CN109493979A - A kind of disease forecasting method and apparatus based on intelligent decision - Google Patents
A kind of disease forecasting method and apparatus based on intelligent decision Download PDFInfo
- Publication number
- CN109493979A CN109493979A CN201811235396.2A CN201811235396A CN109493979A CN 109493979 A CN109493979 A CN 109493979A CN 201811235396 A CN201811235396 A CN 201811235396A CN 109493979 A CN109493979 A CN 109493979A
- Authority
- CN
- China
- Prior art keywords
- time series
- disease
- patient
- series data
- data matrix
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/80—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Abstract
The disease forecasting method and apparatus based on intelligent decision that the embodiment of the invention discloses a kind of, wherein the disease forecasting method based on intelligent decision includes: to obtain the corresponding treatment data of patient's unique encodings after desensitization from medical data base, and establish time series data matrix according to treatment data；According to the data characteristics of time series data matrix, the time series models for being used for disease forecasting are chosen；It is predicted according to disease of the time series models to patient, obtains the disease forecasting result of patient.Suitable time series models are chosen by handling patient medical data using the present invention, finally according to medical data and time series models, disease forecasting is carried out to patient, promotes disease forecasting accuracy, the disease prevention and treatment to patient provide guidance.
Description
Technical field
The present invention relates to data processing fields, and in particular to a kind of disease forecasting method and apparatus based on intelligent decision.
Background technique
Modern medical method is to prevent disease by early intervention rather than treated after diagnosis, in this way can be with
Largely patient is freed from disease.Traditionally, doctor or doctor's application risk calculator are sent out to assess disease
A possibility that exhibition, specifically includes the essential informations such as population in use statistics, medical condition, orbit to calculate and develop certain disease
A possibility that sick.This evaluation process only considers seldom variable, such as age, weight, height, gender, but disease forecasting
In the process, involved variable runs far deeper than these variables.Therefore, traditional prediction technique makes the accuracy rate of disease forecasting
It is very low, and the substantial connection of disease and time cannot be shown, and then provide guidance to medical prevention.How comprehensive to examine
Consider a variety of variables for being related in disease progression, and during these variables are applied to disease forecasting, obtains high precision
The disease forecasting of degree is as a result, be a problem to be solved.
Summary of the invention
The embodiment of the present invention provides a kind of disease forecasting method and apparatus based on intelligent decision, can be by curing to patient
Treat data to be handled, choose suitable time series models, finally according to medical data and time series models, to patient into
Row disease forecasting is able to ascend disease forecasting accuracy, and the disease prevention and treatment to patient provide guidance.
The first aspect of the embodiment of the present invention provides a kind of disease forecasting method based on intelligent decision, described to be based on intelligence
Can the disease forecasting method of decision include:
The corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base, and according to the treatment data
Establish time series data matrix；
According to the data characteristics of the time series data matrix, the time series models for being used for disease forecasting are chosen；
It is predicted according to disease of the time series models to the patient, obtains the disease forecasting knot of the patient
Fruit.
In an alternative scenario, the data characteristics according to the time series data matrix are chosen for disease forecasting
Time series models, comprising:
Determine whether the time series data matrix is steady；
If it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
Determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
Models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization；
Obtain best ARIMA model.
In an alternative scenario, it carries out predicting it according to disease of the time series models to the patient described
Before, the method also includes:
The treatment data for being used for test and verification is obtained, test set time series data matrix and verifying collection time series data square are established
Battle array；
The test set time series data matrix is imported in the best ARIMA model, prediction result is obtained；
The prediction result and verifying collection time series data matrix are compared, model prediction deviation is obtained, it is described
Model prediction deviation includes the coefficient of determination or bayesian information criterion；
If the model prediction deviation is less than predetermined deviation threshold value, it is determined that use the best ARIMA model.
It is in an alternative scenario, described that time series data matrix is established according to the treatment data, comprising:
The treatment data is subjected to feature extraction, obtains patient body condition information therein；
The patient body condition information is recorded according to the time, obtains time series data；
The corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, obtains timing
Data matrix.
It is in an alternative case, described to record the patient body situation according to the time, comprising:
The patient body condition information is matched with every kind of disease information in a variety of disease informations, acquisition and institute
State the matching similarity of every kind of disease information；
If the matching similarity is greater than the first preset threshold, determine that patient has the illness of this kind of disease to be inclined to；
If the kinds of Diseases with illness tendency of the patient are more than default species number, by the patient body situation
Information is recorded according to short period interval, and the time interval less than a week is divided between the short period.
In an alternative scenario, described to be predicted according to disease of the time series models to the patient, it obtains
The disease forecasting result of the patient includes:
According to the time series data matrix of the time series models combination patient, the disease forecasting knot of the patient is obtained
Fruit, the disease forecasting result include time, type and the probability of the possible illness of the patient.
In an alternative scenario, the method also includes:
If the probability of illness for predicting the patient for first kind disease is greater than the second preset threshold, reduce record institute
The time interval for stating patient body information obtains new time series data matrix；
The new time series data matrix predicts the disease of the patient in conjunction with the best ARIMA model,
Obtain new disease forecasting result；
If the new disease forecasting result, which also characterizes the patient, is greater than second for the probability of illness of first kind disease
Preset threshold, it is determined that the patient has the risk for suffering from first kind disease；
Prophylactic treatment is carried out to the patient for first kind disease.
The second aspect of the embodiment of the present invention provides a kind of disease forecasting device, and the disease forecasting device includes:
Unit is established, for obtaining the corresponding treatment data of patient's unique encodings after desensitization, and root from medical data base
Time series data matrix is established according to the treatment data；
Selection unit chooses the time sequence for being used for disease forecasting for the data characteristics according to the time series data matrix
Column model；
Predicting unit obtains the trouble for predicting according to disease of the time series models to the patient
The disease forecasting result of person.
The third aspect of the embodiment of the present invention provides a kind of electronic device, including processor, memory, communication interface, with
And one or more programs, one or more of programs are stored in the memory, and are configured by the processing
Device executes, and described program is included the steps that for executing the instruction in first aspect either method.
Fourth aspect of the embodiment of the present invention provides a kind of computer readable storage medium, and storage is used for electronic data interchange
Computer program, wherein the computer program make computer execute first aspect either method described in step finger
It enables.
