Elderly home health data processing method and elderly home health care system

Abstract

The present application relates to the field of data processing technology, and discloses a method for processing home health data of the elderly and a home health care system for the elderly. First, the home health data of the target object is obtained through a millimeter wave radar, and the home assessment data of the target object is obtained based on the home health data. Then the medical examination data of the target object is obtained, and the inspection and assessment data of the target object is obtained based on the medical examination data. Then, a dynamic weight ratio is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight ratio. Finally, the initial review time is adjusted based on the current weight ratio, the home assessment data and the inspection and assessment data to obtain the recommended review time. The method accurately adjusts the review time by dynamically matching the home health monitoring and medical examination data, so as to more accurately provide a review time recommendation that meets the actual health status of the elderly, and to a certain extent optimize the use of medical resources.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method for processing elderly home health data and an elderly home health care system. Background Art

[0002] The incidence of cardiovascular and cerebrovascular diseases is high among the elderly. Many patients need home care and regular check-ups in the hospital to ensure that their condition is monitored and intervened in a timely manner.

[0003] In the related art, doctors usually arrange a fixed review time after the examination. These times are often set according to the conditions of most patients, such as two weeks, three months, or even some chronic patients may arrange a review every six months. However, many patients tend to rely on the doctor's advice and strictly adhere to the fixed review cycle, but ignore the reality that the condition may change at any time. In this way, when the condition changes acutely or worsens, the patient fails to review in time, often missing the best time for treatment, causing the condition to further deteriorate or complicate.

[0004] Therefore, how to accurately recommend review times to the elderly is a technical problem that needs to be solved urgently in related technologies. Summary of the invention

[0005] The present application provides a method for processing home health data of the elderly and a home health care system for the elderly, which solves the technical problem of how to accurately recommend a review time to the elderly in the related technology, and achieves the technical effect of being able to accurately recommend a review time to the elderly.

[0006] In order to achieve the above objectives, the main technical solutions adopted in this application include:

[0007] In a first aspect, an embodiment of the present application provides a method for processing health data of an elderly person at home, the method comprising:

[0008] The target object's sleep process at home is monitored by a millimeter wave radar to obtain the target object's home health data, and feature extraction is performed based on the home health data to obtain the target object's home assessment data;

[0009] Acquire medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain examination evaluation data of the target object; wherein the medical examination data corresponds to a medical examination time, and the medical examination time is the time when the medical examination of the target object occurs;

[0010] Perform dynamic weight matching according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data;

[0011] The initial review time of the target object is adjusted based on the current weight ratio, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the initial review time is the time recommended by the doctor for the target object to be reviewed.

[0012] Optionally, the medical examination data includes examination image data and report text data, and the report text data is used to describe the health status of the target object; and extracting features based on the medical examination data to obtain the examination evaluation data of the target object includes:

[0013] Performing convolution processing on the inspection image data to obtain image feature data;

[0014] Performing vector representation on the report text data to obtain text feature data;

[0015] Placing the image feature data between the two text feature data for feature stacking to obtain inspection feature data;

[0016] A health assessment is performed on the target object based on the examination feature data to obtain the examination assessment data.

[0017] Optionally, the medical examination data includes report text data, and before dynamically weighting the medical examination data and obtaining a current weighting ratio between the home assessment data and the examination assessment data according to the latest monitoring time of the home health data, the method further includes:

[0018] Extracting the initial review time from the report text data;

[0019] The dynamic weight matching is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data, including:

[0020] Determining a first duration between the latest monitoring time and the initial review time, and a second duration between the latest monitoring time and the medical examination time;

[0021] The current weight ratio is obtained by dynamically weighting the first duration and the second duration.

[0022] Optionally, the current weight ratio is a ratio between the home weight of the home assessment data and the inspection weight of the inspection assessment data;

[0023] Wherein, if the first duration is greater than the second duration, the inspection weight is greater than the home weight;

[0024] If the first duration is less than the second duration, the inspection weight is less than the home weight.

[0025] Optionally, the monitoring of the target object's sleep process at home by using a millimeter wave radar to obtain the target object's home health data includes:

[0026] During the target object's sleep at home, the millimeter wave radar is used to perform breathing monitoring, heart rate monitoring and blood pressure monitoring respectively to obtain breathing time series data, heart rate time series data and blood pressure time series data of the target object;

[0027] The multi-channel home health data is constructed using the respiratory time series data, the heartbeat time series data and the blood pressure time series data.

[0028] Optionally, extracting features based on the home health data to obtain home assessment data of the target object includes:

[0029] Extracting features from the respiratory time series data to obtain respiratory assessment data of the target object;

[0030] Extracting features from the heartbeat time series data to obtain heartbeat evaluation data of the target object;

[0031] Performing feature extraction on the blood pressure time series data to obtain blood pressure assessment data of the target object;

[0032] The breathing assessment data, the heartbeat assessment data and the blood pressure assessment data are weighted to obtain the health status data of the target object as the home assessment data.

[0033] Optionally, adjusting the initial review time of the target object based on the current weight ratio, the home assessment data and the inspection assessment data to obtain a recommended review time includes:

[0034] Performing health prediction on the target object based on the current weight ratio, the home assessment data and the examination assessment data to obtain a health prediction result;

[0035] The initial review time of the target object is adjusted according to the health prediction result to obtain a recommended review time.

[0036] Optionally, performing health prediction on the target object based on the current weight ratio, the home assessment data and the examination assessment data to obtain a health prediction result includes:

[0037] If the current weight ratio is greater than a preset threshold, a first health prediction model is selected, and the home assessment data and the examination assessment data are input into the first health prediction model for health prediction to obtain a corresponding health prediction result;

[0038] If the current weight ratio is less than or equal to a preset threshold, a second health prediction model is selected, and the home assessment data and the examination assessment data are input into the second health prediction model for health prediction to obtain a corresponding health prediction result.

