CN110211690A - Disease risks prediction technique, device, computer equipment and computer storage medium - Google Patents
Disease risks prediction technique, device, computer equipment and computer storage medium Download PDFInfo
- Publication number
- CN110211690A CN110211690A CN201910319422.8A CN201910319422A CN110211690A CN 110211690 A CN110211690 A CN 110211690A CN 201910319422 A CN201910319422 A CN 201910319422A CN 110211690 A CN110211690 A CN 110211690A
- Authority
- CN
- China
- Prior art keywords
- disease
- data
- type
- predicted
- association
- 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
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 title claims abstract description 698
- 201000010099 disease Diseases 0.000 title claims abstract description 697
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 238000013136 deep learning model Methods 0.000 claims abstract description 19
- 230000000306 recurrent effect Effects 0.000 claims description 32
- 238000013528 artificial neural network Methods 0.000 claims description 30
- 238000012545 processing Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 238000013506 data mapping Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 9
- 230000005856 abnormality Effects 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000007405 data analysis Methods 0.000 abstract description 4
- 238000012549 training Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000009826 distribution Methods 0.000 description 5
- 210000002569 neuron Anatomy 0.000 description 5
- 238000004891 communication Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000002265 prevention Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 206010018364 Glomerulonephritis Diseases 0.000 description 1
- 206010019280 Heart failures Diseases 0.000 description 1
- 206010061216 Infarction Diseases 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000009975 flexible effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000007574 infarction Effects 0.000 description 1
- 206010022000 influenza Diseases 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000001717 pathogenic effect Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Strategic Management (AREA)
- Medical Informatics (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Business, Economics & Management (AREA)
- Pathology (AREA)
- Tourism & Hospitality (AREA)
- Epidemiology (AREA)
- Operations Research (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
Abstract
This application discloses a kind of disease risks prediction technique, device and computer storage mediums, are related to data analysis technique field, can more fully be analyzed disease data, and the precision of disease risks prediction is improved.The described method includes: obtaining the disease sample data with relevance, the disease sample data include the disease data and disease association data of each disease type, and each disease type label is recorded in the disease sample data；The disease sample data with relevance are input in deep learning model and are trained, disease risks prediction model is constructed, the disease risks prediction model is used for risk profile result of the predictive disease data on each disease type；When receiving disease data to be predicted, it is based on the disease risks prediction model, exports risk profile result of the disease data to be predicted on each disease type.
Description
Technical field
The present invention relates to data analysis technique fields, set particularly with regard to disease risks prediction technique, device, computer
Standby and computer storage medium.
Background technique
Disease risks prediction is the vital task of each Disease Control and Prevention Center, provinces and cities, and key problem will be prediction individual in future one
Section the time in suffer from certain disease (or certain event occurs) risk probability, thus for prediction risk probability to disease into
Row prevention, and control the development of disease.
Disease control can be defined according to some crowd, such as full crowd, heart infarction inpatients etc., predict mesh for some
Mark, such as heart failure, death etc., set specific time window, including the time point made prediction, and the time that will be predicted
Window predicts the probability of happening of target.Traditional disease risks prediction is based primarily upon disease control expert to the existing data of Disease Control and Prevention Center
Carry out simple manual analysis, disease predicted, this method has biggish limitation, to the dependence of disease control expert compared with
Height, in addition, existing data be it is very limited, be difficult to carry out comprehensive manual analysis, cause disease risks prediction to exist inclined
Difference.
Summary of the invention
In view of this, the present invention provides the storages of a kind of disease risks prediction technique, device, computer equipment and computer
Medium, main purpose are to solve to carry out manual analysis with limitation to Disease Center data with existing currently based on disease control expert
Property, leading to disease risks prediction, there are offset issues.
According to the present invention on one side, a kind of disease risks prediction technique is provided, this method comprises:
The disease sample data with relevance are obtained, the disease sample data include the disease number of each disease type
Accordingly and disease association data, and each disease type label is recorded in the disease sample data；
The disease sample data with relevance are input in deep learning model and are trained, disease wind is constructed
Dangerous prediction model, the disease risks prediction model are used for risk profile knot of the predictive disease data on each disease type
Fruit；
When receiving disease data to be predicted, it is based on the disease risks prediction model, exports the disease to be predicted
Risk profile result of the data on each disease type.
Further, the deep learning model is Recognition with Recurrent Neural Network, and the Recognition with Recurrent Neural Network includes multilayered structure,
The described disease sample data with relevance are input in deep learning model is trained, and building disease risks are pre-
Surveying model includes:
Coding layer structure extraction disease sample data by the Recognition with Recurrent Neural Network are influencing each disease type
Characteristic parameter on association factor；
The disease is continued to optimize according to each disease type label by the decoding layer of the Recognition with Recurrent Neural Network
Characteristic parameter of the sample data on the association factor for influencing each disease type, obtains disease sample data in each disease class
Eigenmatrix in type；
Summarize the disease sample data on each disease type by the full articulamentum of the Recognition with Recurrent Neural Network
Eigenmatrix obtains prediction weight of the disease sample data on each disease type；
It is pre- on each disease type according to the disease sample data by the classification layer of the Recognition with Recurrent Neural Network
Weight is surveyed, generates risk profile of the disease sample data on each disease type as a result, building disease risks prediction model.
Further, the disease sample data of the acquisition with relevance include:
The medical data of history of sample of users, the medical number of the history are called by the preset interface that multiple data sources provide
According to the middle multiple key word informations for recording and having medical disease；
Each disease type is extracted from the medical data of the history according to multiple key word informations of the medical disease
Disease data and disease association data, obtain the disease sample data with relevance.
Further, multiple key word informations according to the medical disease are extracted from the medical data of the history
The disease data and disease association data of each disease type, obtain having the disease sample data of relevance include:
According to multiple key word informations of the medical disease, the disease type in the medical data and influence are determined
The association factor of disease type；
It is medical from the history based on the association factor of disease type and influence disease type in the medical data
The disease data and disease association data of each disease type are extracted in data；
The disease data of each disease type and the disease association data are closed according to data mapping ruler
Connection processing, obtains the disease sample data with relevance.
Further, described to close the disease data of each disease type and the disease according to data mapping ruler
Connection data are associated processing, obtain having the disease sample data of relevance include:
It is extracted from the disease data of each disease type and the disease association data and is suitable for each disease
Relating attribute feature；
Based on the relating attribute suitable for various diseases, according to data mapping ruler by each disease type
Disease data and the disease association data are associated processing, obtain the disease sample data with relevance.
Further, it is based on the disease risks prediction model described, exports the disease data to be predicted each
After risk profile result on disease type, the method also includes:
The disease data to be predicted is carried out abnormality detection according to the history disease data uploaded in preset time period.
