Method and device for monitoring health state of home user, storage medium and terminal

Abstract

The invention discloses a method and device for monitoring the health state of a home user, a storage medium and a terminal, relates to the field of data processing and the field of medical health, and mainly aims to solve the problem of poor monitoring effectiveness of the health state of the home user in the prior art. Comprising the following steps: acquiring electrocardio data of a home user collected by wearing equipment; the electrocardio data are processed on the basis of a combined prediction model which completes model training, an electrocardio prediction result is obtained, and the combined prediction model is constructed on the basis of a random forest model, a support vector machine model and a convolutional neural model; and determining a monitoring result of the home user based on an associated state prediction result of the home user and the electrocardiogram prediction result, wherein the associated state prediction result comprises a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result and a pulse state prediction result.

Description

Technical Field

[0001] The present application relates to the fields of data processing and medical health, and in particular to a method and device, storage medium, and terminal for monitoring the health status of home users. Background Art

[0002] With the rise of smart medical technology, home-based elderly care has become the choice of more and more families. At this time, a home-based elderly care monitoring system can be developed based on smart AI technology to achieve medical purposes such as disease auxiliary diagnosis, health management, and remote consultation for home-based elderly care users.

[0003] Currently, existing home-based elderly care platforms typically monitor users' health status, for example, by implementing threshold monitoring for blood pressure, heart rate, and blood oxygen levels. However, simple threshold monitoring results are inflexible and inaccurate, failing to accurately monitor the health of elderly care users with varying medical conditions. Therefore, an AI-based approach to monitoring the health of home-based users is urgently needed to address these issues. Summary of the Invention

[0004] In view of this, the present application provides a method and device, storage medium, and terminal for monitoring the health status of home users, the main purpose of which is to solve the problem of poor effectiveness of existing monitoring of the health status of home users.

[0005] According to one aspect of the present application, a method for monitoring the health status of a home user is provided, comprising:

[0006] Obtain ECG data collected by wearable devices from users at home;

[0007] Processing the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model;

[0008] The monitoring result of the home user is determined based on the association state prediction result of the home user and the electrocardiogram prediction result, wherein the association state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result and a pulse state prediction result.

[0009] Furthermore, the ECG data is processed based on the combined prediction model that has completed model training to obtain an ECG prediction result, which includes:

[0010] Obtain ECG sample data and build a combined prediction model consisting of two random forest models, a support vector machine model, and a convolutional neural model;

[0011] Performing model training on the ECG sample data based on the combined prediction model, and completing the model training of the combined prediction model when the model loss value of the combined prediction model is less than a preset loss threshold;

[0012] Among them, the input samples and the first output samples in the ECG sample data are used to perform model training on the first random forest model, the support vector machine model and the convolutional neural model, and the second output samples in the ECG sample data and the first output samples are used to perform model training on the second random forest model.

[0013] Furthermore, the ECG data is processed based on the combined prediction model that has completed model training to obtain an ECG prediction result, which includes:

[0014] Using the ECG data as input parameters for the first random forest model, the support vector machine model, and the convolutional neural model to complete model training, to obtain processed output parameters;

[0015] The output parameters are used as input parameters of the second random forest model to complete model training to obtain the electrocardiogram prediction result.

[0016] Furthermore, before determining the monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result, the method further includes:

[0017] Obtaining sleep data, blood pressure data, blood sugar data, and pulse data of the home user collected by the wearable device;

[0018] Determine the sleep state prediction result of the sleep data, the blood pressure state prediction result of the blood pressure data, the blood sugar state prediction result, and the pulse state prediction result of the pulse data according to the state prediction strategy;

[0019] The state prediction strategy includes at least one of a threshold state comparison strategy, a prediction model processing strategy, and an abnormal state identification strategy.

[0020] Furthermore, the method further comprises:

[0021] Generate voice broadcast data according to the monitoring result, and send the voice broadcast data to the monitoring terminal of the home user;

[0022] Receive voice feedback data from the home user or the associated user regarding the voice broadcast data, and process the monitoring result based on the voice feedback data.

[0023] Furthermore, the method further comprises:

[0024] Obtaining preset companionship task information of the home user, where the preset companionship task information is configured for the home user and / or associated users according to time conditions and voice broadcast conditions;

[0025] When the time condition or voice broadcast condition in the preset accompanying task information is detected to be triggered, the broadcast content in the preset accompanying task information is retrieved for broadcasting.

