CN109567742A - A kind of paediatrics state of an illness early warning method - Google Patents
A kind of paediatrics state of an illness early warning method Download PDFInfo
- Publication number
- CN109567742A CN109567742A CN201811211557.4A CN201811211557A CN109567742A CN 109567742 A CN109567742 A CN 109567742A CN 201811211557 A CN201811211557 A CN 201811211557A CN 109567742 A CN109567742 A CN 109567742A
- Authority
- CN
- China
- Prior art keywords
- module
- data
- illness
- disease
- prediction
- 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.)
- Withdrawn
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient ; user input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2503/00—Evaluating a particular growth phase or type of persons or animals
- A61B2503/06—Children, e.g. for attention deficit diagnosis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2560/00—Constructional details of operational features of apparatus; Accessories for medical measuring apparatus
- A61B2560/02—Operational features
- A61B2560/0242—Operational features adapted to measure environmental factors, e.g. temperature, pollution
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Physiology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention belongs to state of an illness early warning technology fields, a kind of paediatrics state of an illness early warning method is disclosed, the paediatrics state of an illness early warning system includes: physiological data collection module, environmental factor acquisition module, central control module, contrast module, prediction module, pathological analysis module, alarm module, data memory module, display module.The present invention carries out synthesis by self-characteristic of the prediction module to multivariate time series and effectively analyzes, and improves the accuracy of time series forecasting;Simultaneously, make system when obtaining pathological data from different storage equipment by pathological analysis module, improve the efficiency for carrying out data access, data transmission and data storage, further, by adjusting the corresponding threshold value screened by the degree of correlation and data type, the efficiency for improving pathological analysis is conducive to carry out efficient diagnosis in time to children's illness.
Description
Technical field
The invention belongs to state of an illness early warning technology field more particularly to a kind of paediatrics state of an illness early warning methods.
Background technique
Common pediatric disease has following several: pneumonia asthma cough (infantile pneumonia) be the most common pediatric disease and youngster
Virgin the first dead reason.Clinic is to generate heat, cough, cough up phlegm, out of breath, nose is incited as cardinal symptom, and the visible opening mouth and lifting shoulder due to dyspnea of severe one is exhaled
Inhale the diseases such as difficulty, pale complexion, cyanotic lips;Flu (acute upper respiratory infection) is the common disease of children, accounts for paediatrics door
The 70%~80% of the amount of examining.It is susceptible by ailment said due to cold or exposure since children's native endowment is weak, therefore can fall ill throughout the year.6 months to 3
Year is onset peak, and 90% or more is virus with pathogenic microorganisms, can generally be generated heat repeatedly 3 days or so, mostly obvious with night,
Infant is not only taken medicine difficulty, and infusion aggravates pain again, and parent subjects the dual torment of the human body and spirit;Diarrhea (diarrhoeal diseases) abdomen
Rushing down disease is cause of disease more than one group, multifactor caused disease, it is to cause infantile malnutrition, dysplasia and the weight of death
Want one of reason.The death rate of the disease is decreased obviously due to the improvement of nutritional status of children and medical condition in China,
But its disease incidence is still higher, area especially poor in condition, for this purpose, diarrhoeal diseases is one of the disease of China's keypoint control;
Baby anorexia refers to the symptom based on children's longer-term appetite stimulator or anorexia, it is not a kind of independent disease, suffers from
Person is general, and symptom is few, is exactly long-term do not feel like eating, and detest is ingested, and appetite has belch, general considerably less than normal child of the same age
Evil, gastral cavity ruffian, the diseases such as uncomfortable of defecating, or with diseases such as lustreless complexion, partially thin, the dry happiness drinks of body, but spirit is fair, and activity is as usual.
A small number of infants are since certain chronic diseases, such as peptic ulcer, chronic hepatitis, tuberculosis, prolonged constipation, zinc deficiency cause, greatly
Be mostly due to undesirable eating habit, unreasonable dietary regimen, bad feed environment and parent and child it is psychological because
Caused by element;Fever caused by exogenous pathogens refers to that the number that upper and lower respiratory tract infection occurs within 1 year is frequent repeatedly, exceeds normal range (NR).Its
The cause of disease includes: native endowment weakness, the improper feeding day after tomorrow, nurses that inappropriate, environment influences, infection, weakness due to chronic disease etc. are multiple repeatedly
Aspect.However, the efficiency of existing paediatrics prognosis is lower, prediction result is also inaccurate;Meanwhile pathology cannot be carried out in time
Analysis, pathological analysis low efficiency influence the diagnosis to children disease.
In conclusion problem of the existing technology is: the efficiency of existing paediatrics prognosis is lower, and prediction result is not yet
Accurately;Meanwhile pathology cannot be analyzed in time, pathological analysis low efficiency influences the diagnosis to children disease.
Summary of the invention
In view of the problems of the existing technology, the present invention provides a kind of paediatrics state of an illness early warning methods.
The invention is realized in this way a kind of paediatrics state of an illness early warning system includes:
Physiological data collection module, environmental factor acquisition module, central control module, contrast module, prediction module, pathology
Analysis module, alarm module, data memory module, display module;
Physiological data collection module, connect with central control module, for acquiring children physiology index by Medical Devices
Data information;
Environmental factor acquisition module, connect with central control module, for acquiring the density of population, air matter by network
Amount, water quality data information;
Central control module, with physiological data collection module, environmental factor acquisition module, contrast module, prediction module, disease
Analysis module, alarm module, data memory module, display module connection are managed, it is normal for controlling modules by single-chip microcontroller
Work;
Contrast module is connect with central control module, is compared for will acquire physical signs with early warning monitoring rank
Judge whether normal;
Prediction module is connect with central control module, for right according to the collected data by multivariate time series model
Children disease carries out predicted operation;
Pathological analysis module, connect with central control module, for carrying out disease to children disease according to history pathological data
Reason analysis;
Alarm module is connect with central control module, is used for through network mail or short message to abnormal index data or in advance
It surveys result and carries out push warning;
Data memory module is connect with central control module, for physiological data, the environment by memory storage acquisition
Factor data information, prediction result;
Display module is connect with central control module, physiological data, the ring for the acquisition by display display acquisition
Border factor data information, prediction result.
A kind of paediatrics state of an illness early warning method the following steps are included:
Step 1 acquires children physiology marker data information using Medical Devices by physiological data collection module;Pass through
Environmental factor acquisition module acquires the density of population, air quality, water quality data information using network;
Step 2, central control module scheduling contrast module, which will acquire physical signs and compare with early warning monitoring rank, to be sentenced
It is disconnected whether normal;
Step 3 according to the collected data carries out in advance children disease using multivariate time series model by prediction module
Survey operation;Pathological analysis is carried out to children disease according to history pathological data by pathological analysis module;
Step 4 pushes abnormal index data or prediction result using network mail or short message by alarm module
Warning;
Step 5, by data memory module using memory storage acquisition physiological data, environmental factor data information,
Prediction result;And the data information of display display acquisition is utilized by display module.
Further, the prediction module prediction technique is as follows:
(1) seasonality and Trend Decomposition algorithm based on local regression, are trend sequence by multivariate time series Seasonal decomposition method
Column, cyclic sequence and irregular sequence;
(2) trend sequence is predicted by linearly or nonlinearly regression algorithm to obtain trend sequence predicted value;
(3) cyclic sequence for introducing external factor and each phase variable of history respectively is predicted to obtain based on built-up pattern
Multiple cyclic sequence initial predictions are merged initial prediction using feedforward neural network to obtain cyclic sequence prediction
Value;
(4) it sums up trend sequence predicted value and cyclic sequence predicted value to obtain multivariate time series predicted value.
Further, the step (1) specifically includes:
Seasonality and Trend Decomposition algorithm based on local regression, by multivariate time series X={ X1..., XNBe decomposed into
Gesture sequence T={ T1..., TN, cyclic sequence S={ S1..., SNAnd irregular sequence I={ I1..., IN, meet, Xt=Tt+
St+It;Give up irregular sequence I;
Wherein, t=1 ..., N;N is multivariate time series length;Irregular sequence I indicates to be superimposed in multivariate time series X
Noise.
