Wearable device-based vital sign management system and method

Abstract

The application provides a kind of wearable device-based vital sign management system and method, belong to medical health technical field, wearable device gathers target object's vital sign time series data, motion state time series data and environment time series data;Data processing terminal selects and motion state time series data and environment time series data corresponding influence coefficient sequence;According to influence coefficient sequence and pre-determined initial threshold value, determine time series-based flow updating threshold value;According to flow updating threshold value, vital sign time series data is detected for abnormal data, and / or, according to vital sign time series data, motion state time series data, environment time series data and flow updating threshold value, target object is managed for vital sign.The application utilizes flow updating threshold value to carry out abnormal data detection or vital sign management, can realize the accurate analysis of sign under special condition or environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical health, and particularly relates to a vital sign management system and method based on a wearable device. BACKGROUND

[0002] In the field of medical health, it is an increasingly growing demand to accurately monitor and analyze human activity and vital sign time series data. In order to realize dynamic monitoring of human vital sign signals, a wearable vital sign monitoring system becomes an ideal product for people. With the large-scale popularization of short-distance wireless communication technology, more and more intelligent wearable devices have entered people's daily life. The wearable vital sign monitoring system is not only a powerful tool for future smart life and work, but also an important terminal for collecting and processing user data in the network era.

[0003] The existing smart watches, wristbands and other wearable devices are prone to vital sign signal distortion under motion conditions or vibration environments, and it is difficult to realize accurate analysis of vital signs under special conditions or environments. SUMMARY

[0004] The present application provides a vital sign management system and method based on a wearable device, which solves the defects that the existing wearable devices are prone to vital sign signal distortion under motion conditions or vibration environments, and it is difficult to realize accurate analysis of vital signs under special conditions or environments.

[0005] The present application provides a vital sign management system based on a wearable device, comprising a wearable device and a data processing terminal in communication connection with the wearable device.

[0006] The wearable device is used to collect vital sign time series data, motion state time series data and environment time series data of an environment where the target object is located, and transmit them to the data processing terminal, wherein the vital sign time series data and the environment time series data are both multi-parameter time series data.

[0007] The data processing terminal is used to select an influence coefficient sequence corresponding to the motion state time series data and the environment time series data, the influence coefficient being used to represent the influence of the target object on vital signs in different motion states and different environments; to determine a time series-based streaming update threshold value according to the influence coefficient sequence and a pre-determined initial threshold value; to perform abnormal data detection on the vital sign time series data according to the streaming update threshold value, and / or to perform vital sign management on the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the streaming update threshold value.

[0008] As an embodiment, the vital sign management of the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the stream updating threshold value comprises:

[0009] The motion state time series data and the environment time series data are weighted to the vital sign time series data based on a time attention mechanism to obtain weighted vital sign time series data;

[0010] The vital sign time series data or the weighted vital sign time series data are associated analyzed based on principal component analysis and grey correlation degree analysis, and the parameter time series data with a correlation degree higher than a threshold value are taken as target vital sign time series data;

[0011] The weighted vital sign time series data and / or the target vital sign time series data are detected for abnormal data according to the stream updating threshold value to determine the overall condition of the vital sign of the target object;

[0012] The vital sign time series data, the weighted vital sign time series data and / or the target vital sign time series data are input into a pre-trained prediction model to obtain a vital sign prediction result of the target object;

[0013] If the overall condition of the vital sign of the target object or the vital sign prediction result is abnormal, an alarm information is initiated.

[0014] As an embodiment, the wearable device comprises an adaptive feedback device, the adaptive feedback device comprises a control module and at least one self-help module, the control module is used to control the self-help module to perform a preset self-help measure in response to an alarm information initiated by the data processing terminal or in response to a received adaptive feedback measure command.

[0015] As an embodiment, the self-help module comprises at least one of a heating module, an electric stimulation module and a microneedle drug delivery module, the microneedle drug delivery module comprises a micro-airbag pressurization module and a flexible microneedle patch, and the flexible microneedle patch is integrally integrated with a carrier of the wearable device.

[0016] As an embodiment, the heating module outputs a heating power or a cooling power based on a difference between a preset temperature and an environment temperature, a difference between the preset temperature and a body temperature of the target object and a preset first control coefficient to realize temperature regulation, the electric stimulation module dynamically adjusts an electric stimulation parameter based on an electromyographic signal of the target object, and the micro-airbag pressurization module outputs a heating power or airbag pressurization data based on a difference between a preset heating temperature and a temperature of the flexible microneedle patch, a difference between a preset air pressure and an airbag air pressure and a preset first control coefficient.