[0026] Furthermore, the method further comprises:

[0027] When the voice companionship instruction of the home user is received, the companionship execution operation and the companionship execution object corresponding to the voice companionship instruction are parsed, and the companionship execution object is driven to perform the companionship action according to the companionship execution operation.

[0028] According to another aspect of the present application, a device for monitoring the health status of a home user is provided, comprising:

[0029] The acquisition module is used to obtain the ECG data collected by the wearable device at home;

[0030] A processing module, configured to process the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model;

[0031] A monitoring module is used to determine the monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result, the association status prediction result including the sleep status prediction result, blood pressure status prediction result, blood sugar status prediction result and pulse status prediction result.

[0032] Furthermore, the device further comprises:

[0033] A construction module is used to obtain ECG sample data and build a combined prediction model including two random forest models, a support vector machine model, and a convolutional neural model;

[0034] A training module, configured to perform model training on the ECG sample data based on the combined prediction model, and complete the model training of the combined prediction model when the model loss value of the combined prediction model is less than a preset loss threshold;

[0035] Among them, the input samples and the first output samples in the ECG sample data are used to perform model training on the first random forest model, the support vector machine model and the convolutional neural model, and the second output samples in the ECG sample data and the first output samples are used to perform model training on the second random forest model.

[0036] Furthermore, the processing module is specifically used to use the ECG data as input parameters of the first random forest model, the support vector machine model and the convolutional neural model to complete model training, and obtain processed output parameters; and use the output parameters as input parameters of the second random forest model to complete model training to obtain the ECG prediction results.

[0037] Furthermore,

[0038] The acquisition module is also used to obtain the sleep data, blood pressure data, blood sugar data and pulse data of the home user collected by the wearable device; determine the sleep state prediction result of the sleep data, the blood pressure state prediction result of the blood pressure data, the blood sugar state prediction result and the pulse state prediction result of the pulse data according to the state prediction strategy; wherein the state prediction strategy includes at least one of a threshold state comparison strategy, a prediction model processing strategy and an abnormal state identification strategy.

[0039] Furthermore, the device further comprises:

[0040] A generating module, configured to generate voice broadcast data according to the monitoring result, and send the voice broadcast data to the monitoring terminal of the home user;

[0041] The receiving module is used to receive the voice feedback data provided by the home user or the associated user regarding the voice broadcast data, and process the monitoring result based on the voice feedback data.

[0042] Furthermore,

[0043] The acquisition module is specifically used to obtain the preset companion task information of the home user, and the preset companion task information is configured for the home user and / or associated users according to time conditions and voice broadcast conditions; when it is detected that the time conditions or voice broadcast conditions in the preset companion task information are triggered, the broadcast content in the preset companion task information is retrieved for broadcast.

[0044] Furthermore, the device further comprises:

[0045] The parsing module is used to parse the companionship execution operation and companionship execution object corresponding to the voice companionship instruction when receiving the voice companionship instruction of the home user, and drive the companionship execution object to perform the companionship action according to the companionship execution operation.

[0046] According to another aspect of the present application, a storage medium is provided, in which at least one executable instruction is stored, and the executable instruction enables a processor to perform operations corresponding to the above-mentioned method for monitoring the health status of a home user.

[0047] According to another aspect of the present application, a computer device is provided, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;

[0048] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned method for monitoring the health status of a home user.

[0049] By means of the above technical solution, the technical solution provided by the embodiment of the present application has at least the following advantages:

[0050] The present application provides a method and device, storage medium, and terminal for monitoring the health status of a home user. Compared with the prior art, the embodiments of the present application collect electrocardiogram (ECG) data of a home user by acquiring a wearable device; the ECG data is processed based on a combined prediction model for which model training has been completed to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model; the monitoring result of the home user is determined based on the associated state prediction result of the home user and the ECG prediction result, wherein the associated state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result, and a pulse state prediction result, thereby achieving the purpose of multi-dimensional physical state monitoring of the home user, meeting the demand for companion monitoring of the health status of the home user, greatly enhancing the companion smart medical effect for the home user, and thereby improving the accuracy and effectiveness of monitoring the health status of the home user.