[0039] In a second aspect, an embodiment of the present application provides a home health care system for the elderly, the system comprising:

[0040] A home feature extraction module is used to monitor the sleep process of the target object at home through a millimeter wave radar to obtain the home health data of the target object, and perform feature extraction based on the home health data to obtain the home assessment data of the target object;

[0041] An examination feature extraction module is used to obtain medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain examination evaluation data of the target object; wherein the medical examination data corresponds to a medical examination time, and the medical examination time is the time when a medical examination is performed on the target object;

[0042] A dynamic weight matching module, used to perform dynamic weight matching according to the medical examination time and the latest monitoring time of the home health data, to obtain the current weight matching between the home assessment data and the examination assessment data;

[0043] A review time adjustment module is used to adjust the initial review time of the target object based on the current weight ratio, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the initial review time is the time when the target object needs to be reviewed mentioned in the medical examination data.

[0044] Optionally, the system includes a millimeter wave monitoring device and a computer device communicatively connected to the millimeter wave monitoring device;

[0045] The millimeter wave monitoring device is used to monitor the target object's sleep process at home to obtain the target object's home health data;

[0046] The computer device is used to perform feature extraction based on the home health data to obtain the home assessment data of the target object; obtain the medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain the examination assessment data of the target object; perform dynamic weight matching according to the medical examination time corresponding to the medical examination data and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data; adjust the initial review time of the target object based on the current weight matching, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the medical examination time is the time when the medical examination of the target object occurs; and the initial review time is the time when the target object mentioned in the medical examination data needs to be reviewed.

[0047] In an embodiment of the present application, the sleep process of the target object at home is first monitored by a millimeter wave radar to obtain the home health data of the target object, and feature extraction is performed based on the home health data to obtain the home assessment data of the target object. Then the medical examination data of the target object is obtained, and feature extraction is performed based on the medical examination data to obtain the inspection and evaluation data of the target object. Then, a dynamic weight ratio is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight ratio between the home assessment data and the inspection and evaluation data. Finally, the initial review time of the target object is adjusted based on the current weight ratio, the home assessment data and the inspection and evaluation data to obtain the recommended review time. This method accurately adjusts the review time by dynamically matching home health monitoring and medical examination data to more accurately provide review time recommendations that meet the actual health status of the elderly, and to a certain extent optimize the use of medical resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1a A schematic diagram of a home health care system for the elderly provided in an embodiment of this specification;

[0050] Figure 1b A flowchart of a method for processing elderly home health data provided in an embodiment of this specification;

[0051] Figure 2aA flowchart of a method for processing elderly home health data provided in an embodiment of this specification;

[0052] Figure 2b A schematic diagram of the medical examination data processing process provided in the embodiments of this specification;

[0053] Figure 3 A flow chart of a method for processing elderly home health data provided in an embodiment of this specification;

[0054] Figure 4 A flowchart of the process of obtaining home health data provided by the embodiments of this specification;

[0055] Figure 5a A flow chart of a method for processing elderly home health data provided in an embodiment of this specification;

[0056] Figure 5b A schematic diagram of a health status data acquisition process provided in an embodiment of this specification;

[0057] Figure 6 A flow chart of a method for processing elderly home health data provided in an embodiment of this specification;

[0058] Figure 7 A flow chart of a method for processing elderly home health data provided in an embodiment of this specification;

[0059] Figure 8 A schematic diagram of a home health care system for the elderly provided in an embodiment of this specification;

[0060] Fig. 9 A schematic diagram of a home health care system for the elderly provided in an embodiment of this specification;

[0061] Fig.10 A schematic diagram of the structure of a computer device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution 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 part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0063] Elderly people who need home care due to cardiovascular and cerebrovascular diseases may quietly worsen their condition, especially during sleep, posing a potential threat to their health. Therefore, monitoring the sleep process of elderly patients can detect abnormalities in time, and go to the hospital for a reexamination before the condition worsens, thereby minimizing the risk of sudden health problems. On the other hand, if the monitoring results show that the patient's condition is stable, the reexamination time can be appropriately delayed according to the specific circumstances to avoid unnecessary medical treatment, thereby making more effective use of limited medical resources. Smart wearable devices are used in related technologies. Although they can accurately monitor vital signs such as heart rate and blood pressure, the sense of restraint and discomfort when wearing them often prevents the elderly from getting a good rest, seriously affecting their comfort and compliance.

[0064] In order to overcome these problems, a method for processing elderly home health data is proposed in this application scenario example. This method can be applied to the elderly home health care system. Figure 1a The system includes a millimeter wave monitoring device 101, a platform management terminal 102 and a user terminal 103, wherein the millimeter wave monitoring device 101 can be installed in the bedroom of the elderly, the platform management terminal 102 is deployed in a computer device, and the user terminal 103 can be an application in a smart phone. The millimeter wave monitoring device 101, the user terminal 103 and the platform management terminal 102 can communicate with each other. After the user successfully registers on the user terminal 103, the millimeter wave monitoring device 101 can be added to the user terminal 103 and managed, such as starting and stopping monitoring the sleep process of the elderly, uploading the medical examination data of the elderly (including the review time arranged by the doctor) to the platform management terminal 102 and viewing the health assessment results of the elderly.

[0065] After the millimeter wave monitoring device 101 starts to monitor the health of the elderly, it uses millimeter wave radar technology to monitor the elderly's sleep process, collects home health data, and sends the home health data to the platform management terminal 102. The platform management terminal 102 obtains home assessment data based on the elderly's home health data. At the same time, the platform management terminal 102 also obtains inspection and assessment data based on the elderly's medical examination data. Then, the home assessment data and the inspection and assessment data are comprehensively considered, and the weight ratio between the home assessment data and the inspection and assessment data is dynamically adjusted by comparing the medical examination time and the latest monitoring time of the home health data to obtain the health prediction result, and the review time of the elderly is adjusted according to the health prediction result.

[0066] The innovation of this scenario example lies in the combination of home health monitoring and medical examination data, and the adjustment of review time through dynamic weight ratio, aiming to more accurately recommend review times for the elderly and optimize the use of medical resources.

