WO2017193497A1 - 基于融合模型的智能化健康管理服务器、***及其控制方法 - Google Patents

基于融合模型的智能化健康管理服务器、***及其控制方法 Download PDF

Info

Publication number
WO2017193497A1
WO2017193497A1 PCT/CN2016/096213 CN2016096213W WO2017193497A1 WO 2017193497 A1 WO2017193497 A1 WO 2017193497A1 CN 2016096213 W CN2016096213 W CN 2016096213W WO 2017193497 A1 WO2017193497 A1 WO 2017193497A1
Authority
WO
WIPO (PCT)
Prior art keywords
health management
feature
state vector
data
source heterogeneous
Prior art date
Application number
PCT/CN2016/096213
Other languages
English (en)
French (fr)
Inventor
包磊
Original Assignee
包磊
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 包磊 filed Critical 包磊
Publication of WO2017193497A1 publication Critical patent/WO2017193497A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/40—Support for services or applications
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/22—Social work or social welfare, e.g. community support activities or counselling services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks

Definitions

  • the invention relates to the field of intelligent medical treatment, in particular to an intelligent health management server, a system and a control method thereof based on a fusion model.
  • a typical wearable somatosensory network node platform can accurately collect physiological signals through biosensors, and wirelessly transmit the data processed by the microcontroller to the intelligent terminal, and all sensor data is collected by the intelligent terminal, and Further processing, convergence, and then transmitted to the central monitoring server via wireless LAN, Bluetooth or 3G/4G network.
  • the core problem can be attributed to the acquisition, storage, transmission, analysis and utilization of health information.
  • the advent of various mobile monitoring instruments has shown that research on mobile medical systems has yielded considerable results.
  • Analytical processing for example, only analyzing heart rate, brain waves, etc., cannot comprehensively analyze human health status, behavioral habits, etc., and thus cannot provide reliable treatment means and prevention means, and the user experience is poor.
  • the technical problem to be solved by the embodiments of the present invention is to provide an intelligent health management server, system and control method based on the fusion model, which overcomes the fact that the basic sensing information in the prior art is too single to provide a comprehensive and reliable evaluation method and Defects in preventive measures.
  • an embodiment of the present invention provides an intelligent health management control method based on a fusion model, including:
  • the health management decision is outputted to provide a feedback intervention training mechanism based on the health management decision.
  • the method further comprises storing the multi-source heterogeneous sensing feature signal and the feature state vector set to a personal profile, the personal profile corresponding to a single user;
  • the steps of obtaining the diagnosis/prediction strategy include:
  • the step of analyzing the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision includes:
  • the step of acquiring the multi-source heterogeneous sensing feature signal includes:
  • the step of acquiring multi-source heterogeneous sensing fusion data includes:
  • the multi-source heterogeneous sensing raw data is normalized to generate multi-source heterogeneous sensing fusion data.
  • the present invention also provides an intelligent health management system based on a fusion model, the system comprising a data collection device, a health management server, and an interactive terminal;
  • the data collection device is configured to collect multi-source heterogeneous sensing raw data
  • the health management server includes:
  • a feature acquisition module configured to acquire a multi-source heterogeneous sensing feature signal from the multi-source heterogeneous sensing raw data
  • a state vector identification module configured to identify a feature state vector set for reflecting a human health state according to the multi-source heterogeneous sensing feature signal, where the feature state vector set includes a motion state vector, an emotional state vector, a sleep state vector, and Position state vector
  • a decision module configured to obtain a diagnosis/prediction strategy, and analyze the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision
  • the interactive terminal is configured to output the health management decision, and provide a feedback intervention training mechanism based on the health management decision for the user.
  • the health management server further includes means for storing the multi-source heterogeneous sensing feature signal and the feature state vector set, the personal profile corresponding to a single user;
  • the decision module further includes:
  • a personalized data acquisition module for acquiring personalized health management data in the cloud
  • a policy generating module configured to generate a diagnosis/prediction strategy according to the personal profile and the cloud personalized health management data
  • the decision fusion module is configured to fuse the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision.
  • the decision fusion module further includes:
  • a health status sample module configured to calculate a multi-source information feature vector set of the user according to the profile
  • a health state decision module configured to perform fusion analysis on the multi-source information feature vector set and the feature state vector set according to the diagnosis/prediction policy, to generate health management and abnormal intervention means for the user.
  • the feature acquiring module includes:
  • the original data receiving module is configured to receive the multi-source heterogeneous sensing raw data collected by the data collection device;
  • the data fusion module is configured to normalize the multi-source heterogeneous sensing raw data to generate multi-source heterogeneous sensing fusion data.
  • a feature fusion module configured to extract a feature set of the multi-source heterogeneous sensing fusion data according to a preset data processing algorithm, as a multi-source heterogeneous sensing feature signal.
  • the present invention also provides an intelligent health management server based on a fusion model, the server comprising:
  • a feature acquisition module configured to acquire a multi-source heterogeneous sensing feature signal
  • a state vector identification module configured to identify a feature state vector set for reflecting a human health state according to the multi-source heterogeneous sensing feature signal, where the feature state vector set includes a motion state vector, an emotional state vector, a sleep state vector, and Position state vector
  • a decision module configured to obtain a diagnosis/prediction strategy, analyze and analyze the feature state vector set according to the diagnosis/prediction strategy, generate a health management decision, and output a health management decision through an interactive terminal to provide a user based on the health management Decision feedback intervention training mechanism.
  • Embodiments of the present invention have the following beneficial effects: identifying a motion state vector from a multi-source heterogeneous sensing feature signal by exploring a mapping relationship between multi-source heterogeneous sensing information and motion state, emotional state, sleep state, and position state
  • the four feature state vector sets, the emotional state vector, the sleep state vector and the position state vector can realize multi-scale and high-precision quantitative calibration, thereby being able to integrate the health of the decision-making individual
  • Conditions and treatments or preventive measures should be taken to guide users to achieve optimal and reliable health regulation and improve user experience.
  • FIG. 1 is a flow chart of a first embodiment of an intelligent health management control method based on a fusion model provided by the present invention
  • FIG. 2 is a flow chart of a second embodiment of an intelligent health management control method based on a fusion model provided by the present invention
  • FIG. 3 is a flowchart of a third embodiment of an intelligent health management control method based on a fusion model provided by the present invention
  • FIG. 4 is a schematic diagram of a motion state vector identification method according to a preferred embodiment of the present invention.
  • FIG. 5 is a schematic diagram of an emotional state vector identification method according to another preferred embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of a first embodiment of an intelligent health management system based on a fusion model provided by the present invention
  • FIG. 7 is a schematic structural diagram of a second embodiment of an intelligent health management system based on a fusion model provided by the present invention.
  • FIG. 8 is a schematic structural diagram of a third embodiment of an intelligent health management system based on a fusion model provided by the present invention.
  • FIG. 1 is a flowchart of a first embodiment of an intelligent health management control method based on a fusion model provided by the present invention, where the method includes:
  • Step S11 Acquire a multi-source heterogeneous sensing feature signal.
  • the step of acquiring the multi-source heterogeneous sensing feature signal includes a process of data fusion.
  • the step of data fusion may include: collecting multi-source heterogeneous sensing raw data by using a sensor; normalizing the multi-source heterogeneous sensing raw data format to generate multi-source heterogeneous sensing fusion data.
  • the multi-source heterogeneous sensing raw data may include GPS data, acceleration data, ECG data, pulse data, skin temperature data, respiratory data, etc. These data formats are various, and after data fusion, all data is integrated into a unified format. Easy data analysis.
  • the step of acquiring the multi-source heterogeneous sensing feature signal further includes a process of feature fusion.
  • the step of the feature fusion may include: acquiring the multi-source heterogeneous sensing fusion data; extracting the feature set of the multi-source heterogeneous sensing fusion data according to a preset data processing algorithm, as the multi-source heterogeneous sensing characteristic signal. Since multi-source heterogeneous sensing fusion data contains more information, some of which are not related to health management, or have low credibility and need to be screened out, various methods can be used for multi-source heterogeneous sensing fusion. The data is processed to extract feature sets.
  • High-quality, high-confidence multi-source heterogeneous sensing fusion data can be extracted by signal confidence enhancement techniques such as anti-jamming algorithms and signal quality assessment, such as noise suppression and artifact cancellation based on mathematical morphology and empirical mode decomposition, based on Signal quality assessment of constrained timing estimation models, etc. Then, by using differential threshold, wavelet analysis, data mining, neural network and other algorithms, the feature set of multi-source heterogeneous sensing fusion data can be collaboratively analyzed and used as multi-source heterogeneous sensing characteristic signal.
  • signal confidence enhancement techniques such as anti-jamming algorithms and signal quality assessment, such as noise suppression and artifact cancellation based on mathematical morphology and empirical mode decomposition, based on Signal quality assessment of constrained timing estimation models, etc. Then, by using differential threshold, wavelet analysis, data mining, neural network and other algorithms, the feature set of multi-source heterogeneous sensing fusion data can be collaboratively analyzed and used as multi-source heterogeneous sensing characteristic signal.
  • Step S12 Identify a feature state vector set for reflecting a human health state according to the multi-source heterogeneous sensing feature signal, where the feature state vector set includes a motion state vector, an emotional state vector, a sleep state vector, and a position state vector.
  • the motion state vector, the emotion state vector, the sleep state vector, and the position state vector belong to the decision level information, and according to the four state vector information, the body state, the mental state, and the daily behavior habits of the human body can be accurately reflected.