[0051] 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. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present application. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0053] Figure 1 A flow chart of a method for monitoring the health status of a home user provided in an embodiment of the present application is shown;

[0054] Figure 2 A flowchart of a combined prediction model prediction method provided by an embodiment of the present application is shown;

[0055] Figure 3 A block diagram of a device for monitoring the health status of a home user provided in an embodiment of the present application is shown;

[0056] Figure 4 A structural diagram of a computer device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0057] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0058] Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0059] Based on this, in one embodiment, the embodiment of the present application provides a method for monitoring the health status of home users, and the method is illustrated by applying it to a computer device such as a server. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, such as smart medical systems, digital medical platforms, etc.

[0060] The present application embodiment provides a method for monitoring the health status of a home user, such as Figure 1 As shown, the method includes:

[0061] 101. Obtain ECG data collected by wearable devices from users at home.

[0062] In the embodiment of the present application, the current execution end acts as the server end of the intelligent medical system to exchange data with a wearable device. The wearable device can be a detection device carried by a home user on the body, used to detect blood pressure, heart rate, etc. The wearable device can be a health bracelet, health anklet, or a detection device such as a smart watch worn by the user, and the embodiment of the present application does not make specific restrictions. Among them, the electrocardiogram data is data for detecting the heart rate of the home user, which can be stored in the form of data or electrocardiogram, and the embodiment of the present application does not make specific restrictions.

[0063] It should be noted that the ECG data in the embodiment of the present application is collected in real time, and the collection frequency can be set based on the monitoring needs of home users, including but not limited to 5 minutes or 10 minutes, etc., which is not specifically limited in the embodiment of the present application.

[0064] 102. Process the ECG data based on the combined prediction model for which model training has been completed to obtain an ECG prediction result.

[0065] In an embodiment of the present application, after the current execution end obtains the ECG data, it predicts the ECG data based on the combined prediction model that has completed model training to obtain the ECG prediction result. At this time, the combined prediction model is constructed based on the random forest model, the support vector machine model and the convolutional neural model, that is, when the combined prediction model is constructed, the ECG data as input data is simultaneously input into the random forest model, the support vector machine model and the convolutional neural model for processing, and the three outputs obtained are re-input as input data into the random forest model for processing to obtain the final ECG prediction result. Among them, the two random forest models used can be the same or different, that is, the first random forest model is a single-input single-output random forest model, and the second random forest model is a multi-input single-output random forest model. This embodiment of the present application does not make specific limitations.

[0066] 103. Determine a monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result.

[0067] In the embodiment of the present application, after obtaining the ECG prediction result, in order to conduct an overall monitoring of the health status of the home user, the current execution end obtains the associated state prediction result of the home user, so as to combine it with the ECG prediction result to determine the monitoring result of the home user. Among them, the associated state prediction result includes the sleep state prediction result, the blood pressure state prediction result, the blood sugar state prediction result and the pulse state prediction result. The sleep state prediction result is used to characterize the result of predicting the sleep condition, which can be obtained by predicting the state based on the sleep time. The blood pressure state prediction result is used to characterize the result of predicting the blood pressure condition, which can be obtained by predicting the state based on the blood pressure value. The blood sugar state prediction result is used to characterize the result of predicting the blood sugar condition, which can be obtained by predicting the state based on the blood sugar value. The pulse state prediction result is used to characterize the result of predicting the pulse condition, which can be predicted based on the pulse strength and pulse count. The embodiment of the present application does not make specific limitations.

[0068] An embodiment of the present application provides a method for monitoring the health status of a home user. Compared with the prior art, the embodiment of the present application collects the ECG data of the home user by acquiring a wearable device; processes the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, and the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model; determines the monitoring result of the home user based on the associated state prediction result of the home user and the ECG prediction result, and the associated state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result, and a pulse state prediction result, thereby achieving the purpose of multi-dimensional physical state monitoring of the home user, meeting the demand for companion monitoring of the health status of the home user, greatly enhancing the companion smart medical effect for the home user, and thus improving the accuracy and effectiveness of monitoring the health status of the home user.