As it can be seen that in the disease forecasting method based on intelligent decision described in the embodiment of the present invention, first from medical data
Library obtains the corresponding treatment data of patient's unique encodings after desensitization, and establishes time series data matrix according to treatment data, then
According to the data characteristics of time series data matrix, the time series models for being used for disease forecasting are chosen, finally according to time series mould
Type predicts the disease of patient, obtains the disease forecasting result of patient.In this process, pass through the place to medical data
Reason, establishes time series data matrix, contacts so that disease forecasting is established with the time, then according to the data characteristics of time series data matrix
Appropriate time series models are chosen, obtain disease forecasting as a result, the accuracy of disease forecasting result is improved, to the disease of patient
Disease, which prevents and treats, provides guidance.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of disease forecasting method flow schematic diagram based on intelligent decision provided in an embodiment of the present invention；
Fig. 2 is the flow diagram of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention；
Fig. 3 is the flow diagram of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention；
Fig. 4 is the flow diagram of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention；
Fig. 5 is a kind of structural schematic diagram of electronic device provided in an embodiment of the present invention；
Fig. 6 is a kind of structural block diagram of disease forecasting device provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Referenced herein " embodiment " is it is meant that a particular feature, structure, or characteristic described can wrap in conjunction with the embodiments
Containing at least one embodiment of the present invention.It is identical that each position in the description shows that the phrase might not be each meant
Embodiment, nor the independent or alternative embodiment with other embodiments mutual exclusion.Those skilled in the art explicitly and
Implicitly understand, embodiment described herein can be combined with other embodiments.
It describes in detail below to the embodiment of the present invention.
Referring to Fig. 1, Fig. 1 is a kind of disease forecasting method flow signal based on intelligent decision in the embodiment of the present invention
Figure, as shown in Figure 1, the disease forecasting method based on intelligent decision includes:
101, the corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base, and according to the treatment
Data establish time series data matrix.
In internet society, the health data of patient, including reservation register, examine middle data, payment data or health examination
Deng can be recorded or be stored by internet.Library is established according to these medical treatment big datas, for health monitoring, trouble
Person's treatment, pharmaceutical developments or medical insurance policies formulation etc., all have highly important value.
After establishing library according to the health data of patient, in order to guarantee the privacy of patient, need to the private data of patient into
Row desensitization process is specifically included to personal information truthful datas such as such as name, identification card number, cell-phone number, medical insurance card number, customer IDs
It is transformed, is used with providing test.Simultaneously in order to distinguish different patient and its correspondence, need to be arranged for each patient unique
Coding, it is corresponding to obtain the patient according to unique encodings.
After obtaining treatment data, time series data matrix can be established.Time series data refers to time series data, is same unification
The data column that index records in chronological order.The matrix of time data composition is time series data matrix.Include in medical data
The diagnostic data and physical examination data of patient, specific include weight, height, working environment, the living environment, physical examination knot of patient again
The information such as fruit, common index, diagnosis construct time series data matrix, available patient with the long-term health data of patient
Physical condition information is characterized, and diagnosis information is label, establishes time series data matrix, specific as shown in table 1:
1 time series data matrix of table
After the crucial word description for obtaining patient body condition information, it is also necessary to quantification treatment is carried out to these information, so as to
It imports model and carries out quantitative analysis.For the quantification treatment of physical condition information, a quantization table can establish, include original number
According to and its corresponding quantized values.A quantized values can be obtained according to each single item body index, it can also be according to all
Body index obtains a quantized values.The method for obtaining quantized values, which may is that, seeks measured parameter value and normal parameter values
The difference of the maximum value of range and the difference of minimum value, and according to and the difference of maximum value and the difference of minimum value acquire product, as quantization
Numerical value.It is as shown in formula 1:
F1=(f1-fmin)*(f1-max)(1)
Wherein F1 indicates quantized values, and f1 indicates measured parameter value, fminIndicate the minimum value of normal parameter values range, fmax
Indicate the maximum value of normal parameter values range.When measured parameter value is not within the scope of normal parameter values, F1 is positive number, works as survey
When amount parameter value is within the scope of normal parameter values, F1 is negative, therefore can be easy to whether observation measured parameter value is in
Normal parameter values range.
Optionally, for this label data of diagnosis information, an affecting parameters be can be used as, for correcting quantization number
Value.If diagnosis information is obtained, then according to diagnosis information while obtaining the physical condition information of patient
Grading, for quantized values assign weight.Such as in table 1,2018.5-2018.7 has been diagnosed as type 1 diabetes to patient, that
One can be arranged for F1 and make a definite diagnosis weight α, if to patient being the diagnostic result of " doubtful ", one can be arranged for F1
Doubtful weight β, wherein α > β.
Optionally, for this label data of diagnosis information, a guide parameters be can also be used as, for patient
The guidance of predictive disease type, such as 2018.5-2018.7 have been diagnosed as type 1 diabetes to patient, then the trouble can be predicted
Person suffers from the probability of type II diabetes.
102, according to the data characteristics of the time series data matrix, the time series models for being used for disease forecasting are chosen.
There is stationarity in time series data, fluctuate up and down around a constant and fluctuation range is limited, that is, have constant
Mean value and constant variance, then it represents that time series data is steady；If there is apparent trend or periodicity, then show time series data injustice
Surely.In addition, if time series data disorderly random fluctuation completely, it is extractable without any information, then show that the time series data is
White noise sequence.For stable time series data, autoregressive moving-average model (arma modeling, Auto- can be used
Regressive and Moving Average Model) it is used as prediction model, it, then can be with for the time series data of non-stationary
Using autoregression integral moving average model (Autoregressive Integrated Moving Average Model,
ARIMA model) it is used as prediction model, ARIMA model can carry out difference, logarithmic transformation, square root transformation etc. to time series data
Processing makes time series reach tranquilization, and then carries out models fitting to time series data, obtains most appropriate prediction model.This
Outside, (Extreme Gradient Boosting, XGBoost) algorithm or artificial neural network can also be promoted using limit gradient
Network scheduling algorithm carries out disease forecasting for time series data.