  • Step S13 Acquire a diagnosis/prediction strategy, and analyze the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision.
  • diagnostic/predictive strategies can be generated based on health data and operational management data from different sources, such as homes, hospitals, and service centers. Analyze multi-modal, multi-dimensional feature state vector sets, including physiological information, psychological information, spatial information, and motion information It can combine the influence of subjective and objective factors on health status to generate more refined and reliable health management decisions.
  • Step S14 Output the health management decision to provide a feedback intervention training mechanism based on the health management decision.
  • the health management decision can be output to the user through an interaction mode of the human-machine interface such as sound, light, electricity, touch, vision, hearing, etc.; after the user performs feedback training according to the health management decision, step S11 is repeated to form a health management Closed loop.
  • the intelligent health management control method based on the fusion model provided by the invention explores the mapping relationship between multi-source heterogeneous sensing information and motion state, emotional state, sleep state and position state, from the multi-source heterogeneous sensing characteristic signal Identifying four feature state vector sets, such as motion state vector, emotional state vector, sleep state vector and position state vector, can achieve multi-scale and high-precision quantitative calibration, thereby being able to integrate the health status of the decision-making individual and the treatment or prevention measures to be taken. Guide users to achieve optimized and reliable health control and enhance user experience.
  • FIG. 2 is a flowchart of a second embodiment of an intelligent health management control method based on a fusion model provided by the present invention, where the method includes:
  • Step S21 Acquire a multi-source heterogeneous sensing feature signal, and store the multi-source heterogeneous sensing feature signal and the feature state vector set to a personal profile, where the personal profile corresponds to a single user. Specifically, the personal profile stores all of the user's previous multi-dimensional heterogeneous sensing feature signals and feature state vector sets to form the user's personal health file.
  • Step S22 Identify, according to the multi-source heterogeneous sensing feature signal, a feature state vector set for reflecting a human health state, where the feature state vector set includes a motion state vector, an emotional state vector, a sleep state vector, and a position state vector.
  • Step S23 Acquire cloud personalized health management data.
  • the cloud personalized health management data is stored in the cloud platform and can include health data and professional knowledge from different sources such as families, hospitals, and service centers.
  • Step S24 Generate a diagnosis/prediction strategy according to the personal profile and the cloud personalized health management data.
  • the diagnosis/prediction strategy can only reflect the individual differences according to the preset professional knowledge experience, which may result in inaccurate results and affect the user experience.
  • Dynamic generation of diagnostic/predictive strategies can improve the personalization, accuracy and reliability of diagnostic/predictive results.
  • Step S25 Perform fusion analysis of the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision.
  • Step S26 Output the health management decision to provide a feedback intervention training mechanism based on the health management decision.
  • the health management decision can be output to the user through an interaction mode of the human-machine interface such as sound, light, electricity, touch, vision, hearing, etc.; after the user performs feedback training according to the health management decision, step S21 is repeated to form a health management Closed loop.
  • the intelligent health management control method based on the fusion model provided by the embodiment of the invention can provide personalized and refined real-time long-term health management for the above patients, enhance the regulation function of the autonomous nervous system, and enhance the stress response level of the central nervous system. .
  • the method will be described below by taking a chronic disease patient as an example.
  • FIG. 3 it is a flowchart of a third embodiment of an intelligent health management control method based on a fusion model provided by the present invention, where the method includes:
  • Step S31 Acquire a multi-source heterogeneous sensing feature signal, and store the multi-source heterogeneous sensing feature signal and the feature state vector set to a personal file, where the personal profile corresponds to a single user.
  • Step S32 Identify, according to the multi-source heterogeneous sensing feature signal, a feature state vector set for reflecting a human health state, where the feature state vector set includes a motion state vector, an emotion state vector, a sleep state vector, and a position state vector.
  • Step S33 Acquire cloud personalized health management data.
  • Step S34 Generate a diagnosis/prediction strategy according to the personal profile and the cloud personalized health management data.
  • Step S35 Calculate the multi-source information feature vector set of the user according to the personal profile.
  • Step S36 Perform fusion analysis on the multi-source information feature vector set according to the diagnosis/prediction strategy to generate a health management and abnormal state intervention means for the user.
  • Step S37 Output the health management and abnormal intervention means to the user, so that the user performs feedback training according to the health management and abnormal intervention means.
  • the health management decision can be output to the user through the interaction mode of the human-machine interface such as sound, light, electricity, touch, vision, hearing, etc.; after the user performs feedback training according to the health management and the abnormal intervention means, step S31 is repeated to form Closed loop of health management.
  • the embodiment of the invention discovers the health management rules related to the individual patient through data mining means, and adopts corresponding intervention means to guide the patient to realize the optimized and differentiated health regulation; further, the health management service can be realized through the mining of the operation management data. Quality control and optimization.
  • the method of identifying the motion state vector is as shown in FIG.
  • the acceleration sensor configured by the wearable device or the smart phone is used to monitor the movement state of the human body, and the daily exercise behavior of the human body is associated with the physical and mental health of the individual to monitor the daily movement of the human body, for scientific development. Exercise and fitness programs are important to improve your health. Acceleration sensor based research theory in the field of human motion monitoring and commonly used research methods, after the acceleration signal is preprocessed by high-pass filter to remove signal DC offset, etc., extract the variance, mean, maximum, minimum, wavelet in the time window.
  • the characteristics of the transform coefficient, the Fourier frequency characteristic, and the like, and the feature evaluation value is compared with a preset threshold to identify the daily motion of the human body.
  • the method is based on the multi-level layer classifier structure.
  • the adaptive feature extraction algorithm is used to train multiple classifiers.
  • One classifier uses acceleration and gyroscope to identify complex poses through data fusion.
  • the other classifier uses acceleration only. Identify simple gestures and get health indicators such as daily exercise volume and energy consumption.
  • the method of identifying the emotional state vector is as shown in FIG.
  • this method by calculating heart rate variability, heart and respiratory rhythm patterns, collecting self-evaluation and life event scales, combined with subjective and objective individual information, based on D-S evidence theory decision-making layer fusion model to achieve multi-level hierarchical psychology Quantitative assessment of stress.
  • the method for identifying a sleep state vector includes: time domain/frequency domain/geometric/nonlinear features and body pose feature sets based on heart rate (or pulse rate) variability, through data fusion and statistics
  • the analysis method approximates the clinical sleep assessment index (respiratory disorder/micro-awakening index, sleep stage, etc.) and obtains a sleep state vector.
  • the location state vector identification method discards the practice of updating the location by the GPRS long-connection server and the short message instruction, and adaptively switches the location report and tracking through the motion state to implement the accurate location service.
  • FIG. 6 is a schematic structural diagram of a first embodiment of an intelligent health management system based on a fusion model provided by the present invention. As shown in FIG. 6, the system includes a data collection device 600, a health management server 700, and an interactive terminal 800.
  • the data collection device 600 is configured to collect multi-source heterogeneous sensing raw data.
  • the data collection device 600 can be any one or more of the wearable devices and/or portable devices, such as clothes, hats, glasses, bracelets, watches, shoes, mobile phones, tablets, etc., which are compact and power-saving.
  • Durable digitally record human signs and perceptions in a lossless, real-time manner, and seamlessly interface with medical resources through mobile Internet, cloud computing and big data analytics to achieve full monitoring of individual health and efficient use of medical resources.
  • the health management server 700 includes a feature acquisition module 710, a state vector identification module 720, and a decision module 730.
  • the feature obtaining module 710 is configured to obtain the multi-source heterogeneous sensing feature signal from the multi-source heterogeneous sensing raw data.
  • the feature acquiring module 710 includes: an original data receiving module 711, configured to receive the multi-source heterogeneous sensing raw data collected by the data collecting device 600, and a data fusion module 712, configured to transmit the multi-source heterogeneous
  • the original data is normalized to generate multi-source heterogeneous sensing fusion data
  • the feature fusion module 713 is configured to extract the feature set of the multi-source heterogeneous sensing fusion data according to a preset data processing algorithm, as a multi-source difference The sensing characteristic signal.
  • the multi-source heterogeneous sensing raw data may include GPS data, acceleration data, ECG data, pulse data, skin temperature data, respiratory data, etc., and the data formats are various, and the data fusion module 712 is adopted. After the data is merged, all the data is merged into a unified format for data analysis.
  • the multi-source heterogeneous sensing fusion data contains more information, some of which are not related to health management, or the credibility is not high, and needs to be screened out, so various methods can be used for multi-source heterogeneous transmission.
  • Sense fusion data is processed to extract feature sets.
  • the feature fusion module 713 can extract high-quality and high-confidence multi-source heterogeneous sensing fusion data through signal confidence enhancement techniques such as anti-interference algorithm and signal quality evaluation, for example, noise suppression and pseudo-based on mathematical morphology and empirical mode decomposition. Difference elimination, signal quality evaluation based on constrained timing estimation model, etc. Then, by using differential threshold, wavelet analysis, data mining, neural network and other algorithms, the feature set of multi-source heterogeneous sensing fusion data can be collaboratively analyzed and used as multi-source heterogeneous sensing characteristic signal.