Further, it is described according to the history disease data uploaded in preset time period to the disease data to be predicted into
Row abnormality detection includes:
The history disease data uploaded in the preset time period is divided into the history disease data of multiple time cycles；
The deviation between the history disease data of the multiple time cycle and the disease data to be predicted is calculated separately,
Obtain the data deviation in each time cycle；
The time cycle quantity that the data deviation is greater than the first default value is obtained, if the time cycle quantity is super
Second default value out then determines that the disease data to be predicted is abnormal data, deletes the disease data to be predicted each
Risk profile result on a disease type；
Otherwise, it is determined that the disease data to be predicted is normal data, the disease data to be predicted is saved in each disease
Risk profile result in sick type.
According to the present invention on the other hand, a kind of disease risks prediction meanss are provided, described device includes:
Acquiring unit, for obtaining the disease sample data with relevance, the disease sample data include each disease
The disease data and disease association data of sick type, and each disease type label is recorded in the disease sample data；
Construction unit is instructed for the disease sample data with relevance to be input in deep learning model
Practice, constructs disease risks prediction model, the disease risks prediction model is for predictive disease data in each disease type
Risk profile result；
Predicting unit exports institute for being based on the disease risks prediction model when receiving disease data to be predicted
State risk profile result of the disease data to be predicted on each disease type.
Further, the deep learning model is Recognition with Recurrent Neural Network, and the Recognition with Recurrent Neural Network includes multilayered structure,
The construction unit includes:
Extraction module is influencing respectively for the coding layer structure extraction disease sample data by the Recognition with Recurrent Neural Network
Characteristic parameter on the association factor of a disease type；
Optimization module, for continuous according to each disease type label by the decoding layer of the Recognition with Recurrent Neural Network
Optimize characteristic parameter of the disease sample data on the association factor for influencing each disease type, obtains disease sample data
Eigenmatrix on each disease type；
Summarizing module summarizes the disease sample data each for the full articulamentum by the Recognition with Recurrent Neural Network
Eigenmatrix on disease type obtains prediction weight of the disease sample data on each disease type；
Generation module, for the classification layer by the Recognition with Recurrent Neural Network according to the disease sample data in each disease
Prediction weight in sick type generates risk profile of the disease sample data on each disease type as a result, building disease wind
Dangerous prediction model.
Further, the acquiring unit includes:
Calling module, the preset interface for being provided by multiple data sources call the medical data of the history of sample of users,
Record has multiple key word informations of medical disease in the medical data of the history；
Module is obtained, for extracting from the medical data of the history according to multiple key word informations of the medical disease
The disease data and disease association data of each disease type, obtain the disease sample data with relevance.
Further, the acquisition module includes:
It determines submodule, for multiple key word informations according to the medical disease, determines in the medical data
Disease type and the association factor for influencing disease type；
Extracting sub-module, for based in the medical data disease type and influence disease type association because
Son extracts the disease data and disease association data of each disease type from the medical data of the history；
Be associated with submodule, for according to data mapping ruler by the disease data of each disease type and the disease
Associated data is associated processing, obtains the disease sample data with relevance.
Further, the association submodule, specifically for from each disease type disease data and the disease
The relating attribute feature for being suitable for each disease is extracted in sick associated data；
The association submodule is specifically also used to reflect based on the relating attribute suitable for various diseases according to data
It penetrates rule and the disease data of each disease type and the disease association data is associated processing, obtain that there is association
The disease sample data of property.
Further, described device further include:
Detection unit exports the disease data to be predicted and exists for being based on the disease risks prediction model described
After risk profile result on each disease type, according to the history disease data uploaded in preset time period to described to pre-
Disease data is surveyed to carry out abnormality detection.
Further, the detection unit includes:
Division module, for the history disease data uploaded in the preset time period to be divided into multiple time cycles
History disease data；
Computing module, for calculating separately the history disease data and the disease number to be predicted of the multiple time cycle
Deviation between obtains the data deviation in each time cycle；
Determination module, the time cycle quantity for being greater than the first default value for obtaining the data deviation, if described
Time cycle quantity exceeds the second default value, then determines that the disease data to be predicted is abnormal data, deletes described to pre-
Survey risk profile result of the disease data on each disease type；
The determination module, if being also used to the time cycle quantity without departing from the second default value, determine it is described to
Predictive disease data are normal data, save risk profile result of the disease data to be predicted on each disease type.
Another aspect according to the present invention provides a kind of computer equipment, including memory and processor, the storage
The step of device is stored with computer program, and the processor realizes disease risks prediction technique when executing the computer program.
Another aspect according to the present invention provides a kind of computer storage medium, is stored thereon with computer program, institute
State the step of disease risks prediction technique is realized when computer program is executed by processor.
By above-mentioned technical proposal, the present invention provides a kind of disease risks prediction technique and device, is closed by obtaining to have
The disease sample data of connection property, due to including the disease data and disease association of each disease type in the disease sample data
Data fully take into account the relation factor of disease data, by will have the disease sample data of relevance to be input to depth
Practise model in be trained, construct disease risks prediction model, be based on disease risks prediction model, treat predictive disease data into
Row prediction, obtains risk profile result of the disease data to be predicted on each disease type.With in the prior art be based on disease control
Expert carries out manual analysis to Disease Center data with existing and compares to carry out the mode of disease risks prediction, synthetic disease of the present invention
The relation factor of data and disease data more fully analyzes disease data, improves the accurate of disease data prediction
Property, carry out risk situation of the predictive disease data on each disease type by using disease risks prediction model, without relying on
Disease control expert can risk situation of the tentative prediction disease data on each disease type, improve disease risks prediction it is flexible
Property.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows a kind of flow diagram of disease risks prediction technique provided in an embodiment of the present invention；
Fig. 2 shows the flow diagrams of another disease risks prediction technique provided in an embodiment of the present invention；
Fig. 3 shows a kind of structural schematic diagram of disease risks prediction meanss provided in an embodiment of the present invention；
Fig. 4 shows the structural schematic diagram of another disease risks prediction meanss provided in an embodiment of the present invention.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
The embodiment of the invention provides a kind of disease risks prediction techniques, can more fully be divided disease data
Analysis improves the precision of disease risks prediction, as shown in Figure 1, this method comprises:
101, the disease sample data with relevance are obtained.
Wherein, it is generated when the disease sample data with relevance are gone to a doctor for the sample of users that existing medical treatment is recorded each
The disease data of disease type, the content of disease sample data may include two parts here, and a part is each disease type
Disease data, can specifically include but be not limited to disease type, area, time, sample of users information, genius morbi information
Etc. data, the disease association data that another part is disease data on the association factor for influencing each disease type specifically may be used
To include but is not limited to the affiliated zone data of disease, weather data, Baidu's exponent data etc..