Optionally, according to the data characteristics of time series data matrix, time series models of the selection for disease forecasting include:
Determine whether time series data matrix is steady；If it is not, then pre-processing to time series data matrix, the time series data of tranquilization is obtained
Matrix；Determine that it corresponds to order according to the characteristics of time series data matrix of tranquilization；Using specific ARIMA model and tranquilization
Time series data matrix carry out models fitting；Obtain best ARIMA model.
When determining whether time series data matrix is steady, the method that can be used includes: directly to draw becoming for time series
Gesture figure, sees Trend judgement；Or draw auto-correlation and partial autocorrelation figure, the autocorrelogram of stable sequence
(Autocorrelation) and partial correlation figure (Partial Correlation) otherwise hangover or be truncation；Or pass through
Unit root test whether there is unit root in checking sequence, be exactly nonstationary time series if there is unit root.Unstable sequence
Column can be stationary sequence by differential conversion.K order difference is exactly to subtract each other between two sequential values of k phase.Time series
After data are handled by tranquilization, ARIMA (p, d, q) model translates into ARMA (p, q) model.Wherein in ARIMA (p, d, q), p
For autoregression item, q is rolling average item number, the difference number that d is done when becoming steady by time series.Using auto-correlation function
Or partial autocorrelation function differentiates the coefficient and order of ARMA (p, q) model, then carries out parameter Estimation to arma modeling, can adopt
Method for parameter estimation includes the methods of maximal possibility estimation or least square product.When obtaining multiple temporal models, constantly
Debugging is compared, and the superiority and inferiority degree of contrast model is carried out by test of fitness of fot statistic, obtains best ARIMA model.
Optionally, before being predicted according to disease of the time series models to patient, this method further include: used
In the treatment data of test and verification, test set time series data matrix and verifying collection time series data matrix are established；When by test set
Sequence data matrix imports in best ARIMA model, obtains prediction result；Prediction result and verifying collection time series data matrix are carried out
Comparison obtains model prediction deviation, and model prediction deviation includes the coefficient of determination or bayesian information criterion；If model prediction deviation
Less than predetermined deviation threshold value, it is determined that use best ARIMA model.
Before predicting according to disease of the time series models to patient, the prediction of Check-Out Time series model is needed
Effect obtains test result by test set, then compare the actual result that test result is concentrated with verifying, obtains
Model prediction deviation illustrates that time series models performance is excellent if model prediction deviation is less than predetermined deviation threshold value, can
Using the best ARIMA model.
Optionally, patient body situation is recorded according to the time, comprising: by patient body condition information and a variety of diseases
Every kind of disease information in sick information is matched, and the matching similarity with every kind of disease information is obtained；If matching similarity is big
In the first preset threshold, then determine that patient has the illness of this kind of disease to be inclined to；If the kinds of Diseases with illness tendency of patient
More than default species number, then patient body condition information is recorded according to short period interval, be divided between short period small
In the time interval in a week.
Specifically, the patient body condition information got be it is very abundant, between every kind of physical condition information again mutually
Influence is blended, will increase the difficulty of patient disease prediction.Therefore, by the physical condition information of patient and a variety of disease informations
In each disease information matched, and it is similar to the matching of each disease information to calculate patient body condition information
Degree, such as patient's first has physical condition information in 5 to be in outside normal range value, wherein 4 meet the symptom of A class disease, and A class
A total of 8 kinds of performances symptom of disease, then the matching similarity of the physical condition information of patient's first and A class disease are as follows: 4 ÷ 8*
100%=50%, i.e. patient meet the performance symptom number ÷ of such disease such disease and show symptom number * 100% in total.If
The similarity of patient and certain disease is greater than preset threshold, such as 80%, illustrates that patient has the tendency that suffering from such disease, if suffered from
Person has the tendency that suffering from default species number disease, illustrates that the patient is primary part observation formation, need to record according to short period interval
The patient body condition information, so as to the physical condition of the patient body more closely be monitored, to make more accurate disease
Prediction.Wherein, default species number can be any quantity greater than 0；It is divided between short period between the time no more than a week
Every, such as 7 days, 3 days etc..
In addition, when carrying out disease forecasting using ARIMA model, since the physical condition of people is variation, and ARIMA mould
Type is predicted according to data rule, then big in body situation of change, when so that huge change occurring for data rule,
The correspondence parameter of ARIMA model also will do it adjustment.Therefore, in order to make each time ARIMA model all with the body of the period
Situation data match, it should according to physical condition information update ARIMA model.The method of update includes: fixed duration change
Method.I.e. each fixed duration interval updates an ARIMA model, and fixed duration interval can be 1 year, 2 years etc..Or using change
Point prediction method.I.e. in the segmentation that physical condition data are carried out to the partial stack time, ARIMA model is imported, if when non-superimposed
Between prediction deviation caused by partial data be greater than predetermined deviation value, then height is set by the starting point of non-superimposed time, from height
Start, obtain new ARIMA model parameter, updates ARIMA model.
As it can be seen that in the embodiment of the present application, by specifically studying time series data matrix data, choosing disease forecasting
Time series models, and after having chosen time series models, training set and test set data are obtained, to time series models
Prediction accuracy is assessed, and further determines that best model.Simultaneously, it is contemplated that the variability of time series data matrix, to the time
Series model is updated and adjusts, and further determined more acurrate and effective time series models, so as to subsequent carry out disease
When disease forecasting, forecasting accuracy is improved.
103, predict that the disease for obtaining the patient is pre- according to disease of the time series models to the patient
Survey result.
The illness of patient usually has certain regularity, such as certain age brackets are the high incidence periods of certain class disease, certain
A little crowds are that the group of people at high risk of certain class disease can be to patient by the long-term record and Model Matching to patient body situation
The disease of illness predicted.
Optionally, it is predicted according to disease of the time series models to patient, obtains the disease forecasting result packet of patient
It includes: according to the time series data matrix of time series models combination patient, obtaining the disease forecasting of patient as a result, disease forecasting result
Time, type and probability including the possible illness of patient.
Specifically, according to time series data matrix and time series models, the body amount of patient on a timeline can be obtained
Change data fluctuations situation, when data fluctuations are to some value, show patient, therefore, can sick time to patient into
Row prediction.And fluctuation situation of the time series data matrix of patient on time series models, it can also be with the parameter of certain class disease
Fluctuation situation of the value on time series models is matched, then the type and probability of patients' disease can be predicted.