  • signal confidence enhancement techniques such as anti-interference algorithm and signal quality evaluation, for example, noise suppression and pseudo-based on mathematical morphology and empirical mode decomposition.
  • the state vector identification module 720 is configured to identify a feature state vector set for reflecting a human health state according to the multi-source heterogeneous sensing feature signal, where the feature state vector set includes a motion state vector, an emotional state vector, and a sleep state vector. And position status vector.
  • the motion state vector, the emotion state vector, the sleep state vector, and the position state vector belong to the decision level information, and according to the four state vector information, the body state, the mental state, and the daily behavior habits of the human body can be accurately reflected.
  • the decision module 730 is configured to obtain a diagnosis/prediction strategy, and analyze the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision. For example, the decision module 730 can generate diagnostic/predictive policies based on health data and operational management data from different sources, such as homes, hospitals, service centers, and the like. Through fusion analysis of multi-modal, multi-dimensional feature state vector sets, including physiological information, psychological information, spatial information and motion information, combined with the influence of subjective and objective factors on health status, generate more refined and reliable health management decision making.
  • the interaction terminal 800 is configured to output the health management decision, and provide a feedback intervention training mechanism based on the health management decision for the user.
  • the interactive terminal 800 can be the same electronic device as the data collection device, such as a mobile phone, a tablet, or the like.
  • the interactive terminal 800 can output the health management decision to the user through an interaction mode of the human-machine interface such as sound, light, electricity, touch, vision, hearing, etc.; after the user performs feedback training according to the health management decision, the data collection device 600 collects again.
  • Real-time multi-source heterogeneous sensing of raw data forms a closed loop of health management.
  • the intelligent health management system based on the fusion model provided by the invention explores multi-source heterogeneous sensing signals
  • the relationship between the information and the motion state, the emotional state, the sleep state, and the position state, and the four feature state vectors of the motion state vector, the emotion state vector, the sleep state vector, and the position state vector are identified from the multi-source heterogeneous sensing feature signal.
  • Multi-scale and high-precision quantitative calibration can be realized, which can integrate the health status of decision-making individuals and the treatment or prevention measures that should be taken to guide users to achieve optimized and reliable health regulation and improve user experience.
  • FIG. 7 is a schematic structural diagram of a second embodiment of an intelligent health management system based on a fusion model provided by the present invention.
  • the health management server 700 also includes a personal profile 740.
  • a profile 740 is used to store the multi-source heterogeneous sensing feature signal and the feature state vector set, the personal profile corresponding to a single user. Specifically, the personal profile stores all of the user's previous multi-dimensional heterogeneous sensing feature signals and feature state vector sets to form the user's personal health file.
  • the decision module 730 further includes:
  • the personalized data obtaining module 731 is configured to obtain cloud personalized health management data.
  • the cloud personalized health management data is stored in the cloud platform and can include health data and professional knowledge from different sources such as families, hospitals, and service centers.
  • the policy generation module 732 is configured to generate a diagnosis/prediction policy according to the personal profile and the cloud personalized health management data.
  • the diagnosis/prediction strategy can only reflect the individual differences according to the preset professional knowledge experience, which may result in inaccurate results and affect the user experience.
  • the policy generation module 732 dynamically generates a diagnosis/prediction strategy by integrating the personal profile and the personalized health management data, which can improve the personalization, accuracy, and reliability of the diagnosis/prediction result.
  • the decision fusion module 733 is configured to fuse the feature state vector set according to the diagnosis/prediction strategy to generate a health management decision.
  • the intelligent health management system based on the fusion model provided by the embodiment of the invention can provide personalized and refined real-time long-term health management for the above patients, enhance the regulation function of the autonomous nervous system, and enhance the stress response level of the central nervous system. The method will be described below by taking a chronic disease patient as an example.
  • FIG. 8 is a schematic structural diagram of a third embodiment of an intelligent health management system based on a fusion model provided by the present invention.
  • the decision fusion module 733 further includes:
  • the chronic disease sample module 733A is configured to calculate a chronic disease management sample of the user according to the personal profile 740.
  • the chronic disease decision module 733B is configured to fuse the chronic disease management sample and the feature state vector according to the diagnosis/prediction strategy to generate a chronic disease intervention means for the user.
  • the chronic disease intervention means is output to the user through the interactive terminal 800, so that the user performs feedback training according to the chronic disease intervention means.
  • the embodiment of the invention discovers the chronic disease management law related to the individual patient through data mining means, and adopts corresponding intervention means to guide the patient to achieve optimal and differentiated health regulation; further, the health management service can be realized through mining the operation management data. Quality control and optimization.
  • the state vector identification module 720 can include a motion state vector identification module, an emotional state vector recognition module, a sleep state vector recognition module, and a position state vector identification module.
  • the method of identifying the motion state vector identification module is as shown in FIG.
  • the acceleration sensor configured by the wearable device or the smart phone is used to monitor the movement state of the human body, and the daily exercise behavior of the human body is associated with the physical and mental health of the individual to monitor the daily movement of the human body.
  • the exercise and fitness program is important to improve your health.
  • Research theory and common research in the field of human motion monitoring based on acceleration sensors The method is to preprocess the acceleration signal through a high-pass filter to remove the signal DC offset, and extract the variance, mean, maximum, minimum, wavelet transform coefficient, Fourier frequency characteristics, etc. in the time window, and then use the The feature evaluation value is compared with a preset threshold to identify the daily movement of the human body.
  • the method is based on the multi-level layer classifier structure.
  • the adaptive feature extraction algorithm is used to train multiple classifiers.
  • One classifier uses acceleration and gyroscope to identify complex poses through data fusion.
  • the other classifier uses acceleration only. Identify simple gestures and get health indicators such as daily exercise volume and energy consumption.
  • the method of identifying the emotional state vector module is as shown in FIG. 6.
  • FIG. 6 by calculating heart rate variability, heart and respiratory rhythm patterns, collecting self-evaluation and life event scales, combined with subjective and objective individual information, based on D-S evidence theory decision-making layer fusion model to achieve multi-level hierarchical psychological stress quantitative assessment.
  • the sleep state vector module approximates the clinical time based on the time domain/frequency domain/geometry/non-linear characteristics of the heart rate (or pulse rate) variability and the human body pose feature set by data fusion and statistical analysis methods.
  • the sleep assessment index (respiratory disorder/micro-awakening index, sleep staging, etc.) obtains a sleep state vector.
  • the location state vector identification module discards the GPRS long-link server and the short message instruction update location, and adaptively switches the location report and tracking through the motion state to implement the accurate location service.
  • the intelligent health management system based on the fusion model provided by the invention adopts a modular structure hardware circuit optimization design scheme of a general structure, including a central processing unit, a sensing unit, an interface and a human-computer interaction unit, etc., and can provide personalized services for users. Customized solutions to reduce system complexity and power consumption.
  • the highly integrated motion, space and physiological sensing sensing unit realizes the addition/angular speed, magnetic field strength and spatial position under the premise of satisfying low power consumption, low noise, stability and reliability.
  • the acquisition, processing, storage, and transmission of sensing information such as electrocardiogram, pulse, and breathing, and outputting corresponding control signals.
  • the invention realizes the cross-platform multi-intelligent terminal software based on the HTML5 framework, optimizes the function module by platformizing the software system, reduces the demand for the RAM/ROM, improves the running speed, and realizes the basic information, the motion information, the position information and the emotional information of the user. , data analysis, storage, synchronization, collaboration, security and sharing of sleep information.
  • the embodiment of the invention optimizes the design of the software and hardware system architecture by integrating the smart wearable device with multi-source heterogeneous sensing information, and integrates a new anti-interference algorithm, a signal quality evaluation algorithm and an eigenvalue extraction algorithm to obtain high quality and high quality.
  • Trusted sensing data combined with professional knowledge and experience to build a decision-level data fusion model, to achieve intelligent assessment of sleep state, motion state, position state and emotional state and personalized feedback training mechanism, while relying on the cloud computing platform center to establish individuals
  • the health management system for chronic patient groups represented by the elderly and sub-health groups, establishes individual patient health records through long-term follow-up records, collects individual physiological data and daily living habits information; uses information extraction, information fusion, and high Uyghur medical data mining and other technologies for data analysis, to provide patients with refined, personalized health management guidance; thereby improving the quality of life, giving early warning tips for sudden disease risks.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Health & Medical Sciences (AREA)
  • Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Multimedia (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Primary Health Care (AREA)
  • Child & Adolescent Psychology (AREA)
  • Computing Systems (AREA)
  • Medical Informatics (AREA)
  • Marketing (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本发明实施例公开了一种基于融合模型的智能化健康管理服务器、***及其控制方法。健康管理服务器从数据采集装置获取多源异构传感特征信号；根据所述多源异构传感信号识别用于反映人体健康状态的特征状态向量集；获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策；交互终端输出所述健康管理决策，并为用户提供基于所述健康管理决策的反馈干预训练机制。本发明提供的实施例，可实现多尺度高精准量化标定，从而能够融合决策个体的健康状况和应采取的干预或预防手段，指导用户实现优化可靠的健康调控，提升用户体验。