[0067] According to an embodiment of the present application, an embodiment of a method for processing health data of the elderly at home 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 a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0068] See also Figure 1b In this embodiment, a method for processing health data of the elderly at home is provided, and the steps include:

[0069] S110. Monitor the target object's sleep process at home by using a millimeter wave radar to obtain the target object's home health data, and perform feature extraction based on the home health data to obtain the target object's home assessment data.

[0070] S120: Obtain medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain examination evaluation data of the target object.

[0071] S130, dynamically weight matching is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data.

[0072] S140: Adjust the initial review time of the target object based on the current weight ratio, home assessment data and inspection assessment data to obtain a recommended review time.

[0073] Among them, the target object can be used to characterize the elderly who suffer from a specified type of disease, and the specified type of disease can be a chronic disease that requires regular review, such as cardiovascular and cerebrovascular diseases. Home health data can be time series data of vital signs of the elderly during sleep at home. Home assessment data can be used to describe the health status assessment results of the elderly at home, such as the home assessment data can describe the risk of serious cardiovascular and cerebrovascular diseases of the target object predicted based on the home health data. Medical examination data can include medical text and image data, and correspond to the medical examination time, which is the time when the medical examination of the target object occurs. Examination assessment data can be used to describe the health status assessment results of the elderly during the medical examination, such as the examination assessment data can describe the risk of serious cardiovascular and cerebrovascular diseases of the target object predicted based on the medical examination data. The initial review time can be the time when the target object needs to be reviewed by the doctor.

[0074] Specifically, when the target object is sleeping at home, the millimeter wave radar is used to continuously monitor the target object and obtain the reflected signal of the target object's sleeping space. By processing the reflected signal, the target object's breathing, heartbeat or blood pressure data can be obtained, and these data can constitute the target object's home health data. It can be understood that home health data belongs to time series data.

[0075] In some embodiments, a prediction model can be used to process home health data. First, a recurrent neural network can be used to extract features from home health data to obtain home health data features, and then the home health data features are input into the encoding layer for vectorization to obtain home health data vectors, and then the home health data vectors are input into the average pooling layer for dimensionality reduction, and the fully connected layer is processed based on the output of the average pooling layer to obtain home assessment data. The training data set of the prediction model can be a large amount of home health data on cardiovascular and cerebrovascular diseases. Each data corresponds to a training label, and the loss function can be a mean square error loss function (MSE).

[0076] Furthermore, a large bimodal model of text and image (such as InternVL2) fine-tuned with medical-related data can be used to evaluate the medical examination data of the target object. Exemplarily, the large bimodal model of text and image can be used to extract features from the image data in the medical examination data to obtain image features, and the large bimodal model of text and image can be used to encode the text data in the medical examination data to obtain text features. Secondly, the image features and text features are spliced ​​to obtain the text and image features. Finally, the medical examination data of the target object is evaluated according to the text and image features to obtain the examination evaluation data. The training data set of the large bimodal model of text and image can be a large amount of medical examination data of cardiovascular and cerebrovascular diseases, including examination image data and corresponding report text data.

[0077] Furthermore, considering the home assessment data and the inspection assessment data comprehensively, the health condition of the target object is evaluated, and the weight ratio of the two needs to be reasonably determined. It can be understood that the closer the latest monitoring time is to the medical examination time, the greater the weight of the inspection assessment data should be. Conversely, the farther away from the medical examination time, the greater the weight of the home assessment data should be. Exemplarily, let the time length between the medical examination time and the initial review time be T, the time length between the latest monitoring time and the medical examination time be t, and the weight ratio between the home assessment data and the inspection assessment data is shown in Table 1. In this table, according to the relationship between t and T, the current weight ratio between the home assessment data and the inspection assessment data can be obtained.

[0078] Table 1 Weight ratio table

[0079]

[0080] Finally, based on the current weight ratio, the home assessment data and the inspection assessment data are weighted to obtain the health prediction result of the current object. The health prediction result can be used to characterize the comprehensive health status of the target object. The initial review time of the target object can be adjusted according to the health prediction result to obtain the recommended review time. Exemplarily, if the health prediction result is lower than the preset threshold value one, it means that the probability of the target object suffering from severe cardiovascular and cerebrovascular diseases in the future is low, and it can be suggested to extend the initial review time by a certain length of time as the recommended review time, or continue to use the initial review time as the recommended review time; if the health prediction result is higher than the preset threshold value one and lower than the preset threshold value two, it means that the probability of the target object suffering from cardiovascular and cerebrovascular diseases in the future is at a medium level, and it is suggested to maintain the initial review time unchanged; if the health prediction result is higher than the preset threshold value two, it means that the probability of the target object suffering from severe cardiovascular and cerebrovascular diseases in the future is high, and the condition has worsened since the medical examination time, and it can be suggested that the target object arrange a review as soon as possible. It can be understood that the initial review time of the target object can be adjusted according to the health prediction result, and a more detailed division than in the above example can be adopted, so that the initial review time can be extended or shortened by a more accurate time.

[0081] In the above embodiment, the sleep process of the target object at home is first monitored by a millimeter wave radar to obtain the home health data of the target object, and feature extraction is performed based on the home health data to obtain the home assessment data of the target object. Then the medical examination data of the target object is obtained, and feature extraction is performed based on the medical examination data to obtain the inspection and evaluation data of the target object. Then, a dynamic weight ratio is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight ratio between the home assessment data and the inspection and evaluation data. Finally, the initial review time of the target object is adjusted based on the current weight ratio, the home assessment data and the inspection and evaluation data to obtain the recommended review time. This method accurately adjusts the review time by dynamically matching the home health monitoring and medical examination data, so as to more accurately provide a review time recommendation that meets the actual health status of the elderly, and to a certain extent optimize the use of medical resources.

[0082] In some embodiments, see Figure 2a Medical examination data includes examination image data and report text data. The report text data is used to describe the health status of the target object. Feature extraction is performed based on the medical examination data to obtain the examination evaluation data of the target object, including:

[0083] S210, performing convolution processing on the inspection image data to obtain image feature data.