Optionally, this method further include: preset if predicting patient and being greater than second for the probability of illness of first kind disease
Threshold value then reduces the time interval of record patient body information, obtains new time series data matrix；By new time series data matrix
The disease of patient is predicted in conjunction with best ARIMA model, obtains new disease forecasting result；If new disease forecasting result
Also characterization patient is greater than the second preset threshold for the probability of illness of first kind disease, it is determined that patient, which has, suffers from first kind disease
Risk；Prophylactic treatment is carried out to patient for first kind disease.
Specifically, if predicting patient according to time series models is greater than second in advance for the probability of illness of first kind disease
If threshold value, then needing to carry out primary part observation to patient for such disease, shorten the time interval of record patient body information,
New time series data matrix is obtained, so that previous disease forecasting result is verified and supplemented, if second is predicted
As a result still characterization patient for the probability of illness of first kind disease is greater than the second preset threshold, then it is determined that patient, which has, suffers from the
The risk of a kind of disease.Prophylactic treatment can be carried out to patient for first kind disease, reduce the probability of patient.
As it can be seen that in the embodiment of the present application, it is corresponding to obtain patient's unique encodings after desensitization from medical data base first
Treatment data, and time series data matrix is established according to treatment data, then according to the data characteristics of time series data matrix, chooses and use
In the time series models of disease forecasting, is finally predicted according to disease of the time series models to patient, obtain patient's
Disease forecasting result.In this process, by the processing to medical data, time series data matrix is established, so that disease forecasting
It establishes and contacts with the time, appropriate time series models are then chosen according to the data characteristics of time series data matrix, obtain disease
Prediction result improves the accuracy of disease forecasting result, and the disease prevention and treatment to patient provide guidance.
Referring to Fig. 2, Fig. 2 is the stream of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention
Journey schematic diagram, as shown, the disease forecasting method based on intelligent decision in the present embodiment includes:
201, the corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base；
202, the treatment data is subjected to feature extraction, obtains patient body condition information therein；
203, the patient body condition information is recorded according to the time, obtains time series data；
204, the corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, is obtained
Time series data matrix；
205, determine whether the time series data matrix is steady；
206, if it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
207, determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
208, models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization, obtained most
Good ARIMA model, as the time series models for disease forecasting；
209, predict that the disease for obtaining the patient is pre- according to disease of the time series models to the patient
Survey result.
In embodiments of the present invention, by obtaining the patient body condition information in treatment data, believe in conjunction with medical treatment
Breath, establishes time series data matrix, then carries out tranquilization processing to time series data, determines time series data order of matrix number, and adopt
Models fitting is carried out with ARIMA model and the time series data matrix of tranquilization, obtains best ARIMA model, as disease
The time series models of prediction.This process improves the efficiency and accuracy of access time series model, enables model
For disease forecasting, and then improve the accuracy rate of disease forecasting.
Referring to Fig. 3, Fig. 3 is the stream of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention
Journey schematic diagram, as shown, the disease forecasting method based on intelligent decision in the present embodiment includes:
301, the corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base, and according to the treatment
Data establish time series data matrix；
302, determine whether the time series data matrix is steady；
303, if it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
304, determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
305, models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization, obtained most
Good ARIMA model；
306, the treatment data for being used for test and verification is obtained, ordinal number when establishing test set time series data matrix and verifying collection
According to matrix；
307, the test set time series data matrix is imported in the best ARIMA model, obtains prediction result；
308, the prediction result and verifying collection time series data matrix are compared, obtain model prediction deviation,
The model prediction deviation includes the coefficient of determination or bayesian information criterion；
If 309, the model prediction deviation is less than predetermined deviation threshold value, it is determined that made using the best ARIMA model
For the time series models for disease forecasting；
310, predict that the disease for obtaining the patient is pre- according to disease of the time series models to the patient
Survey result.
In embodiments of the present invention, by pre-processing to time series data matrix, the time series data square of tranquilization is obtained
Battle array, then determines best ARIMA model according to the time series data matrix of tranquilization.For the best ARIMA model got, adopt
Its prediction deviation is verified with the time series data matrix of test set and verifying collection, if prediction deviation is less than predetermined deviation threshold
Value, it is determined that use the best ARIMA model as the time series models for being used for disease forecasting.This process to use most
Good ARIMA model is further verified, so that the prediction deviation for the best ARIMA model chosen is small, meets forecast demand,
Promote the efficiency and accuracy rate of disease forecasting.
Referring to Fig. 4, Fig. 4 is the stream of another disease forecasting method based on intelligent decision provided in an embodiment of the present invention
Journey schematic diagram, as shown, the disease forecasting method based on intelligent decision in the present embodiment includes:
401, the corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base；
402, the treatment data is subjected to feature extraction, obtains patient body condition information therein；
403, the patient body condition information is matched with every kind of disease information in a variety of disease informations, is obtained
With the matching similarity of every kind of disease information；
If 404, the matching similarity is greater than the first preset threshold, determine that patient has the illness of this kind of disease to be inclined to；
If 405, the kinds of Diseases with illness tendency of the patient are more than default species number, by the patient body
Condition information is recorded according to short period interval, is obtained time series data, is divided between the short period no more than a week
Time interval；
406, the corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, is obtained
Time series data matrix；
407, according to the time series data matrix of the time series models combination patient, the disease forecasting of the patient is obtained
As a result, the disease forecasting result includes time, type and the probability of the possible illness of the patient；
If the probability of illness for 408, predicting the patient for first kind disease is greater than the second preset threshold, reduce note
The time interval for recording the patient body information obtains new time series data matrix；
409, the new time series data matrix is carried out in conjunction with disease of the best ARIMA model to the patient pre-
It surveys, obtains new disease forecasting result；
If 410, the new disease forecasting result also characterizes the patient and is greater than for the probability of illness of first kind disease
Second preset threshold, it is determined that the patient has the risk for suffering from first kind disease；
411, prophylactic treatment is carried out to the patient for first kind disease.