[0017] As an embodiment, the team analysis and decision-making platform is further configured to acquire the vital sign time series data, the motion state time series data and the environment time series data uploaded by the wearable device or the data processing terminal, and perform vital sign management on the target object based on a target scene data model matched with a current application scene of the target object, historical monitoring data of the target object and / or vital sign parameters of the same team as the target object.

[0018] As an embodiment, the team analysis and decision-making platform is provided with a plurality of scene data models, and members corresponding to each scene data model are taken as team members, and correspondingly, the matching step of the target scene data model comprises:

[0019] determining vital sign principal components and motion state principal components of the target object and the team members respectively based on principal component analysis;

[0020] determining gray correlation degrees between the vital sign principal components and the motion state principal components of the target object and the team members respectively based on gray correlation analysis;

[0021] taking a scene data model corresponding to a team member with the highest gray correlation degree as the target scene data model of the target object.

[0022] As an embodiment, the team analysis and decision-making platform is further configured to perform rescue / scheduling measures according to vital sign management data of the target object.

[0023] As an embodiment, the vital sign time series data comprises time series data of multiple parameters of heart rate, respiration, blood oxygen, blood pressure, body temperature and sweat of the target object, and the environment time series data comprises time series data of multiple parameters of temperature, humidity, altitude, air pressure and location.

[0024] The application further provides a vital sign management method based on a wearable device, comprising:

[0025] collecting vital sign time series data, motion state time series data of a target object and environment time series data of an environment where the target object is located, wherein the vital sign time series data and the environment time series data are both multi-parameter time series data;

[0026] selecting an influence coefficient sequence corresponding to the motion state time series data and the environment time series data, the influence coefficient being used to represent the influence of the target object on vital signs in different motion states and different environments;

[0027] determine a time series-based streaming update threshold value according to the influence coefficient sequence and a predetermined initial threshold value;

[0028] perform abnormal data detection on the vital sign time series data according to the streaming update threshold value, and / or perform vital sign management on the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the streaming update threshold value.

[0029] The wearable device-based vital sign management system and method provided by the application can realize precise vital sign analysis under special conditions or environments by selecting an influence coefficient sequence corresponding to the motion state time series data and the environment time series data to determine a streaming update threshold value, and using the streaming update threshold value to perform abnormal data detection on vital sign time series data or vital sign management on a target object. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0031] Figure 1 is one of the structural schematic diagrams of the wearable device-based vital sign management system provided by the application.

[0032] Figure 2 is a structural schematic diagram of the wearable device provided by the application.

[0033] Figure 3 is a process schematic diagram of the data processing terminal performing deep fusion analysis and abnormal monitoring provided by the application.

[0034] Figure 4 is a structural schematic diagram of the adaptive feedback device provided by the application.

[0035] Figure 5 is the second structural schematic diagram of the wearable device-based vital sign management system provided by the application.

[0036] Figure 6 is a structural schematic diagram of the team-based analysis and decision-making platform provided by the application.

[0037] Figure 7 is a process schematic diagram of the wearable device-based vital sign management method provided by the application. DETAILED DESCRIPTION

[0038] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the accompanying drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0039] It should be noted that all the actions of acquiring signals, information or data in the present application are performed under the premise of complying with the corresponding data protection regulations and policies of the place and obtaining the authorization given by the owner of the corresponding device.

[0040] Figure 1 is one of the structural schematic diagrams of the vital sign management system based on the wearable device provided by the present application, as shown in Figure 1 The present application further provides a vital sign management system based on a wearable device, which comprises a wearable device 10 and a data processing terminal 20 in communication connection with the wearable device. The data processing terminal 20 can be integrated on the wearable device 10 and in wired or wireless communication connection with the wearable device 10, or can be separately arranged from the wearable device 10 and in wireless communication connection with the wearable device 10.

[0041] The wearable device 10 is used to collect time series data of vital signs, time series data of motion states and time series data of the environment of the target object and transmit them to the data processing terminal, wherein the time series data of vital signs and the time series data of the environment are both multi-parameter time series data.

