CN111681761B - Situation-oriented health risk identification method and system - Google Patents
Situation-oriented health risk identification method and system Download PDFInfo
- Publication number
- CN111681761B CN111681761B CN202010548021.2A CN202010548021A CN111681761B CN 111681761 B CN111681761 B CN 111681761B CN 202010548021 A CN202010548021 A CN 202010548021A CN 111681761 B CN111681761 B CN 111681761B
- Authority
- CN
- China
- Prior art keywords
- health
- sensor
- sensor group
- situation
- physiological
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Abstract
The invention discloses a situation-oriented health risk identification method, which belongs to the technical field of health monitoring and comprises the following steps: establishing a situation-oriented knowledge graph conceptual model for health risk identification; acquiring sensor parameters so as to establish a knowledge graph instance model for situation-oriented health risk identification; and identifying health risks based on a semantic similarity algorithm, and driving the health care robot to provide care services. The technical scheme comprehensively utilizes the technologies of a sensor network, cloud computing, edge computing, a knowledge graph and the like, exerts the advantages of situation awareness, edge cloud quick response and the like, overcomes the semantic heterogeneity of sensors, sensor parameters, health conditions and care services, improves the accuracy of health risk identification, can actively intervene by using a health care robot in case of abnormal health conditions, finds and provides a first aid in emergency conditions, and improves the daily care quality of chronic diseases such as cardiovascular and cerebrovascular diseases.
Description
Technical Field
The invention belongs to the technical field of health monitoring, and particularly relates to a situation-oriented health risk identification method and system.
Background
According to the report 2018 on cardiovascular and cerebrovascular diseases, the mortality rate of cardiovascular and cerebrovascular diseases in China is in an increasing trend; the number of cardiovascular patients is 2.9 million, 1300 million stroke, 1100 million coronary heart disease, 500 million pulmonary heart disease, 450 million heart failure, 250 million rheumatic heart disease, 200 million congenital heart disease and 2.45 million hypertension. A great deal of medical practice shows that even sudden heart diseases of heart and cerebral vessels can be prevented from death by 70-80% of patients if subtle symptoms can be monitored in advance and intervention measures under emergency conditions are taken.
In order to adapt to the rapid development of the aging society of China, patients with cardiovascular and cerebrovascular diseases need to be monitored in daily life for a long time, can give an alarm immediately when facing health risk emergency, and find and give first aid at the first time. At present, the health care robot gradually walks into hospitals, nursing homes and care centers, and work of care doctors and care personnel is reduced. The monitoring effect is often limited in health monitoring practice, and the main reasons are: (1) physiological parameters alone are not sufficient to characterize the health condition of a patient, for example, an increase in heart rate is considered abnormal when identified as a parameter alone, and normal when the patient is running. The physiological parameters are closely related to the situation of the user, and the correlation plays an important role in accurate identification of health risks and timely early warning; (2) the health care robot in a hospital or an old community scene serves a plurality of cardiovascular and cerebrovascular patients at the same time, and how to improve the service quality of the health care robot is of great importance.
Chinese utility model patent publication No. CN209377561U, granted announcement day 2019-9-13, disclose a long-range electrocardio monitoring node with situation perception ability, this node compact structure, low-power consumption when realizing accurate acquisition patient electrocardiosignal, can discern the situation that current patient is located to can improve the degree of accuracy that the ECG signal acquireed according to the electrode that leads of patient's exercise intensity automatic selection reliability, thereby provide solid powerful help for the more accurate objective analysis ECG data of later stage cardiology department doctor.
Chinese patent publication No. 108024726, published as 2018-5-11 entitled "pulse oximeter with context added on patient monitor," provides a system and method for comparing a first patient's pulse oximeter data with a second patient's pulse oximeter data previously stored in a database. The method includes acquiring a first patient's pulse oximeter data for a first patient and inputting one or more markers for respective segments of the first patient's pulse oximeter data, wherein the one or more markers correspond to a period of time beginning from a time at which patient context data was acquired.
The Chinese invention application publication No. CN109697196A, publication date 2019-4-30, entitled "a situation modeling method, device and equipment", provides a modeling method, which comprises the following steps: identifying a scene in a context aware application; identifying an object in a scene and determining a contextual attribute of the object; establishing a corresponding relation between the situation attribute and the original data; the raw data includes: time data, values and identification data; generating events, services and rules in the scene according to the situation attributes; and generating a model file of the context-aware application according to the events, services and rules in the scene.