In embodiments of the present invention, the patient body condition information in treatment data is obtained, then according to patient body shape
Condition information determines that patient suffers from the disease the species number of tendency, if patient has the tendency for suffering from the disease more than types of forecast number,
The physical condition information of the user is recorded according to short period interval, is generated time series data and is put to the proof, is used for time series models, obtains
Obtain prediction result.If the probability that prediction result shows that patient suffers from certain disease type is excessively high, reduce record patient body information
Time interval, obtain new time series data matrix, and finally obtain new disease forecasting as a result, according to new disease forecasting knot
Fruit determines whether patient has the risk for suffering from such disease, and makes prophylactic treatment.In this process, by record user's body
The adjustment of body condition information time interval, so that model carries out more accurate prediction to patient disease, and then it is pre- to improve disease
Survey the accuracy rate of result.
Fig. 5 is a kind of structural schematic diagram of electronic device provided in an embodiment of the present invention, as shown in figure 5, the electronic device
Including processor, memory, communication interface and one or more programs, wherein said one or multiple programs are stored in
In above-mentioned memory, and it is configured to be executed by above-mentioned processor, above procedure includes the instruction for executing following steps:
The corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base, and according to the treatment data
Establish time series data matrix；
According to the data characteristics of the time series data matrix, the time series models for being used for disease forecasting are chosen；
It is predicted according to disease of the time series models to the patient, obtains the disease forecasting knot of the patient
Fruit.
As it can be seen that electronic device obtains the corresponding treatment data of patient's unique encodings after desensitization from medical data base first,
And time series data matrix is established according to treatment data, then according to the data characteristics of time series data matrix, choose pre- for disease
The time series models of survey are finally predicted according to disease of the time series models to patient, obtain the disease forecasting of patient
As a result.In this process, by the processing to medical data, time series data matrix is established, so that disease forecasting is built with the time
Vertical connection, then chooses appropriate time series models according to the data characteristics of time series data matrix, obtain disease forecasting as a result,
The accuracy of disease forecasting result is improved, the disease prevention and treatment to patient provide guidance.
In a possible example, in the data characteristics according to the time series data matrix, chooses and be used for disease
In terms of the time series models of prediction, described program includes the instruction for executing following steps:
Determine whether the time series data matrix is steady；
If it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
Determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
Models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization, is obtained best
ARIMA model, as the time series models for disease forecasting.
In a possible example, predicted described according to disease of the time series models to the patient
Before, described program includes the instruction for executing following steps:
The treatment data for being used for test and verification is obtained, test set time series data matrix and verifying collection time series data square are established
Battle array；
The test set time series data matrix is imported in the best ARIMA model, prediction result is obtained；
The prediction result and verifying collection time series data matrix are compared, model prediction deviation is obtained, it is described
Model prediction deviation includes the coefficient of determination or bayesian information criterion；
If the model prediction deviation is less than predetermined deviation threshold value, it is determined that use the best ARIMA model as use
In the time series models of disease forecasting.
In a possible example, it is described time series data matrix is established according to the treatment data in terms of, the journey
Sequence includes the instruction for executing following steps:
The treatment data is subjected to feature extraction, obtains patient body condition information therein；
The patient body condition information is recorded according to the time, obtains time series data；
The corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, obtains timing
Data matrix.
In a possible example, in terms of the patient body situation is recorded according to the time, described program
Including the instruction for executing following steps:
The patient body condition information is matched with every kind of disease information in a variety of disease informations, acquisition and institute
State the matching similarity of every kind of disease information；
If the matching similarity is greater than the first preset threshold, determine that patient has the illness of this kind of disease to be inclined to；
If the kinds of Diseases with illness tendency of the patient are more than default species number, by the patient body situation
Information is recorded according to short period interval, and the time interval no more than a week is divided between the short period.
In a possible example, predicts, obtain according to disease of the time series models to the patient
In terms of the disease forecasting result for obtaining the patient, described program includes the instruction for executing following steps:
According to the time series data matrix of the time series models combination patient, the disease forecasting knot of the patient is obtained
Fruit, the disease forecasting result include time, type and the probability of the possible illness of the patient.
In a possible example, described program packet also includes the instruction for executing following steps:
If the probability of illness for predicting the patient for first kind disease is greater than the second preset threshold, reduce record institute
The time interval for stating patient body information obtains new time series data matrix；
The new time series data matrix predicts the disease of the patient in conjunction with the best ARIMA model,
Obtain new disease forecasting result；
If the new disease forecasting result, which also characterizes the patient, is greater than second for the probability of illness of first kind disease
Preset threshold, it is determined that the patient has the risk for suffering from first kind disease；
Prophylactic treatment is carried out to the patient for first kind disease.
Fig. 6 is the functional unit composition block diagram of disease forecasting device 600 involved in the embodiment of the present invention.The disease is pre-
It surveys device 600 and is applied to electronic device, the disease forecasting device includes:
Unit 601 is established, for obtaining the corresponding treatment data of patient's unique encodings after desensitization from medical data base, and
Time series data matrix is established according to the treatment data；
Selection unit 602 chooses the time for being used for disease forecasting for the data characteristics according to the time series data matrix
Series model；
Predicting unit 603, for being predicted according to disease of the time series models to the patient, described in acquisition
The disease forecasting result of patient.
Herein, it should be noted that the above-mentioned specific works for establishing unit 601, selection unit 602 and predicting unit 603
Process referring to above-mentioned steps 101-103 associated description.Details are not described herein.
As can be seen that in embodiments of the present invention, disease forecasting device obtains the trouble after desensitization from medical data base first
The corresponding treatment data of person's unique encodings, and time series data matrix is established according to treatment data, then according to time series data matrix
Data characteristics, choose the time series models for being used for disease forecasting, finally according to time series models to the disease of patient into
Row prediction, obtains the disease forecasting result of patient.In this process, by the processing to medical data, time series data is established
Matrix contacts so that disease forecasting is established with the time, the appropriate time is then chosen according to the data characteristics of time series data matrix
Series model obtains disease forecasting as a result, improving the accuracy of disease forecasting result, mentions to the disease prevention and treatment of patient
For guidance.
In an alternative case, in the data characteristics according to the time series data matrix, the time for being used for disease forecasting is chosen
In terms of series model, the selection unit 602 is specifically used for:
Determine whether the time series data matrix is steady；
If it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
Determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
Models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization, is obtained best
ARIMA model, as the time series models for disease forecasting.