[0084] S220: Perform vector representation on the report text data to obtain text feature data.

[0085] S230: placing the image feature data between two text feature data for feature stacking to obtain inspection feature data.

[0086] S240: Perform a health assessment on the target object based on the inspection feature data to obtain inspection assessment data.

[0087] In some cases, if the image feature data and the text feature data are directly spliced, no matter whether the image feature data is located above or below the text feature data, the image and text information cannot be effectively combined, resulting in the model lacking pertinence and contextual information when processing the image. For example, the image feature data is on top and the text feature data is on the bottom. This may cause the model to fail to fully understand the subsequent text features when processing the image, and thus fail to extract relevant information from the image in a targeted manner. For example, the text feature data is on top and the image feature data is on the bottom. This may cause the model to fail to fully utilize the previous text features when processing the image, and thus fail to effectively extract relevant information from the image. Therefore, in order to take into account both fully understanding the text features and fully utilizing the text features, the image feature data is placed between the two text feature data for feature stacking.

[0088] Specifically, see Figure 2b First, the graph convolutional neural network 201 can be used to perform convolution processing on the inspection image data to obtain image feature data; secondly, the convolutional neural network 202 is used to vectorize the report text data to obtain text feature data; then the image feature data and the text feature data are input into the feature stacking layer 203, and the feature stacking layer 203 places the image feature data between the two text feature data for feature stacking to obtain inspection feature data; finally, the inspection feature data is input into the average pooling layer 204 for dimensionality reduction, and the fully connected layer 205 is processed based on the output of the average pooling layer 204 to obtain inspection evaluation data.

[0089] In the above embodiment, the inspection image data is first convoluted to obtain image feature data; secondly, the report text data is vectorized to obtain text feature data; then the image feature data is placed between two text feature data for feature stacking to obtain inspection feature data; finally, the health assessment of the target object is performed based on the inspection feature data to obtain inspection evaluation data. Feature stacking allows the prediction model to review text information before processing the image, thereby extracting key information more specifically, reducing information loss caused by insufficient image labeling, and helping to obtain more accurate inspection evaluation data.

[0090] In some embodiments, see Figure 3, medical examination data includes report text data. Before dynamically weighting the medical examination time and the latest monitoring time of the home health data to obtain the current weight ratio between the home assessment data and the examination assessment data, it also includes:

[0091] S310, extracting the initial review time from the report text data.

[0092] Dynamic weight matching is performed based on the medical examination time and the latest monitoring time of home health data to obtain the current weight matching between home assessment data and examination assessment data, including:

[0093] S320, determining a first duration between the latest monitoring time and the initial review time, and a second duration between the latest monitoring time and the medical examination time.

[0094] S330: Perform dynamic weight matching according to the first duration and the second duration to obtain a current weight matching.

[0095] Specifically, the initial review time is first extracted from the report text data and recorded as , and record the time of medical examination as , the latest monitoring time is recorded as ; Then the first duration between the latest monitoring time and the initial review time can be determined as , the second duration between the latest monitoring time and the medical examination time ; According to the dynamic weight ratio of the first duration and the second duration, The larger the weight is, that is, the longer the time between the latest monitoring time and the medical examination time is, the smaller the weight of the examination evaluation data should be, while the weight of the home evaluation data should be greater. For example, the calculation formula for determining the weight of the home evaluation data is:

[0096]

[0097] The weight calculation formula for the inspection evaluation data is:

[0098]

[0099] The ratio of W1 to W2 is the current weight ratio of the home assessment data and the inspection assessment data.

[0100] In the above embodiment, the initial review time is first extracted from the report text data; secondly, the first duration between the latest monitoring time and the initial review time, and the second duration between the latest monitoring time and the medical examination time are determined; finally, a dynamic weight ratio is performed based on the first duration and the second duration to obtain the current weight ratio. By dynamically adjusting the weight ratio between the home assessment data and the examination assessment data according to the time difference, the role of data from different sources can be balanced, making the health assessment of the target object more accurate.

[0101] In some embodiments, the current weight ratio is the ratio between the home weight of the home assessment data and the inspection weight of the inspection assessment data; wherein, if the first duration is greater than the second duration, the inspection weight is greater than the home weight; if the first duration is less than the second duration, the inspection weight is less than the home weight.

[0102] Specifically, as mentioned above, the first time between the latest monitoring time and the initial review time is , the second duration between the latest monitoring time and the medical examination time When the first duration is greater than the second duration, that is When , the inspection weight is greater than the home weight; when the first duration is greater than the second duration, that is , the inspection weight is less than the home weight.

[0103] In the above embodiment, by adjusting the weights of the home assessment data and the examination assessment data according to the time difference, the effects of data from different sources can be balanced, making the health assessment of the target object more accurate, which is conducive to recommending a more accurate review time to the elderly.

[0104] In some embodiments, see Figure 4 , the target object's sleep process at home is monitored by millimeter wave radar to obtain the target object's home health data, including:

[0105] S510. When the target object is sleeping at home, the respiration, heartbeat and blood pressure are monitored by millimeter wave radar to obtain the respiration time series data, heartbeat time series data and blood pressure time series data of the target object.

[0106] S520: Construct multi-channel home health data using respiratory time series data, heart rate time series data, and blood pressure time series data.