It should be noted that the reliability in order to guarantee disease sample data, the acquisition channels of disease sample data can be with
Including but not limited to disease website, institutional databases etc., and the disease sample data can be implementation disease data,
It can be the disease data of history full dose, here without limiting.
For the embodiment of the present invention, each channel is during uploading disease sample data, for the ease of to disease sample
Notebook data distinguishes, and for each disease type, can all carry corresponding disease type label, and each disease type
Disease data all can with have an impact the disease type association factor on disease association data, for example, upload disease
Data affiliated area is Chongqing, then disease association data will include the weather data in Chongqing, the public sentiment data etc. of Chongqing City.Institute
The disease association data on the association factor of disease type and the disease during uploading disease sample data, can be will affect
The disease data of sick type is uploaded together.
102, the disease sample data with relevance are input in deep learning model and are trained, construct disease
Sick risk forecast model.
Wherein, disease risks prediction model for risk profile of the predictive disease data on each disease type as a result,
Due to deep learning model can probability scenarios of the training sample data on each disease type, risk profile result here
User probability value on each disease type, or the risk on each disease type etc. can be corresponded to for disease data
Grade, the embodiment of the present invention is without limiting.For example, probability value of the setting user on each disease type is distributed as 0-1, more become
Being bordering on 1, to illustrate that disease data corresponds to user's incidence probability on this kind of disease type higher, and user is arranged in each disease type
On risk class from high to low be respectively grade A, grade B, grade C, more multi-grade can also be set certainly, and risk class is got over
It is higher that height illustrates that disease data corresponds to user's incidence probability on this kind of disease type.
For the embodiment of the present invention, deep learning model here can be Recognition with Recurrent Neural Network model, can specifically lead to
Disease sample data of the repetition training with relevance are crossed to construct the network structure of disease risks prediction model, the network structure
The disease sample data of input can be trained, and provide correct Input output Relationship, that is, be directed to disease sample number
Can all there be the risk profile accordingly exported as a result, namely disease sample data are in each disease according on each disease type
Probability scenarios in type.
The structure of specific Recognition with Recurrent Neural Network model can pass through coding layer, decoding layer, full articulamentum and classification layer knot
Structure realizes that coding layer here, decoding layer are equivalent to the hidden layer of Recognition with Recurrent Neural Network, generally include multilayer, for extracting disease
Characteristic parameter of the sick sample data on the association factor for influencing each disease type, and optimize and form disease sample data each
Eigenmatrix on a disease type.During hidden layer repetition training data, each hidden layer have oneself weight, partially
Value and activation primitive are set, for the ease of data sharing, usually sets identical value for all layers of weight and bias, each
Hidden layer during training data all by disease association aggregation of data into disease data, to be inferred to disease type.
Of course for reducing, output feature vector is excessive, and two full articulamentums can be set, and passes through several hidden layers in training data
After training, the characteristic of training output is integrated.
103, when receiving disease data to be predicted, it is based on the disease risks prediction model, is exported described to be predicted
Risk profile result of the disease data on each disease type.
For the embodiment of the present invention, disease data to be predicted is the disease data of the unidentified illness type acquired at no distant date,
And equally also disease to be predicted can be obtained by disease risks prediction model comprising disease association data in the disease data
Risk profile result of the data on each disease type.
In order to which the disease data to disease type corresponding to risk profile result is further analyzed, strong wind can also be directed to
Dangerous prediction result treats predictive disease data in conjunction with disease association data and carries out correlation analysis, due in disease association data
The pathogenetic each association factor of disease that has an impact is recorded, the association factor by analyzing each disease type treats predictive disease number
According to effect tendency, thus the degree of correlation before obtaining disease data to be predicted and disease association data, corresponding degree of correlation
Higher disease association data should be taken seriously, that is to say, that disease data of the bright disease association data to this kind of disease type
Influence power it is larger, the disease association data lighter for degree of correlation illustrate disease association data to the shadow of this kind of disease type
Sound is smaller, can ignore to a certain extent.
The embodiment of the present invention provides a kind of disease risks prediction technique, has the disease sample number of relevance by obtaining
According to due to, including the disease data and disease association data of each disease type, being fully taken into account in the disease sample data
The relation factor of disease data is instructed by will there are the disease sample data of relevance to be input in deep learning model
Practice, construct disease risks prediction model, be based on disease risks prediction model, treat predictive disease data and predicted, obtain to
Risk profile result of the predictive disease data on each disease type.With in the prior art be based on disease control expert to Disease Center
Data with existing carries out manual analysis and compares to carry out the mode of disease risks prediction, synthetic disease data of the present invention and disease number
According to relation factor disease data is more fully analyzed, the accuracy of disease data prediction is improved, by using disease
Sick risk forecast model carrys out risk situation of the predictive disease data on each disease type, can be first without relying on disease control expert
Risk situation of the predictive disease data on each disease type is walked, the flexibility of disease risks prediction is improved.
The embodiment of the invention provides another disease risks prediction techniques, can more fully be divided disease data
Analysis improves the precision of disease risks prediction, as shown in Figure 2, which comprises
201, the medical data of history of sample of users are called by the preset interface that multiple data sources provide, the history is just
Examine multiple key word informations that record in data has medical disease.
Wherein, multiple data sources may include but be not limited to disease website, database, medical institutions etc., different data
Source can all provide the preset interface to match with the data source during upload history goes to a doctor data, can by preset interface
To call the medical data of history of sample of users.
For the embodiment of the present invention, record has multiple key word informations of medical disease in the medical data of history here,
Multiple key word informations can be specifically embodied from the following aspects, about the key word information of genius morbi, for example, disease
Type, disease symptoms, disease duration etc., about the key word information of medical sample of users personal information, such as sample of users
Name, age, occupation etc., about the associated key word information of medical data, for example, medical region, medical area weather, just
Examine the public sentiment factor etc. in region.
202, the association factor based on disease type and influence disease type in the medical data, from the history
The disease data and disease association data of each disease type are extracted in medical data.
Since history record in data of going to a doctor has the medical data of great amount of samples user, how to it is a large amount of go to a doctor data into
Row processing needs based on the disease type in medical data and influences the association factor of disease type, to extract each
The disease data and disease association data of disease type.
It should be noted that the association factor of above-mentioned each disease type, can go through for what is provided by the doctor that goes to a doctor
It is obtained in the medical data of history, can also be obtained for disease type by search accordingly, association factor is obtained here
Take mode without limiting.