[0042] The data processing terminal 20 is used to select an influence coefficient sequence corresponding to the time series data of motion states and the time series data of the environment, the influence coefficient being used to represent the influence of the target object on vital signs in different motion states and different environments; to determine a time series-based streaming update threshold value according to the influence coefficient sequence and a pre-determined initial threshold value; to perform abnormal data detection on the time series data of vital signs according to the streaming update threshold value, and / or to perform vital sign management on the target object according to the time series data of vital signs, the time series data of motion states, the time series data of the environment and the streaming update threshold value.

[0043] The abnormal data detection on the time series data of vital signs according to the streaming update threshold value can determine the data collection stability of the wearable device, and if abnormal data is continuously detected, it is determined that the stability of the wearable device 10 is insufficient. The vital sign management on the target object includes determining, predicting and warning the overall condition of the vital signs of the target object.

[0044] It is understood that the present invention determines the streaming update threshold by selecting the influence coefficient sequence corresponding to the motion state time series data and the environmental time series data, and uses the streaming update threshold to detect abnormal data in the vital signs time series data or to manage the vital signs of the target object, thereby enabling accurate analysis of vital signs under special conditions or environments.

[0045] Based on the above embodiments, as an optional embodiment, the vital signs time series data includes time series data of multiple parameters among the target object's heart rate, respiration, blood oxygen, blood pressure, body temperature, and sweat, and the environmental time series data includes time series data of multiple parameters among temperature, humidity, altitude, air pressure, and location.

[0046] like Figure 2 As shown, the wearable device integrates a vital signs and motion status monitoring unit, an environmental signal monitoring unit, a processing chip, and a multidimensional data transmission unit. The vital signs and motion status monitoring unit includes, but is not limited to, an ECG sensor, a respiration sensor, a blood oxygen sensor, a blood pressure sensor, a temperature sensor, a sweat sensor, and an accelerometer. The ECG sensor, respiration sensor, blood oxygen sensor, blood pressure sensor, temperature sensor, and sweat sensor are used to collect vital signs parameters such as heart rate, respiration, blood oxygen, blood pressure, body temperature, and sweat of the target object. The accelerometer is used to collect the motion status of the target object. The environmental signal monitoring unit includes, but is not limited to, a temperature sensor, a humidity sensor, a barometric altimeter, an ambient light sensor, and a positioning system, used to collect parameters such as temperature, humidity, altitude, barometric altimeter, light, and position of the environment in which the target object is located. The processing chip processes the real-time data collected by the vital signs and motion status monitoring unit and the environmental signal monitoring unit to obtain multi-parameter, multi-dimensional time-series data, which is then transmitted to the multidimensional data transmission unit.

[0047] Preferably, wearable devices include, but are not limited to, smart bracelets, smart hats, and monitoring vests, used to collect real-time physiological data of the target subject, such as heart rate, respiration, blood pressure, blood sugar, blood oxygen saturation, temperature, electromyography, electroencephalography, and sweat; environmental data such as temperature, humidity, altitude, air pressure, and location; behavioral data such as movement speed, movement time, steps, and cadence; as well as personal basic information, medical history records, and previous health management data entered by the target subject.

[0048] Time-series data collected by wearable devices needs to be preprocessed to improve monitoring accuracy. This preprocessing can be performed on the wearable device or on the data processing terminal. This embodiment of the invention will be described using the data processing terminal as an example.

[0049] Preprocessing includes noise removal, outlier detection, and outlier coefficient calculation. The outlier coefficient is used to provide a reference for the wearability stability of wearable devices and the validity of sensor data.

[0050] Motion state sensors such as accelerometers and environmental signal monitoring units do not need to be in close contact with the target object's skin; data preprocessing can be performed directly through methods such as low-pass filtering, zero-bias error correction, and Kalman filtering.

[0051] Sensors used to collect vital signs of a target object need to be in close contact with the object's skin. They are affected to some extent by the body's movement or sweat, resulting in problems such as electrode detachment and slippage, which affect the validity of the data.

[0052] This invention, based on Sequence Anomaly Detection (Statistical Process Control Over the Threshold, SPOT), introduces human motion states and influence coefficients corresponding to different states. I ms This is used to detect abnormal data, such as human bodies in a static state. I ms = 1; Human body in jogging state I ms =1.08; Human body in a state of rapid motion I ms =1.21, etc. An influence coefficient library can be built based on a large amount of measured data. After the wearable device detects the vital signs, motion state, and environmental signals of the target object, the corresponding influence coefficient can be selected from the library.