The Chinese invention application publication No. CN109446700A, publication No. 2019-3-8, entitled "modeling and executing method in runtime for intelligent home context awareness service", and provides a modeling method and an executing method in runtime for an intelligent home context awareness service knowledge graph instance model aiming at establishing an intelligent home context awareness service knowledge graph conceptual model.
The Chinese invention application publication No. CN106625714A, publication date 2017-8-10, entitled "a monitoring robot for detecting the health condition of the elderly", tracks the actions of the elderly by shooting and comprehensively judges whether the elderly have abnormal conditions by using an intelligent bracelet, and when the abnormal conditions occur, the robot body can timely know the abnormal conditions and is connected with a medical network platform through an alarm module and a community network communication module, so that doctors in community hospitals can timely rescue and treat the elderly according to the physiological data.
The Chinese invention application publication No. CN105078449A, publication date 2015-11-25, entitled "health service robot-based Alzheimer's disease monitoring system", comprises a health service robot, an intelligent terminal and cloud services, wherein the health service robot comprises a robot body, a main control unit, a human-computer interaction unit and a medical detection unit; the intelligent terminal and the tablet computer are connected through a cloud server of the mobile internet. The tablet personal computer collects the voice information of children of the dementia patient, so that the simulated emotion communication between the senile dementia patient and the children is realized; and evaluating the sleep quality by adopting a method of combining the energy characteristics and the least square support vector machine.
Based on the above description, the situation-oriented health risk identification method and system still belong to a weaker field, and are mainly represented by: lack of or insufficient integration with the situation, relying on physiological parameter monitoring makes it difficult to accurately identify health risks; the situation data comprises physiological parameter monitoring and acquisition, health risk identification and health care service response, which are an organic whole, and different acquisition devices have different parameter types in practical application and need to overcome the problem of semantic heterogeneity in a common knowledge map model; the existing health care robot integrating monitoring and service only aims at patients with a single disease type, and is difficult to meet the requirement that hospitals, nursing homes and care centers care a plurality of patients with different requirements at the same time; the health care robot needs to provide proper and timely care service, and complex reasoning is not beneficial to discovering and giving first aid at first time.
Based on the above analysis, the present case has been made.
Disclosure of Invention
The invention aims to provide a situation-oriented health risk identification method and system, which comprehensively utilize technologies such as a sensor network, cloud computing, edge computing and a knowledge graph, give play to advantages such as situation awareness and edge cloud quick response, overcome semantic heterogeneity of sensors, sensor parameters, health conditions and care services, improve accuracy of health risk identification, actively intervene by using a health care robot in case of abnormal health conditions, find and provide rescue in case of emergency, and improve quality of daily care for chronic diseases such as cardiovascular and cerebrovascular diseases.
In order to achieve the above purpose, the solution of the invention is:
a situation-oriented health risk identification method comprises the following steps:
step 1, establishing a situation-oriented knowledge graph conceptual model for health risk identification;
step 2, collecting sensor parameters, and establishing a situation-oriented knowledge graph instance model for health risk identification;
and 3, identifying health risks based on a semantic similarity algorithm, and driving the health care robot to provide care services.
In the step 1, the established model specifically comprises defining concepts of users, sensors, sensor parameters, health conditions and care services, and the corresponding relationships of the concepts comprise use, provision, assignment and triggering;
wherein the user comprises a patient; the sensors comprise a virtual sensor, a physiological sensor, an environment sensor and a motion sensor; the sensor parameters comprise personal information, physiological parameters, environmental parameters and motion parameters; the health condition comprises personal information condition, physiological parameter condition, environmental parameter condition and motion parameter condition; the care services comprise abnormal services (comprising abnormal reminding, meal reminding and movement reminding), emergency services (comprising emergency alarm, drug delivery and emergency medical treatment); usage indicates user usage of the sensor; providing a value of a monitoring parameter indicative of a sensor providing the sensor; the valuation indicates that the monitored parameter of the sensor changes the health condition of the user; triggering means that the health condition triggers a health service.
The specific content of the step 2 is as follows:
step 21, configuring a virtual sensor group, a physiological sensor group, an environmental sensor group, a motion sensor group and a care service set of a user according to requirements according to specific monitored chronic diseases, and respectively recording as: virtual sensor group
Physiological sensor group
Environment sensor group
Motion sensor group
Service set 1 ,Service 2 ,…,Service k }；
Step 22, collecting the following situation data of the user through the virtual sensor group, the physiological sensor group, the environment sensor group and the motion sensor group: personal information
Physiological parameter
Environmental parameter
And motion parameters
And obtaining the current health condition S' according to the personal information, the physiological parameters, the environmental parameters and the interval range of the motion parameters, thereby establishing a situation-oriented knowledge graph instance model for health risk identification.