For the embodiment of the present invention, a large amount of medical data in the process of processing, will record in medical data
Genius morbi when sample of users is medical, based on it is medical when various diseases genius morbi and doctor provide or search for
The association factor of each disease type of influence arrived goes to a doctor from history and extracts each disease type data in data and disease is closed
Join data.
203, according to data mapping ruler by the disease data of each disease type and the disease association data into
Row association process obtains the disease sample data with relevance.
For the embodiment of the present invention, since the disease data and disease association data of each disease type are all around each
Disease data acquired in a disease type, then single from the disease data of each disease type or disease association data
It is difficult to get the comprehensive data arranged around disease type, in order to more fully realize the disease data to each disease type
It is predicted, can be by the way that the disease data of each disease type and disease association data be associated processing, and then guarantee
The reliability of disease forecasting
For the embodiment of the present invention, specifically it is associated by the disease data of each disease type and disease association data
During processing, it is different that associated attributive character is likely to be suited for for various disease type, for example, for influenza number
According to, it is more heavily weighted toward Weather property and time attribute, it specifically can be from the disease data and disease association number of each disease type
According to the middle relating attribute feature extracted suitable for each disease；Based on the relating attribute for being suitable for various diseases, reflected according to data
It penetrates rule and the disease data of each disease type and the disease association data is associated processing, obtain that there is relevance
Disease sample data.
204, the disease sample data with relevance are input in deep learning model and are trained, construct disease
Sick risk forecast model.
For the embodiment of the present invention, specific deep learning model can be multilayered structure Recognition with Recurrent Neural Network, every layer
Structure has the function of that different input/output arguments is different to realize, by Recognition with Recurrent Neural Network come to carrying disease label
Disease sample data with relevance carry out repetition training, obtain wind of the disease data of user on each disease type
Dangerous prediction result.
Specifically, Recognition with Recurrent Neural Network can be 4 layers of structure composition, and first layer structure is coding layer structure, the coding layer
Structure can be made of multilevel encoder, it may for example comprise 64 neurons, 2 bidirectional circulating layers of 32 neurons and 16 follow
The unidirectional ply of ring neuron, encoder here are arranged to can handle the arbitrary sequence that maximum length is setting value, and
All circulation neurons can use GRU structure in encoder, and GRU structure is simple, be determined by updating door and resetting door
The degree of dependence of state before, thus the remote Dependence Problem of final result, by coding layer structure extraction disease sample data in shadow
Ring the characteristic parameter on the association factor of each disease type；Second layer structure is decoding layer, which may include one
Individual circulation layer there are 64 length to remember LSTM unit, in conjunction with attention mechanism, so that network is primarily upon input feature vector
Signal portion, improve classification performance, disease sample data are continued to optimize according to each disease type label by decoding layer and are existed
The characteristic parameter on the association factor of each disease type is influenced, feature of the disease sample data on each disease type is obtained
Matrix；Third layer structure is full articulamentum, which may include 256 neurons, by trained distributed nature table
Show and be mapped to sample labeling space, ensemble learning global feature summarizes disease sample data in each disease by full articulamentum
Eigenmatrix in type obtains prediction weight of the disease sample data on each disease type；Four-layer structure is classification
Softmax output category label can be used in layer, the classification layer, which can become (0,1) for input mapping
Numerical value, the numerical value can be understood as probability, take maximum probability as classification results, be existed by classification layer according to disease sample data
Prediction weight on each disease type generates risk profile of the disease sample data on each disease type as a result, building
Disease risks prediction model.
205, when receiving disease data to be predicted, it is based on the disease risks prediction model, is exported described to be predicted
Risk profile result of the disease data on each disease type.
Wherein, disease data to be predicted is user's disease data of unidentified illness type, likewise, in order to improve to user
The risk profile result accuracy of disease type also carried in addition to carrying genius morbi data in disease data to be predicted
There are disease association data, for example, weather data, public sentiment data etc..
The disease risks prediction model of the embodiment of the present invention is equivalent to the Recognition with Recurrent Neural Network using sequence to sequence for one
A undivided input data is mapped to another output label, it is contemplated that disease data to be predicted to disease association data according to
Lai Xing, Recognition with Recurrent Neural Network can influence each disease class according to disease data with the dependence between capture time and spectrum
Characteristic parameter on the association factor of type, Lai Jinhang disease data correspond to the prediction task of disease type, so that each layer exports meeting
Dependent on upper one layer of calculating, to improve the precision of the risk profile result of disease data.
206, abnormal inspection is carried out to the disease data to be predicted according to the history disease data uploaded in preset time period
It surveys.
It is understood that since the disease data of each disease type is all in certain distribution, if exceeded
The distribution, then illustrating disease data disease data to be predicted, there may be exceptions, in order to avoid the risk to disease data
Prediction result is verified, can be according in preset time after the risk profile result for exporting disease data to be predicted
The history disease data of biography is treated prediction data and is carried out abnormality detection, which can characterize the distribution of disease data
Range illustrates that disease data to be predicted may if disease data to be predicted exceeds the distribution of history disease data
For abnormal data, there may be errors for the risk profile result obtained based on disease risks model, if disease data to be predicted
Distribution in history disease data then illustrates that disease data to be predicted there is no exception, is obtained based on disease risks model
To risk profile result can apply.
It specifically, can be by calculating separately multiple time cycles in order to guarantee the accuracy of risk profile result verification
History disease data and disease data to be predicted between deviation, obtain the data deviation in each time cycle, here
The bigger explanation data to be predicted of data deviation may be higher for the probability of abnormal data, due to being provided with multiple time cycles, leads to
It crosses and obtains the time cycle quantity that data deviation is greater than the first default value, which is deviation threshold, for super
The disease data in the time cycle of the deviation threshold meets the determinating reference of abnormal data out, if time cycle quantity exceeds
Second default value, then explanation meets the time cycle foot of abnormal data all, it is possible to determine that disease data to be predicted is abnormal number
According to deleting risk profile of the disease data to be predicted on each disease type as a result, if time cycle quantity is less than
Two default values then illustrate that the time cycle for meeting normal data is more, determine that disease data to be predicted for normal data, saves
Risk profile result of the disease data to be predicted on each disease type.
Further, the specific implementation as Fig. 1 the method, the embodiment of the invention provides a kind of predictions of disease risks
Device, as shown in figure 3, described device includes: acquiring unit 31, construction unit 32, predicting unit 33.