[0053] The initial threshold is obtained based on a large amount of measured data and can be expressed as: t 0 Streaming update threshold I t The update formula for time series data is shown below:

[0054]

[0055] in, q It is the probability of an extreme event. n It is the number of all monitoring time points. N t It is more than t 0 The peak number, σ It is the extreme value index and γ It is a scale parameter.

[0056] Time-series-based streaming threshold update I tIt can effectively detect abnormal data in vital sign time series data.

[0057] Based on abnormal data and streaming update thresholds I t It can be used to calculate the anomaly coefficient. A c The calculation formula is as follows:

[0058]

[0059] in, X t At a certain point in time t The actual observed value at that location, I t At a certain point in time t The streaming update threshold at that location.

[0060] By observing abnormal fluctuations in a single parameter (higher abnormality coefficient) and abnormal fluctuations in other parameters (lower abnormality coefficient) within a short period of time, the instability of the physiological sign detection electrodes of the wearable device can be preliminarily determined, thus issuing a warning of insufficient wearability stability.

[0061] It is understood that the monitoring parameters of this invention include heart rate, respiration, blood oxygen, blood pressure, body temperature, sweat, and exercise status. It can monitor not only common heart rate, respiration, and body temperature signals, but also complex sweat signals, overcoming the shortcomings of existing vital sign monitoring systems that have relatively limited functionality and monitoring range, failing to meet the needs of comprehensive and continuous monitoring. This invention uses wearable, close-fitting clothing as a carrier and, based on algorithm optimization, can effectively avoid data deviations caused by strenuous exercise, preventing high false alarms in physiological anomaly warnings and the omission of genuine abnormal data, thus improving the quality of real-time data collected by wearable devices.

[0062] Based on the above embodiments, as an optional embodiment, the wearable device based on clothing carrier sends the collected vital signs, movement status, and environmental time series data to the data processing terminal in real time. The data processing terminal is used for in-depth fusion analysis of multi-dimensional time series data, anomaly monitoring, anomaly prediction, and anomaly alarm to achieve management of the vital signs of the target object.

[0063] The data processing terminal normalizes the time series data of vital signs, motion status, and environment to ensure they have the same time scale, and performs corresponding noise reduction filtering and signal amplification. Then, it extracts features to obtain time series data of vital sign features, motion status features, and environmental features. Based on the streaming update thresholds corresponding to different motion statuses and environmental information, it can determine anomalies and trigger threshold alarms for single or multiple vital signs.

[0064] As Figure 3 shown, the vital sign management of the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the flow updating threshold comprises the following steps.

[0065] The motion state time series data and the environment time series data are weighted to the vital sign time series data based on a time attention mechanism to obtain weighted vital sign time series data.

[0066] The vital sign time series data or the weighted vital sign time series data are analyzed based on principal component analysis and grey correlation degree analysis, and the parameter time series data with a correlation degree higher than a threshold are taken as target vital sign time series data.

[0067] The weighted vital sign time series data and / or the target vital sign time series data are detected for abnormal data according to the flow updating threshold to determine the overall vital sign condition of the target object.

[0068] The vital sign time series data, the weighted vital sign time series data and / or the target vital sign time series data are input into a pre-trained prediction model to obtain a vital sign prediction result of the target object.

[0069] If the overall vital sign condition of the target object or the vital sign prediction result is abnormal, an alarm information is initiated.

[0070] Optionally, the vital sign time series data or the weighted vital sign time series data are excluded for abnormal data, and then standardized calculation is performed, a covariance matrix is calculated according to the standardized data, and principal component analysis is performed according to the covariance matrix.

[0071] The formula of the standardized calculation is as follows:

[0072]

[0073] Among them, X is the abnormal coefficient excluded in the time series A c data greater than 0.95, μ is the abnormal coefficient excluded A c data greater than 0.95, and ε is the abnormal coefficient excluded A c data greater than 0.95, X stand is the standardized data matrix.

