Remote monitoring and nursing early warning system based on deep learning

Abstract

The invention, which relates to the technical field of remote monitoring and early warning, discloses a remote monitoring and nursing early warning system based on deep learning, comprising a remote monitoring and nursing early warning platform. The remote monitoring nursing early warning platform is in communication connection with a data acquisition module, a feature extraction module, a nursing early warning evaluation module, an early warning grading response module and a user interaction module, and all the modules are in electric signal connection; the data acquisition module acquires a physiological data set of a patient in real time through a remote monitoring tool. According to the remote monitoring nursing early warning system based on deep learning, a large amount of patient physiological data is analyzed and processed through application of the deep learning technology, potential rules and modes are mined from a large amount of data by learning normal physiological modes and identifying abnormal modes, the health condition change of the patient is analyzed, and the early warning effect is improved. The accuracy of nursing monitoring of the patient is improved, and the subtle change of the illness state of the patient is better captured, so that earlier intervention and treatment are realized.

Description

Technical Field

[0001] The present invention relates to the field of remote monitoring and early warning technology, and in particular to a remote monitoring, nursing and early warning system based on deep learning. Background Art

[0002] With the continuous development of medical informatization and intelligence, telemedicine technology has become an important trend in the medical field. Deep learning, as an important technology of artificial intelligence, can process and analyze large amounts of medical data, extract useful information, and provide support for medical decision-making. Combined with deep learning for early warning of remote monitoring and care, it can achieve accurate analysis and prediction of patients' physiological data. Once an abnormality is found, an early warning will be issued immediately to remind patients and medical staff to take necessary measures to improve the accuracy and timeliness of the early warning.

[0003] In the existing technology, when remotely monitoring the physiological data of patients, the response measures required are different due to the different severity of abnormal conditions. Therefore, how to evaluate the severity of the patient's abnormal condition and issue corresponding warning signals so that medical staff can take appropriate response measures is the problem we need to solve.

[0004] Based on the above problems, it is urgent to propose a remote monitoring and nursing early warning system based on deep learning. Summary of the invention

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a remote monitoring nursing early warning system based on deep learning, including a remote monitoring nursing early warning platform, wherein the remote monitoring nursing early warning platform is communicatively connected with a data acquisition module, a feature extraction module, a nursing early warning evaluation module, an early warning graded response module and a user interaction module, wherein the modules are connected by electrical signals;

[0006] The data acquisition module collects the patient's physiological data in real time through remote monitoring tools (smart wearable devices, sensors), and pre-processes the collected raw data to improve the quality and availability of the data;

[0007] The feature extraction module extracts feature information reflecting the patient's physiological status and health trend from the preprocessed data, constructs a nursing warning feature set, and provides a key basis for subsequent warning evaluation;

[0008] The nursing warning evaluation module uses a deep learning algorithm to construct a nursing warning model based on the feature information in the nursing warning feature set, predicts and classifies the patient's health status, evaluates the patient's health status, and assesses the patient's physiological health trend;

[0009] The early warning graded response module sets different early warning levels according to the results of the nursing early warning evaluation module, matches the corresponding early warning response mechanism, and performs quantitative analysis on the prediction results to ensure the accuracy and timeliness of the early warning;

[0010] The user interaction module provides a user interaction interface to display monitoring results, warning information and health advice, so that medical staff and patients can easily view monitoring data, warning information and response measures.

[0011] Preferably, in the data acquisition module, the process of collecting and preprocessing the patient's physiological data includes:

[0012] Connect the data acquisition module with the remote monitoring tool to monitor the patient's physiological data, where the remote monitoring tool is a smart wearable device and sensor to obtain the patient's physiological data, including heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation, blood sugar, blood lipids, and liver and kidney function indicators;

[0013] Perform pre-processing operations such as cleaning, denoising, and formatting of collected patient physiological data. Remove invalid, duplicate, or abnormal values ​​in the data through data cleaning to ensure data accuracy and consistency. Perform denoising on the data to improve data quality, identify and remove noise components in the data, and format the data to convert the data into a unified format and standard.

[0014] A data warehouse is built in the remote monitoring, nursing and early warning platform, and the pre-processed patient physiological data is stored in the data warehouse for subsequent analysis and use.