[0069] In another embodiment of the present application, for further definition and explanation, the step of processing the ECG data based on the combined prediction model for which model training has been completed to obtain an ECG prediction result includes:

[0070] Obtain ECG sample data and build a combined prediction model consisting of two random forest models, a support vector machine model, and a convolutional neural model;

[0071] Model training is performed on the electrocardiogram sample data based on the combined prediction model, and when the model loss value of the combined prediction model is less than a preset loss threshold, the model training of the combined prediction model is completed.

[0072] In order to predict ECG data based on artificial intelligence models, the current execution end pre-acquires ECG sample data and builds a combined prediction model so that the combined prediction model can be trained based on the ECG sample data. The combined prediction model includes two random forest models, a support vector machine model, and a convolutional neural model. Figure 2 As shown, a random forest network is combined with a support vector machine model and a convolutional neural network model in parallel and trained simultaneously. Another random forest tree model serves as the final integrated network, taking the outputs of the three networks as input for model training. In this case, the input samples and the first output samples in the ECG sample data are used to train the first random forest model, the support vector machine model, and the convolutional neural network model, while the second output samples in the ECG sample data and the first output samples are used to train the second random forest model.

[0073] It should be noted that in the embodiment of the present application, the first random forest model and the second random forest model may be the same or different. In some embodiments, the random forest model includes N trees, and the prediction result of the i-th tree is h i (x), the output prediction result is The model expression of the convolutional neural model is N = (W-F+2P) / S+1, where N is the output data, W is the input data, F is the convolution kernel, P is the padding value, and S is the step size. The model expression of the support vector machine model is Among them, the original optimization problem of the support vector machine model is expressed as:

[0074]

[0075] Among them, W is the weight vector used to determine the direction of the regression hyperplane, b is the bias term, is a slack variable that allows data points to deviate outside the ε error band, and C is a regularization parameter that balances the model complexity (||W|| 2 ) and training error ε is an insensitive parameter that defines the width of the error band. When the deviation between the predicted value and the true value is within ε, no loss is calculated. The Lagrange multiplier method is used to transform the problem into a dual problem, and the kernel function K(x i ,x j ), which is converted to:

[0076] Among them, the kernel function K(x i ,x j ) Different kernel functions can be used for different data input items and data sets. For example, when there is only heartbeat data, a linear kernel can be used. If factors such as blood pressure and medical history are introduced, a polynomial kernel can be used. It is usually necessary to adjust the parameters of the kernel function through cross-validation and other methods to achieve the best model performance, which is not specifically limited in the embodiments of the present application.

[0077] Finally, M, N, Q are taken as data input and trained using the second random forest network.

[0078] In another embodiment of the present application, for further definition and explanation, the step of processing the ECG data based on the combined prediction model for which model training has been completed to obtain an ECG prediction result includes:

[0079] Using the ECG data as input parameters for the first random forest model, the support vector machine model, and the convolutional neural model to complete model training, to obtain processed output parameters;

[0080] The output parameters are used as input parameters of the second random forest model to complete model training to obtain the electrocardiogram prediction result.

[0081] In order to achieve the goal of ECG prediction based on artificial intelligence, thereby improving the accuracy of ECG monitoring for home users, after the final training is completed, the current execution end uses the collected ECG data as the input of the random forest tree model for decision prediction to obtain Y1, and uses the ECG data as the input of the convolutional neural model for prediction to obtain Y2. At the same time, the same ECG data is used as the input of the support vector machine regression model for prediction to obtain Y3. Then, after obtaining Y1, Y2, and Y3, they are re-input into the random forest model as output parameters for prediction to obtain the final ECG prediction result, as shown in the figure. Figure 2 shown.

[0082] In another embodiment of the present application, for further limitation and explanation, before the step of determining the monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result, the method further includes:

[0083] Obtaining sleep data, blood pressure data, blood sugar data, and pulse data of the home user collected by the wearable device;

[0084] The sleep state prediction result of the sleep data, the blood pressure state prediction result of the blood pressure data, the blood sugar state prediction result, and the pulse state prediction result of the pulse data are determined according to a state prediction strategy.

[0085] In order to achieve all-round monitoring of the physical condition of home users, thereby improving the effectiveness of monitoring home users, the current execution end can also collect home users' sleep data, blood pressure data, blood sugar data, and pulse data through wearable devices. At this time, the wearable device that collects the above data can be the same device or a device with different collection functions. For example, a blood sugar detection wearable device can collect real-time blood sugar, and a blood pressure wearable device can collect real-time blood pressure, etc. The embodiments of this application do not make specific limitations.