The specific content of the step 3 is as follows:
step 31, calculating the semantic similarity between the current health condition and the health condition reference standard by using the following formula:
wherein f denotes the common number of concepts of the current health status S 'and the health status reference standard S, g denotes the different number of concepts of the current health status S' and the health status reference standard S, w i 、w j Is a weight and has a value of 0 ≦ w i ,w j ≤1；sima(s' i ,s i ) Representing a current health concept s' i And the concept s of the corresponding health condition reference standard i 0 ≦ sima (s' i ,s i )≤1，1≤i≤f，1≤j≤g；
Step 32, comparing the semantic similarity obtained by the step 31 with the abnormal health condition threshold value delta abnormal Health emergency threshold δ urgent And comparing, identifying the health risks, and driving the health care robot to provide care service by the edge cloud according to the identification result.
In the above step 31, the atomic concept similarity sima (s' i ,s i ) The calculation method comprises the following steps:
firstly, the attributes of the atomic concepts of the health condition and the health condition reference standard are divided into a text type and an interval type, wherein the text type refers to personal information acquired by a virtual sensor, and the interval type refers to parameter data acquired by a physiological sensor group, an environmental sensor group and a motion sensor group respectively;
then, the following processes are respectively performed:
establishing concept s 'for text-type attributes' i 、s i Attribute vector of (1), where s' i Is present and s i Note of attribute location also present as 1, s' i Exist but s i Attribute location of absence is recorded as 0, s' i 、s i The text type attribute similarity calculation formula is as follows:
wherein the content of the first and second substances,
is corresponding s' i 、s i Is a unit vector, ζ is a weight, Dep (s' i )，Dep(s i ) Respectively represent s' i 、s i Conceptual depth in S', S;
for the intertype attribute, s 'is assumed' i ＝[s' a ,s' b ]，s i ＝[s a ,s b ]And let s' a ,s' b ,s a ,s b Is ranked as s 1 ,s 2 ,s 3 ,s 4 Then s' i 、s i The interval type attribute similarity calculation formula is as follows:
wherein sgn () is a sign function;
and finally, weighting and summing the text type attribute similarity and the interval type attribute similarity to obtain the attribute-based atomic concept similarity, which is as follows:
wherein the content of the first and second substances,
a m ,b n respectively are a text type attribute coefficient and an interval type attribute coefficient; m, N are respectively the current health status concept s' i And the concept s of the corresponding health condition reference standard i Article ofNumber of local type attributes and interval type attributes.
In the above step 32, if the semantic similarity Sim (S', S) is smaller than the abnormal health condition threshold δ abnormal Health risk is identified as "normal"; if the semantic similarity Sim (S', S) is larger than or equal to the health abnormal condition threshold value delta abnormal And is less than the health emergency threshold δ urgent If the health risk is identified as abnormal, the edge cloud-driven health care robot provides abnormal services (including abnormal reminding, meal reminding and exercise reminding); if the semantic similarity Sim (S', S) is greater than or equal to the health emergency threshold delta urgent And if the health risk is identified as 'emergency', emergency services (including emergency alarm, drug delivery and emergency medical treatment) are provided by the edge cloud-driven health care robot.
A situation-oriented health risk identification system comprises a mobile intelligent terminal with a built-in virtual sensor group, a physiological sensor group, an environment sensor group, a motion sensor group, an edge cloud and a health care robot, wherein the virtual sensor group is used for acquiring personal information of a user; the physiological sensor group is used for collecting heart rate, electrocardio, blood pressure and blood oxygen; the environment sensor group is used for collecting temperature, humidity and air quality; the motion sensor group is used for acquiring the position, the motion and the body posture of the user; the edge cloud collects the data through the edge server, establishes a situation-oriented knowledge graph instance model for health risk identification, identifies health risks and drives the health care robot to provide care services.
The physiological sensor group adopts a heart rate sensor, an electrocardio sensor, a blood pressure sensor and a blood oxygen sensor; the environment sensor group adopts a temperature sensor, a humidity sensor and a PM2.5 sensor; the motion sensor adopts an acceleration sensor, a gyroscope, a gravity sensor and a direction sensor.