In an alternative case, described before being predicted according to disease of the time series models to the patient
Selection unit 602 also particularly useful for:
The treatment data for being used for test and verification is obtained, test set time series data matrix and verifying collection time series data square are established
Battle array；
The test set time series data matrix is imported in the best ARIMA model, prediction result is obtained；
The prediction result and verifying collection time series data matrix are compared, model prediction deviation is obtained, it is described
Model prediction deviation includes the coefficient of determination or bayesian information criterion；
If the model prediction deviation is less than predetermined deviation threshold value, it is determined that use the best ARIMA model as use
In the time series models of disease forecasting.
In an alternative case, described to establish the tool of unit 601 in terms of establishing time series data matrix according to the treatment data
Body is used for:
The treatment data is subjected to feature extraction, obtains patient body condition information therein；
The patient body condition information is recorded according to the time, obtains time series data；
The corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, obtains timing
Data matrix.
In an alternative case, described to establish unit 601 in terms of being recorded the patient body situation according to the time
It is specifically used for:
The patient body condition information is matched with every kind of disease information in a variety of disease informations, acquisition and institute
State the matching similarity of every kind of disease information；
If the matching similarity is greater than the first preset threshold, determine that patient has the illness of this kind of disease to be inclined to；If institute
The kinds of Diseases with illness tendency for stating patient are more than default species number, then by the patient body condition information according to short-term
Time interval is recorded, and the time interval no more than a week is divided between the short period.
In an alternative case, it is predicted according to disease of the time series models to the patient, described in acquisition
In terms of the disease forecasting result of patient, the predicting unit 603 is specifically used for:
According to the time series data matrix of the time series models combination patient, the disease forecasting knot of the patient is obtained
Fruit, the disease forecasting result include time, type and the probability of the possible illness of the patient.
In an alternative case, the disease forecasting device 600 further includes determination unit 601, is specifically used for:
If the probability of illness for predicting the patient for first kind disease is greater than the second preset threshold, reduce record institute
The time interval for stating patient body information obtains new time series data matrix；
The new time series data matrix predicts the disease of the patient in conjunction with the best ARIMA model,
Obtain new disease forecasting result；
If the new disease forecasting result, which also characterizes the patient, is greater than second for the probability of illness of first kind disease
Preset threshold, it is determined that the patient has the risk for suffering from first kind disease；
Prophylactic treatment is carried out to the patient for first kind disease.
The embodiment of the present invention also provides a kind of computer storage medium, wherein computer storage medium storage is for electricity
The computer program of subdata exchange, the computer program make computer execute any as recorded in above method embodiment
Some or all of method step, above-mentioned computer include mobile terminal.
The embodiment of the present invention also provides a kind of computer program product, and above-mentioned computer program product includes storing calculating
The non-transient computer readable storage medium of machine program, above-mentioned computer program are operable to that computer is made to execute such as above-mentioned side
Some or all of either record method step in method embodiment.The computer program product can be a software installation
Packet, above-mentioned computer includes mobile terminal.
In several embodiments provided herein, it should be understood that disclosed device, it can be by another way
It realizes.For example, the apparatus embodiments described above are merely exemplary, such as the division of said units, it is only a kind of
Logical function partition, there may be another division manner in actual implementation, such as multiple units or components can combine or can
To be integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Coupling, direct-coupling or communication connection can be through some interfaces, the indirect coupling or communication connection of device or unit,
It can be electrical or other forms.
Above-mentioned unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If above-mentioned integrated unit is realized in the form of SFU software functional unit and when sold or used as an independent product,
It can store in a computer-readable access to memory.Based on this understanding, the technical solution of the application substantially or
Say that all or part of the part that contributes to existing technology or the technical solution can embody in the form of software products
Out, which is stored in a memory, including some instructions are used so that a computer equipment (can
For personal computer, server or network equipment etc.) execute all or part of step of each embodiment above method of the application
Suddenly.And memory above-mentioned includes: USB flash disk, read-only memory (Read-Only Memory, ROM), random access memory
The various media that can store program code such as (Random Access Memory, RAM), mobile hard disk, magnetic or disk.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can store in a computer-readable memory, memory
It may include: flash disk, ROM, RAM, disk or CD etc..
The embodiment of the present invention has been described in detail above, specific case used herein to the principle of the application and
Embodiment is expounded, the description of the example is only used to help understand the method for the present application and its core ideas；
At the same time, for those skilled in the art can in specific embodiments and applications according to the thought of the application
There is change place, in conclusion the contents of this specification should not be construed as limiting the present application.
Claims (10)
1. a kind of disease forecasting method based on intelligent decision, which is characterized in that the described method includes:
The corresponding treatment data of patient's unique encodings after desensitization is obtained from medical data base, and is established according to the treatment data
Time series data matrix；
According to the data characteristics of the time series data matrix, the time series models for being used for disease forecasting are chosen；
It is predicted according to disease of the time series models to the patient, obtains the disease forecasting result of the patient.
2. the method according to claim 1, wherein the data characteristics according to the time series data matrix,
Choose the time series models for being used for disease forecasting, comprising:
Determine whether the time series data matrix is steady；
If it is not, then pre-processing to the time series data matrix, the time series data matrix of tranquilization is obtained；
Determine that it corresponds to order according to the characteristics of time series data matrix of the tranquilization；
Models fitting is carried out using specific ARIMA model and the time series data matrix of the tranquilization, obtains best ARIMA mould
Type, as the time series models for disease forecasting.
3. according to the method described in claim 2, it is characterized in that, it is described according to the time series models to the patient
Disease predicted before, the method also includes:
The treatment data for being used for test and verification is obtained, test set time series data matrix and verifying collection time series data matrix are established；
The test set time series data matrix is imported in the best ARIMA model, prediction result is obtained；
The prediction result and verifying collection time series data matrix are compared, model prediction deviation, the model are obtained
Prediction deviation includes the coefficient of determination or bayesian information criterion；
If the model prediction deviation is less than predetermined deviation threshold value, it is determined that use the best ARIMA model as disease
The time series models of disease forecasting.
4. according to the method described in claim 3, it is characterized in that, described establish time series data square according to the treatment data
Battle array, comprising:
The treatment data is subjected to feature extraction, obtains patient body condition information therein；
The patient body condition information is recorded according to the time, obtains time series data；
The corresponding medical treatment information of the patient body condition information is obtained, in conjunction with the time series data, obtains time series data
Matrix.