[0074] The weighted vital sign time series data includes exercise training state time series data and physical and mental health state time series data, the principal components of the vital signs include characteristic vector matrices of five or more than five parameters such as heart rate, respiration, blood pressure, blood oxygen saturation, temperature, electromyogram, exercise speed, ambient temperature, altitude, air pressure, the principal components of the exercise training state include characteristic vector matrices of five or more than five parameters such as heart rate, respiration, blood oxygen saturation, temperature, electromyogram, sweat ion concentration, sweat pH, sweat lactic acid value, exercise speed, exercise time, step count, step frequency, ambient temperature, altitude, air pressure, and the principal components of the physical and mental health state include characteristic vector matrices of five or more than five parameters such as heart rate, respiration, blood pressure, blood glucose, sweat ion concentration, sweat pH, sweat lactic acid value, ambient temperature, air pressure.

[0075] The correlation degrees between factors in the principal components of the vital signs, the principal components of the exercise training state and the principal components of the physical and mental health state are evaluated based on grey correlation degree analysis, and multiple parameters with high correlation degrees are used to reflect the vital sign state of the target object.

[0076] Optionally, the trained prediction model is a long short-term memory autoencoder, the prediction model is trained to predict normal vital sign time series data, and the difference between the predicted value and the actual value is compared, if the difference value is greater than a preset threshold, it is determined that the vital sign of the target object is abnormal. The prediction model can also combine a streaming update threshold to improve the accuracy of detection.

[0077] Based on the above embodiment, as an optional embodiment, as shown in Figure 4 The wearable device includes an adaptive feedback device, the adaptive feedback device includes a control module and at least one self-help module, the control module is used to respond to the alarm information initiated by the data processing terminal or respond to the received adaptive feedback measure command, and control the self-help module to execute the preset self-help measure.

[0078] The adaptive feedback device of the wearable smart device based on the clothing carrier receives the alarm information and the adaptive feedback measure command from the data processing terminal through the signal receiving module, completes the instruction requirements by the control module, and completes the instruction requirements issued by the control module by the self-help module.

[0079] Optionally, the self-help module includes at least one of a heating module, an electric stimulation module, and a microneedle drug delivery module, the microneedle drug delivery module including a micro-airbag pressurization module and a flexible microneedle patch integrated with the carrier of the wearable device. The wearable device is a clothing carrier, and the heating module, the electric stimulation module, and the microneedle drug delivery module are all flexible structures. The flexible electric heating module is woven from flexible electric heating fibers and integrated with the clothing carrier at the front chest and back of the body, and is connected by fiber wires and a control module and a matching power supply module, which can realize heating and warming self-help measures when the human body loses temperature. The flexible electric stimulation module is woven from flexible conductive fibers and integrated with the clothing carrier at different body surface positions such as the arms, abdomen, chest, etc. of the body, and is connected by flexible fiber wires and a control module and a matching power supply module, which relieves muscle fatigue through electric stimulation. The flexible microneedle drug delivery module is integrated with the clothing carrier by a flexible microneedle patch, which can load different drug molecules, including but not limited to cardiovascular disease drugs, inflammation drugs, hemostatic drugs, etc., and is connected by flexible fiber wires and a control module and a matching power supply module, which promotes controllable drug delivery through airbag pressurization and electric heating dissolving microneedles.

[0080] Optionally, the heating module outputs heating power or cooling power based on the difference between the preset temperature and the ambient temperature, the difference between the preset temperature and the body temperature of the target object, and a preset first control coefficient to achieve temperature regulation. The temperature control formula of the heating module is as follows:

[0081]

[0082] wherein, u(t) is the controller output (heating / cooling power), e(t 1 ) is the temperature difference (difference between the set temperature and the ambient temperature), e(t 2 ) is the temperature difference (difference between the set temperature and the body temperature), ρ is the ambient temperature coefficient, θ is the body temperature coefficient K p1 、K i1 、K d1 are the ambient temperature proportional, integral, and differential control coefficients, respectively; K p2 、K i2 、K d2 are the body temperature proportional, integral, and differential control coefficients, respectively.

[0083] The electric stimulation module dynamically adjusts the electric stimulation parameters based on the electromyogram (Root Mean Square, RMS) of the target object, and the RMS calculation formula is as follows:

[0084]

[0085] Wherein, x i For rejecting sampling values with abnormal coefficients greater than 0.95, N is the number of sampling points.