[0015] Preferably, in the feature extraction module, the process of constructing the nursing warning feature set includes:

[0016] Access the data warehouse, extract the patient's physiological data from the data warehouse according to the patient ID, and perform feature analysis on the patient's physiological data to extract feature information reflecting the patient's physiological state and health trend, which are physiological features and biochemical features. The physiological features include heart rate, blood pressure, body temperature, respiratory rate, and blood oxygen saturation features, and the biochemical features include blood sugar, blood lipids, and liver and kidney function index features;

[0017] Determine the standard threshold of each feature information based on historical data and the patient's personalized care standards, and distinguish abnormal features from normal features based on the standard threshold;

[0018] Combined with the determined standard threshold of physiological characteristics, statistical analysis is performed on the extracted physiological characteristic data, the physiological state trend index is calculated, and the trend of changes in the patient's physiological state is analyzed; combined with the determined standard threshold of biochemical characteristics, the extracted biochemical characteristic data is analyzed, the biochemical state evaluation index is calculated, and whether the patient's biochemical indicators are within the normal range is evaluated;

[0019] The extracted feature information and the set standard threshold are integrated to construct a nursing warning feature set, which contains all the features that have an important impact on the warning evaluation and their corresponding standard thresholds.

[0020] Preferably, the standard threshold range of the physiological characteristic is specifically:

[0021] Normal range of heart rate: The resting heart rate of an adult is 60-100 beats / minute; Normal range of blood pressure: Systolic pressure (high pressure) 90-139 mmHg, Diastolic pressure (low pressure) 60-89 mmHg; Normal temperature: 36.3℃-37.2℃; Normal range of respiratory rate: 12-20 times / minute for a normal adult in a resting state; Normal range of blood oxygen saturation: 95%-100%;

[0022] The biochemical characteristic standard threshold range is specifically:

[0023] Normal range of blood sugar: The normal range of fasting blood sugar is 3.9-6.1mmol / L, the blood sugar 1 hour after a meal is 6.7-9.4mmol / L, and the maximum is no more than 11.1mmol / L. The blood sugar 2 hours after a meal is less than 7.8mmol / L; Normal range of blood lipids: The normal range of total cholesterol is less than 5.2mmol / L, and the normal range of triglycerides is less than 1.7mmol / L. For people at risk of cardiovascular disease, low-density lipoprotein cholesterol is controlled below 2.6mmol / L, and the appropriate range for normal people is less than 3.4mmol / L. High-density lipoprotein cholesterol is greater than 1.0mmol / L for men and greater than 1.3mmol / L for women;

[0024] Normal range of liver and kidney function indicators: For liver function, the normal range of alanine aminotransferase (ALT) is 0-40U / L, the normal range of aspartate aminotransferase (AST) is 0-40U / L, the normal range of total bilirubin (TBIL) is 3.4-17.1μmol / L, for kidney function, blood creatinine (Scr): the normal range for males is 53-106μmol / L, for females is 44-97μmol / L, the normal range of urea nitrogen (BUN) is 2.9-7.2mmol / .

[0025] Preferably, the calculation formula of the physiological state trend index is:

[0026] ;

[0027] in, It is a physiological state trend index, which is used to evaluate the changing trend of the patient's overall physiological state. is the actual measured value of the ith physiological characteristic, representing the actual measured value of each physiological characteristic, Corresponding to heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, respectively. is the standard threshold of the ith physiological feature, representing the corresponding standard threshold of the physiological feature, Corresponding to the standard thresholds of heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, It is the base of the natural logarithm and is used in the exponential function to enhance the impact of feature deviation threshold. The value range is ;

[0028] The calculation formula of the biochemical status assessment index is:

[0029] ;

[0030] in, It is a biochemical status assessment index, which is used to assess whether the patient's biochemical indicators are within the normal range. is the actual measured value of the jth biochemical characteristic, representing the actual measured value of each biochemical characteristic, Corresponding to blood sugar, blood lipids and liver and kidney function indicators, is the standard threshold of the jth biochemical feature, represents the corresponding biochemical feature standard threshold, Corresponding to the standard thresholds of blood sugar, blood lipids and liver and kidney function indicators, The value range is .

[0031] Preferably, in the nursing early warning evaluation module, the evaluation process of the patient's physiological health trend includes:

[0032] Comprehensively consider the feature information in the nursing warning feature set, divide the nursing warning feature set into a training set and a test set. The training set is used for model training and learning, and the test set is used to evaluate the performance of the model. According to the characteristics of the nursing warning feature set, the convolutional neural network structure is designed.