Acquiring unit 31 can be used for obtaining the disease sample data with relevance, and the disease sample data include
The disease data and disease association data of each disease type, and each disease type is recorded in the disease sample data
Label；
Construction unit 32 can be used for for the disease sample data with relevance being input in deep learning model
It is trained, constructs disease risks prediction model, the disease risks prediction model is for predictive disease data in each disease
Risk profile result in type；
Predicting unit 33 can be used for when receiving disease data to be predicted, be based on the disease risks prediction model,
Export risk profile result of the disease data to be predicted on each disease type.
A kind of disease risks prediction meanss provided by the invention have the disease sample data of relevance by obtaining, by
Including the disease data and disease association data of each disease type in the disease sample data, disease number is fully taken into account
According to relation factor, be trained by will there are the disease sample data of relevance to be input in deep learning model, construct
Disease risks prediction model is based on disease risks prediction model, treats predictive disease data and predicted, obtain disease to be predicted
Risk profile result of the data on each disease type.With in the prior art be based on disease control expert to Disease Center data with existing
It carries out manual analysis to compare to carry out the mode of disease risks prediction, the association of synthetic disease data of the present invention and disease data
Because usually more fully being analyzed disease data, the accuracy of raising disease data prediction is pre- by using disease risks
It surveys model and carrys out risk situation of the predictive disease data on each disease type, it can tentative prediction disease without relying on disease control expert
Risk situation of the sick data on each disease type improves the flexibility of disease risks prediction.
As the further explanation of disease risks prediction meanss shown in Fig. 3, Fig. 4 is another kind according to embodiments of the present invention
The structural schematic diagram of disease risks prediction meanss, as shown in figure 4, described device further include:
Detection unit 34 can be used for exporting the disease to be predicted based on the disease risks prediction model described
Data are after the risk profile result on each disease type, according to the history disease data uploaded in preset time period to institute
Disease data to be predicted is stated to carry out abnormality detection.
Further, the detection unit 34 includes:
Division module 341, when the history disease data that can be used for upload in the preset time period is divided into multiple
Between the period history disease data；
Computing module 342 can be used for calculating separately the history disease data of the multiple time cycle with described to pre-
The deviation between disease data is surveyed, the data deviation in each time cycle is obtained；
Determination module 343 can be used for obtaining the time cycle quantity that the data deviation is greater than the first default value, such as
Time cycle quantity described in fruit exceeds the second default value, then determines that the disease data to be predicted is abnormal data, deletes institute
State risk profile result of the disease data to be predicted on each disease type；
The determination module 343 determines if can be also used for the time cycle quantity without departing from the second default value
The disease data to be predicted is normal data, saves risk profile of the disease data to be predicted on each disease type
As a result.
Further, the deep learning model is Recognition with Recurrent Neural Network, and the Recognition with Recurrent Neural Network includes multilayered structure,
The construction unit 32 includes:
Extraction module 321 can be used for the coding layer structure extraction disease sample data by the Recognition with Recurrent Neural Network
Characteristic parameter on the association factor for influencing each disease type；
Optimization module 322 can be used for the decoding layer by the Recognition with Recurrent Neural Network according to each disease type
Label continues to optimize characteristic parameter of the disease sample data on the association factor for influencing each disease type, obtains disease
Eigenmatrix of the sample data on each disease type；
Summarizing module 323 can be used for summarizing the disease sample number by the full articulamentum of the Recognition with Recurrent Neural Network
According to the eigenmatrix on each disease type, prediction weight of the disease sample data on each disease type is obtained；
Generation module 324 can be used for the classification layer by the Recognition with Recurrent Neural Network according to the disease sample data
Prediction weight on each disease type generates risk profile of the disease sample data on each disease type as a result, structure
Build disease risks prediction model.
Further, the acquiring unit 31 includes:
Calling module 311, the preset interface that can be used for providing by multiple data sources call the history of sample of users just
Data are examined, record there are multiple key word informations of medical disease in the medical data of the history；
Module 312 is obtained, can be used for being counted according to multiple key word informations of the medical disease from the history is medical
According to the middle disease data and disease association data for extracting each disease type, the disease sample data with relevance are obtained.
Further, the acquisition module 312 includes:
It determines submodule 3121, can be used for multiple key word informations according to the medical disease, determine described medical
The association factor of disease type and influence disease type in data；
Extracting sub-module 3122 can be used for based on the disease type in the medical data and influence disease type
Association factor extracts the disease data and disease association data of each disease type from the medical data of the history；
Be associated with submodule 3123, can be used for according to data mapping ruler by the disease data of each disease type with
The disease association data are associated processing, obtain the disease sample data with relevance.
Further, the association submodule 3123, specifically can be used for the disease data from each disease type
It is suitable for the relating attribute feature of each disease with extraction in the disease association data；
The association submodule 3123 specifically can be also used for pressing based on the relating attribute suitable for various diseases
The disease data of each disease type and the disease association data are associated processing according to data mapping ruler, obtained
Disease sample data with relevance.
It should be noted that other of each functional unit involved by a kind of disease risks prediction meanss provided in this embodiment
Corresponding description, can be with reference to the corresponding description in Fig. 1 and Fig. 2, and details are not described herein.
It is deposited thereon based on above-mentioned method as depicted in figs. 1 and 2 correspondingly, the present embodiment additionally provides a kind of storage medium
Computer program is contained, which realizes above-mentioned disease risks prediction technique as depicted in figs. 1 and 2 when being executed by processor.
Based on this understanding, the technical solution of the application can be embodied in the form of software products, which produces
Product can store in a non-volatile memory medium (can be CD-ROM, USB flash disk, mobile hard disk etc.), including some instructions
With so that computer equipment (can be personal computer, server or the network equipment an etc.) execution the application is each
Method described in implement scene.
Based on above-mentioned method as shown in Figure 1 and Figure 2 and Fig. 3, virtual bench embodiment shown in Fig. 4, in order to realize
Above-mentioned purpose, the embodiment of the present application also provides a kind of computer equipments, are specifically as follows personal computer, server, network
Equipment etc., the entity device include storage medium and processor；Storage medium, for storing computer program；Processor is used for
Computer program is executed to realize above-mentioned disease risks prediction technique as depicted in figs. 1 and 2.
Optionally, which can also include user interface, network interface, camera, radio frequency (Radio
Frequency, RF) circuit, sensor, voicefrequency circuit, WI-FI module etc..User interface may include display screen
(Display), input unit such as keyboard (Keyboard) etc., optional user interface can also connect including USB interface, card reader
Mouthful etc..Network interface optionally may include standard wireline interface and wireless interface (such as blue tooth interface, WI-FI interface).
It will be understood by those skilled in the art that the entity device structure of disease risks prediction meanss provided in this embodiment is simultaneously
The restriction to the entity device is not constituted, may include more or fewer components, perhaps combines certain components or different
Component layout.