Description

基于融合模型的智能化健康管理服务器、***及其控制方法 技术领域
本发明涉及智能化医疗领域，尤其涉及一种基于融合模型的智能化健康管理服务器、***及其控制方法。
背景技术
近年来，为实现国家层面医疗模式的战略转变，大力发展低廉、微型、便捷、智能的分布式医疗和个体化医疗手段，对于广大人民福祉的改善、医疗卫生事业可持续发展和社会和谐稳定具有极其重要的意义。分布式医疗将重心下移到基层医院、社康中心甚至家庭，把服务从治疗前移到预防和早期诊断，把治病模式转变为治未病模式，将以往只有大医院才能使用的医疗仪器廉价化、微型化、操作简化，目前这种理念已被贯彻，在学术界和产业界的研究、开发如火如荼，方兴未艾。而个体化医疗将是未来的发展方向，利用随身携带、穿戴的智能硬件和数码产品，结合移动互联网和大数据分析，为每个人定制健康管理方案，实时监测身体状况，把服务端延伸到健康和亚健康人群，把治病模式转变为健康监护模式。这两种医疗模式代表了当前和未来的发展趋势，也对相应医疗仪器提出了全新的技术挑战，同时意味着巨大的学科发展机会和市场应用前景。
近年来，穿戴式技术、通信技术、云计算技术、大数据技术等飞速发展，为可移动式医疗***的发展带来了新的曙光。典型的可穿戴式躯感网节点平台能够通过生物传感器精确地采集生理信号，通过微控制器处理过的数据，以无线的方式传输到智能终端上，所有的传感器数据由智能终端负责收集，并进一步处理，融合，然后通过无线局域网、蓝牙或3G/4G网络传送到中央监控服务器。其核心问题可以归结为健康信息的获取、存储、传输、分析和利用。各种可移动式监测仪器的相继面世，说明可移动式医疗***的研究已经有了相当成果。
但是，目前面世的这些可移动式医疗***，往往只对单一维度的传感信息进行 分析处理，例如只分析心率、脑电波等，无法全面地分析人体健康状态、行为习惯等，因而无法提供可靠的治疗手段和预防手段，用户体验较差。
技术问题
本发明实施例所要解决的技术问题在于，提供一种基于融合模型的智能化健康管理服务器、***及其控制方法，克服现有技术中基础传感信息过于单一因而无法提供全面可靠的评估手段和预防手段的缺陷。
问题的解决方案
技术解决方案
为了解决上述技术问题，本发明实施例提供了一种基于融合模型的智能化健康管理控制方法，包括：
获取多源异构传感信号及其特征值；
根据所述多源异构传感信号提取用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策；
输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。
其中，所述方法还包括将所述多源异构传感特征信号和所述特征状态向量集存储至个人档案，所述个人档案对应于单个用户；
所述获取诊断/预测策略的步骤包括：
获取云端个性化健康管理数据；
根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略。
其中，所述根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策的步骤包括：
根据所述个人档案计算所述用户的多源信息特征向量集；
根据所述诊断/预测策略融合分析所述多源信息特征向量集，生成针对所述用户的健康管理和异常干预手段。
其中，所述获取多源异构传感特征信号的步骤包括：
获取多源异构传感融合数据；
根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。
其中，所述获取多源异构传感融合数据的步骤包括：
采集多源异构传感原始数据；
将所述多源异构传感原始数据归一化，生成多源异构传感融合数据。
另一方面，本发明还提供了一种基于融合模型的智能化健康管理***，所述***包括数据采集装置、健康管理服务器和交互终端；
所述数据采集装置用于采集多源异构传感原始数据；
所述健康管理服务器包括：
特征获取模块，用于从所述多源异构传感原始数据中获取多源异构传感特征信号；
状态向量识别模块，用于根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
决策模块，用于获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策；
所述交互终端用于输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。
其中，所述健康管理服务器还包括用于存储所述多源异构传感特征信号和所述特征状态向量集，所述个人档案对应于单个用户；
所述决策模块进一步包括：
个性化数据获取模块，用于获取云端个性化健康管理数据；
策略生成模块，用于根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略；
决策融合模块，用于根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策。
其中，所述决策融合模块进一步包括：
健康状态样本模块，用于根据所述个人档案计算所述用户的多源信息特征向量集；
健康状态决策模块，用于根据所述诊断/预测策略融合分析所述多源信息特征向量集和所述特征状态向量集，生成针对所述用户的健康管理和异常干预手段。
其中，所述特征获取模块包括：
原始数据接收模块，用于接收所述数据采集装置采集的多源异构传感原始数据；
数据融合模块，用于将所述多源异构传感原始数据归一化，生成多源异构传感融合数据。
特征融合模块，用于根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。
另外，本发明还提供了一种基于融合模型的智能化健康管理服务器，所述服务器包括：
特征获取模块，用于获取多源异构传感特征信号；
状态向量识别模块，用于根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
决策模块，用于获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策，并通过交互终端输出述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。
发明的有益效果
有益效果
实施本发明实施例，具有如下有益效果：通过探究多源异构传感信息与运动状态、情绪状态、睡眠状态及位置状态的映射关系，从多源异构传感特征信号中识别运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量这四个特征状态向量集，可实现多尺度高精准量化标定，从而能够融合决策个体的健康状 况和应采取的治疗或预防手段，指导用户实现优化可靠的健康调控，提升用户体验。
对附图的简要说明
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明提供的基于融合模型的智能化健康管理控制方法的第一实施例流程图；
图2是本发明提供的基于融合模型的智能化健康管理控制方法的第二实施例流程图；
图3是本发明提供的基于融合模型的智能化健康管理控制方法的第三实施例流程图；
图4是本发明一个优选实施例提供的运动状态向量识别方法的示意图；
图5是本发明另一个优选实施例提供的情绪状态向量识别方法的示意图；
图6是本发明提供的基于融合模型的智能化健康管理***的第一实施例结构示意图；
图7是本发明提供的基于融合模型的智能化健康管理***的第二实施例结构示意图；
图8是本发明提供的基于融合模型的智能化健康管理***的第三实施例结构示意图。
实施该发明的最佳实施例
本发明的最佳实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
请参见图1，是本发明提供的基于融合模型的智能化健康管理控制方法的第一实施例流程图，该方法包括：
步骤S11、获取多源异构传感特征信号。
具体地，获取多源异构传感特征信号的步骤包含了数据融合的过程。数据融合的步骤可以包括：通过传感器采集多源异构传感原始数据；将所述多源异构传感原始数据格式归一化，生成多源异构传感融合数据。其中多源异构传感原始数据可以包括GPS数据、加速度数据、心电数据、脉搏数据、皮温数据、呼吸数据等，这些数据格式多样，通过数据融合后，使所有数据融合为统一格式，便于数据分析。
具体地，获取多源异构传感特征信号的步骤还包含了特征融合的过程。特征融合的步骤可以包括：获取多源异构传感融合数据；根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。由于多源异构传感融合数据中包含了较多信息，其中有些信息与健康管理无关，或者可信度不高，需要加以筛除，因此可以采用各种方法对多源异构传感融合数据进行处理，提取特征集。通常可以通过抗干扰算法、信号质量评估等信号可信度增强技术提取高质量高可信多源异构传感融合数据，例如基于数学形态法及经验模式分解的噪声抑制及伪差消除、基于约束时序估计模型的信号质量评估等。然后可以通过运用差分阈值、小波分析、数据挖掘、神经网络等算法，协同分析与推算多源异构传感融合数据的特征集作为多源异构传感特征信号。
