Elderly nursing scheme generation method based on dynamic data health state evaluation

Abstract

The invention provides an elderly nursing scheme generation method based on dynamic data health state evaluation. The method comprises the steps that vital sign data and life behavior data of a first target object are acquired in real time through acquisition equipment; inputting the vital sign data into a health detection model to obtain a health assessment result; the life behavior data are input into the behavior detection model, a behavior capability evaluation result is obtained, and behavior capability levels of the behavior capability evaluation result comprise self-care, semi-self-care and disability; and inputting the health assessment result and the behavior ability assessment result into a nursing planning model to obtain a target nursing scheme, generating scheduling information according to the target nursing scheme, and sending the scheduling information to a target client. The method solves the problems that in the prior art, the state of the elderly is periodically evaluated, and an evaluation model is based on fixed indexes and weights and is difficult to adapt to fast change of the health state of the elderly, so that the evaluation result is inaccurate, and the provided nursing service effect is poor.

Description

Technical Field

[0001] The present invention relates to the field of health monitoring. Specifically, it relates to a method, device, computer-readable storage medium, and health monitoring system for generating an elderly care plan based on the assessment of health status using dynamic data. Background Art

[0002] In recent years, the elderly population has continued to grow rapidly, showing several significant characteristics: a large elderly population, a rapid aging process, and a significant trend of advanced aging.

[0003] The assessment of the capabilities of the elderly is crucial for accurately providing elderly care services. However, the existing technologies have the following problems:

[0004] 1. Limitations in the design of the assessment model: Traditional assessment models are based on fixed indicators and weights and lack the ability to dynamically adjust, making it impossible to adapt to new data characteristics and changing trends. As time goes by and new data is continuously added, the status and environment of the assessment object may change, and the fixed model may gradually lose its accuracy and effectiveness.

[0005] 2. Disconnection between assessment and service: Many assessment systems only focus on data collection and the generation of assessment results, lacking effective connection with subsequent services. For example, after the assessment is completed, corresponding service suggestions and resource docking are not directly provided for the elderly.

[0006] 3. Lack of dynamics: Traditional assessment systems usually conduct one-time or periodic assessments and it is difficult to track the status changes of the assessment object in real time. For example, in the field of the assessment of the capabilities of the elderly, once the health status and service needs of the assessment object suddenly change, the existing assessment systems cannot respond and adjust in a timely manner.

[0007] In summary, in the existing technology, the periodic assessment of the elderly's status based on a trained model is difficult to accurately predict the health status of the elderly, and the service is disconnected from the assessment result, making it difficult to effectively associate them. Summary of the Invention

[0008] The main objective of the present application is to provide a method, device, computer-readable storage medium, and health monitoring system for generating an elderly care plan based on the assessment of health status using dynamic data, so as to at least solve the problem in the existing technology that the elderly's status is periodically assessed and the assessment model is based on fixed indicators and weights, making it difficult to adapt to the relatively fast changes in the health status of the elderly, resulting in inaccurate assessment results and thus poor effects of the provided care services.

[0009] To achieve the above object, according to one aspect of the present application, a method for generating an elderly care plan based on dynamic data health status assessment is provided, including: obtaining the vital sign data and life behavior data of a first target object in real time through a collection device; inputting the vital sign data into a health detection model to obtain a health assessment result, where the health detection model is trained based on a first classification model and a first regression model through first historical data, and the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold; inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result, where the behavior detection model is trained based on a second classification model and a second regression model through second historical data, and the second historical data is the historical behavior data of the second target object, and the behavior ability level of the behavior ability assessment result includes self-care, semi-self-care, and disability; inputting the health assessment result and the behavior ability assessment result into a care planning model to obtain a target care plan, generating scheduling information according to the target care plan and sending it to the target client, where the care planning model is trained based on a third classification model, a third regression model, and a fourth regression model through third historical data, and the third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical care plan, and user feedback data of the second target object.

[0010] Optionally, inputting the vital sign data into the health detection model to obtain a health assessment result includes: obtaining the vital sign data of the second target object within a first preset period to obtain first historical data, obtaining the vital sign data of the first target object within a first preset period to obtain fourth historical data, and the end time of the first preset period is the current time; predicting the maximum and minimum values of the vital sign data of the first target object at the current time according to the first historical data and the fourth historical data through the first regression model to obtain a first target range; determining the potential disease category of the first target object according to the first target range and the vital sign data through the first classification model, and generating a health assessment result according to the potential disease category.

[0011] Optionally, input the life behavior data into the behavior detection model to obtain the behavior ability assessment, including: determining the behavior ability level of the first target object according to the life behavior data through the second classification model to obtain the first ability level; obtaining the behavior ability level of the second target object within the first preset time period to obtain the fifth historical data, and obtaining the life behavior data of the second target object within the first preset time period to obtain the second historical data. The end time of the first preset time period is the current time; predicting the change trend of the behavior ability level of the first target object within the second preset time period based on time series analysis according to the second historical data, the fifth historical data, and the first ability level through the second regression model to obtain the target sequence. The start time of the second preset time period is the current time; determining the behavior ability level of the first target object at the current time based on the first ability level and the target sequence to obtain the second ability level, and generating a behavior ability assessment result according to the second ability level.

[0012] Optionally, input the health assessment result and the behavior ability assessment result into the care planning model to obtain the target care plan, including: determining the required service category of the first target object according to the health assessment result and the behavior ability assessment result through the third classification model to obtain the first target category. The required service category includes meal assistance and medical assistance; obtaining the target care plan and user feedback data of the first target object within the first preset time period to obtain the sixth historical data; predicting the required service category of the first target object at the current time according to the sixth historical data through the third regression model to obtain the second target category; predicting the required service frequency and required service intensity of the first target object according to the first target category, the second target category, and the user feedback data in the sixth historical data through the fourth regression model; generating the target care plan according to the first target category, the second target category, the required service frequency, and the required service intensity.

[0013] Optionally, obtain the vital sign data and life behavior data of the target object in real time through the acquisition device, including: collecting the physiological indicators of the first target object through the wearable device to obtain the first vital sign data; collecting the activity trajectory, indoor environment parameters, and life behavior of the first target object through the smart home device, and performing a quantitative score to obtain the first behavior data; obtaining the life habit data of the first target object through the first preset path to obtain the second behavior data; obtaining the case data and physical examination data of the first target object through the second preset path to obtain the second vital sign data; integrating and data cleaning the first vital sign data and the second vital sign data to obtain the vital sign data; integrating and data cleaning the first behavior data and the second behavior data to obtain the life behavior data.

[0014] Optionally, the living behaviors of the first target object are collected by a smart home device and quantitatively scored, including: determining the chewing ability score, swallowing ability score, and autonomous feeding coordination score of the first target object based on the video data of the first target object in the performing state collected by the smart home device; determining the balance ability score, limb flexibility score, muscle endurance score, and flexible coordination score of the first target object based on the video data of the first target object in the moving state collected by the smart home device; determining the digital cognition score, time-space cognition score, and social understanding score of the first target object based on the video data of the first target object in the social state collected by the smart home device.

[0015] Optionally, before inputting the vital sign data into the health detection model to obtain the health assessment result, the method further includes: obtaining the vital sign data of the second target object within a third preset period to obtain the seventh historical data, obtaining the vital sign data of the second target object within a fourth preset period to obtain the eighth historical data, and the end moment of the third preset period is the start moment of the fourth preset period; constructing a first target data set according to the seventh historical data and the eighth historical data, and training an alternative regression model according to the first target data set through a cross-validation mechanism to obtain a first regression model, wherein the cross-validation mechanism calculates the loss function of the alternative regression model according to the test set in one or more ways including the sum of squared errors, mean squared error, and mean absolute error; processing the seventh historical data according to the first regression model to obtain a plurality of second target ranges, and determining the potential disease category of the first target object according to the eighth historical data; constructing a second target data set according to the second target range, the seventh historical data, and the potential disease category, and training an alternative classification model according to the second target data set through a cross-validation mechanism to obtain a first classification model, wherein the cross-validation mechanism calculates the loss function of the alternative classification model according to the test set in one or more ways including precision, recall, and F1 value.