It can also include operating system, network communication module in storage medium.Operating system is that the above-mentioned computer of management is set
The program of standby hardware and software resource, supports the operation of message handling program and other softwares and/or program.Network communication mould
Block leads to for realizing the communication between each component in storage medium inside, and between other hardware and softwares in the entity device
Letter.
Through the above description of the embodiments, those skilled in the art can be understood that the application can borrow
It helps software that the mode of necessary general hardware platform is added to realize, hardware realization can also be passed through.Pass through the skill of application the application
Art scheme, compared with currently available technology, the relation factor of synthetic disease data of the present invention and disease data comes to disease number
According to more fully being analyzed, the accuracy of disease data prediction is improved, predicts disease by using disease risks prediction model
Risk situation of the sick data on each disease type, without rely on disease control expert can tentative prediction disease data in each disease
Risk situation in sick type improves the flexibility of disease risks prediction.
It will be appreciated by those skilled in the art that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, module in attached drawing or
Process is not necessarily implemented necessary to the application.It will be appreciated by those skilled in the art that the mould in device in implement scene
Block can according to implement scene describe be distributed in the device of implement scene, can also carry out corresponding change be located at be different from
In one or more devices of this implement scene.The module of above-mentioned implement scene can be merged into a module, can also be into one
Step splits into multiple submodule.
Above-mentioned the application serial number is for illustration only, does not represent the superiority and inferiority of implement scene.Disclosed above is only the application
Several specific implementation scenes, still, the application is not limited to this, and the changes that any person skilled in the art can think of is all
The protection scope of the application should be fallen into.
Claims (10)
1. a kind of disease risks prediction technique, which is characterized in that the described method includes:
Obtain the disease sample data with relevance, the disease sample data include the disease data of each disease type with
And disease association data, and each disease type label is recorded in the disease sample data；
The disease sample data with relevance are input in deep learning model and are trained, building disease risks are pre-
Model is surveyed, the disease risks prediction model is used for risk profile result of the predictive disease data on each disease type；
When receiving disease data to be predicted, it is based on the disease risks prediction model, exports the disease data to be predicted
Risk profile result on each disease type.
2. the method according to claim 1, wherein the deep learning model be Recognition with Recurrent Neural Network, it is described
Recognition with Recurrent Neural Network includes multilayered structure, described that the disease sample data with relevance are input to deep learning model
In be trained, building disease risks prediction model include:
By the coding layer structure extraction disease sample data of the Recognition with Recurrent Neural Network in the association for influencing each disease type
Characteristic parameter in the factor；
The disease sample is continued to optimize according to each disease type label by the decoding layer of the Recognition with Recurrent Neural Network
Characteristic parameter of the data on the association factor for influencing each disease type, obtains disease sample data on each disease type
Eigenmatrix；
Summarize feature of the disease sample data on each disease type by the full articulamentum of the Recognition with Recurrent Neural Network
Matrix obtains prediction weight of the disease sample data on each disease type；
It is weighed by the classification layer of the Recognition with Recurrent Neural Network according to prediction of the disease sample data on each disease type
Weight generates risk profile of the disease sample data on each disease type as a result, building disease risks prediction model.
3. the method according to claim 1, wherein described obtain the disease sample data packet with relevance
It includes:
The medical data of history of sample of users are called by the preset interface that multiple data sources provide, the history is gone to a doctor in data
Record has multiple key word informations of medical disease；
The disease of each disease type is extracted from the medical data of the history according to multiple key word informations of the medical disease
Sick data and disease association data obtain the disease sample data with relevance.
4. according to the method described in claim 3, it is characterized in that, multiple key word informations according to the medical disease
The disease data and disease association data that each disease type is extracted from the medical data of the history, obtain with relevance
Disease sample data include:
According to multiple key word informations of the medical disease, determines the disease type in the medical data and influence disease
The association factor of type；
Based on the association factor of disease type and influence disease type in the medical data, from the medical data of the history
The middle disease data and disease association data for extracting each disease type；
The disease data of each disease type and the disease association data are associated place according to data mapping ruler
Reason, obtains the disease sample data with relevance.
5. according to the method described in claim 4, it is characterized in that, it is described according to data mapping ruler by each disease class
The disease data of type and the disease association data are associated processing, obtain having the disease sample data of relevance include:
It is suitable for being associated with for each disease with extraction in the disease association data from the disease data of each disease type
Attributive character；
Based on the relating attribute suitable for various diseases, according to data mapping ruler by the disease of each disease type
Data and the disease association data are associated processing, obtain the disease sample data with relevance.
6. method according to any one of claims 1-5, which is characterized in that predicted described based on the disease risks
Model exports the disease data to be predicted after the risk profile result on each disease type, the method also includes:
The disease data to be predicted is carried out abnormality detection according to the history disease data uploaded in preset time period.
7. according to method described in right 6, which is characterized in that described according to the history disease data pair uploaded in preset time period
The disease data to be predicted, which carries out abnormality detection, includes:
The history disease data uploaded in the preset time period is divided into the history disease data of multiple time cycles；
The deviation between the history disease data of the multiple time cycle and the disease data to be predicted is calculated separately, is obtained
Data deviation in each time cycle；
The time cycle quantity that the data deviation is greater than the first default value is obtained, if the time cycle quantity is beyond the
Two default values then determine that the disease data to be predicted is abnormal data, delete the disease data to be predicted in each disease
Risk profile result in sick type；
Otherwise, it is determined that the disease data to be predicted is normal data, the disease data to be predicted is saved in each disease class
Risk profile result in type.