The edge server adopts a Cortex A8 processor chip, a 250G Byte solid state disk, a Linux operating system Xen VMM, a fusion gateway (Bluetooth, ZigBee, WIFI, 6LoWPAN) and a 10M/100M network card.
After the scheme is adopted, compared with the traditional health risk identification method (or system), the situation-oriented health risk identification method (or system) provided by the invention has the following advantages:
(1) the system architecture of the invention adopts the sensors to collect situation data (including physiological parameters), the edge cloud identifies health risks, and the health care robot provides proper and timely health services, so that the system architecture has lower response delay compared with the system architecture which relies on a traditional cloud platform and does not need long-distance network transmission;
(2) the situation-oriented knowledge graph conceptual model for health risk identification is used, the semantic relation among the sensors, the sensor parameters, the health condition and the care service is determined, and the situation-oriented knowledge graph conceptual model is different from the problem that the information organization of the traditional health care service system lacks sufficient knowledge support;
(3) the knowledge graph instance model of situation-oriented health risk identification is used, the health risk is identified based on semantic similarity, the accuracy is higher than that of traditional health risk identification, and the time efficiency is higher than that of rule reasoning;
(4) the invention can be suitable for hospitals, nursing homes and nursing centers, and can meet the condition of simultaneous nursing of different chronic disease patients by customizing appropriate sensors and health risk reference standards.
Drawings
FIG. 1 is a knowledge graph conceptual model of context-oriented health risk identification of the present invention;
FIG. 2 is a schematic diagram of the creation of a conceptual model of a knowledge-graph.
Detailed Description
The technical solution and the advantages of the present invention will be described in detail with reference to the accompanying drawings.
The invention provides a situation-oriented health risk identification method, which comprises the following steps:
step 1, establishing a Situation-oriented knowledge graph conceptual model of health risk identification by matching with the knowledge graph concept model shown in fig. 1 and fig. 2, wherein the Situation-oriented knowledge graph conceptual model specifically comprises defining concepts of a User (User), a Sensor (Sensor), Sensor parameters (SensorParameter), health conditions (health) and care services (Service), and the corresponding relationships of the concepts comprise using (Use), providing (provision), assigning (Assign) and triggering (trigger);
wherein the User (User) comprises a patient; the Sensor (Sensor) comprises a virtual Sensor (VirtualSensor) (comprising software of a built-in mobile phone), a physiological Sensor (Biosensor) (comprising a heart rate Sensor, an electrocardio Sensor, a blood pressure Sensor and a blood oxygen Sensor), an environment Sensor (environmental Sensor) (comprising a temperature Sensor, a humidity Sensor and a PM2.5 Sensor) and a motion Sensor (motionSensor) (comprising a GPS Sensor, an acceleration Sensor, a gyroscope, a gravity Sensor and a direction Sensor); sensor parameters (SensorParameter) include personal information (VirtualParameter) (including age, gender, weight, height, family genetic history, dietary preferences), physiological parameters (physiological parameter) (including heart rate, electrocardiogram, blood pressure, blood oxygen), environmental parameters (EnvironmentParameter) (including temperature, humidity, PM2.5 air mass), motion parameters (MotionParameter) (including position, motion, and body posture); the health condition (configuration) includes a personal information condition (configuration of physiological parameter), a physiological parameter condition (configuration of physiological parameter), an environmental parameter condition (configuration of environmental parameter), a motion parameter condition (configuration of motion parameter); the care Service (Service) comprises an Abnormal Service (Abnormal Service) (comprising an Abnormal Alert (Abnormal Alert), a Meal Alert (Meal Recommendation), a movement Alert (Abnormal Alert)), an Emergency Service (Emergency Service) (comprising an Emergency Alarm (Emergency Alert), a Drug Delivery (Drug Delivery), an Emergency Medical Treatment (Emergency Medical Treatment)); use of
Indicating that the user is using the sensor; provide for
A value indicative of a monitored parameter of the sensor; valuation
Indicating that the monitored parameter of the sensor changes the health condition of the user; triggering
Indicating that the health condition triggers a health service.