[0016] According to another aspect of the present application, there is provided an apparatus for generating an elderly care plan based on dynamic data health status assessment. The apparatus includes: a first acquisition unit for real-time acquiring the vital sign data and life behavior data of a first target object through an acquisition device; a first input unit for inputting the vital sign data into a health detection model to obtain a health assessment result. The health detection model is trained based on a first classification model and a first regression model through first historical data, and the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold; a second input unit for inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained based on a second classification model and a second regression model through second historical data, and the second historical data is the historical behavior data of the second target object. The behavior ability level of the behavior ability assessment result includes self-care, semi-self-care, and disability; a third input unit for inputting the health assessment result and the behavior ability assessment result into a nursing plan model to obtain a target nursing plan, generating scheduling information according to the target nursing plan and sending it to a target client. The nursing plan model is trained based on a third classification model, a third regression model, and a fourth regression model through third historical data, and the third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object.

[0017] According to still another aspect of the present application, there is provided a computer-readable storage medium. The computer-readable storage medium includes a stored program, wherein when the program runs, it controls the device where the computer-readable storage medium is located to execute any one of the methods.

[0018] According to yet another aspect of the present application, there is provided a health monitoring system, including: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include those for executing any one of the methods.

[0019] Applying the technical solution of the present application in the above-mentioned method for generating an elderly care plan based on dynamic data health status assessment, first, the vital sign data and life behavior data of the first target object are obtained in real time through a collection device; then, the vital sign data is input into a health detection model to obtain a health assessment result. The health detection model is trained based on a first classification model and a first regression model using first historical data, and the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold; thereafter, the life behavior data is input into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained based on a second classification model and a second regression model using second historical data, and the second historical data is the historical behavior data of the second target object. Among them, the behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability; finally, the health assessment result and the behavior ability assessment result are input into a care planning model to obtain a target care plan. Scheduling information is generated according to the target care plan and sent to the target client. The care planning model is trained based on a third classification model, a third regression model, and a fourth regression model using third historical data, and the third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical care plan, and user feedback data of the second target object. The health detection model, behavior detection model, and care planning model of the present application all combine a classification model and a regression model. Among them, the regression model is introduced to predict the change trend of state indicators based on historical data, while the classification model is used to determine the current state. Through the classification model and the regression model, it is not only possible to timely judge the current abnormal state of the elderly but also predict future risks, and then personalize the care plan for the elderly based on the current state and future trends. Moreover, the prediction based on the regression model can ensure that the criteria for abnormal determination change in real time with the change of the state. This method solves the problem in the prior art that the elderly's state is evaluated periodically and the evaluation model is based on fixed indicators and weights, which is difficult to adapt to the rapid change of the elderly's health state, resulting in inaccurate evaluation results and poor care service effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 FIG. shows a hardware structure block diagram of a mobile terminal for a method of generating an elderly care plan based on dynamic data health status assessment provided in an embodiment of the present application;

[0021] Figure 2 FIG. shows a flowchart of a method of generating an elderly care plan based on dynamic data health status assessment provided in an embodiment of the present application;

[0022] Figure 3 FIG. shows a structure block diagram of an apparatus for generating an elderly care plan based on dynamic data health status assessment provided in an embodiment of the present application.

[0023] Among them, the above-mentioned drawings include the following reference numerals:

[0024] 102, processor; 104, memory; 106, transmission device; 108, input / output device. Detailed implementation manners

[0025] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The following will describe the present application in detail with reference to the drawings and in combination with the embodiments.

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification, claims and drawings of the present application are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances for the embodiments of the present application described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily need to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0028] As introduced in the background art, it is difficult to accurately predict the health status of the elderly by periodically evaluating the status of the elderly according to the trained model in the prior art, and the service is separated from the evaluation result and it is difficult to effectively associate them. To solve the problem that in the prior art, the status of the elderly is evaluated periodically and the evaluation model is based on fixed indicators and weights, which is difficult to adapt to the relatively fast change of the health status of the elderly, resulting in inaccurate evaluation results and thus poor nursing service effects, the embodiments of the present application provide a method, device, computer-readable storage medium and health monitoring system for generating an elderly care plan based on dynamic data health status evaluation.

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention.

[0030] The method embodiments provided in the embodiments of the present application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 is a hardware structure block diagram of a mobile terminal for a method of generating an elderly care plan based on dynamic data health status assessment according to an embodiment of the present invention. As Figure 1 shown, the mobile terminal may include one or more ( Figure 1 only one is shown in Figure 1 ) processors 102 (the processors 102 may include, but are not limited to, processing devices such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Among them, the above-mentioned mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown in Figure 1 is only schematic and does not limit the structure of the above-mentioned mobile terminal. For example, the mobile terminal may further include more or fewer components than

[0031] shown in

[0032] Figure 1 shown, or have a different configuration from

[0031] shown.

[0031] The memory 104 can be used to store computer programs. For example, software programs and modules of application software, such as the computer program corresponding to the method of generating an elderly care plan based on dynamic data health status assessment in the embodiments of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implements the above-mentioned method. The memory 104 may include a high-speed random access memory, and may further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely set relative to the processor 102, and these remote memories may be connected to the mobile terminal through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and their combinations. The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by a communication provider of the mobile terminal. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency (Radio Frequency, abbreviated as RF) module, which is used to communicate with the Internet wirelessly.

[0032] In this embodiment, a method for generating an elderly care plan based on the assessment of the health status of dynamic data running on a mobile terminal, a computer terminal, or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0033] Figure 2 is a flowchart of a method for generating an elderly care plan based on the assessment of the health status of dynamic data according to an embodiment of the present application. As Figure 2 shown, the method includes the following steps:

[0034] Step S201, obtaining the vital sign data and life behavior data of the first target object in real time through a collection device;

[0035] Specifically, the vital sign data (such as heart rate, blood pressure, sleep quality, number of activity steps) and life behavior data (such as daily living time, eating habits, activity range) of the elderly are collected in real time through non-contact sensor technologies (such as millimeter wave radar, environmental sensors, intelligent cameras, etc.) and wearable devices (such as smart bracelets, smart mattresses, etc.). At the same time, medical data from health care institutions and community health service centers, as well as the behavior data of the elderly and their families on the APP, are also integrated to form a comprehensive dataset of the health and life status of the elderly.

[0036] Step S202, inputting the vital sign data into a health detection model to obtain a health assessment result. The health detection model is trained through first historical data based on a first classification model and a first regression model. The first historical data is the historical sign data of a second target object whose similarity to the first target object is greater than a first threshold;

[0037] Specifically, the health detection model is trained using a supervised learning method. In the health detection model, a first classification model (such as a random forest classifier) and a first regression model (such as a support vector machine regression) are used to train through the first historical data (that is, the sign data and health assessment results of the second target object with similar characteristics to the first target object) to predict the health assessment result, and the above-mentioned health assessment result is generated by combining the real-time collected vital sign data through the trained health detection model.

[0038] Step S203, inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained through second historical data based on a second classification model and a second regression model. The second historical data is the historical behavior data of the second target object. Among them, the behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability;

[0039] Specifically, a supervised learning method is used to train a behavior detection model. In the behavior detection model, it is trained based on a second classification model (such as logistic regression classification) and a second regression model (such as linear regression) through second historical data (i.e., the behavior data and behavior ability evaluation results of the second target object), and is used to evaluate the behavior ability level. And the above-mentioned behavior evaluation result is generated through the trained behavior detection model combined with the above-mentioned life behavior data collected in real time.