Further, the step (2) specifically includes:
By including but is not limited to least square method, one of integrating autoregressive moving-average model linearly or nonlinearly
Regression algorithm models trend sequence, and the trend sequence predicted value T at (N+1) moment is calculatedN+1。
Further, the step (3) specifically includes:
The influence to be dealt with externality using the mixed Gauss model in built-up pattern;Certainly using the integration in built-up pattern
Regressive averaging model handles the influence of each phase variable of history, and the cyclic sequence initial prediction at (N+1) moment is calculated
SN+1。
Further, the pathological analysis module analysis method is as follows:
1) a kind of prediction model of disease is obtained, wherein the prediction model is based on a kind of history disease of disease
The model of the degree of correlation foundation of data and a kind of disease is managed, and the degree of correlation is by the history pathological data
It is obtained after progress covariance matrix processing and data type Screening Treatment;
2) pathological data of tester is obtained, and the pathological data is brought into a kind of prediction model of disease, really
The analysis result of the pathological data of the fixed tester.
It is further, described before the prediction model for obtaining a kind of disease, comprising:
Obtain a kind of history pathological data of disease, wherein described one is included at least in the history pathological data
The history feature data of several illnesss of kind disease;
Covariance matrix processing is carried out to the history feature data for each illness for including in the history pathological data, and
Characteristic value collection is obtained based on covariance matrix processing result, wherein a characteristic value in the characteristic value collection and one
Illness is corresponding, and a corresponding illness of characteristic value and a kind of degree of correlation of disease;
Each illness that the degree of correlation meets the first preset condition is filtered out, the first set of disorders is obtained;
From each illness that first set of disorders includes, the disease that data type meets the second preset condition is filtered out
Disease obtains the second set of disorders;
Based on each illness and corresponding history feature data for including in the second feature set, described one is established
The prediction model of kind disease.
Further, the history feature data for each illness for including in the history pathological data carry out at covariance matrix
Reason, and characteristic value collection is obtained based on covariance matrix processing result, comprising:
History feature data based on each illness determine the history feature data mean value of each illness respectively;
Based on each difference of each illness corresponding history feature data and corresponding history feature data mean value, obtain
Obtain the corresponding difference value vector of each illness;
Calculate separately the product vector of the difference value vector of every two illness, and calculate separately the difference of each illness to
The product vector of amount and itself;
The element mean value for each element that each product vector includes is calculated separately, and is based on each product vector pair
The element mean value answered obtains a kind of Eigen Covariance matrix of disease;
By obtaining the Eigen Covariance matrix to a kind of Eigen Covariance matrix progress matrixing of disease
Corresponding characteristic value collection.
Advantages of the present invention and good effect are as follows: the present invention utilizes the spy of multivariate time series itself by prediction module
Property, it is broken down into trend sequence and cyclic sequence, and predicted respectively;In cyclic sequence prediction, the time is comprehensively considered
The external factor of sequence influences and each phase variable of history, uses mixed Gauss model respectively and integrates autoregressive moving-average model
Cyclic sequence is modeled and is predicted, recycles feedforward neural network that its prediction result is carried out effective integration, thus to more
The self-characteristic of elementary time sequence carries out synthesis and effectively analyzes, and improves the accuracy of time series forecasting;Meanwhile passing through pathology
Analysis module is first passed through according to each illness and corresponding history feature data that include in history pathological data is obtained to going through
History pathological data carry out covariance processing, filter out with a kind of above-mentioned higher illness of the disease degree of correlation, reduce analysis use
Illness quantity, then, by data type Screening Treatment, eliminate to pathological analysis generate interference illness, further,
The quantity for reducing the illness that analysis uses makes system when obtaining pathological data from different storage equipment, improves progress
The efficiency that data access, data transmission and data store further is screened by adjusting by the degree of correlation and data type
Corresponding threshold value, improve the efficiency of pathological analysis, be conducive to carry out efficient diagnosis in time to children's illness.
Detailed description of the invention
Fig. 1 is that the present invention implements the paediatrics state of an illness early warning method flow diagram provided.
Fig. 2 is that the present invention implements the paediatrics state of an illness early warning system structural block diagram provided.
In Fig. 2: 1, physiological data collection module;2, environmental factor acquisition module;3, central control module;4, mould is compared
Block;5, prediction module;6, pathological analysis module;7, alarm module;8, data memory module;9, display module.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to
Limit the present invention.