Step 2, collecting sensor parameters, thereby establishing a situation-oriented knowledge graph instance model for health risk identification, which specifically comprises the following contents:
step 21, configuring a virtual sensor group, a physiological sensor group, an environmental sensor group, a motion sensor group and a care service set of a user according to requirements according to specific monitored chronic diseases, and respectively recording as: virtual sensor group
Physiological sensor group
Environment sensor group
Motion sensor group
Service set 1 ,Service 2 ,…,Service k }；
Step 22, collecting the following situation data of the user through the virtual sensor group, the physiological sensor group, the environment sensor group and the motion sensor group: personal information
Physiological parameter
Environmental parameter
And motion parameters
Obtaining the current health status according to the personal information and the interval range of the physiological parameters, the environmental parameters and the motion parametersCondition S', thereby establishing a knowledge graph instance model for situation-oriented health risk identification;
step 3, identifying health risks based on a semantic similarity algorithm, and driving a health care robot to provide care services; the method specifically comprises the following steps:
step 31, calculating the semantic similarity between the current health condition and the health condition reference standard by using the following formula:
wherein f denotes the common number of concepts of the current health status S 'and the health status reference standard S, g denotes the different number of concepts of the current health status S' and the health status reference standard S, w i 、w j Is a weight and has a value of 0 ≦ w i ,w j ≤1；sima(s' i ,s i ) Representing a current health concept s' i And corresponding health reference criteria concept s i 0 ≦ sima (s' i ,s i )≤1，1≤i≤f，1≤j≤g；
Wherein the atom concept similarity sima (s' i ,s i ) The calculation method comprises the following steps:
firstly, dividing the atomic conceptual attributes of the health condition and the health condition reference standard into a text type and an interval type, wherein the text type refers to personal information acquired by a virtual sensor of a built-in mobile phone, and the interval type refers to parameter data acquired by a physiological sensor group, an environmental sensor group and a motion sensor group respectively;
then the following treatments are respectively carried out:
establishing concept s 'for text-type attributes' i 、s i Attribute vector of (1), where s' i Is present and s i Note of attribute location also present as 1, s' i Exist but s i Attribute location of absence is recorded as 0, s' i 、s i The text type attribute similarity calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,
is corresponding s' i 、s i Is a unit vector, ζ is a weight, Dep (s' i )，Dep(s i ) Respectively represent s' i 、s i Conceptual depth in S', S;
for the interzone type attribute, assume s' i ＝[s' a ,s' b ]，s i ＝[s a ,s b ]And let s' a ,s' b ,s a ,s b Is ranked as s 1 ,s 2 ,s 3 ,s 4 S 'to' i 、s i The interval type attribute similarity calculation formula is as follows:
wherein sgn () is a sign function;
and finally, weighting and summing the text type attribute similarity and the interval type attribute similarity to obtain the attribute-based atomic concept similarity, which is as follows:
wherein the content of the first and second substances,
a m ,b n respectively are a text type attribute coefficient and an interval type attribute coefficient; m, N are respectively the current health status concept s' i And the concept s of the corresponding health condition reference standard i The number of text type attributes and interval type attributes.
Step 32, comparing the semantic similarity obtained by the step 31 with the abnormal health condition threshold value delta abnormal Health emergency threshold δ urgent Comparing, identifying health risks, and providing care services by the edge cloud-driven health care robot according to identification results, wherein the care services comprise abnormal services (including abnormal reminding, meal reminding and movement reminding) and emergency services (including emergency alarm, drug delivery and emergency medical treatment);
specifically, if the semantic similarity Sim (S', S) is smaller than the abnormal health condition threshold δ abnormal (0<Sim(S',S)<δ abnormal ) Health risk is identified as "normal"; if the semantic similarity Sim (S', S) is larger than or equal to the health abnormal condition threshold value delta abnormal And is less than the health emergency threshold δ urgent (δ abnormal ≤Sim(S',S)<δ urgent ) If the health risk is identified as abnormal, the edge cloud-driven health care robot provides abnormal services (including abnormal reminding, meal reminding and exercise reminding); if the semantic similarity Sim (S', S) is greater than or equal to the health emergency threshold delta urgent (δ urgent Sim (S', S)), the health risk is identified as "emergency", and emergency services (including emergency alerts, drug delivery, emergency medical visits) are provided by the edge cloud driven health care robot.
The invention also provides a situation-oriented health risk identification system, which comprises a mobile intelligent terminal with a built-in virtual sensor group, a physiological sensor group, an environmental sensor group, a motion sensor group, an edge cloud (including an edge server) and a health care robot; the virtual sensor group collects personal information of a user, including age, sex, weight, height, family genetic history and diet preference; the physiological sensor group collects heart rate, electrocardio, blood pressure and blood oxygen through sensors; the environment sensor group acquires temperature, humidity and air quality through a sensor; the motion sensor group acquires the position, the motion and the body posture of a user through a sensor; the edge cloud collects the data through a server (namely an edge server) positioned at the edge of the network, establishes a knowledge graph instance model for situation-oriented health risk identification, identifies health risks and drives the health care robot to provide care services.