[0033] Use the training set data combined with the convolutional neural network to build a nursing early warning model, learn and identify the patient's health status pattern, adjust the model's parameters through the forward propagation and back propagation algorithms to minimize the loss function, and use the test set data to evaluate the performance of the model. According to the verification results, adjust the model's parameters and structure to improve the generalization ability of the model. In the nursing early warning model, the input layer accepts the physiological data sequence, and the output layer outputs the patient's health status prediction results;

[0034] The nursing early warning model is used in combination with the patient's physiological data to evaluate the patient's health status, and the early warning evaluation coefficient is calculated in combination with the physiological status trend index and biochemical status assessment index to comprehensively evaluate the patient's physiological health status.

[0035] Preferably, the calculation formula of the early warning evaluation coefficient is:

[0036] ;

[0037] in, It is the early warning evaluation coefficient, which is used to comprehensively evaluate the patient's physiological health status. is the weight coefficient of the physiological state trend index, is the weight coefficient of the biochemical status assessment index, and Used for balance and right The contribution of It is a physiological state trend index, which quantifies the patient's physiological state. It is a biochemical status assessment index that quantifies the patient's biochemical status. is the threshold parameter used to adjust The sensitivity is set according to actual clinical needs. The value range is between 0 and 1.

[0038] Preferably, in the early warning graded response module, the process of setting the early warning level and matching the early warning response mechanism includes:

[0039] Receive the output results from the nursing early warning evaluation module, including the patient's health status prediction, early warning evaluation coefficient, and related physiological and biochemical characteristics data;

[0040] According to the results of the nursing early warning evaluation module, different early warning levels are set, and the early warning levels are divided into low warning level, medium warning level, high warning level and emergency warning level according to the early warning evaluation coefficient. Each early warning level corresponds to different risk levels and urgency levels.

[0041] Analyze historical data to match corresponding warning thresholds for each warning level, and design corresponding response mechanisms for each warning level, including notifying medical staff, initiating emergency treatment procedures, and arranging further examinations or treatments.

[0042] Preferably, the plurality of warning levels correspond to the plurality of warning thresholds, specifically satisfying the following relationship:

[0043] Low alert level: , indicating that the patient is in good physiological condition; the patient's physiological and biochemical status is close to normal, the risk is low, routine monitoring and regular evaluation are required, and no immediate intervention is required;

[0044] Medium warning level: , indicating that the patient's physiological state is slightly abnormal; the patient's physiological or biochemical state is slightly abnormal, which requires attention, increased monitoring frequency, and consideration of further diagnostic tests;

[0045] High alert level: , indicating that the patient's physiological state is moderately abnormal; the patient's physiological or biochemical state is moderately abnormal and may require medical intervention. Immediately assess the patient's condition and prepare for medical intervention;

[0046] Emergency warning level: , indicating that the patient's physiological state is seriously abnormal and requires emergency intervention. The patient's physiological or biochemical state is seriously abnormal and requires emergency medical intervention. Immediately start the emergency treatment process and notify medical staff for emergency treatment;

[0047] in, is the early warning evaluation coefficient, is the lower threshold corresponding to the low warning level and the upper threshold corresponding to the medium warning level, is the lower threshold corresponding to the medium warning level and the upper threshold corresponding to the high warning level, is the lower threshold corresponding to the high warning level and the upper threshold corresponding to the emergency warning level, , , .

[0048] Preferably, the user interaction module specifically includes:

[0049] According to the needs of the remote monitoring and nursing early warning system, a user interaction interface including monitoring data display area, early warning information prompt area, and health advice area is designed;

[0050] The user interaction module receives real-time data from the remote monitoring nursing early warning platform, including the patient's physiological characteristics and biochemical characteristics monitoring results, and updates the patient's physiological data to the user interaction interface in real time;

[0051] When the warning condition is triggered, the user interaction module receives the warning information and displays it in the warning information prompt area on the user interaction interface, and issues a flashing reminder. The user clicks on the warning information to view detailed warning details, including the warning level, triggering conditions, and recommended response measures;

[0052] Provide personalized health advice and intervention measures based on the patient's monitoring data and warning level to help patients take appropriate actions to improve their health, and display health advice in the health advice area on the interface.