8. a kind of disease risks prediction meanss, which is characterized in that described device includes:
Acquiring unit, for obtaining the disease sample data with relevance, the disease sample data include each disease class
The disease data and disease association data of type, and each disease type label is recorded in the disease sample data；
Construction unit is trained for the disease sample data with relevance to be input in deep learning model,
Disease risks prediction model is constructed, the disease risks prediction model is used for wind of the predictive disease data on each disease type
Dangerous prediction result；
Predicting unit is based on the disease risks prediction model for when receiving disease data to be predicted, output it is described to
Risk profile result of the predictive disease data on each disease type.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer storage medium, is stored thereon with computer program, which is characterized in that the computer program is located
The step of reason device realizes method described in any one of claims 1 to 7 when executing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319422.8A CN110211690A (en) | 2019-04-19 | 2019-04-19 | Disease risks prediction technique, device, computer equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319422.8A CN110211690A (en) | 2019-04-19 | 2019-04-19 | Disease risks prediction technique, device, computer equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110211690A true CN110211690A (en) | 2019-09-06 |
Family
ID=67786074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910319422.8A Pending CN110211690A (en) | 2019-04-19 | 2019-04-19 | Disease risks prediction technique, device, computer equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110211690A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110993103A (en) * | 2019-11-28 | 2020-04-10 | 阳光人寿保险股份有限公司 | Method for establishing disease risk prediction model and method for recommending disease insurance product |
| CN111063436A (en) * | 2019-11-25 | 2020-04-24 | 泰康保险集团股份有限公司 | Data processing method and device, storage medium and electronic terminal |
| 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 |
| CN111274305A (en) * | 2020-01-15 | 2020-06-12 | 平安医疗健康管理股份有限公司 | Three-dimensional picture generation method and device, computer equipment and storage medium |
| CN111681759A (en) * | 2020-05-29 | 2020-09-18 | 芯海科技（深圳）股份有限公司 | Chronic disease risk analysis device, equipment and storage medium |
| 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 |
| CN111899883A (en) * | 2020-09-29 | 2020-11-06 | 平安科技（深圳）有限公司 | Disease prediction device, method, apparatus and storage medium for small sample or zero sample |
| CN111933303A (en) * | 2020-09-30 | 2020-11-13 | 平安科技（深圳）有限公司 | Event prediction method and device, electronic equipment and storage medium |
| CN111986763A (en) * | 2020-09-03 | 2020-11-24 | 平安国际智慧城市科技股份有限公司 | Disease data analysis method and device, electronic device and storage medium |
| CN112053779A (en) * | 2020-06-15 | 2020-12-08 | 深圳博脑医疗科技有限公司 | Construction method and construction device of disease detection model and terminal equipment |
| CN112117015A (en) * | 2020-10-28 | 2020-12-22 | 平安科技（深圳）有限公司 | Sepsis early warning device, method, apparatus and storage medium |
| CN112132625A (en) * | 2020-09-27 | 2020-12-25 | 平安医疗健康管理股份有限公司 | Health risk-based medical expense prediction system |
| CN112216400A (en) * | 2020-10-10 | 2021-01-12 | 黑龙江省疾病预防控制中心 | Method and system for predicting food-borne disease pathogenic factors based on big data |
| CN112435747A (en) * | 2020-11-26 | 2021-03-02 | 新智数字科技有限公司 | Disease risk prediction method, device, equipment and computer readable medium |
| CN112465011A (en) * | 2020-11-25 | 2021-03-09 | 平安医疗健康管理股份有限公司 | Project risk prediction method and system based on project research and development process |
| CN112750532A (en) * | 2019-10-30 | 2021-05-04 | 宏碁股份有限公司 | Feature identification method and electronic device |
| WO2021151291A1 (en) * | 2020-05-26 | 2021-08-05 | 平安科技（深圳）有限公司 | Disease risk analysis method, apparatus, electronic device, and computer storage medium |
| CN113657550A (en) * | 2021-08-31 | 2021-11-16 | 平安医疗健康管理股份有限公司 | Patient marking method, device, equipment and storage medium based on hierarchical calculation |
| CN113707323A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Disease prediction method, device, equipment and medium based on machine learning |
| CN113782216A (en) * | 2021-09-15 | 2021-12-10 | 平安科技（深圳）有限公司 | Method and device for determining incapability weight, electronic equipment and storage medium |
| CN114898888A (en) * | 2022-07-15 | 2022-08-12 | 武汉大学 | Medical data processing method and device, computer equipment and readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108288502A (en) * | 2018-04-11 | 2018-07-17 | 平安科技（深圳）有限公司 | Disease forecasting method and device, computer installation and readable storage medium storing program for executing |
| CN109346183A (en) * | 2018-09-18 | 2019-02-15 | 山东大学 | Disease diagnosing and predicting system based on Recognition with Recurrent Neural Network model RNN |
| CN109493972A (en) * | 2018-10-30 | 2019-03-19 | 平安医疗健康管理股份有限公司 | Data processing method, device, server and storage medium based on prediction model |
| US20190267113A1 (en) * | 2016-10-31 | 2019-08-29 | Preferred Networks, Inc. | Disease affection determination device, disease affection determination method, and disease affection determination program |
-
2019
- 2019-04-19 CN CN201910319422.8A patent/CN110211690A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190267113A1 (en) * | 2016-10-31 | 2019-08-29 | Preferred Networks, Inc. | Disease affection determination device, disease affection determination method, and disease affection determination program |
| CN108288502A (en) * | 2018-04-11 | 2018-07-17 | 平安科技（深圳）有限公司 | Disease forecasting method and device, computer installation and readable storage medium storing program for executing |
| CN109346183A (en) * | 2018-09-18 | 2019-02-15 | 山东大学 | Disease diagnosing and predicting system based on Recognition with Recurrent Neural Network model RNN |
| CN109493972A (en) * | 2018-10-30 | 2019-03-19 | 平安医疗健康管理股份有限公司 | Data processing method, device, server and storage medium based on prediction model |
Non-Patent Citations (2)
| Title |
|---|
| 付加锋 等: "《华北地区气候灾害对能源和交通的影响与适应评估》", 31 May 2017, 中国环境科学出版社 * |
| 邱泉 等: "《城市灾害与疾病防控》", 30 June 2016, 光明日报出版社 * |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112750532B (en) * | 2019-10-30 | 2024-01-19 | 宏碁股份有限公司 | Feature recognition method and electronic device |
| CN112750532A (en) * | 2019-10-30 | 2021-05-04 | 宏碁股份有限公司 | Feature identification method and electronic device |
| CN111063436A (en) * | 2019-11-25 | 2020-04-24 | 泰康保险集团股份有限公司 | Data processing method and device, storage medium and electronic terminal |
| CN110993103A (en) * | 2019-11-28 | 2020-04-10 | 阳光人寿保险股份有限公司 | Method for establishing disease risk prediction model and method for recommending disease insurance product |
| 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 |
| CN111274305A (en) * | 2020-01-15 | 2020-06-12 | 平安医疗健康管理股份有限公司 | Three-dimensional picture generation method and device, computer equipment and storage medium |