In the present embodiment, the virtual sensor group, the physiological sensor group, the environmental sensor group, the motion sensor group, and the health service set are customized according to the specific monitored subject. The physiological sensor group adopts a heart rate sensor, an electrocardio sensor, a blood pressure sensor and a blood oxygen sensor; the environment sensor group adopts a temperature sensor, a humidity sensor and a PM2.5 sensor; the motion sensor adopts a position sensor, an acceleration sensor, a gyroscope, a gravity sensor and a direction sensor; the edge server adopts a Cortex A8 processor chip, a 250G Byte solid state disk, a Linux operating system Xen VMM, a fusion gateway (Bluetooth, ZigBee, WIFI, 6LoWPAN) and a 10M/100M network card. The fusion gateway, namely four coordinators with built-in Bluetooth, ZigBee, WIFI and 6LoWPAN short-range protocols, uniformly processes received various sensor data according to a predefined data packet format, and realizes data fusion, external network access and data forwarding.
The system comprises the following working steps:
step S1, starting a virtual sensor on the mobile intelligent terminal, and registering personal information of the user, wherein the personal information comprises age, sex, weight, height, family genetic history and diet preference;
step S2, starting a physiological sensor group which comprises a plurality of physiological sensor sensors and collects heart rate, electrocardio, blood pressure and blood oxygen;
step S3, starting an environment sensor group, which comprises a plurality of environment sensors and collects temperature, humidity and air quality;
step S4, starting a motion sensor group which comprises a plurality of motion sensors and collects the position, the motion and the body posture of the user;
step S5, the edge server collects personal information, physiological parameters, environmental parameters and motion parameters collected by the mobile intelligent terminal, the physiological sensor group, the environmental sensor group and the motion sensor group;
step S6, the edge cloud establishes a situation-oriented knowledge graph instance model for health risk identification according to the situation-oriented knowledge graph conceptual model for health risk identification and by combining the data acquired in the step S5;
step S7, recognizing health risks based on semantic similarity by the edge cloud, and driving the health care robot to provide care services according to recognition results, wherein the care services comprise abnormal services (including abnormal reminding, meal reminding and movement reminding) and emergency services (including emergency alarm, drug delivery and emergency medical treatment); the specific process is as follows:
step S71, calculating the semantic similarity between the current health condition and the health condition reference standard by using formulas (1) - (5) through an edge cloud;
step S72, comparing the semantic similarity obtained in step S71 with the abnormal health condition threshold value delta abnormal Health emergency threshold δ urgent And (3) comparison: for semantic similarity less than threshold δ abnormal Health risk is judged as "normal"; for semantic similarity greater than or equal to threshold δ abnormal And is less than the threshold value delta urgent If the health risk is judged to be abnormal, the edge cloud-driven health care robot provides abnormal services (including abnormal reminding, meal reminding and exercise reminding); for semantic similarity greater than or equal to threshold δ urgent And if the health risk is judged to be 'emergency', the health care robot is driven by the edge cloud to provide emergency services (including emergency alarm, drug delivery and emergency medical treatment).
The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the protection scope of the present invention.