[0053] The present invention provides a remote monitoring and nursing early warning system based on deep learning. It has the following beneficial effects:

[0054] 1. This remote monitoring and nursing early warning system based on deep learning analyzes and processes a large amount of patient physiological data through the application of deep learning technology, and mines potential laws and patterns from a large amount of data by learning normal physiological patterns and identifying abnormal patterns, analyzing changes in patients' health status, improving the accuracy of patient nursing monitoring, and better capturing subtle changes in patients' conditions, thereby achieving earlier intervention and treatment.

[0055] 2. This remote monitoring and nursing early warning system based on deep learning receives and processes monitoring data in real time. Once an abnormality is detected, it immediately triggers the early warning mechanism, enabling medical staff to respond quickly, greatly reducing the risks caused by delayed discovery of the disease and providing patients with more timely and effective nursing services. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a module diagram of a remote monitoring and nursing early warning system based on deep learning of the present invention;

[0057] Figure 2 A flowchart for constructing a nursing warning feature set of the present invention;

[0058] Figure 3 A flow chart for evaluating the physiological health trend of a patient of the present invention;

[0059] Figure 4 This is a flow chart for setting the warning level and matching the warning response mechanism of the present invention. DETAILED DESCRIPTION

[0060] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

[0061] The first embodiment, as Figure 1 , Figure 2 As shown, the present invention provides a technical solution: a remote monitoring nursing early warning system based on deep learning, including a remote monitoring nursing early warning platform, the remote monitoring nursing early warning platform is communicatively connected with a data acquisition module, a feature extraction module, a nursing early warning evaluation module, an early warning graded response module and a user interaction module, wherein the modules are connected by electrical signals;

[0062] The data acquisition module collects the patient's physiological data in real time through remote monitoring tools (smart wearable devices, sensors), and pre-processes the collected raw data to improve the quality and availability of the data. The data acquisition module is connected to the remote monitoring tool to monitor the patient's physiological data, where the remote monitoring tool is a smart wearable device and a sensor to obtain the patient's physiological data, including heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation, blood sugar, blood lipids, and liver and kidney function indicators. The collected patient's physiological data is cleaned, denoised, and formatted for pre-processing operations. The invalid, repeated, or abnormal values ​​in the data are removed through data cleaning to ensure the accuracy and consistency of the data. The data is denoised to improve the data quality, and the noise components in the data are identified and removed. The data is formatted to convert the data into a unified format and standard to make the data easier to process and analyze, including converting the data into a specific format (such as CSV, JSON, etc.), and standardizing or normalizing the data to ensure that the data is comparable and consistent in different scenarios. A data warehouse is built in the remote monitoring nursing early warning platform, and the pre-processed patient physiological data is stored in the data warehouse for subsequent analysis and use;

[0063] The feature extraction module extracts feature information reflecting the patient's physiological state and health trend from the preprocessed data, constructs a nursing early warning feature set, and provides a key basis for subsequent early warning evaluation. The data warehouse is accessed, and the patient's physiological data is extracted from the data warehouse according to the patient ID. The feature analysis of the patient's physiological data is performed to extract feature information reflecting the patient's physiological state and health trend, which are physiological features and biochemical features. Among them, the physiological features include heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation features, and the biochemical features include blood sugar, blood lipids and liver and kidney function index features. According to historical data and the patient's personalized nursing standards, the standard threshold of each feature information is determined, and abnormal features and normal features are distinguished based on the standard threshold. Combined with the determined standard threshold of physiological features, the extracted physiological feature data is statistically analyzed, the physiological state trend index is calculated, and the patient's physiological state change trend is analyzed. Combined with the determined standard threshold of biochemical features, the extracted biochemical feature data is analyzed, and the biochemical state evaluation index is calculated to evaluate whether the patient's biochemical indicators are within the normal range. The extracted feature information and the set standard threshold are integrated to construct a nursing early warning feature set. The nursing early warning feature set contains all features that have an important impact on the early warning evaluation and their corresponding standard thresholds.