[0086] The micro air bag pressurization module outputs heating power or air bag pressurization data based on the difference between the preset heating temperature and the temperature of the flexible microneedle patch, the difference between the preset air pressure and the air bag pressure, and the preset first control coefficient. The micro air bag pressurization module controls the drug delivery based on the temperature control algorithm and the micro air bag pressurization control algorithm, and the specific formula is as follows:

[0087]

[0088] Wherein, u(P,T) is the controller output (micro air bag inflation / electric heating power), e P (t) is the air pressure difference (the difference between the set air pressure and the air bag pressure), e T (t) is the temperature difference (the difference between the heating set temperature and the microneedle patch temperature), is the integral of the air pressure difference is the integral of the temperature difference; ψbarometric pressure coefficient, τ is the temperature coefficient, K Pp 、K Pi 、K Pd are the air pressure proportional, integral, and differential control coefficients, respectively; K Tp 、K Ti 、K Td are the temperature proportional, integral, and differential control coefficients, respectively.

[0089] It can be understood that the present application compares the multi-dimensional signs and environmental signal data at different times through the wearable smart device, forms a comparison and analysis of the body conditions before and after the adaptive feedback unit by the data processing terminal, and triggers self-help measures after the threshold alarm, which expands the conventional vital sign monitoring and early warning to real-time adaptive feedback measures.

[0090] On the basis of the above embodiments, as an optional embodiment, as shown in Figure 5 The present application also includes a team analysis and decision platform, wearable devices and data processing terminals for monitoring individuals. The team analysis and decision platform can access wearable devices and data processing terminals of a plurality of different individual users (such as target objects in the present application), obtain multi-source, multi-dimensional vital sign information data, fusion information results, and whether there are abnormal conditions, and form a vital sign information parameter set for each individual user. According to different scenarios such as smart aging, intensive care, daily exercise training, military training, battlefield rescue, and extreme environment rescue, individuals are divided into a plurality of teams, and the vital signs of one or more groups of people are centrally managed and data is shared.

[0091] As shown in Figure 6 The team analysis and decision platform is also used for data comparison and analysis of the parameter set, and evaluation of the health condition. The comparison and analysis includes taking the scene data model as a reference, and each type of scene model in the scene model matching unit includes: 1. Major disease data model, disease prevention data model, and exercise training data model, which can match the current application scenario of the user for multi-dimensional data comparison and analysis; 2. Taking historical monitoring data as a reference, multi-dimensional data time series comparison and analysis; 3. Taking the vital sign information parameter set of different users as a reference, individual multi-dimensional data horizontal comparison. Based on the current vital sign information parameter, the health condition is evaluated, the possible vital sign changes are predicted, and rescue scheduling is further implemented.

[0092] Specifically, the team analysis and decision platform is used to obtain the vital sign time series data, the exercise state time series data, and the environment time series data uploaded by the wearable device or the data processing terminal, and perform vital sign management on the target object based on the target scene data model matched with the current application scenario of the target object, the historical monitoring data of the target object, and / or the vital sign parameters of the same team as the target object.

[0093] Optionally, the team analysis and decision platform is provided with a plurality of scene data models, and members corresponding to each of the scene data models are taken as team members. Correspondingly, the matching step of the target scene data model includes the following steps.

[0094] The principal component analysis is used to determine the vital sign principal components and exercise state principal components of the target object and the team members, respectively.

[0095] The grey correlation degree analysis is used to determine the grey correlation degree between the vital sign principal components and the exercise state principal components of the target object and the team members, respectively.

[0096] The scene data model corresponding to the team member with the highest gray correlation degree is taken as the target scene data model of the target object.

[0097] Optionally, the team-based analysis and decision platform is further configured to execute rescue / scheduling measures according to the vital sign management data of the target object.

[0098] The vital sign state principal component, the exercise training state principal component, and the physical and mental health principal component of the team member are matched with the vital sign state, the exercise training state, and the physical and mental health state in different scenes extracted from the scene model, the correlation coefficient and the correlation degree between the different state principal component data sets of the team member and the state data sets of the model are analyzed through gray correlation degree analysis, and the scene data model with the highest correlation degree is matched.

[0099] The calculation formula of the correlation coefficient is as follows:

[0100]

[0101] Among them, ξ i,j is the gray correlation coefficient between the different state principal component data sets of the team member and the state data sets of the model; Δ min is the minimum value of the absolute value of the numerical difference between two sequences (the sub-sequence of the different state principal component data sets of the team member and the mother sequence of the state data sets of the model); Δ max is the maximum value of the absolute value of the numerical difference between two sequences (the sub-sequence of the different state principal component data sets of the team member and the mother sequence of the state data sets of the model); ρ is the resolution coefficient; Δ t(i,j) is the absolute value of the numerical difference between two sequences, i.e., the absolute value of the difference between the observation values of the sub-sequence and the mother sequence at the same observation time (the same row).