[0107] Specifically, the breathing and heart rate monitoring of the target object is performed by the millimeter wave radar, and the main steps of obtaining the breathing rate and heart rate of the target object include:

[0108] 1) First, the millimeter wave radar is used to collect the unmanned reflection signal of the target object's sleeping space when no one is in the sleeping space, and the manned reflection signal of the target object's sleeping space when manned is in the sleeping space;

[0109] 2) Mix the unmanned reflected signal and the manned reflected signal with the local oscillator signal (a signal generated by the millimeter-wave radar at a time with the same frequency as the transmitted signal) to obtain the unmanned mixed signal and the manned mixed signal;

[0110] 3) extracting the difference frequency signal between the unattended mixing signal and the attended mixing signal from the low-pass filter;

[0111] 4) Perform FFT (Fast Fourier Transform) spectrum analysis on the difference frequency signal of the unattended mixing signal, and record the complete spectrum information of the target object's sleeping space as the spectrum initial value S1;

[0112] 5) Perform FFT spectrum analysis on the difference frequency signal of the mixed signal, and record the target object's sleeping space and the target object's complete spectrum information as spectrum value S2;

[0113] 6) extracting the spectrum information S of the target object and recording the maximum peak frequency f0 of the amplitude-frequency characteristic of the spectrum information S, where S = S2 – S1;

[0114] 7) Perform FFT transformation on the amplitude-frequency and phase-frequency information at the maximum peak frequency f0 to obtain the amplitude spectrum Sv, calculate the speed information of the object at the distance corresponding to the maximum peak frequency f0, and obtain the breathing speed spectrum f1 and heart rate spectrum f2 of the target object;

[0115] 8) Record the time interval between the start times of the target subject's breathing velocity spectrum f1 twice , find the target object's breathing frequency .

[0116] 9) Record the time interval between the start times of the target object's heart rate spectrum f2 , find the target object's heart rate .

[0117] The millimeter-wave radar is used to monitor the target object's blood pressure, and an example is given to illustrate how to obtain the target object's blood pressure value. First, a preprocessing process for millimeter-wave radar signals is designed. The radar echo phase is extracted and the phase constant is removed by using an extended differential and cross-multiply algorithm (DACM) and average filtering. Then, wavelet filtering technology is used to eliminate high-frequency noise and baseline drift to ensure that high-quality pulse wave signals are obtained. In the neural network part, a two-stage progressive feature fusion and mapping network with an encoder-decoder structure is constructed to better establish the mapping relationship from pulse wave to blood pressure and achieve accurate blood pressure measurement. For example, in the first stage, MultiResUNet is used to extract multi-scale features of the input pulse wave signal, and the long-distance dependencies between feature vectors are captured with the help of the self-attention mechanism to reconstruct a detailed blood pressure waveform; in the second stage, the model sends the deep features of the pulse wave obtained during the encoding process to two output networks composed of a convolutional neural network (CNN) and a long short-term memory network (LSTM), thereby completing feature dimensionality reduction, time series feature analysis, and further realizing feature fusion to construct an accurate mapping of pulse wave features to blood pressure values.

[0118] By using millimeter wave radar to continuously monitor the respiration, heartbeat and blood pressure of the target object, the target object's respiration time series data, heartbeat time series data and blood pressure time series data are obtained. Furthermore, based on the integration of respiration time series data, heartbeat time series data and blood pressure time series data, a multi-channel home health data is constructed. The three types of data can be independent time series data sequences or merged into a matrix for evaluating the health status of the target object.

[0119] In the above embodiment, firstly, during the target object's sleep at home, the millimeter wave radar is used to monitor breathing, heart rate and blood pressure respectively, and the breathing time series data, heart rate time series data and blood pressure time series data of the target object are obtained; secondly, the breathing time series data, heart rate time series data and blood pressure time series data are used to construct multi-channel home health data. Through the non-sensing monitoring technology, an accurate data basis is provided for the evaluation of the target object's home health status.

[0120] In some embodiments, see Figure 5a , extract features based on home health data to obtain home assessment data of the target object, including:

[0121] S610: Extract features from the respiratory time series data to obtain respiratory assessment data of the target object.

[0122] S620: Extract features from the heartbeat time series data to obtain heartbeat evaluation data of the target object.

[0123] S630: Extract features from the blood pressure time series data to obtain blood pressure assessment data of the target object.

[0124] S640: Perform weighted processing on the respiratory assessment data, the heartbeat assessment data and the blood pressure assessment data to obtain the health status data of the target object as home assessment data.

[0125] Among them, respiratory assessment data can be used to describe the respiratory function and degree of dyspnea of ​​the elderly at home. Given the correlation between respiratory conditions and cardiovascular and cerebrovascular diseases, respiratory assessment data can also be used to characterize the impact of the elderly's respiratory conditions on the development of cardiovascular and cerebrovascular diseases, or the probability of causing cardiovascular and cerebrovascular diseases.

[0126] Heart rate assessment data can be used to describe the heart rate, rhythm, and variability of the elderly at home. Given the close relationship between heart rate and cardiovascular and cerebrovascular diseases, heart rate assessment data can also be used to characterize the impact of the elderly's heart rate on the development of cardiovascular and cerebrovascular diseases, or the probability of causing serious cardiovascular and cerebrovascular diseases.

[0127] Blood pressure assessment data can be used to describe the blood pressure level and cardiovascular health status of the elderly at home. Given the close relationship between blood pressure and cardiovascular and cerebrovascular diseases, blood pressure assessment data can also be used to characterize the impact of the elderly's blood pressure on the development of cardiovascular and cerebrovascular diseases, or the probability of causing serious cardiovascular and cerebrovascular diseases.

[0128] In some embodiments, see Figure 5bFirst, the respiratory time series data is input into the recurrent neural network 601 for feature extraction to obtain respiratory feature data, the heartbeat time series data is input into the recurrent neural network 602 for feature extraction to obtain heartbeat feature data, and the blood pressure time series data is input into the recurrent neural network 603 for feature extraction to obtain blood pressure feature data; secondly, the respiratory feature data is input into the respiratory evaluation branch 604, the heartbeat feature data is input into the heartbeat evaluation branch 605, and the blood pressure feature data is input into the blood pressure evaluation branch 606; then each of the above evaluation branches is evaluated according to the input feature data, and the corresponding evaluation data is output. Taking the respiratory evaluation branch 604 as an example, the respiratory feature data is first input into the average pooling layer of the respiratory evaluation branch 604 for dimension reduction, and then the output of the average pooling layer is processed by the fully connected layer to obtain the respiratory evaluation data; the heartbeat evaluation branch 605 and the blood pressure evaluation branch 606 also have the same data processing process. Finally, the respiratory evaluation data, the heartbeat evaluation data and the blood pressure evaluation data are input into the weighted fusion layer 607 for weighted fusion to obtain the health status data of the target object, that is, the home evaluation data. For example, the recurrent neural network can adopt an LSTM (Long Short-Term Memory) network to evaluate the breathing, heart rate and blood pressure of the elderly, and its prediction accuracy can reach more than 80%.