[0040] In step S204, the health evaluation result and the behavior ability evaluation result are input into a nursing plan model to obtain a target nursing plan, and scheduling information is generated according to the target nursing plan and sent to the target client. The nursing plan model is obtained by training based on a third classification model, a third regression model, and a fourth regression model through third historical data. The third historical data is obtained by training the historical health evaluation results, historical behavior ability evaluation results, historical nursing plans, and user feedback data of the second target object.

[0041] Specifically, a supervised learning method is used to train the nursing plan model. The nursing plan model further integrates the output results of the above-mentioned health detection model and behavior detection model, combines the third historical data (including historical health evaluation results, behavior ability evaluation results, nursing plans, and user feedback), and trains a model based on a third classification model (such as K-nearest neighbor classification), a third regression model (such as ridge regression), and a fourth regression model (such as gradient boosting regression) to generate a target nursing plan. And the above-mentioned target nursing plan is generated through the trained nursing plan model combined with the above-mentioned health evaluation result and the above-mentioned behavior evaluation result output in real time. Furthermore, the resources to be scheduled (such as nursing staff, etc.) are determined according to the target nursing plan, and the corresponding scheduling instructions are generated and sent to the target client to complete the scheduling.

[0042] It can be understood that the solution of this application realizes the dynamic assessment of the health status of the elderly and the generation of personalized care plans by real-time monitoring and analyzing the vital sign data and life behavior data of the elderly, combined with machine learning technology. The vital sign data includes but is not limited to heart rate, blood pressure, blood oxygen saturation, etc., which can reflect the physiological health status of the elderly; the life behavior data covers the daily activities, eating habits, social interactions, etc. of the elderly and is an important basis for evaluating the self-care ability of the elderly. Through the combination of the first classification model and the first regression model, the health detection model can accurately predict the health risks of the elderly, such as potential diseases like heart disease and diabetes, which is achieved by analyzing the trend of physical sign changes of similar elderly people in historical data, thus solving the problem of inaccurate prediction of health risks in traditional care plans. The behavior detection model evaluates the behavior ability level of the elderly through the second classification model and the second regression model, providing key information for the formulation of care plans and ensuring the pertinence and effectiveness of care measures. The care planning model comprehensively considers the health assessment results and the behavior ability assessment results. The generated care plan not only includes service categories such as meal assistance and medical assistance, but also predicts the frequency and intensity of services, which is achieved through the deep learning of the third historical data by the third classification model, the third regression model and the fourth regression model, ensuring the personalization and dynamic adjustment ability of the care plan and effectively improving the accuracy and efficiency of elderly care.

[0043] Through this embodiment, first, a collection device is used to obtain the vital sign data and daily behavior data of the first target object in real time; then, the vital sign data is input into a health detection model to obtain a health assessment result. The health detection model is trained based on a first classification model and a first regression model using first historical data, where the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold; thereafter, the daily behavior data is input into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained based on a second classification model and a second regression model using second historical data, where the second historical data is the historical behavior data of the second target object. The behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability; finally, the health assessment result and the behavior ability assessment result are input into a nursing plan model to obtain a target nursing plan. Scheduling information is generated according to the target nursing plan and sent to the target client. The nursing plan model is trained based on a third classification model, a third regression model, and a fourth regression model using third historical data, where the third historical data is obtained by training using the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object. The health detection model, behavior detection model, and nursing plan model of this application all combine a classification model and a regression model. The regression model is introduced to predict the change trend of the state index based on historical data, while the classification model is used to determine the current state. Through the classification model and the regression model, not only can the current abnormal state of the elderly be judged in a timely manner, but also the future risks can be predicted. Furthermore, based on the current state and future trends, a personalized nursing plan for the elderly can be planned, and the prediction based on the regression model can also ensure that the criteria for abnormal determination change in real time with the change of the state. This method solves the problem in the prior art that the elderly's state is evaluated periodically and the evaluation model is based on fixed indicators and weights, making it difficult to adapt to the relatively fast change of the elderly's health state, resulting in inaccurate evaluation results and thus poor nursing service effects.

[0044] In an alternative embodiment, in order to generate the above-mentioned health assessment result, step S202 includes:

[0045] Step S2021, obtain the vital sign data of the second target object within a first preset time period to obtain first historical data, and obtain the vital sign data of the first target object within the first preset time period to obtain fourth historical data. The end time of the first preset time period is the current time;

[0046] Specifically, the system continuously obtains the vital sign data of the first target object from wearable devices and non-contact sensors, including but not limited to heart rate, blood pressure, sleep quality, activity level, etc. At the same time, the historical vital sign data of the second target object with a similar health condition and life background to the first target object within a specific period (the first preset period) is collected as the first historical data.

[0047] It can be understood that the first preset period is determined according to medical expert suggestions and data statistical rules. For example, it can be set to the past month to cover the fluctuations of vital sign data in a month. By selecting the latest first preset period data, the timeliness and accuracy of model prediction are ensured.

[0048] Step S2022, predict the maximum and minimum values of the vital sign data of the first target object at the current moment according to the first historical data and the fourth historical data through the first regression model, and obtain the first target range;

[0049] Specifically, the model may adopt algorithms such as support vector machine regression (SVR), linear regression, neural network regression, etc., and optimize the prediction results by adjusting the model parameters. In specific implementation, the regression model takes the fourth historical data (i.e., the vital sign data of the first target object within the first preset period) as the input, predicts the possible extreme values of the vital sign data at the current moment, and forms a predicted range for the subsequent classification model to judge whether an abnormality occurs.

[0050] Step S2023, determine the potential disease category of the first target object according to the first target range and the vital sign data through the first classification model, and generate a health assessment result according to the potential disease category.

[0051] Specifically, based on the first target range and the real-time vital sign data of the first target object, use the first classification model to identify whether there are abnormal values outside the normal range, and then judge the potential disease category and generate a health assessment result. The classification model may adopt random forest, support vector machine classification (SVC), logistic regression or deep learning algorithms such as convolutional neural network (CNN) or recurrent neural network (RNN), and is trained through a large amount of historical health assessment data to identify patterns and abnormalities in the vital sign data. In application, the classification model compares the real-time vital sign data with the predicted first target range. Since the data has multiple categories, it can then analyze according to the data type outside the normal range to determine the potential disease category or risk category, and then generate the corresponding assessment result.

[0052] Through the above embodiments, through the prediction of the first regression model and the analysis of the first classification model, the system can monitor the vital sign data of the elderly in real time. Once abnormal data is found, it can immediately generate a health assessment result and issue an alarm, effectively preventing and managing diseases and reducing health risks. Specifically, the vital sign data range at the current moment is predicted through the first regression model. This is based on time series analysis, considering the continuity and trend of the data, and can more accurately reflect the current health status of the elderly. The first classification model then identifies the health risks that the elderly may face, such as heart disease, hypertension, etc., according to the predicted range and the actual vital sign data. This is achieved by analyzing the correlation between the physical sign changes and diseases of similar elderly people in historical data, solving the problems of lag and inaccuracy in disease prediction in traditional health assessments.

[0053] In an alternative embodiment, in order to generate the above-mentioned ability assessment result, step S203 includes:

[0054] Step S2031, determine the behavior ability level of the first target object according to the life behavior data through the second classification model to obtain the first ability level;

[0055] Specifically, use the second classification model (such as random forest, support vector machine classification SVC or logistic regression) to train a classification model that can identify specific behavior patterns through the behavior ability level data (the fifth historical data) and life behavior data (the second historical data) of the second target object (other elderly people with similar life backgrounds) within the first preset time period. This model can output the behavior ability level (the first ability level) of the elderly at the current moment based on real-time life behavior data. For example, the model may identify that when the daily activity level of the elderly significantly decreases and the living time becomes irregular, their behavior ability level may decline from self-care to semi-self-care.