[0102] The calculation formula of the correlation degree is as follows:

[0103]

[0104] R i,j is the gray correlation degree between the sub-sequence of the different state principal component data sets of the team member i and the mother sequence of the state data sets of the model j . ξ i,j is the gray correlation coefficient; L is the length of the sequence, i.e., the number of observation data; num is the serial number of the data in the sequence.

[0105] The calculation formula of the overall similarity degree is as follows:

[0106]

[0107] R The overall similarity degree; α, β,..., ω The correlation degree weight coefficient of different parameter sequences, m The sequence logarithm.

[0108] The overall similarity degree between different state principal components and model state data R The value size, the highest correlation degree of the matching scene data model is obtained and is used as the target scene data model.

[0109] The team analysis and decision-making platform can predict the development of the team member's physical state according to the target scene data model, and provide feedback. In addition, the above method can also be used to compare the historical monitoring data sequence and the real-time monitoring data sequence to prevent the recurrence of diseases and the like; the comparison between the sequence data of different users of the team members can be used to compare the team member's internal vital sign state principal component, exercise training state principal component, and physical and mental health principal component, and to propose reasonable deployment arrangements for the team members.

[0110] It can be understood that the team analysis and decision-making platform is provided in the application, which is used for centralized management / data sharing of the vital signs of one or more groups of personnel, and realizes the functions of centralized management, data sharing, efficient response, unified standard, resource optimization and the like.

[0111] The wearable device-based vital sign management method provided by the application will be described below, and the wearable device-based vital sign management method described below can be mutually referred to the wearable device-based vital sign management system described above.

[0112] Figure 7 The flowchart of the wearable device-based vital sign management method provided by the application is shown in FIG. Figure 7 The application further provides a wearable device-based vital sign management method, which comprises the following steps.

[0113] In step S100, the vital sign time sequence data, the motion state time sequence data and the environment time sequence data of the target object are collected, wherein the vital sign time sequence data and the environment time sequence data are both multi-parameter time sequence data.

[0114] Step S200, selecting an influence coefficient sequence corresponding to the motion state time series data and the environment time series data, the influence coefficient being used to represent the influence of the target object on the vital sign in different motion states and different environments.

[0115] Step S300, determining a time series-based streaming update threshold according to the influence coefficient sequence and a predetermined initial threshold.

[0116] Step S400, performing abnormal data detection on the vital sign time series data according to the streaming update threshold, and / or performing vital sign management on the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the streaming update threshold.

[0117] It should be noted that the wearable device-based vital sign management method provided by the present application is implemented based on the wearable device-based vital sign management system according to any of the above embodiments when specifically running, has a technical effect corresponding to the system, and the present embodiment will not be described here.

[0118] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A wearable device based vital sign management system, characterized in that, The wearable device and a data processing terminal in communication connection with the wearable device are comprised; The wearable device is used for collecting time series data of vital signs, time series data of motion states and time series data of environment of a target object, and transmitting to the data processing terminal, wherein the time series data of vital signs and the time series data of environment are both multi-parameter time series data; The data processing terminal is used for selecting an influence coefficient sequence corresponding to the time series data of motion states and the time series data of environment, the influence coefficient is used for representing the influence of the target object on vital signs in different motion states and different environments, determining a time series based flow updating threshold value according to the influence coefficient sequence and a pre-determined initial threshold value, performing abnormal data detection on the time series data of vital signs according to the flow updating threshold value, and / or performing vital sign management on the target object according to the time series data of vital signs, the time series data of motion states, the time series data of environment and the flow updating threshold value; The updating formula of the flow updating threshold value under time series is as follows: ； in, I t For streaming, the threshold is updated. t 0 As the initial threshold, I ms Influence coefficient ，q It is the probability of an extreme event. n It is the number of all monitoring time points. N t It is more than t 0 The peak number, The time series data of motion states and the time series data of environment are weighted processed on the time series data of vital signs based on a time attention mechanism to obtain weighted time series data of vital signs; It is the extreme value index and The weighted time series data of vital signs is associated analyzed based on principal component analysis and grey correlation degree analysis, and parameter time series data with a correlation degree higher than a threshold value is taken as target time series data of vital signs; It is a scale parameter; Based on streaming update threshold I t Calculate the anomaly coefficient A c The calculation formula is as follows: ； wherein, X t is an actual observation value at a time point t ; the vital sign management on the target object according to the vital sign time series data, the motion state time series data, the environment time series data and the flow updating threshold value comprises: The weighted time series data of vital signs and / or the target time series data of vital signs are detected for abnormal data according to the flow updating threshold value to determine the overall condition of vital signs of the target object; The weighted time series data of vital signs and / or the target time series data of vital signs are input into a pre-trained prediction model to obtain a vital sign prediction result of the target object; If the overall condition of vital signs of the target object and the vital sign prediction result are abnormal, an alarm information is initiated. The wearable device comprises an adaptive feedback device, the adaptive feedback device comprises a control module and at least one self-help module, the control module is used for controlling the self-help module to perform a pre-set self-help measure in response to an alarm information initiated by the data processing terminal or in response to a received adaptive feedback measure command. The self-help module comprises at least one of a heating module, an electric stimulation module and a microneedle drug delivery module, the microneedle drug delivery module comprises a micro-airbag pressurizing module and a flexible microneedle patch, and the flexible microneedle patch is integrally integrated with a carrier of the wearable device.