步骤S12、根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量。运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量属于决策级信息，根据这四个状态向量信息，能够较准确地反映人体的身体状态、心理状态及日常的行为习惯。
步骤S13、获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策。例如，可以根据来自家庭、医院、服务中心等不同来源的健康数据及运营管理数据来生成诊断/预测策略。通过融合分析多模态、多维度的特征状态向量集，包括生理信息、心理信息、空间信息及运动信息 ，可以结合主观、客观各因素对健康状态的影响，生成更加精细化可靠化的健康管理决策。
步骤S14、输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。例如，可以通过声、光、电、触觉、视觉、听觉等人机接口的交互模式，将健康管理决策输出给用户；用户根据健康管理决策进行反馈训练后，再重复步骤S11，形成健康管理的闭循环。
本发明提供的基于融合模型的智能化健康管理控制方法，通过探究多源异构传感信息与运动状态、情绪状态、睡眠状态及位置状态的映射关系，从多源异构传感特征信号中识别运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量这四个特征状态向量集，可实现多尺度高精准量化标定，从而能够融合决策个体的健康状况和应采取的治疗或预防手段，指导用户实现优化可靠的健康调控，提升用户体验。
请参见图2，是本发明提供的基于融合模型的智能化健康管理控制方法的第二实施例流程图，该方法包括：
步骤S21、获取多源异构传感特征信号，将所述多源异构传感特征信号和所述特征状态向量集存储至个人档案，所述个人档案对应于单个用户。具体地，个人档案中存储有该用户此前的所有多元异构传感特征信号和特征状态向量集，形成该用户的个人健康档案。
步骤S22、根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量。
步骤S23、获取云端个性化健康管理数据。云端个性化健康管理数据存储在云端平台，可以包括来自家庭、医院、服务中心等不同来源的健康数据及专业知识经验。
步骤S24、根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略。目前常见的移动式健康管理***中，由于没有个人历史数据做参考，诊断/预测策略只能依据预置的专业知识经验，无法反映个体差异性，可能造成结果不准确，影响用户体验。本实施例中通过融合个人档案和个性化健康管理数据 来动态地生成诊断/预测策略，能够提高诊断/预测结果的个性化、精确性和可靠性。
步骤S25、根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策。
步骤S26、输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。例如，可以通过声、光、电、触觉、视觉、听觉等人机接口的交互模式，将健康管理决策输出给用户；用户根据健康管理决策进行反馈训练后，再重复步骤S21，形成健康管理的闭循环。
本发明实施例，通过建立个人长期健康档案，综合专业知识经验与多源异构传感信息，在一定准则下加以自动分析、综合、支配和使用，通过数据级融合、特征级融合和决策级融合，获得对被测对象的一致性解释与描述，以完成所需的健康管理决策任务，通过对信息优化组合导出更多的有效信息，可以向患者本人、亲属及医院提供病人健康状况及健康管理实施效果的评估结果，为专业医务人员进行诊断提供参考。
特别是对于以下患者：1)属于慢性疾病患者，需要长期连续进行监测；2)属于已被医院告知为急性疾病突然发作的潜在患者；3)属于航空、军事、运动员、危险作业等行业人员，工作地点远离医院，且工作场所不适宜携带大型设备，有可能导致事故发生而不能提前预警的人群。本发明实施例提供的基于融合模型的智能化健康管理控制方法可以为以上患者提供个性化、精细化的实时长期健康管理，增强自治神经***的调控功能，增强中央神经***的应激反应水平等。下面将以慢性疾病患者为例说明本方法。
请参见图3，是本发明提供的基于融合模型的智能化健康管理控制方法的第三实施例流程图，该方法包括：
步骤S31、获取多源异构传感特征信号，将所述多源异构传感特征信号和所述特征状态向量集存储至个人档案，所述个人档案对应于单个用户。
步骤S32、根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量。
步骤S33、获取云端个性化健康管理数据。
步骤S34、根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略。
步骤S35、根据所述个人档案计算所述用户的多源信息特征向量集。
步骤S36、根据所述诊断/预测策略融合分析所述多源信息特征向量集，生成针对所述用户的健康管理及异常状态干预手段。
步骤S37、将所述健康管理和异常干预手段输出给所述用户，使所述用户根据所述健康管理和异常干预手段进行反馈训练。例如，可以通过声、光、电、触觉、视觉、听觉等人机接口的交互模式，将健康管理决策输出给用户；用户根据健康管理和异常干预手段进行反馈训练后，再重复步骤S31，形成健康管理的闭循环。
本发明实施例通过数据挖掘手段发现与个体患者相关的健康管理规律，采取对应的干预手段指导患者实现优化、差异化的健康调控；更进一步，还可以通过对运营管理数据的挖掘，实现健康服务质量的控制及优化。
在本发明的一个优选实施例中，运动状态向量的识别方法如图5所示。在该方法中，利用可穿戴设备或智能手机配置的加速度传感器实现对人体运动状态的监测，并将人体日常运动行为与个人的身心健康关联起来实现对人体的日常运动进行监测，对于制定科学的运动健身计划，改善身体健康状况具有重要意义。基于加速度传感器的人体运动监测领域的研究理论以及常用的研究方法，将加速度信号经过高通滤波器去除信号直流偏置等预处理后，提取时间窗内的方差，均值，最大值，最小值，小波变换系数，傅里叶频率特性等特征，再利用该特征评估值跟预先设置的阈值比较来识别人体日常运动。该方法针对多层次层分类器结构，采用自适应特征提取算法实现训练多个分类器，一类分类器采用加速度和陀螺仪，通过数据融合识别复杂姿态，另一类分类器只采用加速度用来识别简单姿态，同时可获得日常运动量和能量消耗等健康指标。
在本发明的一个优选实施例中，情绪状态向量的识别方法如图6所示。该方法中，通过计算心率变异性、心脏及呼吸节奏模式，采集自我评价及生活事件量表，结合主客观个体信息，基于D-S证据理论决策层融合模型实现多级分层心理 压力量化评估。
在本发明的一个优选实施例中，睡眠状态向量的识别方法包括：基于心率(或脉率)变异性的时域/频域/几何/非线性特征及人体姿态特征集，通过数据融合及统计分析方法逼近临床睡眠评估指标(呼吸紊乱/微觉醒指数、睡眠分期等)，得到睡眠状态向量。
在本发明的一个优选实施例中，位置状态向量识别方法摒弃了GPRS长连服务器和短信指令更新位置的做法，通过运动状态自适应切换位置上报及跟踪，实现精准位置服务。
请参见图6，是本发明提供的基于融合模型的智能化健康管理***的第一实施例结构示意图。如图6所示，该***包括数据采集装置600、健康管理服务器700和交互终端800。
数据采集装置600，用于采集多源异构传感原始数据。数据采集装置600可以是任意一种或多种穿戴式设备和/或便携式设备，例如衣服、帽子、眼镜、手环、手表、鞋子、手机、平板电脑等等，这种设备小巧贴身、省电耐用，可以通过无损、实时地数字化记录人的体征和感知信息，并通过移动互联网、云计算和大数据分析，与医疗资源无缝对接，实现对个体健康的充分监护和医疗资源的高效利用。
健康管理服务器700包括特征获取模块710、状态向量识别模块720和决策模块730。
特征获取模块710，用于从所述多源异构传感原始数据中获取多源异构传感特征信号。
具体地，特征获取模块710包括：原始数据接收模块711，用于接收所述数据采集装置600采集的多源异构传感原始数据；数据融合模块712，用于将所述多源异构传感原始数据归一化，生成多源异构传感融合数据；特征融合模块713，用于根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。
其中，多源异构传感原始数据可以包括GPS数据、加速度数据、心电数据、脉搏数据、皮温数据、呼吸数据等，这些数据格式多样，通过数据融合模块712 的数据融合后，使所有数据融合为统一格式，便于数据分析。