[0064] Furthermore, the standard threshold range of physiological characteristics is specifically:

[0065] Normal range of heart rate: The resting heart rate of an adult is 60-100 beats / minute; Normal range of blood pressure: Systolic pressure (high pressure) 90-139 mmHg, Diastolic pressure (low pressure) 60-89 mmHg; Normal temperature: 36.3℃-37.2℃; Normal range of respiratory rate: 12-20 times / minute for a normal adult in a resting state; Normal range of blood oxygen saturation: 95%-100%;

[0066] The standard threshold ranges for biochemical characteristics are as follows:

[0067] Normal range of blood sugar: The normal range of fasting blood sugar is 3.9-6.1mmol / L, the blood sugar 1 hour after a meal is 6.7-9.4mmol / L, and the maximum is no more than 11.1mmol / L. The blood sugar 2 hours after a meal is less than 7.8mmol / L; Normal range of blood lipids: The normal range of total cholesterol is less than 5.2mmol / L, and the normal range of triglycerides is less than 1.7mmol / L. For people at risk of cardiovascular disease, low-density lipoprotein cholesterol is controlled below 2.6mmol / L, and the appropriate range for normal people is less than 3.4mmol / L. High-density lipoprotein cholesterol is greater than 1.0mmol / L for men and greater than 1.3mmol / L for women;

[0068] Normal range of liver and kidney function indicators: For liver function, the normal range of alanine aminotransferase (ALT) is 0-40U / L, the normal range of aspartate aminotransferase (AST) is 0-40U / L, the normal range of total bilirubin (TBIL) is 3.4-17.1μmol / L, for kidney function, serum creatinine (Scr): the normal range for males is 53-106μmol / L, for females is 44-97μmol / L, the normal range of urea nitrogen (BUN) is 2.9-7.2mmol / L;

[0069] Furthermore, the calculation formula of the physiological state trend index is:

[0070] ;

[0071] in, It is a physiological state trend index, which is used to evaluate the changing trend of the patient's overall physiological state. is the actual measured value of the ith physiological characteristic, representing the actual measured value of each physiological characteristic, Corresponding to heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, respectively. is the standard threshold of the ith physiological feature, representing the corresponding standard threshold of the physiological feature, Corresponding to the standard thresholds of heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, It is the base of the natural logarithm and is used in the exponential function to enhance the impact of feature deviation threshold. The value range is , when all eigenvalues ​​are within the standard threshold range, Close to 0, it means that the patient's physiological state is stable. When any characteristic value deviates from its standard threshold, Increase, indicating that the patient's physiological state tends to be abnormal, the index part The influence of feature deviation threshold is amplified, making More sensitive to the combined impact of multiple indicators;

[0072] The calculation formula of biochemical status assessment index is:

[0073] ;

[0074] in, It is a biochemical status assessment index, which is used to assess whether the patient's biochemical indicators are within the normal range. is the actual measured value of the jth biochemical characteristic, representing the actual measured value of each biochemical characteristic, Corresponding to blood sugar, blood lipids and liver and kidney function indicators, is the standard threshold of the jth biochemical feature, represents the corresponding biochemical feature standard threshold, Corresponding to the standard thresholds of blood sugar, blood lipids and liver and kidney function indicators, The value range is , when all eigenvalues ​​are within the standard threshold range, Close to 0, indicating that the patient's biochemical status is normal. When any characteristic value deviates from its standard threshold, Increase, indicating that the patient's biochemical status tends to be abnormal, the square term in the formula The influence of feature deviation threshold is enhanced, making More sensitive to the combined impact of multiple indicators;

[0075] The nursing warning evaluation module uses deep learning algorithms to build a nursing warning model based on the feature information in the nursing warning feature set, predict and classify the patient's health status, evaluate the patient's health status, and assess the patient's physiological health trend;

[0076] The early warning graded response module sets different early warning levels according to the results of the nursing early warning evaluation module, matches the corresponding early warning response mechanism, and conducts quantitative analysis on the prediction results to ensure the accuracy and timeliness of the early warning;

[0077] The user interaction module provides a user interaction interface to display monitoring results, early warning information and health recommendations, allowing medical staff and patients to easily view monitoring data, early warning information and response measures.

[0078] The second embodiment is based on the first embodiment. Figure 3 , Figure 4 As shown in the figure, in the nursing early warning evaluation module, the evaluation process of the patient's physiological health trend includes:

[0079] The feature information in the comprehensive nursing warning feature set is divided into a training set and a test set. The training set is used for model training and learning, and the test set is used to evaluate the performance of the model. According to the characteristics of the nursing warning feature set, a convolutional neural network structure is designed. The nursing warning model is constructed by combining the training set data with the convolutional neural network to learn and identify the patient's health status pattern. The parameters of the model are adjusted through the forward propagation and back propagation algorithms to minimize the loss function. The performance of the model is evaluated using the test set data. The parameters and structure of the model are adjusted according to the verification results to improve the generalization ability of the model. In the nursing warning model, the input layer accepts the physiological data sequence, and the output layer outputs the patient's health status prediction result. The nursing warning model is combined with the patient's physiological data to evaluate the patient's health status. The physiological state trend index and the biochemical state evaluation index are combined to calculate the warning evaluation coefficient to comprehensively evaluate the patient's physiological health status.