[0086] It should be noted that after the current execution end collects the above data, it can determine the sleep state prediction results for sleep data, blood pressure data, blood glucose data, and pulse data according to the state prediction strategy. The state prediction strategy includes at least one of a threshold state comparison strategy, a prediction model processing strategy, and an abnormal state identification strategy. Specifically, the threshold state comparison strategy represents a strategy for determining the state of sleep data, blood pressure data, blood glucose data, and pulse data according to a preset threshold. For example, if the sleep threshold is 4, then when the sleep data is less than 4, a sleep deficiency prediction result is obtained. The preset model processing strategy represents a method for classifying the state of sleep data, blood pressure data, blood glucose data, and pulse data according to a pre-selected trained machine learning model. For example, after training a deep learning model based on blood pressure sample data, the blood pressure data is classified to obtain a classification result of normal blood pressure state. The abnormal state identification strategy represents a method for determining only the values in the sleep data, blood pressure data, blood glucose data, and pulse data that are in an abnormal state. It can be a threshold method or a machine learning model identification method, which is not specifically limited in this embodiment of the application.

[0087] In another embodiment of the present application, for further definition and explanation, the steps further include:

[0088] Generate voice broadcast data according to the monitoring result, and send the voice broadcast data to the monitoring terminal of the home user;

[0089] Receive voice feedback data from the home user or the associated user regarding the voice broadcast data, and process the monitoring result based on the voice feedback data.

[0090] In order to improve the flexibility of health status monitoring for home users, the current execution end can also generate voice broadcast data based on the obtained monitoring results to output to the monitoring terminal for broadcast. Among them, since the monitoring results may include the results predicted based on the electrocardiogram monitoring results and the sleep state prediction results, the blood pressure state prediction results of the blood pressure data, the blood sugar state prediction results and the pulse state prediction results of the pulse data, the current execution end can pre-configure specific broadcast content for different monitoring results. For example, when the monitoring result is insufficient sleep, the pre-configured voice broadcast data can be "Insufficient sleep, please rest as soon as possible", which is not specifically limited in the embodiment of the present application. At this time, the monitoring terminal can be a terminal device that has data interaction with the current execution end and plays voice, such as a mobile phone, a playback speaker, etc., and then broadcasts based on the received voice broadcast data.

[0091] It should be noted that after the monitoring terminal completes the broadcast, the home user or the associated user can reply, and the voice feedback data as feedback is fed back to the current end. At this time, after the current execution end receives the voice feedback data of the home user or the associated user for the voice broadcast data, it processes the monitoring results based on this voice feedback data. For example, if the voice feedback data is "OK, take a rest immediately", the monitoring results can be tagged to start the execution of related smart devices. For example, in the scenario of smart home device interconnection, the closing operation of the smart curtains can be started. This is not specifically limited in the embodiment of the present application. In addition, the current execution end can pre-record the voice information of the home user for identification, and can also record associated users with supervision or kinship for identification to assist the current execution end in automatically executing terminal execution operations such as smart home devices.

[0092] In some professional medical audio and video embodiments, family doctors give chronic disease labels to the elderly, such as hypertension. When the smart wearable device collects the blood pressure, the current execution end can directly predict the blood pressure status based on the blood pressure. At the same time, the obtained blood pressure status prediction results are transmitted to the AIGC audio and video tool to generate corresponding audio and video content. Finally, the audio and video content is displayed to the elderly at home through the connected multimedia peripherals.

[0093] In another embodiment of the present application, for further definition and explanation, the steps further include:

[0094] Obtaining preset companionship task information of the home user;

[0095] When the time condition or voice broadcast condition in the preset accompanying task information is detected to be triggered, the broadcast content in the preset accompanying task information is retrieved for broadcasting.