[0056] Step S2032, obtain the behavior ability level of the second target object within the first preset time period to obtain the fifth historical data, and obtain the life behavior data of the second target object within the first preset time period to obtain the second historical data. The end time of the first preset time period is the current moment;

[0057] Step S2033, predict the change trend of the behavior ability level of the first target object within the second preset time period based on time series analysis according to the second historical data, the fifth historical data, and the first ability level to obtain the target sequence. The start time of the second preset time period is the current moment;

[0058] Specifically, based on time series analysis techniques (such as ARIMA, LSTM, etc.), the second regression model is trained using the second historical data and the fifth historical data to predict the change trend of the elderly's behavior ability level within the second preset time period, obtaining a target sequence. This prediction process takes into account the time dependence in the historical data, which helps to identify the long-term change trend of the behavior ability level, rather than just the current state.

[0059] Step S2034, based on the first ability level and the target sequence, determine the behavior ability level of the first target object at the current moment, obtain the second ability level, and generate a behavior ability evaluation result according to the second ability level.

[0060] Specifically, by combining the current behavior ability level (the first ability level) and the future change trend (the target sequence), the system determines the behavior ability level (the second ability level) of the elderly at the current moment, and generates a behavior ability evaluation result based on this. The evaluation result will include the current behavior ability level of the elderly.

[0061] Through the above embodiments, the generation of the behavior ability evaluation result is achieved by analyzing the elderly's life behavior data and predicting the change trend of their self-care ability. By combining the real-time life behavior data and the classification model of historical data, the system can provide a more accurate and personalized behavior ability evaluation, helping service personnel and family members accurately judge the current self-care level of the elderly and laying a foundation for providing appropriate services. Specifically, the second classification model first quantifies and determines the current behavior ability level of the elderly according to their activity trajectories, indoor environment parameters, and life behaviors, such as diet, exercise, social interaction, etc. Subsequently, based on time series analysis, the second regression model predicts the change of the elderly's behavior ability level in the future by analyzing the relationship between the behavior patterns of similar elderly people in historical data and their self-care ability, solving the instability problem of predicting the elderly's self-care ability in the care plan.

[0062] In order to generate the above-mentioned target care plan, in an optional implementation manner, the above step S204 includes:

[0063] Step S2041, through the third classification model, determine the required service category of the first target object according to the health assessment result and the behavior ability evaluation result, obtain the first target category, and the required service category includes meal assistance and medical assistance;

[0064] Specifically, a third classification model (such as K-nearest neighbor classification, neural network classification, etc.) is trained to determine the current required service category (the first target category) of the elderly according to the health assessment results (such as disease states like hypertension, arrhythmia, etc.) and the behavioral ability assessment results (self-care, semi-self-care, disability levels), such as meal assistance (catering service), medical assistance (medical and health service), etc. This classification process takes into account the comprehensive impact of the health and behavioral ability of the elderly to ensure that the selected service category matches the actual needs of the elderly.

[0065] Step S2042: Obtain the target care plan and user feedback data of the first target object in the first preset time period to obtain the sixth historical data.

[0066] Specifically, the system obtains the implementation situation of the target care plan and user feedback data of the first target object in the first preset time period from the database as the sixth historical data for subsequent prediction and plan optimization. These data include the actual implementation situation of the service, the satisfaction evaluations of the elderly and their families, the feedback of the service personnel, etc., providing a rich information source for model prediction.

[0067] Step S2043: Predict the required service category of the first target object at the current moment according to the sixth historical data through the third regression model to obtain the second target category.

[0068] Specifically, based on the sixth historical data, use the third regression model (such as ridge regression, gradient boosting regression, etc.) to predict the required service category (the second target category) of the elderly at the current moment. This prediction takes into account the health change trend of the elderly, the change of the behavioral ability level, and the historical service preferences to adapt to the dynamics of the elderly's needs.

[0069] Step S2044: Predict the required service frequency and required service intensity of the first target object according to the first target category, the second target category, and the user feedback data in the sixth historical data through the fourth regression model.

[0070] Specifically, through the fourth regression model (such as XGBoost, LightGBM, etc.), combine the first target category, the second target category, and the user feedback data to predict the required service frequency and required service intensity of the elderly for various services. For example, if it is predicted that the elderly need to increase meal assistance services, the model will simultaneously predict the number of meal assistance services per month or per week, and the intensity of each service (such as whether special diet is required, whether feeding service is required, etc.).

[0071] Step S2045: Generate a target care plan according to the first target category, the second target category, the required service frequency, and the required service intensity.

[0072] Specifically, the system takes the first target category, the second target category, the frequency of required services, and the intensity of required services as inputs to generate a personalized target care plan. This plan will detail various services that the elderly need in the next period of time, including service categories, frequencies, intensities, etc., providing clear service guidance for elderly care service institutions and family members.

[0073] Through the above embodiments, by deeply analyzing the health and behavioral ability assessment results of the elderly, combining historical service data and user feedback, the system can generate a more accurate and personalized care plan to ensure that the elderly receive the most suitable services. Specifically, the generation of the target care plan comprehensively considers the health assessment results and behavioral ability assessment results of the elderly, ensuring the comprehensiveness and personalization of care measures. The third classification model first determines the current required service categories of the elderly, such as meal assistance, medical assistance, bathing assistance, etc., based on the assessment results. This is achieved by analyzing the correlation between the service needs of similar elderly people in historical data and their health and behavioral abilities, solving the problem of blindness in the selection of service categories in the care plan. The third regression model and the fourth regression model then predict the current service demand frequency and intensity of the elderly based on historical care plans and user feedback data. This is achieved through deep learning technology, considering the dynamic changes in service demand and user satisfaction, ensuring the timely adjustment and optimization of the care plan.

[0074] In order to obtain the above-mentioned vital sign data and the above-mentioned life behavior data, in an optional implementation manner, the above-mentioned step S201 includes:

[0075] Step S2011, collecting the physiological indicators of the first target object through a wearable device to obtain the first vital sign data;

[0076] Specifically, through wearable devices such as smart bracelets, smart pillows, and smart mattresses, the physiological indicators of the elderly, such as heart rate, blood pressure, sleep quality, and number of exercise steps, are monitored and collected in real time to form the first vital sign data.

[0077] Step S2012, collecting the activity trajectory, indoor environment parameters, and life behaviors of the first target object through a smart home device, and performing a quantitative score to obtain the first behavior data;

[0078] Specifically, smart home devices (such as smart cameras, environmental sensors, smart door locks, etc.) are used to collect the activity trajectory of the elderly, indoor environment parameters (such as temperature, humidity, etc.), and life behaviors (such as turning on and off lights, using electrical appliances, etc.). These data are quantitatively scored to intuitively reflect the activity ability and living habits of the elderly. For example, the smart camera analyzes the activities of the elderly in the kitchen through a behavior recognition algorithm (such as a deep learning model) to evaluate their eating ability and safety risks.

[0079] In step S2013, obtain the living habit data of the first target object through the first preset path to obtain the second behavior data;

[0080] Specifically, the system obtains the living habit data of the elderly, including daily routines, dietary preferences, etc., through the first preset path (such as the data interface with the community service center) as the second behavior data.

[0081] In step S2014, obtain the case data and physical examination data of the first target object through the second preset path to obtain the second vital sign data;

[0082] Similarly, obtain the case data and physical examination data of the elderly through the second preset path (such as the data interface with the hospital) as the second vital sign data to comprehensively understand the health status of the elderly.

[0083] In step S2015, integrate and clean the first vital sign data and the second vital sign data to obtain the vital sign data;

[0084] In step S2017, integrate and clean the first behavior data and the second behavior data to obtain the living behavior data.

[0085] Specifically, the system integrates the first vital sign data and the second vital sign data, as well as the first behavior data and the second behavior data, to form a comprehensive vital sign data and living behavior data set. During the data integration process, data cleaning is performed, including removing noise, filling in missing values, and correcting outliers to ensure the accuracy and usability of the data.

[0086] By collecting vital sign data in real time, the system can immediately detect health abnormalities, such as abnormal heart rate or sudden increase in blood pressure, and immediately notify the family members and medical service providers to achieve early warning and intervention and reduce health risks. Specifically, by using a variety of collection devices to obtain the vital sign data and living behavior data of the elderly in real time, the comprehensiveness and accuracy of the data are ensured. Wearable devices, such as smart watches and health monitoring bracelets, can continuously monitor physiological indicators of the elderly, such as heart rate, blood pressure, and steps. Smart home devices use technologies such as video monitoring and sensors to record the activity trajectories, environmental parameters, and living behaviors of the elderly, such as diet, exercise, and sleep, and perform quantitative scoring, which is achieved by analyzing video data through deep learning technology and can more accurately evaluate the self-care ability and quality of life of the elderly. The living habit data obtained through the first preset path and the case data and physical examination data obtained through the second preset path further enrich the data dimensions and provide more comprehensive information for health and behavioral ability assessment. The data integration and cleaning process ensure the quality of the data and provide a reliable basis for subsequent model training and prediction.

[0087] In an alternative implementation, in order to quantitatively score the living behaviors of the elderly, the above step S2012 includes:

[0088] Step S20121: Based on the video data of the first target object in the performing state collected by the smart home device, determine the chewing ability score, swallowing ability score, and autonomous eating coordination score of the first target object;

[0089] Specifically, the smart camera in the smart home device continuously monitors the living environment of the elderly and records the video data in different states. These video data first need to be preprocessed, including video segmentation, frame extraction, and data augmentation, to adapt to subsequent deep learning model analysis. Among them, the system analyzes the actions of the elderly during the eating process in the video, such as chewing frequency, swallowing interval, and proficiency in using tableware, and combines deep learning models (such as convolutional neural network CNN) to identify the corresponding behavior patterns and quantitatively score. For example, if the system detects that the elderly have a lower chewing frequency and a longer swallowing interval when chewing food, it may indicate a decline in their chewing or swallowing ability.

[0090] Step S20122: Based on the video data of the first target object in the motion state collected by the smart home device, determine the balance ability score, limb flexibility score, muscle endurance score, and flexible coordination score of the first target object;

[0091] Specifically, based on the video data of the elderly in the motion state, such as walking, standing, sitting down, etc., the system uses video analysis techniques (such as skeletal key point detection, motion trajectory analysis) and deep learning models to evaluate their balance ability, limb flexibility, and muscle endurance, etc., and perform quantitative scoring. For example, by analyzing the pace stability of the elderly when walking, their balance ability can be evaluated; by analyzing the limb amplitude when they perform simple stretching exercises, their limb flexibility can be evaluated.

[0092] Step S20123: Based on the video data of the first target object in the social state collected by the smart home device, determine the digital cognition score, time-space cognition score, and social understanding score of the first target object.

[0093] Specifically, by monitoring the behaviors of the elderly in social scenarios through the smart home device, such as conversations with family or friends, handling daily bills, etc., combined with natural language processing technology (NLP) and deep learning models, evaluate their digital cognition ability (such as understanding and calculating numbers), time-space cognition (such as perception of time and positioning in space), and social understanding ability (such as understanding the intentions and rules of social scenarios). For example, by analyzing the reaction speed and accuracy of the elderly when handling bills, their digital cognition ability can be evaluated; by analyzing their expressions and tones when talking with others, their social understanding ability can be evaluated.

[0094] Through the above embodiments, quantifying the scores of life behaviors based on the video data collected by smart home devices is a key link in this technical solution. The chewing ability, swallowing ability, and self-feeding coordination scores reflect the elderly's ability to take care of their own diet, while the balance ability, limb flexibility, and muscle endurance scores reflect the elderly's motor ability and physical function. The digital cognition, time-space cognition, and social understanding scores evaluate the elderly's cognitive ability and social interaction ability. Analyzing the video data through deep learning technology with the scores introducing these detailed data can more comprehensively and accurately reflect the elderly's life behavior ability.

[0095] In order to train the above health detection model, in an optional implementation manner, before inputting the vital sign data into the health detection model to obtain a health assessment result, the above method further includes:

[0096] Step S301: Obtain the vital sign data of the second target object within the third preset time period to obtain the seventh historical data, and obtain the vital sign data of the second target object within the fourth preset time period to obtain the eighth historical data. The end moment of the third preset time period is the start moment of the fourth preset time period;

[0097] Specifically, before inputting the current vital sign data into the health detection model, the system first collects the vital sign data of the second target object (other elderly people with similar health backgrounds) within the third and fourth preset time periods as the seventh historical data (data within the third preset time period) and the eighth historical data (data within the fourth preset time period). Here, the end moment of the third preset time period is exactly the start moment of the fourth preset time period, ensuring the continuity and timeliness of the historical data.

[0098] Step S302: Construct a first target data set based on the seventh historical data and the eighth historical data, and train an alternative regression model according to the first target data set through a cross-validation mechanism to obtain a first regression model. Among them, the ways for the cross-validation mechanism to calculate the loss function of the alternative regression model based on the test set include one or more of the sum of squared errors, mean squared error, and mean absolute error;

[0099] Specifically, construct a first target data set based on the seventh and eighth historical data for training alternative regression models (such as linear regression, support vector regression SVR, etc.). During the training process, use the cross-validation mechanism to evaluate the fitting degree and prediction ability of the model by calculating indicators such as the sum of squared errors (SSE), mean squared error (MSE), and mean absolute error (MAE), and select the model with the highest prediction accuracy as the first regression model for identifying the normal range and abnormal change trends of vital sign data.

[0100] Step S303: Process the seventh historical data according to the first regression model to obtain multiple second target ranges, and determine the potential disease categories of the first target object according to the eighth historical data.

[0101] Specifically, analyze and process the seventh historical data through the first regression model to obtain multiple second target ranges, that is, the prediction intervals of vital sign data in the normal state. Subsequently, based on the data points in the eighth historical data that exceed these prediction intervals, the system can identify the potential disease categories of the second target object. For example, if the blood pressure data of an elderly person frequently exceeds the normal range, the system may predict that they are at risk of hypertension.

[0102] Step S304: Construct a second target data set according to the second target range, the seventh historical data, and the potential disease categories, and train an alternative classification model through a cross-validation mechanism according to the second target data set to obtain a first classification model, where the cross-validation mechanism calculates the loss function of the alternative classification model according to the test set in one or more ways including precision, recall rate, and F1 value.

[0103] Specifically, based on the second target range, the seventh historical data, and the potential disease categories, the system constructs a second target data set for training alternative classification models (such as decision trees, random forests, neural network classification, etc.). Also using the cross-validation mechanism, evaluate the classification performance of the model by calculating indicators such as precision, recall rate, and F1 value, and finally select the model with the best classification effect as the first classification model to predict the potential disease categories for the first target object (the currently evaluated elderly person).

[0104] Through the above embodiments, the training and optimization of the health detection model are realized through deep learning technology combined with the cross-validation mechanism. The training processes of the alternative regression model and the classification model consider the continuity and change trend of the data, and can more accurately predict the range of the elderly's vital sign data and potential disease categories. The cross-validation mechanism evaluates the prediction accuracy of the model by calculating loss functions such as the sum of squared errors, mean squared error, and mean absolute error, and evaluates the classification ability of the model through indicators such as precision, recall rate, and F1 value. This is achieved by splitting the historical data set into a training set and a test set, training and testing the model repeatedly, ensuring the generalization ability and prediction accuracy of the model.