Claims (6)
1. A situation-oriented health risk identification method is characterized by comprising the following steps:
step 1, establishing a situation-oriented knowledge graph conceptual model for health risk identification;
step 2, collecting sensor parameters, and establishing a situation-oriented knowledge graph instance model for health risk identification;
step 3, identifying health risks based on a semantic similarity algorithm, and driving a health care robot to provide care services;
in the step 1, the established model specifically comprises defining concepts of users, sensors, sensor parameters, health conditions and care services, and the corresponding relationships of the concepts comprise use, provision, assignment and triggering;
wherein the user comprises a patient; the sensors comprise a virtual sensor, a physiological sensor, an environment sensor and a motion sensor; the sensor parameters comprise personal information, physiological parameters, environmental parameters and motion parameters; the health condition comprises personal information condition, physiological parameter condition, environmental parameter condition and motion parameter condition; the care service comprises abnormal service and emergency service; usage indicates user usage of the sensor; providing a value of a monitoring parameter indicative of a sensor providing the sensor; the valuation indicates that the monitored parameter of the sensor changes the health condition of the user; triggering means that the health condition triggers a health service;
the specific content of the step 2 is as follows:
step 21, configuring a virtual sensor group, a physiological sensor group, an environmental sensor group, a motion sensor group and a care service set of a user according to requirements according to specific monitored chronic diseases, and respectively recording as: virtual sensor group
Physiological sensor group
Environment sensor group
Motion sensor group
Service set 1 ,Service 2 ,…,Service k }；
22, through the virtual sensor group, the physiological sensor group and the environment sensor groupThe group and the motion sensor group collect the following situation data of the user: personal information
Physiological parameter
Environmental parameter
And motion parameters
Obtaining a current health condition S' according to the personal information, the physiological parameters, the environmental parameters and the interval range of the motion parameters, thereby establishing a situation-oriented knowledge graph instance model for health risk identification;
the specific content of the step 3 is as follows:
step 31, calculating the semantic similarity between the current health condition and the health condition reference standard by using the following formula:
wherein f denotes the common number of concepts of the current health status S 'and the health status reference standard S, g denotes the different number of concepts of the current health status S' and the health status reference standard S, w i 、w f+j Is a weight and has a value of 0 ≦ w i ,w f+j ≤1；sima(s' i ,s i ) Representing a current health concept s' i And corresponding health reference criteria concept s i 0 ≦ sima (s' i ,s i )≤1，1≤i≤f，1≤j≤g；
Step 32, comparing the semantic similarity obtained by the step 31 with the abnormal health condition threshold value delta abnormal Health emergency threshold δ urgent Comparing, identifying health risks, and identifying based on the identificationAnd as a result, the health care robot is driven by the edge cloud to provide the care service.
2. The context-oriented health risk identification method of claim 1, wherein: in the step 31, the atom concept similarity sima (s' i ,s i ) The calculation method comprises the following steps:
firstly, the attributes of the atomic concepts of the health condition and the health condition reference standard are divided into a text type and an interval type, wherein the text type refers to personal information acquired by a virtual sensor, and the interval type refers to parameter data acquired by a physiological sensor group, an environmental sensor group and a motion sensor group respectively;
then, the following processes are respectively performed:
establishing concept s 'for text-type attributes' i 、s i Attribute vector of (1), where s' i Is present and s i Note of attribute location also present as 1, s' i Exist but s i Attribute location of absence is recorded as 0, s' i 、s i The text type attribute similarity calculation formula is as follows:
wherein the content of the first and second substances,
is corresponding s' i 、s i Is a unit vector, ζ is a weight, Dep (s' i )，Dep(s i ) Respectively represent s' i 、s i Conceptual depth in S', S;
for the intertype attribute, s 'is assumed' i ＝[s' a ,s' b ]，s i ＝[s a ,s b ]And let s' a ,s' b ,s a ,s b S after sorting from big to small 1 ,s 2 ,s 3 ,s 4 S 'to' i 、s i The interval type attribute similarity calculation formula is as follows:
wherein sgn () is a sign function;
and finally, weighting and summing the text type attribute similarity and the interval type attribute similarity to obtain the attribute-based atomic concept similarity, which is as follows:
wherein, the first and the second end of the pipe are connected with each other,
a m ,b n respectively are a text type attribute coefficient and an interval type attribute coefficient; m is a current health status concept s' m And the concept s of the corresponding health condition reference standard m Is the current health concept s' n And the concept s of the corresponding health condition reference standard n Number of interval type attributes.
3. The context-oriented health risk identification method of claim 1, wherein: in the step 32, if the semantic similarity Sim (S', S) is smaller than the abnormal health condition threshold δ abnormal Health risk is identified as "normal"; if the semantic similarity Sim (S', S) is larger than or equal to the health abnormal condition threshold value delta abnormal And is less than the health emergency threshold δ urgent If the health risk is identified as abnormal, the edge cloud drives the health care robot to provide abnormal service; if the semantic similarity Sim (S', S) is greater than or equal to the health emergency threshold delta urgent Health risk identified as "Emergency", the health care robot is driven by the edge cloud to provide emergency service.