With reference to the accompanying drawing and specific embodiment is further described application principle of the invention.
As shown in Figure 1, a kind of paediatrics state of an illness early warning method provided by the invention the following steps are included:
Step S101 acquires children physiology marker data information using Medical Devices by physiological data collection module;It is logical
It crosses environmental factor acquisition module and acquires the density of population, air quality, water quality data information using network;
Step S102, central control module scheduling contrast module will acquire physical signs and compare with early warning monitoring rank
Judge whether normal;
Step S103 according to the collected data carries out children disease using multivariate time series model by prediction module
Predicted operation;Pathological analysis is carried out to children disease according to history pathological data by pathological analysis module;
Step S104 pushes away abnormal index data or prediction result using network mail or short message by alarm module
Send warning;
Step S105 is believed by data memory module using the physiological data of memory storage acquisition, environmental factor data
Breath, prediction result;And the data information of display display acquisition is utilized by display module.
As shown in Fig. 2, paediatrics state of an illness early warning system provided by the invention includes: physiological data collection module 1, environment
Factor acquisition module 2, central control module 3, contrast module 4, prediction module 5, pathological analysis module 6, alarm module 7, data
Memory module 8, display module 9.
Physiological data collection module 1 is connect with central control module 3, is referred to for acquiring children physiology by Medical Devices
Mark data information;
Environmental factor acquisition module 2 is connect with central control module 3, for acquiring the density of population, air matter by network
Amount, water quality data information;
Central control module 3, with physiological data collection module 1, environmental factor acquisition module 2, contrast module 4, prediction mould
Block 5, pathological analysis module 6, alarm module 7, data memory module 8, display module 9 connect, each for being controlled by single-chip microcontroller
A module works normally;
Contrast module 4 is connect with central control module 3, is carried out pair for will acquire physical signs and early warning monitoring rank
It is more normal than judging whether;
Prediction module 5 is connect with central control module 3, for passing through multivariate time series model according to the collected data
Predicted operation is carried out to children disease;
Pathological analysis module 6 is connect with central control module 3, for being carried out according to history pathological data to children disease
Pathological analysis;
Alarm module 7 is connect with central control module 3, for by network mail or short message to abnormal index data or
Prediction result carries out push warning;
Data memory module 8 is connect with central control module 3, for physiological data, the ring by memory storage acquisition
Border factor data information, prediction result;
Display module 9 is connect with central control module 3, the physiological data of the acquisition for being acquired by display display,
Environmental factor data information, prediction result.
5 prediction technique of prediction module provided by the invention is as follows:
(1) seasonality and Trend Decomposition algorithm based on local regression, are trend sequence by multivariate time series Seasonal decomposition method
Column, cyclic sequence and irregular sequence;
(2) trend sequence is predicted by linearly or nonlinearly regression algorithm to obtain trend sequence predicted value;
(3) cyclic sequence for introducing external factor and each phase variable of history respectively is predicted to obtain based on built-up pattern
Multiple cyclic sequence initial predictions are merged initial prediction using feedforward neural network to obtain cyclic sequence prediction
Value;
(4) it sums up trend sequence predicted value and cyclic sequence predicted value to obtain multivariate time series predicted value.
Step (1) provided by the invention specifically includes:
Seasonality and Trend Decomposition algorithm based on local regression, by multivariate time series X={ X1..., XNBe decomposed into
Gesture sequence T={ T1..., TN, cyclic sequence S={ S1..., SNAnd irregular sequence I={ I1..., IN, meet, Xt=Tt+
St+It;Give up irregular sequence I;
Wherein, t=1 ..., N;N is multivariate time series length;Irregular sequence I indicates to be superimposed in multivariate time series X
Noise.