[0080] Furthermore, the calculation formula of the early warning evaluation coefficient is:

[0081] ;

[0082] in, It is the early warning evaluation coefficient, which is used to comprehensively evaluate the patient's physiological health status. is the weight coefficient of the physiological state trend index, is the weight coefficient of the biochemical status assessment index, and Used for balance and right The contribution of It is a physiological state trend index, which quantifies the patient's physiological state. It is a biochemical status assessment index that quantifies the patient's biochemical status. is the threshold parameter used to adjust The sensitivity is set according to actual clinical needs. The value range is between 0 and 1. and When they are within the normal range, Close to 1, it means the patient is in good physiological condition. or When it deviates from the normal range, Decrease indicates that the patient's physiological state tends to be abnormal. The closer it is to 0, the worse the patient's physiological health is, and more attention and intervention are needed. The closer the value is to 1, the more stable the patient's physiological state is. The lower the value, the more unstable the patient's physiological state is and the higher the risk;

[0083] In the early warning graded response module, the process of setting the early warning level and matching the early warning response mechanism includes:

[0084] Receive the output results from the nursing early warning evaluation module, including the patient's health status prediction, early warning evaluation coefficient, and related physiological and biochemical characteristics data. According to the results of the nursing early warning evaluation module, set different early warning levels. Combined with the early warning evaluation coefficient, the early warning levels are divided into low warning level, medium warning level, high warning level, and emergency warning level. Each early warning level corresponds to a different degree of risk and urgency. Analyze historical data to match the corresponding early warning threshold for each early warning level, and design a corresponding response mechanism for each early warning level, including notifying medical staff, initiating emergency treatment procedures, and arranging further examinations or treatments.

[0085] Furthermore, multiple warning levels correspond to multiple warning thresholds, which specifically satisfy the following relationship:

[0086] Low alert level: , indicating that the patient is in good physiological condition; the patient's physiological and biochemical status is close to normal, the risk is low, routine monitoring and regular evaluation are required, and no immediate intervention is required;

[0087] Medium warning level: , indicating that the patient's physiological state is slightly abnormal; the patient's physiological or biochemical state is slightly abnormal, which requires attention, increased monitoring frequency, and consideration of further diagnostic tests;

[0088] High alert level: , indicating that the patient's physiological state is moderately abnormal; the patient's physiological or biochemical state is moderately abnormal and may require medical intervention. Immediately assess the patient's condition and prepare for medical intervention;

[0089] Emergency warning level: , indicating that the patient's physiological state is seriously abnormal and requires emergency intervention. The patient's physiological or biochemical state is seriously abnormal and requires emergency medical intervention. Immediately start the emergency treatment process and notify medical staff for emergency treatment;

[0090] in, is the early warning evaluation coefficient, is the lower threshold corresponding to the low warning level and the upper threshold corresponding to the medium warning level, is the lower threshold corresponding to the medium warning level and the upper threshold corresponding to the high warning level, is the lower threshold corresponding to the high warning level and the upper threshold corresponding to the emergency warning level, , , ;

[0091] The user interaction module specifically includes:

[0092] According to the needs of the remote monitoring and nursing early warning system, a user interaction interface is designed, which includes a monitoring data display area, an early warning information prompt area, and a health advice area. The user interaction module receives real-time data from the remote monitoring and nursing early warning platform, including the patient's physiological characteristics and biochemical characteristics monitoring results, and updates the patient's physiological data to the user interaction interface in real time. When the early warning condition is triggered, the user interaction module receives the early warning information and displays it in the early warning information prompt area on the user interaction interface, and issues a flashing reminder. The user clicks on the early warning information to view detailed early warning details, including the warning level, triggering conditions, and recommended response measures. According to the patient's monitoring data and early warning level, personalized health advice and intervention measures are provided to help patients take appropriate actions to improve their health status, and health advice is displayed in the health advice area on the interface.