5. according to the method described in claim 4, it is characterized in that, described remember the patient body situation according to the time
Record, comprising:
The patient body condition information is matched with every kind of disease information in a variety of disease informations, is obtained and described every
The matching similarity of kind disease information；
If the matching similarity is greater than the first preset threshold, determine that patient has the illness of this kind of disease to be inclined to；
If the kinds of Diseases with illness tendency of the patient are more than default species number, by the patient body condition information
It is recorded according to short period interval, the time interval no more than a week is divided between the short period.
6. method described in -5 according to claim 1, which is characterized in that it is described according to the time series models to the patient
Disease predicted that the disease forecasting result for obtaining the patient includes:
According to the time series data matrix of the time series models combination patient, the disease forecasting of the patient is obtained as a result, institute
State time, type and the probability that disease forecasting result includes the possible illness of the patient.
7. according to the method described in claim 6, it is characterized in that, the method also includes:
It is greater than the second preset threshold for the probability of illness of first kind disease if predicting the patient, reduction records the trouble
The time interval of person's biological information obtains new time series data matrix；
The new time series data matrix predicts the disease of the patient in conjunction with the best ARIMA model, obtains
New disease forecasting result；
If it is default greater than second for the probability of illness of first kind disease that the new disease forecasting result also characterizes the patient
Threshold value, it is determined that the patient has the risk for suffering from first kind disease；
Prophylactic treatment is carried out to the patient for first kind disease.
8. a kind of disease forecasting device, which is characterized in that the disease forecasting device includes:
Unit is established, for obtaining the corresponding treatment data of patient's unique encodings after desensitization from medical data base, and according to institute
It states treatment data and establishes time series data matrix；
Selection unit chooses the time series mould for being used for disease forecasting for the data characteristics according to the time series data matrix
Type；
Predicting unit obtains the patient's for predicting according to disease of the time series models to the patient
Disease forecasting result.
9. a kind of electronic device, including processor, memory, communication interface, and one or more programs, one or more
A program is stored in the memory, and is configured to be executed by the processor, and described program includes being used for right of execution
Benefit requires the instruction of the step in 1-7 any means.
10. a kind of computer readable storage medium, storage is used for the computer program of electronic data interchange, wherein the calculating
Machine program makes the instruction of step described in any one of computer perform claim requirement 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811235396.2A CN109493979A (en) | 2018-10-23 | 2018-10-23 | A kind of disease forecasting method and apparatus based on intelligent decision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811235396.2A CN109493979A (en) | 2018-10-23 | 2018-10-23 | A kind of disease forecasting method and apparatus based on intelligent decision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109493979A true CN109493979A (en) | 2019-03-19 |
Family
ID=65692512
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811235396.2A Pending CN109493979A (en) | 2018-10-23 | 2018-10-23 | A kind of disease forecasting method and apparatus based on intelligent decision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109493979A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032670A (en) * | 2019-04-17 | 2019-07-19 | 腾讯科技（深圳）有限公司 | Method for detecting abnormality, device, equipment and the storage medium of time series data |
| CN110782989A (en) * | 2019-09-18 | 2020-02-11 | 平安科技（深圳）有限公司 | Data analysis method, device, equipment and computer readable storage medium |
| CN110893099A (en) * | 2019-11-26 | 2020-03-20 | 江苏大学 | Intelligent early warning method for sudden diseases in sleep state |
| CN111063430A (en) * | 2019-11-04 | 2020-04-24 | 珠海健康云科技有限公司 | Disease prediction method and device |
| CN111161880A (en) * | 2019-12-23 | 2020-05-15 | 平安医疗健康管理股份有限公司 | Medical information classification method and device based on classification model and computer equipment |
| CN111370122A (en) * | 2020-02-27 | 2020-07-03 | 西安交通大学 | Knowledge guidance-based time sequence data risk prediction method and system and application thereof |
| CN111462919A (en) * | 2020-03-31 | 2020-07-28 | 中国科学院软件研究所 | Method and system for predicting insect-borne diseases based on sliding window time sequence model |
| CN111785370A (en) * | 2020-07-01 | 2020-10-16 | 医渡云（北京）技术有限公司 | Medical record data processing method and device, computer storage medium and electronic equipment |
| CN112164448A (en) * | 2020-09-25 | 2021-01-01 | 上海市胸科医院 | Training method, prediction system, method and medium of immunotherapy efficacy prediction model |
| CN113611430A (en) * | 2021-07-28 | 2021-11-05 | 广东省科学院智能制造研究所 | Epidemic situation prediction method and device based on Bayesian neural network |
| CN115470936A (en) * | 2022-09-23 | 2022-12-13 | 广州爱浦路网络技术有限公司 | NWDAF-based machine learning model updating method and device |
| CN117219284A (en) * | 2023-09-11 | 2023-12-12 | 广东德澳智慧医疗科技有限公司 | System with time sequence intelligent medical big data management |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101894309A (en) * | 2009-11-05 | 2010-11-24 | 南京医科大学 | Epidemic situation predicting and early warning method of infectious diseases |
| CN104970775A (en) * | 2015-07-01 | 2015-10-14 | 深圳市前海安测信息技术有限公司 | White plague early symptom monitoring method and system |
| CN107688872A (en) * | 2017-08-20 | 2018-02-13 | 平安科技（深圳）有限公司 | Forecast model establishes device, method and computer-readable recording medium |
-
2018
- 2018-10-23 CN CN201811235396.2A patent/CN109493979A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101894309A (en) * | 2009-11-05 | 2010-11-24 | 南京医科大学 | Epidemic situation predicting and early warning method of infectious diseases |