Step (2) provided by the invention specifically includes:
By including but is not limited to least square method, one of integrating autoregressive moving-average model linearly or nonlinearly
Regression algorithm models trend sequence, and the trend sequence predicted value T at (N+1) moment is calculatedN+1。
Step (3) provided by the invention specifically includes:
The influence to be dealt with externality using the mixed Gauss model in built-up pattern;Certainly using the integration in built-up pattern
Regressive averaging model handles the influence of each phase variable of history, and the cyclic sequence initial prediction at (N+1) moment is calculated
SN+1。
6 analysis method of pathological analysis module provided by the invention is as follows:
1) a kind of prediction model of disease is obtained, wherein the prediction model is based on a kind of history disease of disease
The model of the degree of correlation foundation of data and a kind of disease is managed, and the degree of correlation is by the history pathological data
It is obtained after progress covariance matrix processing and data type Screening Treatment;
2) pathological data of tester is obtained, and the pathological data is brought into a kind of prediction model of disease, really
The analysis result of the pathological data of the fixed tester.
It is provided by the invention before the prediction model for obtaining a kind of disease, comprising:
Obtain a kind of history pathological data of disease, wherein described one is included at least in the history pathological data
The history feature data of several illnesss of kind disease;
Covariance matrix processing is carried out to the history feature data for each illness for including in the history pathological data, and
Characteristic value collection is obtained based on covariance matrix processing result, wherein a characteristic value in the characteristic value collection and one
Illness is corresponding, and a corresponding illness of characteristic value and a kind of degree of correlation of disease;
Each illness that the degree of correlation meets the first preset condition is filtered out, the first set of disorders is obtained;
From each illness that first set of disorders includes, the disease that data type meets the second preset condition is filtered out
Disease obtains the second set of disorders;
Based on each illness and corresponding history feature data for including in the second feature set, described one is established
The prediction model of kind disease.
The history feature data for each illness for including in history pathological data provided by the invention carry out covariance matrix
Processing, and characteristic value collection is obtained based on covariance matrix processing result, comprising:
History feature data based on each illness determine the history feature data mean value of each illness respectively;
Based on each difference of each illness corresponding history feature data and corresponding history feature data mean value, obtain
Obtain the corresponding difference value vector of each illness;
Calculate separately the product vector of the difference value vector of every two illness, and calculate separately the difference of each illness to
The product vector of amount and itself;
The element mean value for each element that each product vector includes is calculated separately, and is based on each product vector pair
The element mean value answered obtains a kind of Eigen Covariance matrix of disease;
By obtaining the Eigen Covariance matrix to a kind of Eigen Covariance matrix progress matrixing of disease
Corresponding characteristic value collection.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (9)
1. a kind of paediatrics state of an illness early warning system, which is characterized in that the paediatrics state of an illness early warning system includes:
Physiological data collection module, environmental factor acquisition module, central control module, contrast module, prediction module, pathological analysis
Module, alarm module, data memory module, display module;
Physiological data collection module, connect with central control module, for acquiring children physiology achievement data by Medical Devices
Information;
Environmental factor acquisition module, connect with central control module, for acquiring the density of population, air quality, water by network
Matter data information;
Central control module, with physiological data collection module, environmental factor acquisition module, contrast module, prediction module, pathology point
Module, alarm module, data memory module, display module connection are analysed, is worked normally for controlling modules by single-chip microcontroller;
Contrast module is connect with central control module, compares judgement for will acquire physical signs and early warning monitoring rank
It is whether normal;
Prediction module is connect with central control module, for passing through multivariate time series model according to the collected data to children
Disease carries out predicted operation;
Pathological analysis module, connect with central control module, for carrying out pathology point to children disease according to history pathological data
Analysis;
Alarm module is connect with central control module, for being tied by network mail or short message to abnormal index data or prediction
Fruit carries out push warning;
Data memory module is connect with central control module, for the physiological data by memory storage acquisition, environmental factor
Data information, prediction result;
Display module is connect with central control module, for by display display acquisition the physiological data of acquisition, environment because
Plain data information, prediction result.
2. a kind of paediatrics state of an illness early warning method as described in claim 1, which is characterized in that the paediatrics state of an illness early stage is pre-
Alarm method the following steps are included:
Step 1 acquires children physiology marker data information using Medical Devices by physiological data collection module;Pass through environment
Factor acquisition module acquires the density of population, air quality, water quality data information using network;
Step 2, central control module scheduling contrast module, which will acquire physical signs and early warning, which guards rank and compare judgement, is
It is no normal;
Step 3 carries out prediction behaviour to children disease according to the collected data using multivariate time series model by prediction module
Make;Pathological analysis is carried out to children disease according to history pathological data by pathological analysis module;
Step 4 carries out push police to abnormal index data or prediction result using network mail or short message by alarm module
Show;
Step 5 utilizes the physiological data, environmental factor data information, prediction of memory storage acquisition by data memory module
As a result;And the data information of display display acquisition is utilized by display module.