| CN111274305B (en) * | 2020-01-15 | 2023-03-31 | 深圳平安医疗健康科技服务有限公司 | Three-dimensional picture generation method and device, computer equipment and storage medium |
| WO2021151291A1 (en) * | 2020-05-26 | 2021-08-05 | 平安科技（深圳）有限公司 | Disease risk analysis method, apparatus, electronic device, and computer storage medium |
| CN111681759A (en) * | 2020-05-29 | 2020-09-18 | 芯海科技（深圳）股份有限公司 | Chronic disease risk analysis device, equipment and storage medium |
| CN111681759B (en) * | 2020-05-29 | 2023-10-27 | 芯海科技（深圳）股份有限公司 | Chronic disease risk analysis device, apparatus, and storage medium |
| CN112053779A (en) * | 2020-06-15 | 2020-12-08 | 深圳博脑医疗科技有限公司 | Construction method and construction device of disease detection model and terminal equipment |
| CN112053779B (en) * | 2020-06-15 | 2024-01-26 | 深圳博脑医疗科技有限公司 | Disease detection model construction method, disease detection model construction device and terminal equipment |
| CN111785370B (en) * | 2020-07-01 | 2024-05-17 | 医渡云（北京）技术有限公司 | Medical record data processing method and device, computer storage medium and electronic equipment |
| 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 |
| CN111986763A (en) * | 2020-09-03 | 2020-11-24 | 平安国际智慧城市科技股份有限公司 | Disease data analysis method and device, electronic device and storage medium |
| CN111986763B (en) * | 2020-09-03 | 2024-05-14 | 深圳平安智慧医健科技有限公司 | Disease data analysis method, device, electronic equipment and storage medium |
| CN112132625A (en) * | 2020-09-27 | 2020-12-25 | 平安医疗健康管理股份有限公司 | Health risk-based medical expense prediction system |
| CN111899883B (en) * | 2020-09-29 | 2020-12-15 | 平安科技（深圳）有限公司 | Disease prediction device, method, apparatus and storage medium for small sample or zero sample |
| CN111899883A (en) * | 2020-09-29 | 2020-11-06 | 平安科技（深圳）有限公司 | Disease prediction device, method, apparatus and storage medium for small sample or zero sample |
| CN111933303B (en) * | 2020-09-30 | 2021-01-15 | 平安科技（深圳）有限公司 | Event prediction method and device, electronic equipment and storage medium |
| CN111933303A (en) * | 2020-09-30 | 2020-11-13 | 平安科技（深圳）有限公司 | Event prediction method and device, electronic equipment and storage medium |
| CN112216400A (en) * | 2020-10-10 | 2021-01-12 | 黑龙江省疾病预防控制中心 | Method and system for predicting food-borne disease pathogenic factors based on big data |
| CN112117015A (en) * | 2020-10-28 | 2020-12-22 | 平安科技（深圳）有限公司 | Sepsis early warning device, method, apparatus and storage medium |
| CN112117015B (en) * | 2020-10-28 | 2023-09-05 | 平安科技（深圳）有限公司 | Sepsis early warning equipment, sepsis early warning method, sepsis early warning device and sepsis early warning storage medium |
| CN112465011A (en) * | 2020-11-25 | 2021-03-09 | 平安医疗健康管理股份有限公司 | Project risk prediction method and system based on project research and development process |
| CN112465011B (en) * | 2020-11-25 | 2022-08-02 | 深圳平安医疗健康科技服务有限公司 | Project risk prediction method and system based on project research and development process |
| CN112435747A (en) * | 2020-11-26 | 2021-03-02 | 新智数字科技有限公司 | Disease risk prediction method, device, equipment and computer readable medium |
| CN113707323A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Disease prediction method, device, equipment and medium based on machine learning |
| CN113657550A (en) * | 2021-08-31 | 2021-11-16 | 平安医疗健康管理股份有限公司 | Patient marking method, device, equipment and storage medium based on hierarchical calculation |
| CN113707323B (en) * | 2021-08-31 | 2024-05-14 | 平安科技（深圳）有限公司 | Disease prediction method, device, equipment and medium based on machine learning |
| CN113782216B (en) * | 2021-09-15 | 2023-10-24 | 平安科技（深圳）有限公司 | Disabling weight determining method and device, electronic equipment and storage medium |
| CN113782216A (en) * | 2021-09-15 | 2021-12-10 | 平安科技（深圳）有限公司 | Method and device for determining incapability weight, electronic equipment and storage medium |
| CN114898888B (en) * | 2022-07-15 | 2022-09-23 | 武汉大学 | Medical data processing method and device, computer equipment and readable storage medium |
| CN114898888A (en) * | 2022-07-15 | 2022-08-12 | 武汉大学 | Medical data processing method and device, computer equipment and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110211690A (en) | Disease risks prediction technique, device, computer equipment and computer storage medium | |
| Dev et al. | A predictive analytics approach for stroke prediction using machine learning and neural networks | |
| Kim et al. | A deep learning model for real-time mortality prediction in critically ill children | |
| Ansarullah et al. | Significance of Visible Non‐Invasive Risk Attributes for the Initial Prediction of Heart Disease Using Different Machine Learning Techniques | |
| Lin et al. | User-level psychological stress detection from social media using deep neural network | |
| Yang et al. | Coronary heart disease optimization system on adaptive-network-based fuzzy inference system and linear discriminant analysis (ANFIS–LDA) | |
| CN110363090A (en) | Intelligent heart disease detection method, device and computer readable storage medium | |
| CN116994709A (en) | Personalized diet and exercise recommendation method and system and electronic equipment | |
| Yang et al. | Multimodal temporal-clinical note network for mortality prediction | |
| CA3110455A1 (en) | Visualization of social determinants of health | |
| Ben-Assuli et al. | Analysing repeated hospital readmissions using data mining techniques | |
| CN118277573B (en) | Pre-hospital emergency text classification labeling method based on ChatGLM model, electronic equipment, storage medium and computer program product | |
| CN113658655B (en) | Physical examination recommendation method, device, storage medium and equipment | |
| CN116010586A (en) | Method, device, equipment and storage medium for generating health advice | |
| CN117557331A (en) | Product recommendation method and device, computer equipment and storage medium | |
| CN117012380A (en) | Event estimation method, device, equipment, storage medium and program product | |
| CN110414562B (en) | X-ray film classification method, device, terminal and storage medium | |
| CN117312514A (en) | Consultation reply method, consultation reply device and computer readable storage medium | |
| Ayad et al. | Predicting abnormalities in laboratory values of patients in the intensive care unit using different deep learning models: comparative study | |
| Asif et al. | Improving the accuracy of diagnosing and predicting coronary heart disease using ensemble method and feature selection techniques | |
| CN116403728B (en) | Data processing device for medical treatment data and related equipment | |
| CN117457159A (en) | Medical main body recommendation method, medical main body recommendation device, computer equipment and storage medium | |
| CN117457185A (en) | Disease prediction method, device, apparatus and storage medium | |
| CN116863116A (en) | Image recognition method, device, equipment and medium based on artificial intelligence | |
| CN114707633B (en) | Feature extraction method, device, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |