multiple transmission channels, perform template processing on the chronic disease resource data transmitted by the multiple transmission channels according to the multiple granularity templates, and transmit them to the multiple information receiving ends accordingly.

Images