2. The wearable device based vital sign management system of claim 1, wherein, ​ 3. The wearable device based vital sign management system of claim 2, wherein, ​ 4. The wearable device based vital sign management system of claim 3, wherein, The heating module outputs heating power or cooling power based on a difference between a preset temperature and an ambient temperature, a difference between the preset temperature and a human body temperature of the target object, and a preset first control coefficient to achieve temperature regulation, the electric stimulation module dynamically adjusts electric stimulation parameters based on an electromyographic signal of the target object, and the micro-air sac pressurization module outputs heating power or air sac pressurization data based on a difference between a preset heating temperature and a temperature of the flexible microneedle patch, a difference between a preset air pressure and an air sac air pressure, and a preset first control coefficient.

5. The wearable device based vital sign management system according to any of claims 1-4, characterized by, The team analysis and decision platform is further configured to obtain the time series data of vital signs, the time series data of motion states, and the time series data of environments uploaded by the wearable device or the data processing terminal, and perform vital sign management on the target object based on a target scene data model matched with a current application scene of the target object, historical monitoring data of the target object, and vital sign parameters of a same team as the target object.

6. The wearable device based vital sign management system of claim 5, wherein, The team analysis and decision platform is provided with a plurality of scene data models, and members corresponding to each scene data model are taken as team members. The team analysis and decision platform is further configured to determine vital sign principal components and motion state principal components of the target object and the team members respectively based on principal component analysis. The team analysis and decision platform is further configured to determine gray correlation degrees between the vital sign principal components and the motion state principal components of the target object and the team members respectively based on gray correlation analysis. The team analysis and decision platform is further configured to take a scene data model corresponding to a team member with the highest gray correlation degree as a target scene data model of the target object.

7. The wearable device based vital sign management system of claim 5, wherein, The team analysis and decision platform is further configured to perform rescue scheduling measures according to vital sign management data of the target object.

8. The wearable device based vital sign management system according to any of claims 1-4, 6-7, wherein, The time series data of vital signs of the target object include time series data of a plurality of parameters such as heart rate, respiration, blood oxygen, blood pressure, body temperature, and sweat, and the time series data of environments include time series data of a plurality of parameters such as temperature, humidity, altitude, air pressure, and location.

9. A wearable device based vital sign management method implemented by the wearable device based vital sign management system of claim 1, wherein, The team analysis and decision platform is further configured to: Collect time series data of vital signs, time series data of motion states, and time series data of environments of a target object, wherein the time series data of vital signs and the time series data of environments are both multi-parameter time series data; Select an influence coefficient sequence corresponding to the time series data of motion states and the time series data of environments, the influence coefficient being used to represent the influence of the target object on vital signs in different motion states and different environments; Determine a time series-based streaming update threshold value according to the influence coefficient sequence and a predetermined initial threshold value; Perform abnormal data detection on the time series data of vital signs according to the streaming update threshold value, and / or perform vital sign management on the target object according to the time series data of vital signs, the time series data of motion states, the time series data of environments, and the streaming update threshold value.

Images