[0105] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0106] The embodiments of the present application further provide an apparatus for generating an elderly care plan based on the assessment of the health status of dynamic data. It should be noted that the apparatus for generating an elderly care plan based on the assessment of the health status of dynamic data in the embodiments of the present application can be used to execute the method for generating an elderly care plan based on the assessment of the health status of dynamic data provided in the embodiments of the present application. The apparatus for realizing the above embodiments and preferred embodiments has been described and will not be repeated here. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the apparatuses described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0107] The following introduces the apparatus for generating an elderly care plan based on the assessment of the health status of dynamic data provided in the embodiments of the present application.

[0108] Figure 3 is a structural block diagram of an apparatus for generating an elderly care plan based on the assessment of the health status of dynamic data according to an embodiment of the present application. As Figure 3 shown, the apparatus includes:

[0109] A first acquisition unit 10, configured to obtain the vital sign data and life behavior data of a first target object in real time through a collection device;

[0110] Specifically, the vital sign data (such as heart rate, blood pressure, sleep quality, number of steps) and life behavior data (such as daily living time, eating habits, activity range) of the elderly are collected in real time through non-contact sensor technologies (such as millimeter wave radar, environmental sensors, intelligent cameras, etc.) and wearable devices (such as smart bracelets, smart mattresses, etc.). At the same time, medical data from health care institutions and community health service centers, as well as the behavior data of the elderly and their families on the APP, are integrated to form a comprehensive dataset of the health and life status of the elderly.

[0111] A first input unit 20, configured to input the vital sign data into a health detection model to obtain a health assessment result. The health detection model is trained through first historical data based on a first classification model and a first regression model. The first historical data is the historical sign data of a second target object whose similarity to the first target object is greater than a first threshold;

[0112] Specifically, the health detection model is trained using a supervised learning method. In the health detection model, a first classification model (such as a random forest classifier) and a first regression model (such as a support vector machine regression) are used to train through the first historical data (that is, the sign data and health assessment results of a second target object with similar characteristics to the first target object) to predict the health assessment result, and the above-mentioned health assessment result is generated by combining the real-time collected vital sign data through the trained health detection model.

[0113] A second input unit 30 for inputting the daily life behavior data into the behavior detection model to obtain a behavior ability evaluation result. The behavior detection model is trained by second historical data based on a second classification model and a second regression model. The second historical data is the historical behavior data of a second target object. Among them, the behavior ability levels of the behavior ability evaluation result include self-care, semi-self-care, and disability.

[0114] Specifically, the behavior detection model is trained using a supervised learning method. In the behavior detection model, it is trained based on a second classification model (such as logistic regression classification) and a second regression model (such as linear regression) through second historical data (i.e., the behavior data and behavior ability evaluation result of the second target object), which is used to evaluate the behavior ability level, and the above-mentioned behavior evaluation result is generated by combining the trained behavior detection model with the real-time collected daily life behavior data.

[0115] A third input unit 40 for inputting the health evaluation result and the behavior ability evaluation result into the nursing plan model to obtain a target nursing plan, and generating scheduling information according to the target nursing plan and sending it to the target client. The nursing plan model is trained by third historical data based on a third classification model, a third regression model, and a fourth regression model. The third historical data is obtained by training the historical health evaluation result, historical behavior ability evaluation result, historical nursing plan, and user feedback data of the second target object.

[0116] Specifically, the nursing plan model is trained using a supervised learning method. The nursing plan model further integrates the output results of the above-mentioned health detection model and behavior detection model, and combines the third historical data (including historical health evaluation result, behavior ability evaluation result, nursing plan, and user feedback) to train a model based on a third classification model (such as K-nearest neighbor classification), a third regression model (such as ridge regression), and a fourth regression model (such as gradient boosting regression) to generate a target nursing plan, and generates the above-mentioned target nursing plan by combining the trained nursing plan model with the real-time output of the above-mentioned health evaluation result and the above-mentioned behavior evaluation result. Furthermore, the resources to be scheduled (such as nursing staff, etc.) are determined according to the target nursing plan, and the corresponding scheduling instructions are generated and sent to the target client to complete the scheduling.

[0117] It is understandable that the solution of the present application realizes the dynamic assessment of the health status of the elderly and the generation of personalized care plans by monitoring and analyzing the vital sign data and life behavior data of the elderly in real time and combining machine learning technology. The vital sign data includes but is not limited to heart rate, blood pressure, blood oxygen saturation, etc., which can reflect the physiological health status of the elderly; the life behavior data covers the daily activities, eating habits, social interactions, etc. of the elderly and is an important basis for evaluating the self-care ability of the elderly. Through the combination of the first classification model and the first regression model, the health detection model can accurately predict the health risks of the elderly, such as potential diseases like heart disease and diabetes. This is achieved by analyzing the trend of physical sign changes of similar elderly people in historical data, thus solving the problem of inaccurate prediction of health risks in traditional care plans. The behavior detection model evaluates the behavior ability level of the elderly through the second classification model and the second regression model, providing key information for the formulation of care plans and ensuring the pertinence and effectiveness of care measures. The care planning model comprehensively considers the health assessment results and the behavior ability assessment results. The generated care plan not only includes service categories such as meal assistance and medical assistance, but also predicts the frequency and intensity of services. This is achieved through the deep learning of the third historical data by the third classification model, the third regression model and the fourth regression model, ensuring the personalization and dynamic adjustment ability of the care plan and effectively improving the accuracy and efficiency of elderly care.

[0118] Through this embodiment, the first acquisition unit acquires the vital sign data and life behavior data of the first target object in real time through the acquisition device; the first input unit inputs the vital sign data into the health detection model to obtain a health assessment result. The health detection model is trained based on the first classification model and the first regression model through the first historical data, and the first historical data is the historical vital sign data of the second target object whose similarity to the first target object is greater than the first threshold; the second input unit inputs the life behavior data into the behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained based on the second classification model and the second regression model through the second historical data, and the second historical data is the historical behavior data of the second target object. The behavior ability level of the behavior ability assessment result includes self-care, semi-self-care, and disability; the third input unit inputs the health assessment result and the behavior ability assessment result into the nursing plan model to obtain a target nursing plan, generates scheduling information according to the target nursing plan, and sends it to the target client. The nursing plan model is trained based on the third classification model, the third regression model, and the fourth regression model through the third historical data. The third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object. The health detection model, behavior detection model, and nursing plan model of this application combine the classification model and the regression model. The regression model is introduced to predict the change trend of the state index according to the historical data, while the classification model is used to determine the current state. Through the classification model and the regression model, not only can the current abnormal state of the elderly be judged in time, but also the future risks can be predicted. Furthermore, based on the current state and future trends, the nursing plan for the elderly can be personalized, and the prediction based on the regression model can also ensure that the criteria for abnormal determination change in real time with the change of the state. This method solves the problem in the prior art that the state of the elderly is evaluated periodically and the evaluation model is based on fixed indicators and weights, which is difficult to adapt to the rapid change of the health state of the elderly, resulting in inaccurate evaluation results and poor nursing service effects.

[0119] In an alternative embodiment, in order to generate the above health assessment result, the above first input unit includes:

[0120] The first acquisition module is used to acquire the vital sign data of the second target object within the first preset period to obtain the first historical data, and acquire the vital sign data of the first target object within the first preset period to obtain the fourth historical data. The end time of the first preset period is the current time;

[0121] The first prediction module is used to predict the maximum and minimum values of the vital sign data of the first target object at the current time according to the first historical data and the fourth historical data through the first regression model to obtain the first target range;

[0122] The first classification module is used to determine the potential disease categories of the first target object based on the first target range and vital sign data through the first classification model, and generate a health assessment result according to the potential disease categories.