[0129] In the above embodiment, firstly, feature extraction is performed on the respiratory time series data to obtain the respiratory assessment data of the target object; secondly, feature extraction is performed on the heartbeat time series data to obtain the heartbeat assessment data of the target object; then feature extraction is performed on the blood pressure time series data to obtain the blood pressure assessment data of the target object; finally, weighted processing is performed on the respiratory assessment data, the heartbeat assessment data and the blood pressure assessment data to obtain the health status data of the target object as the home assessment data. By weighted processing of the assessment data of respiration, heartbeat and blood pressure, the health status of the target object can be comprehensively reflected, providing a multi-dimensional and real-time data basis for the prediction of the health status of the target object, thereby providing strong support for the recommended review time.

[0130] In some embodiments, see Figure 6 , based on the current weight ratio, home assessment data and inspection assessment data, the initial review time of the target object is adjusted to obtain a recommended review time, the steps include:

[0131] S710: Perform health prediction on the target object based on the current weight ratio, home assessment data and examination assessment data to obtain a health prediction result.

[0132] S720: Adjust the initial review time of the target object according to the health prediction result to obtain a recommended review time.

[0133] Among them, the health prediction results can reflect the future trend of the elderly's health. For example, the health prediction results can be the future development trend of the elderly's cardiovascular and cerebrovascular diseases, such as the probability of the elderly developing serious cardiovascular and cerebrovascular diseases in the future, described by any of low, medium, and high.

[0134] Specifically, the health prediction of the target object can be performed through the large model based on the current weight ratio, home assessment data and inspection assessment data. For example, when the ratio of the inspection weight to the home weight is greater than the preset threshold, the first health prediction model is used for health prediction; when the ratio of the inspection weight to the home weight is less than the preset threshold, the second health prediction model is used for health prediction.

[0135] After the prediction is completed, the initial review time of the target object can be adjusted according to the health prediction result. For example, when the health prediction result is "low", it means that the probability of the target object developing serious cardiovascular and cerebrovascular diseases in the future is low, and the condition has improved compared to the medical examination time. It can be recommended to extend the initial review time by 30% as the recommended review time; when the health prediction result is "medium", it means that the probability of the target object developing serious cardiovascular and cerebrovascular diseases in the future is at a medium level, and there is no significant change in the condition compared to the medical examination time. It is recommended to maintain the initial review time unchanged; when the health prediction result is "high", it means that the probability of the target object developing serious cardiovascular and cerebrovascular diseases in the future is high, and the condition has worsened compared to the medical examination time. It can be recommended that the target object arrange a review as soon as possible.

[0136] In the above embodiment, since the medical examination data reflects the health status of the elderly at the time of the medical examination, and the home health data is continuous and real-time updated information, health predictions are dynamically made based on weight ratios, making the health prediction results more accurate, thereby providing more accurate recommended review times through more accurate health prediction results.

[0137] In some embodiments, see Figure 7 , based on the current weight ratio, home assessment data and examination assessment data, the health prediction of the target object is carried out to obtain the health prediction results, including:

[0138] S810: If the current weight ratio is greater than a preset threshold, select the first health prediction model, input the home assessment data and the examination assessment data into the first health prediction model for health prediction, and obtain a corresponding health prediction result.

[0139] S820: If the current weight ratio is less than or equal to the preset threshold, select the second health prediction model, input the home assessment data and the examination assessment data into the second health prediction model for health prediction, and obtain the corresponding health prediction result.

[0140] The first time Tf between the latest monitoring time and the initial review time is the time to review. The second time Tp between the latest monitoring time and the medical examination time is the time to be examined. The ratio between the time to review and the time to be examined is the current weight ratio.

[0141] When the current weight ratio is greater than the preset threshold, the time until reexamination is greater than the time already examined. At this time, the weight of the examination and evaluation data is greater than the weight of the home evaluation data, and the first health prediction model is applicable to this situation.

[0142] When the current weight ratio is less than or equal to the preset threshold, the time until reexamination is less than the time already examined. At this time, the weight of the examination and evaluation data is less than the weight of the home evaluation data, and the second health prediction model is applicable to this situation.

[0143] The second health prediction model is structurally identical to the first health prediction model, but the model parameters are different. Exemplarily, the first health prediction model includes a convolutional layer, a nonlinear unit, a pooling layer, and a fully connected layer. Among them, the convolutional layer is used to extract features; the nonlinear unit is used to correct the features; the pooling layer (such as maximum pooling, average pooling, or sum pooling) is used to reduce the dimension of the corrected features; the fully connected layer is used to comprehensively analyze the reduced-dimensional features and output the probability of the target object developing severe cardiovascular and cerebrovascular diseases through an activation function.

[0144] Specifically, when predicting, the model selection is first performed. Specifically, if the current weight ratio is greater than the preset threshold, the first health prediction model is selected; if the current weight ratio is less than the preset threshold, the second health prediction model is selected. Furthermore, the home assessment data and the examination assessment data are input into the selected model for health prediction to obtain the corresponding health prediction results. Finally, according to the health prediction results of the target object, the recommended review time is output.

[0145] In the above embodiment, a first health prediction model and a second health prediction model are pre-trained for different time periods, that is, between the medical examination time and the initial review time, if the latest monitoring time is closer to the medical examination time, the first health prediction model is selected; if the latest monitoring time is closer to the initial review time, the second health prediction model is selected; this achieves that at different times, by using a more suitable first health prediction model or a second health prediction model, more accurate health prediction results corresponding to different times are obtained.

[0146] See also Figure 8 In this embodiment, a home health care system 900 for the elderly is also provided. The home health care system 900 for the elderly includes:

[0147] The home feature extraction module 910 is used to monitor the sleep process of the target object at home through the millimeter wave radar, obtain the home health data of the target object, and perform feature extraction based on the home health data to obtain the home assessment data of the target object.