另外，由于多源异构传感融合数据中包含了较多信息，其中有些信息与健康管理无关，或者可信度不高，需要加以筛除，因此可以采用各种方法对多源异构传感融合数据进行处理，提取特征集。通常特征融合模块713可以通过抗干扰算法、信号质量评估等信号可信度增强技术提取高质量高可信多源异构传感融合数据，例如基于数学形态法及经验模式分解的噪声抑制及伪差消除、基于约束时序估计模型的信号质量评估等。然后可以通过运用差分阈值、小波分析、数据挖掘、神经网络等算法，协同分析与推算多源异构传感融合数据的特征集作为多源异构传感特征信号。
状态向量识别模块720，用于根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量。运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量属于决策级信息，根据这四个状态向量信息，能够较准确地反映人体的身体状态、心理状态及日常的行为习惯。
决策模块730，用于获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策。例如，决策模块730可以根据来自家庭、医院、服务中心等不同来源的健康数据及运营管理数据来生成诊断/预测策略。通过融合分析多模态、多维度的特征状态向量集，包括生理信息、心理信息、空间信息及运动信息，可以结合主观、客观各因素对健康状态的影响，生成更加精细化可靠化的健康管理决策。
交互终端800，用于输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。交互终端800可以和数据采集装置是同一电子设备，例如手机、平板电脑等。例如，交互终端800可以通过声、光、电、触觉、视觉、听觉等人机接口的交互模式，将健康管理决策输出给用户；用户根据健康管理决策进行反馈训练后，数据采集装置600再采集实时多源异构传感原始数据，形成健康管理的闭循环。
本发明提供的基于融合模型的智能化健康管理***，通过探究多源异构传感信 息与运动状态、情绪状态、睡眠状态及位置状态的映射关系，从多源异构传感特征信号中识别运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量这四个特征状态向量，可实现多尺度高精准量化标定，从而能够融合决策个体的健康状况和应采取的治疗或预防手段，指导用户实现优化可靠的健康调控，提升用户体验。
请参见图7，是本发明提供的基于融合模型的智能化健康管理***的第二实施例结构示意图。
该实施例中，健康管理服务器700还包括个人档案740。个人档案740用于存储所述多源异构传感特征信号和所述特征状态向量集，所述个人档案对应于单个用户。具体地，个人档案中存储有该用户此前的所有多元异构传感特征信号和特征状态向量集，形成该用户的个人健康档案。
决策模块730进一步包括：
个性化数据获取模块731，用于获取云端个性化健康管理数据。云端个性化健康管理数据存储在云端平台，可以包括来自家庭、医院、服务中心等不同来源的健康数据及专业知识经验。
策略生成模块732，用于根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略。目前常见的移动式健康管理***中，由于没有个人历史数据做参考，诊断/预测策略只能依据预置的专业知识经验，无法反映个体差异性，可能造成结果不准确，影响用户体验。本实施例中策略生成模块732通过融合个人档案和个性化健康管理数据来动态地生成诊断/预测策略，能够提高诊断/预测结果的个性化、精确性和可靠性。
决策融合模块733，用于根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策。
本发明实施例，通过建立个人长期健康档案，综合专业知识经验与多源异构传感信息，在一定准则下加以自动分析、综合、支配和使用，通过数据级融合、特征级融合和决策级融合，获得对被测对象的一致性解释与描述，以完成所需的健康管理决策任务，通过对信息优化组合导出更多的有效信息，可以向患者 本人、亲属及医院提供病人健康状况及健康管理实施效果的评估结果，为专业医务人员进行诊断提供参考。
特别是对于以下患者：1)属于慢性疾病患者，需要长期连续进行监测；2)属于已被医院告知为急性疾病突然发作的潜在患者；3)属于航空、军事、运动员、危险作业等行业人员，工作地点远离医院，且工作场所不适宜携带大型设备，有可能导致事故发生而不能提前预警的人群。本发明实施例提供的基于融合模型的智能化健康管理***可以为以上患者提供个性化、精细化的实时长期健康管理，增强自治神经***的调控功能，增强中央神经***的应激反应水平等。下面将以慢性疾病患者为例说明本方法。
请参见图8，是本发明提供的基于融合模型的智能化健康管理***的第三实施例结构示意图。
其中，决策融合模块733进一步包括：
慢性病样本模块733A，用于根据所述个人档案740计算所述用户的慢性病管理样本。
慢性病决策模块733B，用于根据所述诊断/预测策略融合分析所述慢性病管理样本和所述特征状态向量，生成针对所述用户的慢性病干预手段。通过交互终端800将所述慢性病干预手段输出给所述用户，使所述用户根据所述慢性病干预手段进行反馈训练。
本发明实施例通过数据挖掘手段发现与个体患者相关的慢性病管理规律，采取对应的干预手段指导患者实现优化、差异化的健康调控；更进一步，还可以通过对运营管理数据的挖掘，实现健康服务质量的控制及优化。
具体地，状态向量识别模块720可以包括运动状态向量识别模块、情绪状态向量识别模块、睡眠状态向量识别模块和位置状态向量识别模块。
在本发明的一个优选实施例中，运动状态向量识别模块的识别方法如图5所示。在该实施例中，利用可穿戴设备或智能手机配置的加速度传感器实现对人体运动状态的监测，并将人体日常运动行为与个人的身心健康关联起来实现对人体的日常运动进行监测，对于制定科学的运动健身计划，改善身体健康状况具有重要意义。基于加速度传感器的人体运动监测领域的研究理论以及常用的研 究方法，将加速度信号经过高通滤波器去除信号直流偏置等预处理后，提取时间窗内的方差，均值，最大值，最小值，小波变换系数，傅里叶频率特性等特征，再利用该特征评估值跟预先设置的阈值比较来识别人体日常运动。该方法针对多层次层分类器结构，采用自适应特征提取算法实现训练多个分类器，一类分类器采用加速度和陀螺仪，通过数据融合识别复杂姿态，另一类分类器只采用加速度用来识别简单姿态，同时可获得日常运动量和能量消耗等健康指标。
在本发明的一个优选实施例中，情绪状态向量模块的识别方法如图6所示。该实施例中，通过计算心率变异性、心脏及呼吸节奏模式，采集自我评价及生活事件量表，结合主客观个体信息，基于D-S证据理论决策层融合模型实现多级分层心理压力量化评估。
在本发明的一个优选实施例中，睡眠状态向量模块基于心率(或脉率)变异性的时域/频域/几何/非线性特征及人体姿态特征集，通过数据融合及统计分析方法逼近临床睡眠评估指标(呼吸紊乱/微觉醒指数、睡眠分期等)，得到睡眠状态向量。
在本发明的一个优选实施例中，位置状态向量识别模块摒弃了GPRS长连服务器和短信指令更新位置的做法，通过运动状态自适应切换位置上报及跟踪，实现精准位置服务。
本发明提供的基于融合模型的智能化健康管理***采用通用结构的模块化硬件电路优化设计方案，包括中央处理单元、传感单元、接口及人机交互单元等部分，可为用户提供个性化服务定制方案，以降低***复杂性及功耗。通过结构优化、器件选型、工作模式设定等，高度集成运动、空间、生理感知传感单元，在满足低功耗、低噪声、稳定可靠的前提下实现加/角速度、磁场强度、空间位置、心电、脉搏、呼吸等传感信息的获取、处理、存储、传输，并输出相应控制信号。本发明将基于HTML5框架实现跨平台多智能终端软件，通过将软件***平台化，优化功能模块，降低对RAM/ROM需求，改善运行速度，联动实现用户基本信息、运动信息、位置信息、情绪信息、睡眠信息的数据分析、存储、同步、协同、安全及分享等功能。
本发明实施例通过用于采用多源异构传感信息获取的智能可穿戴设备，优化设计软硬件***架构，集成新型抗干扰算法、信号质量评估算法及特征值提取算法，从而获取高质量高可信传感数据，并结合专业知识经验构建决策级数据融合模型，实现睡眠状态、运动状态、位置状态及情绪状态的智能化评估及个性化反馈训练机制，同时依托云计算平台中心，建立个人健康管理***，针对以老年人和亚健康群体为代表的慢性病人群，通过长期跟踪记录，建立患者个体健康档案，采集个体的多种生理数据和日常生活习惯信息；利用信息提取、信息融合、高维医学数据挖掘等技术进行数据分析，为患者提供精细化、个性化的健康管理指导；从而提高生活质量，对突发疾病风险给予预警提示。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于一计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)或随机存储记忆体(Random Access Memory，RAM)等。
以上所揭露的仅为本发明一种较佳实施例而已，当然不能以此来限定本发明之权利范围，本领域普通技术人员可以理解实现上述实施例的全部或部分流程，并依本发明权利要求所作的等同变化，仍属于发明所涵盖的范围。