[0096] In order to achieve the purpose of intelligent companionship for home users, thereby improving the effectiveness of home users' health status monitoring, the current execution end can also obtain the preset companionship task information aggregated by user. At this time, the preset companionship task information is configured for the home user and / or associated users according to the time condition and voice broadcast condition. For example, the preset companionship task information can be a voice content that broadcasts a reminder to take medicine at 12 o'clock, or it can be a specific companionship speech that broadcasts once every hour, such as "Grandma, do you need to turn on the music?", etc., which is not specifically limited in this embodiment of the application. Furthermore, when the time condition or voice broadcast condition in the preset companionship task information is detected to be triggered, the broadcast content in the preset companionship task information is retrieved for broadcasting. In addition, the voice broadcast condition can also be based on the preset companionship task information after monitoring the home user to say a specific sentence such as "Is it time to take medicine?", and if it meets the requirements, the broadcast content in the preset companionship task information is output (such as "It's 12 o'clock, it's time to take medicine") for broadcasting, which is not specifically limited in this embodiment of the application.

[0097] In some companion-style interactive audio and video embodiments, language broadcast conditions can be set according to the elderly's living area, living habits, etc. After obtaining the inquiry statement about the weather of the day, it can be processed through a pre-trained large language model to obtain a reply text. Then, the reply text is transmitted to the AIGC audio and video tool to generate the corresponding audio and video file. Finally, the audio and video file is displayed to the elderly at home through the connected multimedia peripherals.

[0098] In another embodiment of the present application, for further definition and explanation, the steps further include:

[0099] When the voice companionship instruction of the home user is received, the companionship execution operation and the companionship execution object corresponding to the voice companionship instruction are parsed, and the companionship execution object is driven to perform the companionship action according to the companionship execution operation.

[0100] In order to meet the intelligent companionship needs of different home users, the current execution end can also monitor the voice companionship instructions of the home users in real time. At this time, it can be pre-configured as different voice start-up statements, for example, "Smart doctor, I'm a little stuffy" or "Smart doctor, I'm a little bored", etc., which are not specifically limited in the embodiments of the present application. After receiving the voice companionship instruction, the accompanying execution operation and accompanying execution object corresponding to the voice companionship instruction are parsed. At this time, the accompanying execution operations and accompanying execution objects corresponding to different voice companionship instructions can be determined based on the pre-established accompanying operation mapping relationship. The accompanying operation mapping relationship includes the accompanying execution operations and accompanying execution objects corresponding to different voice companionship instructions. The accompanying intention in the voice companionship instruction can also be parsed based on the intelligent artificial model to match the corresponding accompanying execution operation and accompanying execution object based on the accompanying intention. For example, if the voice companionship instruction is "Little doctor, I have chest tightness", the corresponding accompanying execution operation can be to turn on the air purifier, and the corresponding accompanying execution object is the air purifier, so as to drive the accompanying execution object to perform the accompanying action according to the accompanying execution operation, that is, to turn on the air purifier. The embodiments of the present application are not specifically limited.

[0101] In some embodiments, the current execution end can be connected to TTS (Text To Speech), voice recognition function, AIGC (Artificial Intelligence Generated Content) and other applications to achieve the purpose of communicating with home users. In addition, the current execution end can also be connected to mobile application APPs and multimedia devices to establish a medium for interaction with the elderly, and can also be connected to a digital display system to facilitate family doctors and guardians to check the physical condition of the elderly. The current execution end can also be connected to the early warning system. If the elderly person's body is abnormal, an alarm will be automatically generated to inform the family doctor and guardian so that a treatment plan can be taken in time. The implementation of this application is not specifically limited.

[0102] In some embodiments, the current execution end can also display all the aforementioned generated data to the family doctor and the associated users of the home user in order to adjust the companionship task information, such as reducing certain types of labels that have been defined as rumors, such as: microwave heating of food causes cancer, etc., which is not specifically limited in the embodiments of this application.

[0103] The embodiment of the present application provides another method for monitoring the health status of home users, which achieves the purpose of multi-dimensional physical condition monitoring of home users, meets the needs of companion monitoring of the health status of home users, greatly enhances the companion smart medical effect for home users, and thus improves the accuracy and effectiveness of monitoring the health status of home users.