[0148] The examination feature extraction module 920 is used to obtain the medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain the examination evaluation data of the target object; wherein the medical examination data corresponds to the medical examination time, and the medical examination time is the time when the medical examination of the target object occurs.

[0149] The dynamic weight matching module 930 is used to perform dynamic weight matching according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data.

[0150] The review time adjustment module 940 is used to adjust the initial review time of the target object based on the current weight ratio, home assessment data and examination assessment data to obtain a recommended review time; wherein the initial review time is the time when the target object mentioned in the medical examination data needs to be reviewed.

[0151] In some implementations, the medical examination data includes examination image data and report text data, and the report text data is used to describe the health status of the target object; the examination feature extraction module 920 further includes:

[0152] A convolution processing unit, used for performing convolution processing on the inspection image data to obtain image feature data;

[0153] A vector representation unit, used for performing vector representation on the report text data to obtain text feature data;

[0154] A feature stacking unit, used for placing the image feature data between two text feature data for feature stacking to obtain inspection feature data;

[0155] The health assessment unit is used to perform health assessment on the target object based on the inspection feature data to obtain inspection assessment data.

[0156] In some embodiments, the medical examination data includes report text data, and dynamic weight matching is performed according to the medical examination time and the latest monitoring time of the home health data. The dynamic weight matching module 930 also includes:

[0157] An initial review time extraction unit, used for extracting the initial review time from the report text data;

[0158] a duration determination unit, used to determine a first duration between the latest monitoring time and the initial review time, and a second duration between the latest monitoring time and the medical examination time;

[0159] The dynamic weight matching unit is used to perform dynamic weight matching according to the first duration and the second duration to obtain a current weight matching.

[0160] In some embodiments, the dynamic weight matching module 930 also includes a current weight matching unit, and the current weight matching is the ratio between the home weight of the home assessment data and the inspection weight of the inspection assessment data; wherein, if the first duration is greater than the second duration, the inspection weight is greater than the home weight; if the first duration is less than the second duration, the inspection weight is less than the home weight.

[0161] In some implementations, the home feature extraction module 910 further includes:

[0162] A monitoring unit is used to monitor the breathing, heart rate and blood pressure of the target object through millimeter wave radar during the target object's sleep at home, so as to obtain the target object's breathing time series data, heart rate time series data and blood pressure time series data;

[0163] The data construction unit is used to construct multi-channel home health data using respiratory time series data, heart rate time series data and blood pressure time series data.

[0164] In some implementations, the home feature extraction module 910 further includes:

[0165] A breathing feature extraction unit, used to extract features from the breathing time series data to obtain breathing evaluation data of the target object;

[0166] A heartbeat feature extraction unit is used to extract features from the heartbeat time series data to obtain the heartbeat evaluation data of the target object;

[0167] A blood pressure feature extraction unit, used to extract features from the blood pressure time series data to obtain blood pressure assessment data of the target object;

[0168] The weighted processing unit is used to perform weighted processing on the respiratory assessment data, the heartbeat assessment data and the blood pressure assessment data to obtain the health status data of the target object as home assessment data.

[0169] In some implementations, the review time adjustment module 940 further includes:

[0170] A health prediction unit is used to predict the health of the target object based on the current weight ratio, home assessment data and examination assessment data to obtain a health prediction result;

[0171] The review time adjustment unit is used to adjust the initial review time of the target object according to the health prediction result to obtain a recommended review time.

[0172] In some embodiments, the review time adjustment module 940 is also used to select the first health prediction model if the current weight ratio is greater than a preset threshold, and input the home assessment data and the examination assessment data into the first health prediction model for health prediction to obtain a corresponding health prediction result; if the current weight ratio is less than or equal to the preset threshold, select the second health prediction model, and input the home assessment data and the examination assessment data into the second health prediction model for health prediction to obtain a corresponding health prediction result.

[0173] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0174] The elderly home health care system in this embodiment is presented in the form of functional units, where the units refer to ASIC (Application Specific Integrated Circuit) circuits, processors and memories that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0175] In some embodiments, see Fig. 9 The elderly home health care system includes a millimeter wave monitoring device and a computer device connected to the millimeter wave monitoring device for communication.

[0176] The millimeter wave monitoring device 901 is used to monitor the target object's sleep process at home and obtain the target object's home health data.

[0177] Computer device 902 is used to perform feature extraction based on home health data to obtain home assessment data of the target object; obtain medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain examination assessment data of the target object; perform dynamic weight matching according to the medical examination time corresponding to the medical examination data and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the examination assessment data; adjust the initial review time of the target object based on the current weight matching, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the medical examination time is the time when the medical examination of the target object occurs; the initial review time is the time when the target object mentioned in the medical examination data needs to be reviewed.

[0178] See also Fig.10 , Fig.10 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application, such as Fig.10As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Fig.10 A processor 10 is taken as an example.

[0179] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0180] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0181] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0182] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0183] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0184] The embodiment of the present application also provides a computer-readable storage medium. The above method according to the embodiment of the present application can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0185] The embodiment of the present application provides a computer program product, which includes computer instructions, which are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method of any embodiment of the present application.

[0186] Although the embodiments of the present application are described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations are all within the scope defined by the appended claims.

[0187] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0188] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0189] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0190] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate a device for implementing the functions specified in one process or multiple processes in the flowchart and / or one box or multiple boxes in the block diagram.