[0093] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. A remote monitoring and nursing early warning system based on deep learning, including a remote monitoring and nursing early warning platform, characterized in that: The remote monitoring nursing warning platform is communicatively connected with a data acquisition module, a feature extraction module, a nursing warning evaluation module, a warning classification response module and a user interaction module, wherein the modules are electrically connected with each other; The data acquisition module collects the patient's physiological data in real time through a remote monitoring tool and pre-processes the collected raw data; The feature extraction module extracts feature information reflecting the patient's physiological state and health trend from the preprocessed data to construct a nursing warning feature set; The nursing warning evaluation module uses a deep learning algorithm to construct a nursing warning model based on the feature information in the nursing warning feature set, predicts and classifies the patient's health status, evaluates the patient's health status, and assesses the patient's physiological health trend; The early warning graded response module sets different early warning levels according to the results of the nursing early warning evaluation module, and matches the corresponding early warning response mechanism; The user interaction module provides a user interaction interface to display monitoring results, warning information and health suggestions.

2. A remote monitoring and nursing early warning system based on deep learning according to claim 1, characterized in that: In the data acquisition module, the acquisition and preprocessing process of the patient's physiological data includes: Connect the data acquisition module with the remote monitoring tool to monitor the patient's physiological data, where the remote monitoring tool is a smart wearable device and sensor to obtain the patient's physiological data, including heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation, blood sugar, blood lipids, and liver and kidney function indicators; Perform pre-processing operations such as cleaning, denoising, and formatting on the collected patient physiological data. Remove invalid, duplicate, or abnormal values ​​from the data through data cleaning, denoise the data, format the data, and convert the data into a unified format and standard. A data warehouse is built in the remote monitoring, nursing and early warning platform, and the pre-processed patient physiological data is stored in the data warehouse.

3. A remote monitoring and nursing early warning system based on deep learning according to claim 2, characterized in that: In the feature extraction module, the process of constructing the nursing warning feature set includes: Access the data warehouse, extract the patient's physiological data from the data warehouse according to the patient ID, and perform feature analysis on the patient's physiological data to extract feature information reflecting the patient's physiological state and health trend, which are physiological features and biochemical features. The physiological features include heart rate, blood pressure, body temperature, respiratory rate, and blood oxygen saturation features, and the biochemical features include blood sugar, blood lipids, and liver and kidney function index features; Determine the standard threshold of each feature information based on historical data and the patient's personalized care standards, and distinguish abnormal features from normal features based on the standard threshold; Combined with the determined standard threshold of physiological characteristics, statistical analysis is performed on the extracted physiological characteristic data, the physiological state trend index is calculated, and the trend of changes in the patient's physiological state is analyzed; combined with the determined standard threshold of biochemical characteristics, the extracted biochemical characteristic data is analyzed, the biochemical state evaluation index is calculated, and whether the patient's biochemical indicators are within the normal range is evaluated; The extracted feature information and the set standard threshold are integrated to construct a nursing warning feature set.

4. A remote monitoring and nursing early warning system based on deep learning according to claim 3, characterized in that: The standard threshold range of the physiological characteristics is specifically: Normal range of heart rate: resting heart rate for adults is 60-100 beats / minute; normal range of blood pressure: systolic pressure 90-139 mmHg, diastolic pressure 60-89 mmHg; normal body temperature: 36.3℃-37.2℃; normal range of respiratory rate: 12-20 times / minute for normal adults in a resting state; normal range of blood oxygen saturation: 95%-100%; The biochemical characteristic standard threshold range is specifically: Normal range of blood sugar: The normal range of fasting blood sugar is 3.9-6.1mmol / L, the blood sugar 1 hour after a meal is 6.7-9.4mmol / L, and the blood sugar 2 hours after a meal is less than 7.8mmol / L; Normal range of blood lipids: The normal range of total cholesterol is less than 5.2mmol / L, and the normal range of triglycerides is less than 1.7mmol / L. For people at risk of cardiovascular disease, low-density lipoprotein cholesterol is controlled below 2.6mmol / L, and the appropriate range for normal people is less than 3.4mmol / L. High-density lipoprotein cholesterol is greater than 1.0mmol / L for men and greater than 1.3mmol / L for women; Normal range of liver and kidney function indicators: For liver function, the normal range of alanine aminotransferase is 0-40U / L, the normal range of aspartate aminotransferase is 0-40U / L, the normal range of total bilirubin is 3.4-17.1μmol / L, for kidney function, blood creatinine: the normal range for men is 53-106μmol / L, for women is 44-97μmol / L, the normal range of urea nitrogen is 2.9-7.2mmol / L.