Claims (10)

  1. 一种基于融合模型的智能化健康管理控制方法，其特征在于，包括：
    获取多源异构传感信号及其特征值；
    根据所述多源异构传感信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
    获取诊断/预测策略，根据所述诊断/预测策略模型融合、分析所述特征状态向量集，生成健康管理决策；
    输出所述健康管理决策，为用户提供基于所述健康管理决策反馈干预训练机制。
  2. 如权利要求1所述的基于融合模型的智能化健康管理控制方法，其特征在于，所述方法还包括将所述多源异构传感特征信号和所述特征状态向量集存储至个人档案，所述个人档案对应于单个用户；
    所述获取诊断/预测策略的步骤包括：
    获取云端个性化健康管理数据；
    根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略。
  3. 如权利要求2所述的基于融合模型的智能化健康管理控制方法，其特征在于，所述根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策的步骤包括：
    根据所述个人档案计算所述用户的多源信息特征向量集；
    根据所述诊断/预测策略融合分析所述多源信息特征向量集特征状态向量集，生成针对所述用户的健康管理和异常干预手段。
  4. 如权利要求1所述的基于融合模型的智能化健康管理控制方法，其特征在于，所述获取多源异构传感特征信号的步骤包括：
    获取多源异构传感融合数据；
    根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。
  5. 如权利要求4所述的基于融合模型的智能化健康管理控制方法，其特征在于，所述获取多源异构传感融合数据的步骤包括：
    采集多源异构传感原始数据；
    将所述多源异构传感原始数据归一化，生成多源同构融合数据。
  6. 一种基于融合模型的智能化健康管理***，其特征在于，所述***包括数据采集装置、健康管理服务器和交互终端；
    所述数据采集装置用于采集多源异构传感原始数据；
    所述健康管理服务器包括：
    特征获取模块，用于从所述多源异构传感原始数据中获取多源异构传感特征信号；
    状态向量识别模块，用于根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
    决策模块，用于获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策；
    所述交互终端用于输出所述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。
  7. 如权利要求6所述的基于融合模型的智能化健康管理***，其特征在于，所述健康管理服务器还包括用于存储所述多源异构传感特征信号和所述特征状态向量集，所述个人档案对应于单个用户；
    所述决策模块进一步包括：
    个性化数据获取模块，用于获取云端个性化健康管理数据；
    策略生成模块，用于根据所述个人档案和所述云端个性化健康管理数据生成诊断/预测策略；
    决策融合模块，用于根据所述诊断/预测策略融合分析所述特征状 态向量集，生成健康管理决策。
  8. 如权利要求7所述的基于融合模型的智能化健康管理***，其特征在于，所述决策融合模块进一步包括：
    慢性病样本模块，用于根据所述个人档案计算所述用户的多源信息特征向量集；
    慢性病决策模块，用于根据所述诊断/预测策略融合分析所述多源信息和所述特征状态向量集，生成针对所述用户的健康管理和异常状态干预手段。
  9. 如权利要求6所述的基于融合模型的智能化健康管理***，其特征在于，所述特征获取模块包括：
    原始数据接收模块，用于接收所述数据采集装置采集的多源异构传感原始数据；
    数据融合模块，用于将所述多源异构传感原始数据归一化，生成多源异构传感融合数据；
    特征融合模块，用于根据预设的数据处理算法提取所述多源异构传感融合数据的特征集，作为多源异构传感特征信号。
  10. 一种基于融合模型的智能化健康管理服务器，其特征在于，所述服务器包括：
    特征获取模块，用于获取多源异构传感特征信号；
    状态向量识别模块，用于根据所述多源异构传感特征信号识别用于反映人体健康状态的特征状态向量集，所述特征状态向量集包括运动状态向量、情绪状态向量、睡眠状态向量和位置状态向量；
    决策模块，用于获取诊断/预测策略，根据所述诊断/预测策略融合分析所述特征状态向量集，生成健康管理决策，并通过交互终端输出述健康管理决策，为用户提供基于所述健康管理决策的反馈干预训练机制。
PCT/CN2016/096213 2016-05-09 2016-08-22 基于融合模型的智能化健康管理服务器、***及其控制方法 WO2017193497A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610303848.0A CN105827731A (zh) 2016-05-09 2016-05-09 基于融合模型的智能化健康管理服务器、***及其控制方法
CN201610303848.0 2016-05-09

Publications (1)

Publication Number Publication Date
WO2017193497A1 true WO2017193497A1 (zh) 2017-11-16

Family

ID=56529218

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/096213 WO2017193497A1 (zh) 2016-05-09 2016-08-22 基于融合模型的智能化健康管理服务器、***及其控制方法

Country Status (2)

Country Link
CN (1) CN105827731A (zh)
WO (1) WO2017193497A1 (zh)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108932581A (zh) * 2018-06-06 2018-12-04 深圳大学 多物理域信息融合的自主感知方法及***
CN109033279A (zh) * 2018-06-28 2018-12-18 西安华云智联信息科技有限公司 一种数据处理方法及模块、计算平台及工业物联网
CN111241032A (zh) * 2020-01-08 2020-06-05 中国电子科技集团公司电子科学研究院 一种多源异构数据处理方法、装置和存储介质
CN111310198A (zh) * 2020-01-15 2020-06-19 华南理工大学 移动群体感知中异构数据隐私保护及可靠性判断方法
CN112802500A (zh) * 2020-12-31 2021-05-14 周凯 一种面向多源异构文旅大数据的分布式全息数据存储装置
CN112914589A (zh) * 2021-03-02 2021-06-08 钦州市第二人民医院 一种多导睡眠监测无线网帽装置及监测方法
CN114510367A (zh) * 2022-01-11 2022-05-17 电子科技大学 一种安全智能的多源异构数据处理***
CN114785821A (zh) * 2022-03-16 2022-07-22 陇东学院 一种基于Hadoop的城市物联网数据处理***及其方法
CN114818961A (zh) * 2022-05-10 2022-07-29 西安交通大学 一种滚动轴承多模态信息多级融合监测方法及***
CN115054248A (zh) * 2021-12-10 2022-09-16 荣耀终端有限公司 情绪监测方法和情绪监测装置
CN115101188A (zh) * 2022-08-23 2022-09-23 苏州利博软件技术有限公司 一种生化分析仪预测性维护***及辅助决策方法
CN116092673A (zh) * 2023-04-10 2023-05-09 华南理工大学 一种便携式多信息融合分析及干预评估***及其方法
CN116564561A (zh) * 2023-05-11 2023-08-08 亿慧云智能科技（深圳）股份有限公司 基于生理与情感特征的智能语音看护***及看护方法