[0104] Furthermore, as a response to the above Figure 1The embodiment of the present application provides a device for monitoring the health status of a home user, such as Figure 3 As shown, the device includes:

[0105] An acquisition module 21 is used to acquire the ECG data of the user at home collected by the wearable device;

[0106] A processing module 22 is configured to process the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model;

[0107] The monitoring module 23 is used to determine the monitoring result of the home user based on the association state prediction result of the home user and the electrocardiogram prediction result, and the association state prediction result includes the sleep state prediction result, blood pressure state prediction result, blood sugar state prediction result and pulse state prediction result.

[0108] Furthermore, the device further comprises:

[0109] A construction module is used to obtain ECG sample data and build a combined prediction model including two random forest models, a support vector machine model, and a convolutional neural model;

[0110] A training module, configured to perform model training on the ECG sample data based on the combined prediction model, and complete the model training of the combined prediction model when the model loss value of the combined prediction model is less than a preset loss threshold;

[0111] Among them, the input samples and the first output samples in the ECG sample data are used to perform model training on the first random forest model, the support vector machine model and the convolutional neural model, and the second output samples in the ECG sample data and the first output samples are used to perform model training on the second random forest model.

[0112] Furthermore, the processing module is specifically used to use the ECG data as input parameters of the first random forest model, the support vector machine model and the convolutional neural model to complete model training, and obtain processed output parameters; and use the output parameters as input parameters of the second random forest model to complete model training to obtain the ECG prediction results.

[0113] Further,

[0114] The acquisition module is also used to obtain the sleep data, blood pressure data, blood sugar data and pulse data of the home user collected by the wearable device; determine the sleep state prediction result of the sleep data, the blood pressure state prediction result of the blood pressure data, the blood sugar state prediction result and the pulse state prediction result of the pulse data according to the state prediction strategy; wherein the state prediction strategy includes at least one of a threshold state comparison strategy, a prediction model processing strategy and an abnormal state identification strategy.

[0115] Furthermore, the device further comprises:

[0116] A generating module, configured to generate voice broadcast data according to the monitoring result, and send the voice broadcast data to the monitoring terminal of the home user;

[0117] The receiving module is used to receive the voice feedback data provided by the home user or the associated user regarding the voice broadcast data, and process the monitoring result based on the voice feedback data.

[0118] Furthermore,

[0119] The acquisition module is specifically used to obtain the preset companion task information of the home user, and the preset companion task information is configured for the home user and / or associated users according to time conditions and voice broadcast conditions; when it is detected that the time conditions or voice broadcast conditions in the preset companion task information are triggered, the broadcast content in the preset companion task information is retrieved for broadcast.

[0120] Furthermore, the device further comprises:

[0121] The parsing module is used to parse the companionship execution operation and companionship execution object corresponding to the voice companionship instruction when receiving the voice companionship instruction of the home user, and drive the companionship execution object to perform the companionship action according to the companionship execution operation.

[0122] An embodiment of the present application provides a device for monitoring the health status of a home user. Compared with the prior art, the embodiment of the present application collects the electrocardiogram (ECG) data of the home user by acquiring a wearable device; processes the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model; determines the monitoring result of the home user based on the associated state prediction result of the home user and the ECG prediction result, wherein the associated state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result, and a pulse state prediction result, thereby achieving the purpose of multi-dimensional physical state monitoring of the home user, meeting the demand for companion monitoring of the health status of the home user, greatly enhancing the companion smart medical effect for the home user, and thereby improving the accuracy and effectiveness of monitoring the health status of the home user.

[0123] According to one embodiment of the present application, a storage medium is provided, wherein the storage medium stores at least one executable instruction, and the computer executable instruction can execute the method for monitoring the health status of a home user in any of the above method embodiments.

[0124] Figure 4 A schematic structural diagram of a computer device provided according to an embodiment of the present application is shown. The specific embodiment of the present application does not limit the specific implementation of the computer device.

[0125] like Figure 4 As shown, the computer device may include: a processor 302 , a communications interface 304 , a memory 306 , and a communication bus 308 .

[0126] The processor 302 , the communication interface 304 , and the memory 306 communicate with each other via a communication bus 308 .

[0127] The communication interface 304 is used to communicate with other devices such as clients or other servers.

[0128] The processor 302 is used to execute the program 310, and specifically can execute the relevant steps in the above-mentioned embodiment of the method for monitoring the health status of a home user.

[0129] Specifically, the program 310 may include program codes, which include computer operation instructions.