5. A remote monitoring and nursing early warning system based on deep learning according to claim 4, characterized in that: The calculation formula of the physiological state trend index is: ； in, is the physiological state trend index, is the actual measured value of the ith physiological characteristic, representing the actual measured value of each physiological characteristic, Corresponding to heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, respectively. is the standard threshold of the ith physiological feature, representing the corresponding standard threshold of the physiological feature, Corresponding to the standard thresholds of heart rate, blood pressure, body temperature, respiratory rate and blood oxygen saturation, The value range is ; The calculation formula of the biochemical status assessment index is: ； in, is the biochemical status assessment index, is the actual measured value of the jth biochemical characteristic, representing the actual measured value of each biochemical characteristic, Corresponding to blood sugar, blood lipids and liver and kidney function indicators, is the standard threshold of the jth biochemical feature, represents the corresponding biochemical feature standard threshold, Corresponding to the standard thresholds of blood sugar, blood lipids and liver and kidney function indicators, The value range is .

6. A remote monitoring and nursing early warning system based on deep learning according to claim 5, characterized in that: In the nursing early warning evaluation module, the evaluation process of the patient's physiological health trend includes: Comprehensive feature information in the nursing warning feature set is used to divide the nursing warning feature set into a training set and a test set. The training set is used for model training and learning, and the test set is used to evaluate the performance of the model. Use the training set data combined with the convolutional neural network to build a nursing early warning model to learn and identify the patient's health status pattern, and use the test set data to evaluate the performance of the model. In the nursing early warning model, the input layer accepts the physiological data sequence, and the output layer outputs the patient's health status prediction result; The nursing early warning model is used in combination with the patient's physiological data to evaluate the patient's health status, and the early warning evaluation coefficient is calculated in combination with the physiological status trend index and biochemical status assessment index to comprehensively evaluate the patient's physiological health status.

7. A remote monitoring and nursing early warning system based on deep learning according to claim 6, characterized in that: The calculation formula of the early warning evaluation coefficient is: ； in, is the early warning evaluation coefficient, is the weight coefficient of the physiological state trend index, is the weight coefficient of the biochemical status assessment index, and Used for balance and right Contribution is the physiological state trend index, is the biochemical status assessment index, is the threshold parameter, The value range is between 0 and 1.

8. A remote monitoring and nursing early warning system based on deep learning according to claim 7, characterized in that: In the early warning graded response module, the process of setting the early warning level and matching the early warning response mechanism includes: Receive the output results from the nursing early warning evaluation module, including the patient's health status prediction, early warning evaluation coefficient, and related physiological and biochemical characteristics data; According to the results of the nursing early warning evaluation module, different early warning levels are set, and the early warning levels are divided into low warning level, medium warning level, high warning level and emergency warning level according to the early warning evaluation coefficient. Each early warning level corresponds to different risk levels and urgency levels. Analyze historical data to match corresponding warning thresholds for each warning level, and design corresponding response mechanisms for each warning level.

9. A remote monitoring and nursing early warning system based on deep learning according to claim 8, characterized in that: The multiple warning levels correspond to the multiple warning thresholds, which specifically satisfy the following relationship: Low alert level: ; Medium warning level: ; High alert level: ; Emergency warning level: ; in, is the early warning evaluation coefficient, is the lower threshold corresponding to the low warning level and the upper threshold corresponding to the medium warning level, is the lower threshold corresponding to the medium warning level and the upper threshold corresponding to the high warning level, is the lower threshold corresponding to the high warning level and the upper threshold corresponding to the emergency warning level, , , .

10. A remote monitoring and nursing early warning system based on deep learning according to claim 9, characterized in that: The user interaction module specifically includes: According to the needs of the remote monitoring and nursing early warning system, a user interaction interface including monitoring data display area, early warning information prompt area, and health advice area is designed; The user interaction module receives real-time data from the remote monitoring nursing early warning platform, including the patient's physiological characteristics and biochemical characteristics monitoring results, and updates the patient's physiological data to the user interaction interface in real time; When the warning condition is triggered, the user interaction module receives the warning information and displays it in the warning information prompt area on the user interaction interface, and issues a flashing reminder. The user clicks on the warning information to view detailed warning details, including the warning level, triggering conditions, and recommended response measures; Provide personalized health advice and intervention measures based on the patient's monitoring data and warning level, and display health advice in the health advice area on the interface.

Images