[0123] To generate the above-mentioned ability assessment result, in an alternative embodiment, the above-mentioned second input unit includes:

[0124] The second classification module is used to determine the behavior ability level of the first target object according to the life behavior data through the second classification model, and obtain the first ability level;

[0125] The second acquisition module is used to obtain the behavior ability level of the second target object within the first preset time period to obtain the fifth historical data, and obtain the life behavior data of the second target object within the first preset time period to obtain the second historical data. The end time of the first preset time period is the current time;

[0126] The second prediction module is used to predict the change trend of the behavior ability level of the first target object within the second preset time period based on time series analysis according to the second historical data, the fifth historical data and the first ability level through the second regression model, and obtain the target sequence. The start time of the second preset time period is the current time;

[0127] The first determination module is used to determine the behavior ability level of the first target object at the current time based on the first ability level and the target sequence, obtain the second ability level, and generate a behavior ability assessment result according to the second ability level.

[0128] To generate the above-mentioned target care plan, in an alternative embodiment, the above-mentioned second input unit includes:

[0129] The third classification module is used to determine the required service category of the first target object according to the health assessment result and the behavior ability assessment result through the third classification model, and obtain the first target category. The required service category includes meal assistance and medical assistance;

[0130] The third acquisition module is used to obtain the target care plan and user feedback data of the first target object within the first preset time period to obtain the sixth historical data;

[0131] The third prediction module is used to predict the required service category of the first target object at the current time according to the sixth historical data through the third regression model, and obtain the second target category;

[0132] The fourth prediction module is used to predict the required service frequency and required service intensity of the first target object according to the user feedback data in the first target category, the second target category and the sixth historical data through the fourth regression model;

[0133] A generation module for generating a target care plan according to the first target category, the second target category, the frequency of demand services, and the intensity of demand services.

[0134] In an optional implementation manner, in order to obtain the above-mentioned vital sign data and the above-mentioned life behavior data, the above-mentioned first acquisition unit includes:

[0135] A fourth acquisition module for collecting physiological indicators of the first target object through a wearable device to obtain first vital sign data;

[0136] A fifth acquisition module for collecting the activity trajectory, indoor environment parameters, and life behaviors of the first target object through a smart home device, and performing a quantitative score to obtain first behavior data;

[0137] A sixth acquisition module for obtaining the life habit data of the first target object through a first preset path to obtain second behavior data;

[0138] A seventh acquisition module for obtaining the case data and physical examination data of the first target object through a second preset path to obtain second vital sign data;

[0139] A first processing module for integrating and data cleaning the first vital sign data and the second vital sign data to obtain vital sign data;

[0140] A second processing module for integrating and data cleaning the first behavior data and the second behavior data to obtain life behavior data.

[0141] In an optional implementation manner, in order to perform a quantitative score on the life behavior of the elderly, the above-mentioned fifth acquisition module includes:

[0142] A first determination sub-module for determining the chewing ability score, swallowing ability score, and autonomous eating coordination score of the first target object based on the video data of the first target object in the performing state collected by the smart home device;

[0143] A second determination sub-module for determining the balance ability score, limb flexibility score, muscle endurance score, and flexible coordination score of the first target object based on the video data of the first target object in the moving state collected by the smart home device;

[0144] A third determination sub-module for determining the digital cognition score, time-space cognition score, and social understanding score of the first target object based on the video data of the first target object in the social state collected by the smart home device.

[0145] In an optional implementation manner, in order to train the above-mentioned health detection model, the above-mentioned device further includes:

[0146] A second acquisition unit, configured to acquire the vital sign data of a second target object within a third preset period to obtain seventh historical data, and acquire the vital sign data of the second target object within a fourth preset period to obtain eighth historical data, before inputting the vital sign data into a health detection model to obtain a health assessment result, where the end moment of the third preset period is the start moment of the fourth preset period;

[0147] A first construction unit, configured to construct a first target data set according to the seventh historical data and the eighth historical data, and train an alternative regression model according to the first target data set through a cross-validation mechanism to obtain a first regression model, where the cross-validation mechanism calculates the loss function of the alternative regression model according to the test set in one or more ways including sum of squared errors, mean squared error, and mean absolute error;

[0148] A processing unit, configured to process the seventh historical data according to the first regression model to obtain a plurality of second target ranges, and determine the potential disease category of a first target object according to the eighth historical data;

[0149] A second construction unit, configured to construct a second target data set according to the second target ranges, the seventh historical data, and the potential disease category, and train an alternative classification model according to the second target data set through a cross-validation mechanism to obtain a first classification model, where the cross-validation mechanism calculates the loss function of the alternative classification model according to the test set in one or more ways including precision, recall rate, and F1 value.

[0150] The above-mentioned device for generating an elderly care plan based on dynamic data health status assessment includes a processor and a memory. The above-mentioned first acquisition unit, first input unit, second input unit, third input unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned each module is located in different processors in any combination form.

[0151] The processor contains a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be set, and the prediction accuracy of the physical condition of the elderly can be improved by adjusting the kernel parameters.

[0152] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

[0153] An embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored program. When the program runs, it controls the device where the computer-readable storage medium is located to execute the method for generating an elderly care plan based on dynamic data health status assessment.

[0154] An embodiment of the present invention provides a processor. The processor is used to run a program. When the program runs, it executes the method for generating an elderly care plan based on dynamic data health status assessment.

[0155] An embodiment of the present invention provides a health detection system. The health detection system includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements at least the steps of the method for generating an elderly care plan based on dynamic data health status assessment.

[0156] The present application also provides a computer program product. When executed on a data processing device, it is adapted to execute a program initialized with at least the steps of the method for generating an elderly care plan based on dynamic data health status assessment.

[0157] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices. They can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0158] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0159] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and combinations of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce means for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0160] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including instruction means for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or means for implementing the functions specified in one or more of the blocks.

[0162] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0163] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0164] A computer-readable medium includes both permanent and non-permanent, removable and non-removable media and can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0165] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0166] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0167] 1) The method for generating an elderly care plan based on the health status assessment of dynamic data in this application. First, the vital sign data and life behavior data of the first target object are obtained in real time through a collection device. Then, the vital sign data is input into a health detection model to obtain a health assessment result. The health detection model is trained based on a first classification model and a first regression model using first historical data, and the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold. After that, the life behavior data is input into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained based on a second classification model and a second regression model using second historical data, and the second historical data is the historical behavior data of the second target object. Among them, the behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability. Finally, the health assessment result and the behavior ability assessment result are input into a care planning model to obtain a target care plan, and scheduling information is generated according to the target care plan and sent to the target client. The care planning model is trained based on a third classification model, a third regression model, and a fourth regression model using third historical data, and the third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical care plan, and user feedback data of the second target object. The health detection model, behavior detection model, and care planning model in this application all combine a classification model and a regression model. Among them, the regression model is introduced to predict the change trend of state indicators based on historical data, while the classification model is used to determine the current state. Through the classification model and the regression model, not only can the current abnormal state of the elderly be judged in time, but also the future risks can be predicted. Furthermore, the care plan for the elderly can be personalized based on the current state and future trends, and the prediction based on the regression model can also ensure that the criteria for abnormal determination change in real time with the change of the state. This method solves the problem in the prior art that the elderly's state is evaluated periodically and the evaluation model is based on fixed indicators and weights, which is difficult to adapt to the rapid change of the elderly's health status, resulting in inaccurate evaluation results and poor care service effects.