[0191] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0192] These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0193] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0194] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0195] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

[0196] Although the embodiments of the present application have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for processing health data of the elderly at home, characterized in that: The method comprises: The target object's sleep process at home is monitored by a millimeter wave radar to obtain the target object's home health data, and feature extraction is performed based on the home health data to obtain the target object's home assessment data; Acquire medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain examination evaluation data of the target object; wherein the medical examination data corresponds to a medical examination time, which is the time when the medical examination of the target object occurs, and the medical examination data includes report text data; Dynamic weight matching is performed according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the inspection assessment data; specifically: determining a first duration between the latest monitoring time and the initial review time, and a second duration between the latest monitoring time and the medical examination time; performing dynamic weight matching according to the first duration and the second duration to obtain the current weight matching; wherein the current weight matching is the ratio between the home weight of the home assessment data and the inspection weight of the inspection assessment data; if the first duration is greater than the second duration, the inspection weight is greater than the home weight; if the first duration is less than the second duration, the inspection weight is less than the home weight; The initial review time of the target object is adjusted based on the current weight ratio, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the initial review time is the time recommended by the doctor for the target object to be reviewed, and is extracted from the report text data.

2. The method according to claim 1, characterized in that The medical examination data includes examination image data and report text data, and the report text data is used to describe the health condition of the target object.

3. The method according to claim 2, characterized in that The extracting features based on the medical examination data to obtain the examination evaluation data of the target object includes: Performing convolution processing on the inspection image data to obtain image feature data; Performing vector representation on the report text data to obtain text feature data; Placing the image feature data between the two text feature data for feature stacking to obtain inspection feature data; A health assessment is performed on the target object based on the examination feature data to obtain the examination assessment data.

4. The method according to any one of claims 1 to 3, characterized in that: The method of monitoring the sleep process of the target object at home by using the millimeter wave radar to obtain the home health data of the target object includes: During the target object's sleep at home, the millimeter wave radar is used to perform breathing monitoring, heart rate monitoring and blood pressure monitoring respectively to obtain breathing time series data, heart rate time series data and blood pressure time series data of the target object; The multi-channel home health data is constructed using the respiratory time series data, the heartbeat time series data and the blood pressure time series data.

5. The method according to claim 4, characterized in that The extracting features based on the home health data to obtain the home assessment data of the target object includes: Extracting features from the respiratory time series data to obtain respiratory assessment data of the target object; Extracting features from the heartbeat time series data to obtain heartbeat evaluation data of the target object; Performing feature extraction on the blood pressure time series data to obtain blood pressure assessment data of the target object; The breathing assessment data, the heartbeat assessment data and the blood pressure assessment data are weighted to obtain the health status data of the target object as the home assessment data.

6. The method according to claim 1, characterized in that The adjusting the initial review time of the target object based on the current weight ratio, the home assessment data and the inspection assessment data to obtain a recommended review time includes: Performing health prediction on the target object based on the current weight ratio, the home assessment data and the examination assessment data to obtain a health prediction result; The initial review time of the target object is adjusted according to the health prediction result to obtain a recommended review time.

7. The method according to claim 6, characterized in that The health prediction of the target object based on the current weight ratio, the home assessment data and the inspection assessment data to obtain a health prediction result includes: If the current weight ratio is greater than a preset threshold, the first health prediction model is selected, and the home assessment data and the examination assessment data are input into the first health prediction model for health prediction to obtain a corresponding health prediction result.

8. The method according to claim 6, characterized in that The health prediction of the target object based on the current weight ratio, the home assessment data and the inspection assessment data to obtain a health prediction result includes: If the current weight ratio is less than or equal to a preset threshold, a second health prediction model is selected, and the home assessment data and the examination assessment data are input into the second health prediction model for health prediction to obtain a corresponding health prediction result.

9. A home health care system for the elderly, characterized in that: The system comprises: A home feature extraction module is used to monitor the sleep process of the target object at home through a millimeter wave radar to obtain the home health data of the target object, and perform feature extraction based on the home health data to obtain the home assessment data of the target object; An examination feature extraction module is used to obtain the medical examination data of the target object, and perform feature extraction based on the medical examination data to obtain the examination evaluation data of the target object; wherein the medical examination data corresponds to the medical examination time, the medical examination time is the time when the medical examination of the target object occurs, and the medical examination data includes report text data; A dynamic weight matching module is used to perform dynamic weight matching according to the medical examination time and the latest monitoring time of the home health data to obtain the current weight matching between the home assessment data and the inspection and assessment data; specifically: determine the first duration between the latest monitoring time and the initial review time, and the second duration between the latest monitoring time and the medical examination time; perform dynamic weight matching according to the first duration and the second duration to obtain the current weight matching; wherein the current weight matching is the ratio between the home weight of the home assessment data and the inspection weight of the inspection and assessment data; if the first duration is greater than the second duration, the inspection weight is greater than the home weight; if the first duration is less than the second duration, the inspection weight is less than the home weight; A review time adjustment module is used to adjust the initial review time of the target object based on the current weight ratio, the home assessment data and the examination assessment data to obtain a recommended review time; wherein the initial review time is the time when the target object needs to be reviewed mentioned in the medical examination data, and is extracted from the report text data.

10. A home health care system for the elderly, characterized in that: The system includes a millimeter wave monitoring device and a computer device communicatively connected to the millimeter wave monitoring device; The millimeter wave monitoring device is used to monitor the target object's sleep process at home to obtain the target object's home health data; The computer device is used to extract features based on the home health data to obtain home assessment data of the target object; Acquiring medical examination data of the target object, and performing feature extraction based on the medical examination data to obtain examination evaluation data of the target object; Determine a first duration between the latest monitoring time of the home health data and the initial review time, and a second duration between the latest monitoring time and the medical examination time corresponding to the medical examination data; Performing dynamic weight matching according to the first duration and the second duration to obtain a current weight matching between the home assessment data and the inspection assessment data; The initial review time of the target object is adjusted based on the current weight ratio, the home assessment data and the inspection and assessment data to obtain a recommended review time; wherein the medical examination time is the time when the medical examination of the target object occurs; the initial review time is the time when the target object needs to be reviewed mentioned in the medical examination data, and is extracted from the report text data included in the medical examination data; the current weight ratio is the ratio between the home weight of the home assessment data and the inspection weight of the inspection and assessment data; if the first time length is greater than the second time length, the inspection weight is greater than the home weight; if the first time length is less than the second time length, the inspection weight is less than the home weight.

Images