3. paediatrics state of an illness early warning system as described in claim 1, which is characterized in that the prediction module prediction technique is such as
Under:
(1) based on local regression seasonality with Trend Decomposition algorithm, by multivariate time series Seasonal decomposition method be trend sequence, follow
Ring sequence and irregular sequence;
(2) trend sequence is predicted by linearly or nonlinearly regression algorithm to obtain trend sequence predicted value;
(3) cyclic sequence for introducing external factor and each phase variable of history respectively is predicted to obtain based on built-up pattern multiple
Cyclic sequence initial prediction is merged initial prediction using feedforward neural network to obtain cyclic sequence predicted value;
(4) it sums up trend sequence predicted value and cyclic sequence predicted value to obtain multivariate time series predicted value.
4. paediatrics state of an illness early warning system as claimed in claim 3, which is characterized in that the step (1) specifically includes:
Seasonality and Trend Decomposition algorithm based on local regression, by multivariate time series X={ X1..., XNIt is decomposed into trend sequence
Arrange T={ T1..., TN, cyclic sequence S={ S1..., SNAnd irregular sequence I={ I1..., IN, meet, Xt=Tt+St+
It;Give up irregular sequence I;
Wherein, t=1 ..., N;N is multivariate time series length;Irregular sequence I indicates that is be superimposed in multivariate time series X makes an uproar
Sound.
5. paediatrics state of an illness early warning system as claimed in claim 3, which is characterized in that the step (2) specifically includes:
By including but is not limited to least square method, one of integrating autoregressive moving-average model and linearly or nonlinearly return
Algorithm models trend sequence, and the trend sequence predicted value T at (N+1) moment is calculatedN+1。
6. paediatrics state of an illness early warning system as claimed in claim 3, which is characterized in that the step (3) specifically includes:
The influence to be dealt with externality using the mixed Gauss model in built-up pattern;Use the integration autoregression in built-up pattern
Moving average model handles the influence of each phase variable of history, and the cyclic sequence initial prediction S at (N+1) moment is calculatedN+1。
7. paediatrics state of an illness early warning system as described in claim 1, which is characterized in that the pathological analysis module analysis side
Method is as follows:
1) a kind of prediction model of disease is obtained, wherein the prediction model is based on a kind of history pathology number of disease
According to the model established with a kind of degree of correlation of the disease, and the degree of correlation is by carrying out to the history pathological data
It is obtained after covariance matrix processing and data type Screening Treatment;
2) pathological data of tester is obtained, and the pathological data is brought into a kind of prediction model of disease, determines institute
State the analysis result of the pathological data of tester.
8. paediatrics state of an illness early warning system as claimed in claim 7, which is characterized in that described to obtain a kind of the pre- of disease
It surveys before model, comprising:
Obtain a kind of history pathological data of disease, wherein a kind of disease is included at least in the history pathological data
The history feature data of several illnesss of disease;
Covariance matrix processing is carried out to the history feature data for each illness for including in the history pathological data, and is based on
Covariance matrix processing result obtains characteristic value collection, wherein a characteristic value and an illness in the characteristic value collection
It is corresponding, and a corresponding illness of characteristic value and a kind of degree of correlation of disease;
Each illness that the degree of correlation meets the first preset condition is filtered out, the first set of disorders is obtained;
From each illness that first set of disorders includes, the illness that data type meets the second preset condition is filtered out,
Obtain the second set of disorders;
Based on each illness and corresponding history feature data for including in the second feature set, a kind of disease is established
The prediction model of disease.
9. paediatrics state of an illness early warning system as claimed in claim 8, which is characterized in that include in the history pathological data
Each illness history feature data carry out covariance matrix processing, and based on covariance matrix processing result obtain characteristic value
Set, comprising:
History feature data based on each illness determine the history feature data mean value of each illness respectively;
Based on each difference of each illness corresponding history feature data and corresponding history feature data mean value, obtain every
The corresponding difference value vector of one illness;
Calculate separately the product vector of the difference value vector of every two illness, and calculate separately the difference value vector of each illness with
The product vector of itself;
The element mean value for each element that each product vector includes is calculated separately, and corresponding based on each product vector
Element mean value obtains a kind of Eigen Covariance matrix of disease;
By the way that it is corresponding to obtain the Eigen Covariance matrix to a kind of Eigen Covariance matrix progress matrixing of disease
Characteristic value collection.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811211557.4A CN109567742A (en) | 2018-10-17 | 2018-10-17 | A kind of paediatrics state of an illness early warning method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811211557.4A CN109567742A (en) | 2018-10-17 | 2018-10-17 | A kind of paediatrics state of an illness early warning method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109567742A true CN109567742A (en) | 2019-04-05 |
Family
ID=65920611
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811211557.4A Withdrawn CN109567742A (en) | 2018-10-17 | 2018-10-17 | A kind of paediatrics state of an illness early warning method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109567742A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110047592A (en) * | 2019-04-24 | 2019-07-23 | 河北省中医院 | A kind of critical value warning system of medical test and method |
CN112835316A (en) * | 2021-01-06 | 2021-05-25 | 重庆医科大学 | Neonatal sepsis shock prediction system and monitoring equipment |
-
2018
- 2018-10-17 CN CN201811211557.4A patent/CN109567742A/en not_active Withdrawn
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110047592A (en) * | 2019-04-24 | 2019-07-23 | 河北省中医院 | A kind of critical value warning system of medical test and method |
CN112835316A (en) * | 2021-01-06 | 2021-05-25 | 重庆医科大学 | Neonatal sepsis shock prediction system and monitoring equipment |
CN112835316B (en) * | 2021-01-06 | 2022-04-19 | 重庆医科大学 | Neonatal sepsis shock prediction system and monitoring equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108198620A (en) | A kind of skin disease intelligent auxiliary diagnosis system based on deep learning | |
Hoque et al. | Holmes: A comprehensive anomaly detection system for daily in-home activities | |
CN109087691A (en) | A kind of OTC drugs recommender system and recommended method based on deep learning | |
Heydarzadeh et al. | In-bed posture classification using deep autoencoders | |
JP2005502937A (en) | System and method for monitoring behavioral information engineering | |
AS et al. | Estimating the surveillance of liver disorder using classification algorithms | |
CN108806765A (en) | A kind of intelligence baby monitoring system | |
CN109480868A (en) | A kind of intelligence baby monitoring system | |
CN109473173A (en) | A kind of the elderly's Cognitive deficiency assessment system and device based on video | |
CN109567742A (en) | A kind of paediatrics state of an illness early warning method | |
CN115690874A (en) | Old people medicine taking detection method and system based on human body key point detection and Transformer model | |
CN115910346A (en) | Monitoring and early warning system and method for diabetic patient | |
CN114241270A (en) | Intelligent monitoring method, system and device for home care | |
Nadda et al. | Weighted extreme learning machine for dengue detection with class-imbalance classification | |
Sledge et al. | Emergent trend detection in diurnal activity | |
CN113593702A (en) | Pet health early-warning diagnosis and treatment system | |
Chen et al. | The Method of Selecting the Evaluation Model for the State of Children's Growth Based on Machine Learning | |
CN110189803A (en) | The disease risk factor extracting method combined based on cluster with classification | |
Ariyadasa et al. | Knowledge extraction to mitigate child malnutrition in developing countries (Sri Lankan context) | |
Liu | Research on Question Classification Methods in the Medical Field | |
CN106919775B (en) | A kind of processing method and processing device of sleep info | |
CN109887605A (en) | A kind of mechanism long-term care hospital stays pattern analysis method based on model | |
CN110427848A (en) | A kind of psychoanalysis system | |
Barbieri | The determinants of infant and child mortality in Senegal: An analysis of DHS data | |
Kim et al. | Pet dog disease pre-diagnosis system for caregiver with possibilistic C-Means clustering and disease database |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190405 |