[0168] 2) The elderly care plan generation device based on the health status assessment of dynamic data in this application. The first acquisition unit acquires the vital sign data and life behavior data of the first target object in real time through the acquisition device; the first input unit inputs the vital sign data into the health detection model to obtain the health assessment result. The health detection model is trained based on the first classification model and the first regression model through the first historical data, and the first historical data is the historical vital sign data of the second target object whose similarity to the first target object is greater than the first threshold; the second input unit inputs the life behavior data into the behavior detection model to obtain the behavior ability assessment result. The behavior detection model is trained based on the second classification model and the second regression model through the second historical data, and the second historical data is the historical behavior data of the second target object. Among them, the behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability; the third input unit inputs the health assessment result and the behavior ability assessment result into the nursing plan model to obtain the target nursing plan, and generates scheduling information according to the target nursing plan and sends it to the target client. The nursing plan model is trained based on the third classification model, the third regression model, and the fourth regression model through the third historical data. The third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object. The health detection model, behavior detection model, and nursing plan model in this application all combine the classification model and the regression model. Among them, the regression model is introduced to predict the change trend of the state index according to the historical data, while the classification model is used to determine the current state. Through the classification model and the regression model, not only can the current abnormal state of the elderly be judged in time, but also the future risks can be predicted. Furthermore, based on the current state and future trends, the nursing plan for the elderly can be personalized, and the prediction based on the regression model can also ensure that the criteria for abnormal determination change in real time with the change of the state. This method solves the problem in the prior art that the elderly's state is evaluated periodically and the evaluation model is based on fixed indicators and weights, which is difficult to adapt to the rapid change of the elderly's health state, resulting in inaccurate evaluation results and poor nursing service effects.

[0169] The above are only the preferred embodiments of this application and are not used to limit this application. For those skilled in the art, this application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the protection scope of this application.

Claims

1. A method for generating an elderly care plan based on the assessment of the health status of dynamic data, characterized in that, Including: Real-time acquiring the vital sign data and life behavior data of the first target object through an acquisition device; Inputting the vital sign data into a health detection model to obtain a health assessment result, where the health detection model is trained by first historical data based on a first classification model and a first regression model, and the first historical data is the historical vital sign data of a second target object whose similarity to the first target object is greater than a first threshold; Inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result, where the behavior detection model is trained by second historical data based on a second classification model and a second regression model, and the second historical data is the historical behavior data of the second target object, and the behavior ability levels of the behavior ability assessment result include self-care, semi-self-care, and disability; Inputting the health assessment result and the behavior ability assessment result into a nursing planning model to obtain a target nursing plan, generating scheduling information according to the target nursing plan and sending it to a target client, where the nursing planning model is trained by third historical data based on a third classification model, a third regression model, and a fourth regression model, and the third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object.

2. The method according to claim 1, characterized in that, Inputting the vital sign data into a health detection model to obtain a health assessment result, including: Obtaining the vital sign data of the second target object within a first preset time period to obtain the first historical data, and obtaining the vital sign data of the first target object within the first preset time period to obtain fourth historical data, where the end time of the first preset time period is the current time; Predicting the maximum and minimum values of the vital sign data of the first target object at the current time according to the first historical data and the fourth historical data through the first regression model to obtain a first target range; Determining the potential disease category of the first target object according to the first target range and the vital sign data through the first classification model, and generating the health assessment result according to the potential disease category.

3. The method according to claim 1, characterized in that, Inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result, including: Determining the behavior ability level of the first target object according to the life behavior data through the second classification model to obtain a first ability level; Obtaining the behavior ability level of the second target object within a first preset time period to obtain fifth historical data, and obtaining the life behavior data of the second target object within the first preset time period to obtain the second historical data, where the end time of the first preset time period is the current time; Predicting the change trend of the behavior ability level of the first target object within a second preset time period according to the second historical data, the fifth historical data, and the first ability level through the second regression model based on time series analysis to obtain a target sequence, where the start time of the second preset time period is the current time; Determine the behavioral ability level of the first target object at the current moment based on the first ability level and the target sequence, obtain a second ability level, and generate the behavioral ability assessment result according to the second ability level.

4. The method according to claim 1, wherein Input the health assessment result and the behavioral ability assessment result into the nursing plan model to obtain a target nursing plan, including: Determine the required service category of the first target object according to the health assessment result and the behavioral ability assessment result through the third classification model, obtain a first target category, and the required service category includes meal assistance and medical assistance; Obtain the target nursing plan and the user feedback data of the first target object in the first preset time period to obtain sixth historical data; Predict the required service category of the first target object at the current moment according to the sixth historical data through the third regression model to obtain a second target category; Predict the required service frequency and required service intensity of the first target object according to the first target category, the second target category, and the user feedback data in the sixth historical data through the fourth regression model; Generate the target nursing plan according to the first target category, the second target category, the required service frequency, and the required service intensity.

5. The method according to claim 1, wherein Real-time obtain the vital sign data and life behavior data of the target object through a collection device, including: Collect the physiological indicators of the first target object through a wearable device to obtain first vital sign data; Collect the activity trajectory, indoor environment parameters, and life behavior of the first target object through a smart home device, and perform a quantitative score to obtain first behavior data; Obtain the life habit data of the first target object through a first preset path to obtain second behavior data; Obtain the case data and physical examination data of the first target object through a second preset path to obtain second vital sign data; Integrate and perform data cleaning on the first vital sign data and the second vital sign data to obtain the vital sign data; Integrate and perform data cleaning on the first behavior data and the second behavior data to obtain the life behavior data.

6. The method according to claim 5, characterized in that, Collect the life behavior of the first target object through a smart home device and perform a quantitative score, including: Based on the video data in the ongoing state of the first target object collected by the smart home device, determine the chewing ability score, swallowing ability score, and autonomous eating coordination score of the first target object; Based on the video data in the motion state of the first target object collected by the smart home device, determine the balance ability score, limb flexibility score, muscle endurance score, and flexible coordination score of the first target object; Based on the video data in the social state of the first target object collected by the smart home device, determine the digital cognition score, time and space cognition score, and social understanding score of the first target object.

7. The method according to claim 2, wherein Before inputting the vital sign data into the health detection model to obtain the health assessment result, the method further includes: Obtain the vital sign data of the second target object within a third preset time period to obtain seventh historical data, and obtain the vital sign data of the second target object within a fourth preset time period to obtain eighth historical data. The end time of the third preset time period is the start time of the fourth preset time period; Construct a first target data set according to the seventh historical data and the eighth historical data, and train an alternative regression model according to the first target data set through a cross-validation mechanism to obtain the first regression model. Among them, the way the cross-validation mechanism calculates the loss function of the alternative regression model according to the test set includes one or more of the sum of squared errors, mean squared error, and mean absolute error; Process the seventh historical data according to the first regression model to obtain a plurality of second target ranges, and determine the potential disease category of the first target object according to the eighth historical data; Construct a second target data set according to the second target range, the seventh historical data, and the potential disease category, and train an alternative classification model according to the second target data set through a cross-validation mechanism to obtain the first classification model. Among them, the way the cross-validation mechanism calculates the loss function of the alternative classification model according to the test set includes one or more of precision, recall, and F1 value.

8. An elderly care plan generation device based on the assessment of the health status of dynamic data, characterized in that, The device includes: A first acquisition unit for real-time acquiring the vital sign data and life behavior data of a first target object through an acquisition device; A first input unit for inputting the vital sign data into a health detection model to obtain a health assessment result. The health detection model is trained by first historical data based on a first classification model and a first regression model. The first historical data is the historical vital sign data of a second target object with a similarity greater than a first threshold to the first target object; A second input unit for inputting the life behavior data into a behavior detection model to obtain a behavior ability assessment result. The behavior detection model is trained by second historical data based on a second classification model and a second regression model. The second historical data is the historical behavior data of the second target object. Among them, the behavior ability level of the behavior ability assessment result includes self-care, semi-self-care, and disability; A third input unit for inputting the health assessment result and the behavior ability assessment result into a nursing plan model to obtain a target nursing plan, generating scheduling information according to the target nursing plan and sending it to a target client. The nursing plan model is trained by third historical data based on a third classification model, a third regression model, and a fourth regression model. The third historical data is obtained by training the historical health assessment result, historical behavior ability assessment result, historical nursing plan, and user feedback data of the second target object.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program. When the program runs, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 7.

10. A health monitoring system, characterized in that, Include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include those for performing the method according to any one of claims 1 to 7.

Images