| CN104970775A (en) * | 2015-07-01 | 2015-10-14 | 深圳市前海安测信息技术有限公司 | White plague early symptom monitoring method and system |
| CN107688872A (en) * | 2017-08-20 | 2018-02-13 | 平安科技（深圳）有限公司 | Forecast model establishes device, method and computer-readable recording medium |
Non-Patent Citations (3)
| Title |
|---|
| 曾高;焦风;宋修会;路步来;刘来福;栾文忠;梁冶矢;: "时间序列ARIMA模型应用于颅内压监测初步探讨", 中国临床神经外科杂志, no. 10, pages 583 - 585 * |
| 王艳杰;杨蕾;吴文华;任泉;王倩;刘早玲;: "ARIMA模型在乌鲁木齐市儿童呼吸系统疾病发病趋势预测中的应用", 职业与健康, no. 21 * |
| 胡子慧;李亮明;林少华;李凤明;黄汉添;廖巍;: "应用自回归单整移动平均模型预测高血压脑出血患者颅内压的可行性", 广东医学, no. 21 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032670A (en) * | 2019-04-17 | 2019-07-19 | 腾讯科技（深圳）有限公司 | Method for detecting abnormality, device, equipment and the storage medium of time series data |
| CN110032670B (en) * | 2019-04-17 | 2022-11-29 | 腾讯科技（深圳）有限公司 | Method, device and equipment for detecting abnormity of time sequence data and storage medium |
| WO2021052156A1 (en) * | 2019-09-18 | 2021-03-25 | 平安科技（深圳）有限公司 | Data analysis method, apparatus and device, and computer readable storage medium |
| CN110782989A (en) * | 2019-09-18 | 2020-02-11 | 平安科技（深圳）有限公司 | Data analysis method, device, equipment and computer readable storage medium |
| CN111063430A (en) * | 2019-11-04 | 2020-04-24 | 珠海健康云科技有限公司 | Disease prediction method and device |
| CN111063430B (en) * | 2019-11-04 | 2024-01-26 | 珠海健康云科技有限公司 | Disease prediction method and device |
| CN110893099A (en) * | 2019-11-26 | 2020-03-20 | 江苏大学 | Intelligent early warning method for sudden diseases in sleep state |
| CN111161880A (en) * | 2019-12-23 | 2020-05-15 | 平安医疗健康管理股份有限公司 | Medical information classification method and device based on classification model and computer equipment |
| CN111161880B (en) * | 2019-12-23 | 2022-12-02 | 深圳平安医疗健康科技服务有限公司 | Medical information classification method and device based on classification model and computer equipment |
| CN111370122B (en) * | 2020-02-27 | 2023-12-19 | 西安交通大学 | Time sequence data risk prediction method and system based on knowledge guidance and application thereof |
| CN111370122A (en) * | 2020-02-27 | 2020-07-03 | 西安交通大学 | Knowledge guidance-based time sequence data risk prediction method and system and application thereof |
| CN111462919A (en) * | 2020-03-31 | 2020-07-28 | 中国科学院软件研究所 | Method and system for predicting insect-borne diseases based on sliding window time sequence model |
| CN111462919B (en) * | 2020-03-31 | 2023-01-13 | 中国科学院软件研究所 | Method and system for predicting insect-borne diseases based on sliding window time sequence model |
| CN111785370A (en) * | 2020-07-01 | 2020-10-16 | 医渡云（北京）技术有限公司 | Medical record data processing method and device, computer storage medium and electronic equipment |
| CN112164448A (en) * | 2020-09-25 | 2021-01-01 | 上海市胸科医院 | Training method, prediction system, method and medium of immunotherapy efficacy prediction model |
| CN112164448B (en) * | 2020-09-25 | 2021-06-22 | 上海市胸科医院 | Training method, prediction system, method and medium of immunotherapy efficacy prediction model |
| CN113611430A (en) * | 2021-07-28 | 2021-11-05 | 广东省科学院智能制造研究所 | Epidemic situation prediction method and device based on Bayesian neural network |
| CN115470936A (en) * | 2022-09-23 | 2022-12-13 | 广州爱浦路网络技术有限公司 | NWDAF-based machine learning model updating method and device |
| CN117219284A (en) * | 2023-09-11 | 2023-12-12 | 广东德澳智慧医疗科技有限公司 | System with time sequence intelligent medical big data management |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109493979A (en) | A kind of disease forecasting method and apparatus based on intelligent decision | |
| Sopic et al. | Real-time event-driven classification technique for early detection and prevention of myocardial infarction on wearable systems | |
| JP7208224B2 (en) | Selection of Audio Features to Build a Model to Detect Medical Conditions | |
| JP6909078B2 (en) | Disease onset prediction device, disease onset prediction method and program | |
| KR101970947B1 (en) | Apparatus and method for predicting health information using big data | |
| Lim et al. | Analysis of single-electrode EEG rhythms using MATLAB to elicit correlation with cognitive stress | |
| US10912508B2 (en) | Method and system for assessing mental state | |
| CN109864736A (en) | Processing method, device, terminal device and the medium of electrocardiosignal | |
| KR20170061222A (en) | The method for prediction health data value through generation of health data pattern and the apparatus thereof | |
| JP7299427B2 (en) | A method for predicting mental health and providing a mental health solution by learning psychological index data and physical index data based on machine learning, and a mental health evaluation device using the same | |
| US20180211727A1 (en) | Automated Evidence Based Identification of Medical Conditions and Evaluation of Health and Financial Benefits Of Health Management Intervention Programs | |
| CN111933291A (en) | Medical information recommendation device, method, system, equipment and readable storage medium | |
| CN112043252B (en) | Emotion recognition system and method based on respiratory component in pulse signal | |
| CN108461110A (en) | Medical information processing method, device and equipment | |
| CN112641451B (en) | Multi-scale residual error network sleep staging method and system based on single-channel electroencephalogram signal | |
| CN110464303A (en) | Sleep quality appraisal procedure and device | |
| Mortensen et al. | Multi-class stress detection through heart rate variability: A deep neural network based study | |
| CN109994211B (en) | Modeling method for chronic kidney disease worsening risk based on EHR data | |
| CN109493975B (en) | Chronic disease recurrence prediction method, device and computer equipment based on xgboost model | |
| Pale et al. | Importance of methodological choices in data manipulation for validating epileptic seizure detection models | |
| CN114550924A (en) | Artificial intelligence-based serious disease state prediction method, device, equipment and medium | |
| Cheng et al. | Mining discriminative patterns to predict health status for cardiopulmonary patients | |
| CN116344042B (en) | Cognitive reserve intervention lifting method and system based on multi-modal analysis | |
| Al-Mter | Automatic Prediction of Human Age based on Heart Rate Variability Analysis using Feature-Based Methods | |
| KR102407987B1 (en) | Methdo and apparatus for building bio data hub |
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 |