[0130] Processor 302 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in a computer device may be processors of the same type, such as one or more CPUs, or may be processors of different types, such as one or more CPUs and one or more ASICs.

[0131] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0132] The program 310 may be specifically configured to cause the processor 302 to perform the following operations:

[0133] Obtain ECG data collected by wearable devices from users at home;

[0134] Processing the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model;

[0135] The monitoring result of the home user is determined based on the association state prediction result of the home user and the electrocardiogram prediction result, wherein the association state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result and a pulse state prediction result.

[0136] Obviously, those skilled in the art should understand that the modules or steps of the present application described above can be implemented using a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices. Alternatively, they can be implemented using program code executable by the computing device, so that they can be stored in a storage device and executed by the computing device. In some cases, the steps shown or described can be performed in a different order than herein, or they can be made into separate integrated circuit modules, or multiple modules or steps can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0137] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Persons skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for monitoring the health status of a home user, characterized in that: include: Obtain ECG data collected by wearable devices from users at home; Processing the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model; The monitoring result of the home user is determined based on the association state prediction result of the home user and the electrocardiogram prediction result, wherein the association state prediction result includes a sleep state prediction result, a blood pressure state prediction result, a blood sugar state prediction result and a pulse state prediction result.

2. The method according to claim 1, characterized in that The processing of the ECG data by the combined prediction model based on the completed model training to obtain the ECG prediction result includes: Obtain ECG sample data and build a combined prediction model consisting of two random forest models, a support vector machine model, and a convolutional neural model; Performing model training on the ECG sample data based on the combined prediction model, and completing the model training of the combined prediction model when the model loss value of the combined prediction model is less than a preset loss threshold; Among them, the input samples and the first output samples in the ECG sample data are used to perform model training on the first random forest model, the support vector machine model and the convolutional neural model, and the second output samples in the ECG sample data and the first output samples are used to perform model training on the second random forest model.

3. The method according to claim 2, characterized in that The processing of the ECG data by the combined prediction model based on the completed model training to obtain the ECG prediction result includes: Using the ECG data as input parameters for the first random forest model, the support vector machine model, and the convolutional neural model to complete model training, to obtain processed output parameters; The output parameters are used as input parameters of the second random forest model to complete model training to obtain the electrocardiogram prediction result.

4. The method according to claim 1, wherein Before determining the monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result, the method further includes: Obtaining sleep data, blood pressure data, blood sugar data, and pulse data of the home user collected by the wearable device; Determine the sleep state prediction result of the sleep data, the blood pressure state prediction result of the blood pressure data, the blood sugar state prediction result, and the pulse state prediction result of the pulse data according to the state prediction strategy; The state prediction strategy includes at least one of a threshold state comparison strategy, a prediction model processing strategy, and an abnormal state identification strategy.

5. The method according to claim 1, wherein The method further comprises: Generate voice broadcast data according to the monitoring result, and send the voice broadcast data to the monitoring terminal of the home user; Receive voice feedback data from the home user or the associated user regarding the voice broadcast data, and process the monitoring result based on the voice feedback data.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining preset companionship task information of the home user, where the preset companionship task information is configured for the home user and / or associated users according to time conditions and voice broadcast conditions; When the time condition or voice broadcast condition in the preset accompanying task information is detected to be triggered, the broadcast content in the preset accompanying task information is retrieved for broadcasting.

7. The method according to any one of claims 1 to 6, characterized in that The method further comprises: When the voice companionship instruction of the home user is received, the companionship execution operation and the companionship execution object corresponding to the voice companionship instruction are parsed, and the companionship execution object is driven to perform the companionship action according to the companionship execution operation.

8. A device for monitoring the health status of a home user, characterized in that: include: The acquisition module is used to obtain the ECG data collected by the wearable device at home; A processing module, configured to process the ECG data based on a combined prediction model that has completed model training to obtain an ECG prediction result, wherein the combined prediction model is constructed based on a random forest model, a support vector machine model, and a convolutional neural model; A monitoring module is used to determine the monitoring result of the home user based on the association status prediction result of the home user and the electrocardiogram prediction result, the association status prediction result including the sleep status prediction result, blood pressure status prediction result, blood sugar status prediction result and pulse status prediction result.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to claim 1 are implemented.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to claim 1.

Images