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105827731A (zh) * 2016-05-09 2016-08-03 包磊 基于融合模型的智能化健康管理服务器、***及其控制方法
CN106557655A (zh) * 2016-11-17 2017-04-05 包磊 一种基于用户状态的分析推荐方法和装置
US20190000384A1 (en) * 2017-06-30 2019-01-03 Myant Inc. Method for sensing of biometric data and use thereof for determining emotional state of a user
CN110085317A (zh) * 2018-01-26 2019-08-02 上海草家物联网科技有限公司 个人健康指导***和指导方法以及智能餐具
CN108389628A (zh) * 2018-02-28 2018-08-10 张晶心 一种健康指标改善模型确定方法和装置
CN109102890B (zh) * 2018-08-07 2021-11-26 福建医锦智能科技有限公司 分布式医疗检测服务平台、***及其终端
CN109215791A (zh) * 2018-10-31 2019-01-15 深圳市儿童医院 基于多信息决策的健康管理方法、***、设备及存储介质
CN109711678B (zh) * 2018-12-07 2021-02-12 江南机电设计研究所 一种基于机器学习的异构传感器智能任务规划方法
CN110051191B (zh) * 2019-04-11 2021-02-26 南京邮电大学 智能办公学习桌垫及其健康提醒方法
CN110364241A (zh) * 2019-07-17 2019-10-22 孟李中 一种练养防治身心疾病并提供锻炼记录及数据反馈的***
CN110728606A (zh) * 2019-09-04 2020-01-24 重庆特斯联智慧科技股份有限公司 一种基于目标跟踪的社区养老服务远程交互方法和***
CN110970114A (zh) * 2019-11-13 2020-04-07 清华大学 中医健康导引***
CN111427909A (zh) * 2020-01-17 2020-07-17 吾达软件（武汉）股份有限公司 一种基于人工智能的行为大数据与心理健康测试***
CN112712874A (zh) * 2020-12-29 2021-04-27 北京浪潮数据技术有限公司 一种智能诊脉方法、装置及电子设备和存储介质
CN113693572A (zh) * 2021-07-21 2021-11-26 湖北智奥物联网科技有限公司 无创多维度动态健康管理***及装置
CN114081462B (zh) * 2021-11-19 2023-06-09 齐齐哈尔大学 一种基于多维度生理信息的心脏健康监护***
CN115969330B (zh) * 2023-03-20 2023-07-04 安徽星辰智跃科技有限责任公司 一种睡眠情绪活跃水平检测量化的方法、***和装置
CN117524489B (zh) * 2024-01-08 2024-03-08 吉林大学 患者的健康监测管理***及方法
CN117894476B (zh) * 2024-03-12 2024-05-14 西南医科大学附属医院 一种基于云数据的生殖健康管理***

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102087680A (zh) * 2010-11-19 2011-06-08 陈博 创意性生命健康状态集成评价***
CN102715902A (zh) * 2012-06-15 2012-10-10 天津大学 特殊人群的情绪监护方法
WO2014063518A1 (zh) * 2012-10-24 2014-05-01 中兴通讯股份有限公司 远程家庭保健***
CN103970271A (zh) * 2014-04-04 2014-08-06 浙江大学 融合运动和生理传感数据的日常活动识别方法
CN105069728A (zh) * 2015-08-19 2015-11-18 南京邮电大学 一种基于无线传感网的情感推理方法
CN105827731A (zh) * 2016-05-09 2016-08-03 包磊 基于融合模型的智能化健康管理服务器、***及其控制方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102087680A (zh) * 2010-11-19 2011-06-08 陈博 创意性生命健康状态集成评价***
CN102715902A (zh) * 2012-06-15 2012-10-10 天津大学 特殊人群的情绪监护方法
WO2014063518A1 (zh) * 2012-10-24 2014-05-01 中兴通讯股份有限公司 远程家庭保健***
CN103970271A (zh) * 2014-04-04 2014-08-06 浙江大学 融合运动和生理传感数据的日常活动识别方法
CN105069728A (zh) * 2015-08-19 2015-11-18 南京邮电大学 一种基于无线传感网的情感推理方法
CN105827731A (zh) * 2016-05-09 2016-08-03 包磊 基于融合模型的智能化健康管理服务器、***及其控制方法

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108932581A (zh) * 2018-06-06 2018-12-04 深圳大学 多物理域信息融合的自主感知方法及***
CN109033279A (zh) * 2018-06-28 2018-12-18 西安华云智联信息科技有限公司 一种数据处理方法及模块、计算平台及工业物联网
CN111241032A (zh) * 2020-01-08 2020-06-05 中国电子科技集团公司电子科学研究院 一种多源异构数据处理方法、装置和存储介质
CN111310198A (zh) * 2020-01-15 2020-06-19 华南理工大学 移动群体感知中异构数据隐私保护及可靠性判断方法
CN111310198B (zh) * 2020-01-15 2022-08-16 华南理工大学 移动群体感知中异构数据隐私保护及可靠性判断方法
CN112802500A (zh) * 2020-12-31 2021-05-14 周凯 一种面向多源异构文旅大数据的分布式全息数据存储装置
CN112914589A (zh) * 2021-03-02 2021-06-08 钦州市第二人民医院 一种多导睡眠监测无线网帽装置及监测方法
CN115054248B (zh) * 2021-12-10 2023-10-20 荣耀终端有限公司 情绪监测方法和情绪监测装置
CN115054248A (zh) * 2021-12-10 2022-09-16 荣耀终端有限公司 情绪监测方法和情绪监测装置
CN114510367A (zh) * 2022-01-11 2022-05-17 电子科技大学 一种安全智能的多源异构数据处理***
CN114785821A (zh) * 2022-03-16 2022-07-22 陇东学院 一种基于Hadoop的城市物联网数据处理***及其方法
CN114818961A (zh) * 2022-05-10 2022-07-29 西安交通大学 一种滚动轴承多模态信息多级融合监测方法及***
CN114818961B (zh) * 2022-05-10 2024-04-02 西安交通大学 一种滚动轴承多模态信息多级融合监测方法及***
CN115101188A (zh) * 2022-08-23 2022-09-23 苏州利博软件技术有限公司 一种生化分析仪预测性维护***及辅助决策方法
CN116092673A (zh) * 2023-04-10 2023-05-09 华南理工大学 一种便携式多信息融合分析及干预评估***及其方法
CN116092673B (zh) * 2023-04-10 2023-06-16 华南理工大学 一种便携式多信息融合分析及干预评估***及其方法
CN116564561A (zh) * 2023-05-11 2023-08-08 亿慧云智能科技（深圳）股份有限公司 基于生理与情感特征的智能语音看护***及看护方法

Also Published As

Publication number Publication date
CN105827731A (zh) 2016-08-03

Similar Documents

Publication Publication Date Title
WO2017193497A1 (zh) 基于融合模型的智能化健康管理服务器、***及其控制方法
Panicker et al. A survey of machine learning techniques in physiology based mental stress detection systems
US11696714B2 (en) System and method for brain modelling
Albahri et al. IoT-based telemedicine for disease prevention and health promotion: State-of-the-Art
Pham et al. Delivering home healthcare through a cloud-based smart home environment (CoSHE)
Anzanpour et al. Self-awareness in remote health monitoring systems using wearable electronics
Yu et al. Optimization of IoT-based artificial intelligence assisted telemedicine health analysis system
Liaqat et al. WearBreathing: Real world respiratory rate monitoring using smartwatches
CN108778097A (zh) 用于评估心力衰竭的装置和方法
Myroniv et al. Analyzing user emotions via physiology signals
Parra et al. Multimedia sensors embedded in smartphones for ambient assisted living and e-health
WO2018151628A1 (ru) Алгоритм комплексного дистанционного бесконтактного мультиканального анализа психоэмоционального и физиологического состояния объекта по аудио- и видеоконтенту
US20160128638A1 (en) System and method for detecting and quantifying deviations from physiological signals normality
CN109124655A (zh) 精神状态分析方法、装置、设备、计算机介质及多功能椅
Camcı et al. Abnormal respiratory event detection in sleep: A prescreening system with smart wearables
Ahmed et al. IoMT-based biomedical measurement systems for healthcare monitoring: A review
Rasool et al. Maximizing Efficiency in Telemedicine: An IoT-Based Artificial Intelligence Optimization Framework for Health Analysis
WO2022093707A1 (en) Cardiopulmonary health monitoring using thermal camera and audio sensor
CN117064388A (zh) 基于情绪识别实现精神障碍评估分析的***
CN110755091A (zh) 一种个人心理健康监测***及方法
Gan et al. Human-computer interaction based interface design of intelligent health detection using PCANet and multi-sensor information fusion
Alam et al. Autocognisys: Iot assisted context-aware automatic cognitive health assessment
Haque et al. State-of-the-art of stress prediction from heart rate variability using artificial intelligence
Du et al. Non-interference driving fatigue detection system based on intelligent steering wheel
Mantri et al. Real time multimodal depression analysis

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16901452

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 27/03/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16901452

Country of ref document: EP

Kind code of ref document: A1