4. The identification system of the situation-oriented health risk identification method according to claim 1, wherein: the system comprises a mobile intelligent terminal with a built-in virtual sensor group, a physiological sensor group, an environment sensor group, a motion sensor group, an edge cloud and a health care robot, wherein the virtual sensor group is used for acquiring personal information of a user; the physiological sensor group is used for collecting heart rate, electrocardio, blood pressure and blood oxygen; the environment sensor group is used for collecting temperature, humidity and air quality; the motion sensor group is used for acquiring the position, the motion and the body posture of the user; the edge cloud collects data collected by the virtual sensor, the physiological sensor, the environmental sensor and the motion sensor through the edge server, establishes a knowledge graph instance model for situation-oriented health risk identification, identifies health risks and drives the health care robot to provide care service.
5. The identification system of claim 4, wherein: the physiological sensor group adopts a heart rate sensor, an electrocardio sensor, a blood pressure sensor and a blood oxygen sensor; the environment sensor group adopts a temperature sensor, a humidity sensor and a PM2.5 sensor; the motion sensor adopts an acceleration sensor, a gyroscope, a gravity sensor and a direction sensor.
6. The identification system of claim 5, wherein: the edge server adopts a Cortex A8 processor chip, a 250G Byte solid state disk, a Linux operating system Xen VMM, a fusion gateway and a 10M/100M network card.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010548021.2A CN111681761B (en) | 2020-06-16 | 2020-06-16 | Situation-oriented health risk identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010548021.2A CN111681761B (en) | 2020-06-16 | 2020-06-16 | Situation-oriented health risk identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111681761A CN111681761A (en) | 2020-09-18 |
| CN111681761B true CN111681761B (en) | 2022-08-09 |
Family
ID=72455164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010548021.2A Active CN111681761B (en) | 2020-06-16 | 2020-06-16 | Situation-oriented health risk identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111681761B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114432607B (en) * | 2021-12-30 | 2022-09-13 | 安徽上造智能设备科技有限公司 | Protective clothing equipped with intelligent wearable equipment and working method thereof |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559397A (en) * | 2013-10-31 | 2014-02-05 | 合肥学院 | Intelligent service system and method applied to hypertension chronic diseases |
| EP3552177A1 (en) * | 2016-12-12 | 2019-10-16 | Koninklijke Philips N.V. | System and method for facilitating computational analysis of a health condition |
| CN109446700A (en) * | 2018-11-09 | 2019-03-08 | 福州大学 | Modeling and the method for execution when operation towards smart home context-aware services |
-
2020
- 2020-06-16 CN CN202010548021.2A patent/CN111681761B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111681761A (en) | 2020-09-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Azimi et al. | Internet of things for remote elderly monitoring: a study from user-centered perspective | |
| JP5250827B2 (en) | Action history generation method and action history generation system | |
| CN108778097A (en) | Device and method for assessing heart failure | |
| US20140324459A1 (en) | Automatic health monitoring alerts | |
| CN104207753A (en) | Personal health monitoring system | |
| Babu et al. | Cloud-enabled remote health monitoring system | |
| CN106897569A (en) | A kind of method of work of intelligent health terminal system | |
| Sahu et al. | The internet of things in geriatric healthcare | |
| WO2021121226A1 (en) | Electrocardiosignal prediction method and device, terminals, and storage medium | |
| Bott et al. | Towards new scopes: sensor-enhanced regional health information systems | |
| CN114883006A (en) | Health monitoring and dynamic management system based on big data and application | |
| Ahmed | An intelligent healthcare service to monitor vital signs in daily life–a case study on health-iot | |
| CN109378086A (en) | The analysis of family endowment health and fitness information Intellisense and early warning system under environment of internet of things | |
| CN111681761B (en) | Situation-oriented health risk identification method and system | |
| Nurshod et al. | Remote monitoring system architectures in healthcare | |
| Nugent et al. | Managing sensor data in ambient assisted living | |
| Chen et al. | An adaptive sensor data segments selection method for wearable health care services | |
| CA3175840A1 (en) | System and methods utilizing artificial intelligence algorithms to analyze wearable activity tracker data | |
| CN113643814A (en) | Health management scheme recommendation method, device, equipment and storage medium | |
| Panagiotou et al. | A multi: modal decision making system for an ambient assisted living environment | |
| CN115590483B (en) | Smart phone with health measurement system | |
| CN115662631A (en) | AI intelligence discrimination-based nursing home management system | |
| Aghav et al. | Health track | |
| US20210407667A1 (en) | Systems and methods for prediction of unnecessary emergency room visits | |
| Pereira et al. | A movement activity recognition pervasive system for patient monitoring in ambient assisted living |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |