CN109785928A - Diagnosis and treatment proposal recommending method, device and storage medium - Google Patents
Diagnosis and treatment proposal recommending method, device and storage medium Download PDFInfo
- Publication number
- CN109785928A CN109785928A CN201811592838.9A CN201811592838A CN109785928A CN 109785928 A CN109785928 A CN 109785928A CN 201811592838 A CN201811592838 A CN 201811592838A CN 109785928 A CN109785928 A CN 109785928A
- Authority
- CN
- China
- Prior art keywords
- case history
- sample
- diagnosis
- similar
- treatment
- 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.)
- Granted
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000012549 training Methods 0.000 claims description 67
- 238000012545 processing Methods 0.000 claims description 12
- 230000000153 supplemental effect Effects 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 4
- 238000007405 data analysis Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 230000017105 transposition Effects 0.000 claims description 4
- 238000013075 data extraction Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000001934 delay Effects 0.000 abstract description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 13
- 239000003814 drug Substances 0.000 description 10
- 229940079593 drug Drugs 0.000 description 9
- 238000012550 audit Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000001225 therapeutic effect Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000002560 therapeutic procedure Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000012512 characterization method Methods 0.000 description 3
- 238000010835 comparative analysis Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000002512 chemotherapy Methods 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000013467 fragmentation Methods 0.000 description 2
- 238000006062 fragmentation reaction Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 230000005055 memory storage Effects 0.000 description 2
- 238000013146 percutaneous coronary intervention Methods 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000001959 radiotherapy Methods 0.000 description 2
- 230000003014 reinforcing effect Effects 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 239000002671 adjuvant Substances 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention belongs to field of artificial intelligence, disclose a kind of diagnosis and treatment proposal recommending method, comprising: the building database of case history, the database of case history include the first case history sample and mark label；Construct similar case history model and recommended models；The similar case history model and the recommended models are trained according to the data in the database of case history；Obtain the second case history sample；The second case history sample is inputted into the trained similar case history model, export the similar case history of similar to the second case history sample one or more, the obtained similar case history of one or more is inputted into the trained recommended models, exports diagnosis and treatment path corresponding with the second case history sample.The present invention recommends to obtain the diagnosis and treatment scheme an of globality for the current state of an illness of patient, reduces the dependence for medical expert's level, and reducing human factor influences, and avoids the occurrence of and delays the situations such as treatment.The invention also discloses a kind of electronic device and computer readable storage mediums.
Description
Technical field
The present invention relates to field of artificial intelligence more particularly to a kind of diagnosis and treatment proposal recommending method, device and storage to be situated between
Matter.
Background technique
With aging of population and the continuous development of medical information, quality of medical care should be ensured, pursue again clinical
Working efficiency.Medical industry, which is directed to clinical big data and deep learning algorithm, at present many Aided intelligent decision-makings, such as:
Intelligent recommendation inspection, medication recommendation, intelligent answer and follow-up system etc. are carried out by physical signs of patient data, but these are all only examined
Fragmentation scene during treatment, without really with regard to the most crucial diagnosis and treatment side how treated patient and form a globality
Case.Currently, for providing whole diagnosis and treatment scheme, diagnosis and treatment path etc. to patient, mainly still passing through in hospital at home and abroad
Doctor combines patient's state of an illness and is artificially handled with supervisor's experience of expert and theoretical knowledge etc., the medical water dependent on expert
It is flat, when expert is wrong to medical knowledge or to diagnosing patient it is wrong when, easily occur partially for the recommendation of patient's diagnosis and treatment scheme
Difference can generate adverse effect to the solution of patient's problem, delay patient to obtain the opportunity of therapeutic scheme, delay the treatment of patient.
Summary of the invention
The present invention provides a kind of diagnosis and treatment proposal recommending method, device and storage medium, recommends for the current state of an illness of patient
To the diagnosis and treatment scheme of a globality, the dependence for medical expert's level is reduced, is controlled so that patient obtains in the best opportunity
Treatment scheme avoids the occurrence of and delays the situations such as treatment.
To achieve the goals above, the present invention provides a kind of diagnosis and treatment proposal recommending method, comprising the following steps: building case history
Database, the database of case history include the first case history sample and mark label；Construct similar case history model and recommended models；Root
The similar case history model and the recommended models are trained according to the data in the database of case history；Obtain the second case history sample；
The second case history sample is inputted into the trained similar case history model, is exported similar with the second case history sample
The obtained similar case history of one or more is inputted the trained recommended models, output by one or more similar case histories
Diagnosis and treatment path corresponding with the second case history sample.
Preferably, the step of training similar case history model includes: building training set, and the training set includes multiple instructions
Practice sample；Initialize the parameter of the similar case history model, the weighting parameters of the characteristic value including each training sample；Input instruction
Practice the training sample concentrated；Test label corresponding with the training sample is exported by similar case history model；Building loss
Model calculates test label by the loss model and marks the degree of loss between label；Update the ginseng of similar case history model
Number, is according to the following formula updated the weighting parameters of the characteristic value of each training sample:
K '=k+e
Wherein, k indicates the weighting parameters of the characteristic value of training sample before updating；The feature of training sample after k ' expression updates
The weighting parameters of value, e indicate error；
It according to updated similar case history model parameter, inputs next training sample and is trained, until all training
Sample all training, complete an iteration；Judge whether similar case history model training meets termination condition, if meeting terminates item
Part then exports the similar case history model and continues to train if being unsatisfactory for termination condition, wherein the termination condition includes the
One termination condition and/or the second termination condition, first termination condition are the number of iterations that current iteration number is greater than setting,
Second termination condition is that loss function value is less than set target value.
Preferably, the step of exporting one or more similar case histories similar to the second case history sample includes: to obtain
The similarity of first case history sample in the second case history sample and the database of case history；According to the size of similarity, selection
The biggish one or more first case history samples of similarity are as similar case history similar to the second case history sample；Output institute
State similar case history.
Further, it is preferable to which the step of similarity of ground, acquisition the second case history sample and the first case history sample, includes:
Extract the characteristic value of the second case history sample and the first case history sample；Obtain the characteristic value of the second case history sample and described
The similarity of the characteristic value of first case history sample；The second case history sample and the first case history sample are obtained by following formula
Similarity:
Wherein, S indicates the similarity of the second case history sample and the first case history sample；The index of i expression characteristic value；N is indicated
The quantity of characteristic value；The weighting parameters of k expression characteristic value；S indicates the characteristic value and the first case history sample of the second case history sample
The similarity of characteristic value.
Further, it is preferable to which the characteristic value of ground, the characteristic value of the second case history sample and the first case history sample is similar
Degree is obtained by following formula:
Wherein, X indicates the characteristic value of the first case history sample, and Y indicates the characteristic value of the second case history sample, and s (X, Y) indicates X
With the similarity of Y, T indicates transposition, ∑-1Indicate the covariance matrix between X and Y.
Preferably, the step of exporting diagnosis and treatment path corresponding with the second case history sample includes: each of to acquire
The supplemental characteristic information of similar case history；The supplemental characteristic information and the similar case history are inputted into the trained recommendation
Model；Diagnosis and treatment path corresponding with the second case history sample is exported by the recommended models.
Preferably, the step of building database of case history includes: acquisition medical record data；The medical record data is tied
Structureization processing obtains the first case history sample, and marks label to the first case history sample；By the first case history sample and mark
Label storage, forms the database of case history.
Further, the step of carrying out structuring processing to the medical record data includes: to collect to the medical record data
At formation complete data set；Data cleansing is carried out to the complete data set；To after cleaning data set carry out data analysis and
It extracts, the medical record data that garbled data is concentrated；The medical record data obtained by screening is converted, feature vector is formed；It deposits
Described eigenvector is stored up, the first case history sample is formed.
Another aspect of the present invention provides a kind of electronic device, which includes: memory and processor, described
It include diagnosis and treatment scheme recommended program in memory, the diagnosis and treatment scheme recommended program realizes institute as above when being executed by the processor
The step of diagnosis and treatment proposal recommending method stated.
Another aspect of the invention provides a kind of computer readable storage medium, in the computer readable storage medium
Including diagnosis and treatment scheme recommended program, when the diagnosis and treatment scheme recommended program is executed by processor, diagnosis and treatment side as described above is realized
The step of case recommended method.
Compared with the existing technology, the present invention has the following advantages and beneficial effects:
Similar case history model and recommended models of the present invention by building, the current state of an illness based on intended patient, in conjunction with institute
Some history diagnosis and therapy recordings, recommend diagnosis and treatment path optimal out to the intended patient, the diagnosis and treatment path obtain no longer completely according to
Rely in the medical level of medical expert and supervisor's experience etc., reduce the influence of human factor, also, by precisely, efficiently to
Patient recommends to obtain diagnosis and treatment path, avoids patient and misses medical opportunity, delays treatment.
Detailed description of the invention
Fig. 1 is the flow diagram of diagnosis and treatment proposal recommending method of the present invention；
Fig. 2 is the module diagram of diagnosis and treatment scheme recommended program in the present invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
Embodiment of the present invention described below with reference to the accompanying drawings.Those skilled in the art may recognize that
It arrives, it without departing from the spirit and scope of the present invention, can be with a variety of different modes or combinations thereof to described
Embodiment is modified.Therefore, attached drawing and description are regarded as illustrative in nature, and are only used to explain the present invention, rather than are used
In limitation scope of protection of the claims.In addition, in the present specification, attached drawing is drawn not in scale, and identical attached drawing mark
Note indicates identical part.
Fig. 1 is the flow diagram of diagnosis and treatment proposal recommending method of the present invention, as shown in Figure 1, diagnosis and treatment of the present invention
Proposal recommending method the following steps are included:
Step S1, the database of case history is constructed, the database of case history includes the first case history sample and mark label；
Step S2, similar case history model and recommended models are constructed；
Step S3, the similar case history model and the recommended models are trained according to the data in the database of case history；
Step S4, the second case history sample is obtained；
Step S5, the second case history sample is inputted into the trained similar case history model, output and described the
The similar case history of the similar one or more of two case history samples；
Step S6, the obtained similar case history of one or more is inputted into the trained recommended models, output and institute
State the corresponding diagnosis and treatment path of the second case history sample.
The present invention is obtained by the similar case history model and recommended models, the superiority and inferiority in the multiple diagnosis and treatment paths of comparative analysis of building
To optimal diagnosis and treatment path corresponding with the current state of an illness of intended patient.Intelligent recommendation not only is carried out in some fragmentation scenes, and
It is to recommend the diagnosis and treatment scheme of a set of globality for patient, and the determination of final diagnosis and treatment scheme does not depend solely on the management warp of expert
Test with theoretical medicine knowledge, and combine existing historical data to multiple diagnosis and treatment paths carry out comprehensive analysis so that patient can and
When obtain accurately, effectively diagnosis and treatment scheme.
In the present invention, construct the database of case history the step of include: acquisition medical record data, the medical record data includes: electronics
Case history, patient's audit report, medication record, historical therapeutic scheme, the data such as record of being hospitalized may include all history diagnosis and treatment notes
Record；Structuring processing is carried out to the medical record data, obtains the first case history sample, and mark and mark to the first case history sample
Label；The first case history sample and label are stored, the database of case history is formed.
Case history similitude is measured, similarity calculation etc. is carried out to case history, is built upon the medical record data base of structuring
On plinth, it is preferable that the step of carrying out structuring processing to medical record data includes: to integrate to medical record data, forms complete number
According to collection；Data cleansing is carried out to the complete data set, to remove the data for not meeting specification；Data set after cleaning is carried out
Data analysis and extraction, the medical record data that garbled data is concentrated, to filter out satisfactory data；By what is obtained by screening
Medical record data is converted, and feature vector is formed, and characterizes medical record data feature with this feature vector；Described eigenvector is stored,
Form the first case history sample.
When constructing the database of case history, label for labelling is carried out to the first obtained case history sample, in order to according to case history number
Similar case history model and recommended models are trained according to the data in library.The label of mark includes corresponding with the first case history sample
Case treatment path, the treatment path based on the essential information of patient, symptom, check that data etc. determine, the treatment road
Diameter includes: medicining condition, percutaneous coronary intervention (for example, the treatment means such as drug, operation, radiation and chemotherapy), curative effect situation, wind
Danger, treatment cost and follow-up tracking etc., to form the treatment path of complete set.
In the embodiment of the present invention, with the relevant medical record data of the first case history sample characterization history diagnosis and therapy recording, with second
Case history sample characterizes the medical record data based on the current state of an illness of intended patient.What the similar case history model of building was used to obtain and input
The similar one or more first case history samples of second case history sample.After the recommended models of building are located at similar case history model
End, all or part of input of the output of similar case history model as recommended models are used for according to similar one or more the
One case history sample obtains diagnosis and therapy recording corresponding with the second case history sample.
In one embodiment of the present of invention, training the similar case history model the step of include:
Training set is constructed, the training set includes multiple training samples, wherein training sample is selected from the database of case history, packet
Include the first case history sample and corresponding mark label；
Initialize the parameter of the similar case history model, the weighting parameters of the characteristic value including each training sample；
Input a training sample in training set；
Test label corresponding with the training sample is exported by similar case history model；
Loss model is constructed, test label is calculated by the loss model and marks the degree of loss between label；
The parameter of similar case history model is updated, the weighting parameters of the characteristic value of each training sample are carried out more according to the following formula
It is new:
K '=k+e
Wherein, k indicates the weighting parameters of the characteristic value of training sample before updating；The feature of training sample after k ' expression updates
The weighting parameters of value, e indicate error；
It according to updated similar case history model parameter, inputs next training sample and is trained, until all training
Sample all training, complete an iteration；
Judge whether similar case history model training meets termination condition, if meeting termination condition, exports described similar
Case history model continues to train if being unsatisfactory for termination condition, wherein the termination condition includes the first termination condition and/or the
Two termination conditions, first termination condition are the number of iterations that current iteration number is greater than setting, second termination condition
It is less than set target value for loss function value.
In the present invention, the number of iterations of setting can be according to the complexity (for example, GPU calculates power, calculating time etc.) of operation
It is determined with data sample (data volume, quality including data sample etc.), approaches 0% infinitely to lose curve as mesh
Mark.Loss function can be cross entropy loss function, quadratic loss function etc. in loss model.
Similarly, the training of recommended models is exported most by gradient descent algorithm using the smallest penalty values as recommended models
The standard in excellent treatment path.
After the similar case history model of training and recommended models, the medical record data of the characterization current goal patient state of an illness, packet are obtained
It includes: each dimension letter such as patient's essential information, household heredity factors, vital sign, audit report, history medication, state of an illness historical development
Breath carries out structuring processing to the medical record data of current goal patient, obtains the second case history sample.
The current state of an illness of intended patient is characterized using the second case history sample, by similar case history model, obtains likeness in form case history, this
Invention an alternative embodiment in, export it is similar to the second case history sample one or more similar case history the step of wrap
It includes:
Obtain the similarity of the first case history sample in the second case history sample and the database of case history；
According to the size of similarity, select the biggish one or more first case history samples of similarity as with described second
The similar similar case history of case history sample；
Export the similar case history.
Wherein, the first case history sample is arranged successively from big to small according to the similarity of itself and the second case history sample, according to
Similarity size therefrom chooses the biggish one or more first case history samples of similarity (for example, if choosing first case history
Sample then selects the corresponding first case history sample of maximum similarity, if the multiple first case history samples of selection, according to similarity from
10 the first case history samples that small successively selection is located at before similarity arrangement list are arrived greatly), similar case history list is formed, by phase
It is exported like the first case history sample in case history list as similar case history, and is used for input recommended models.
In the present invention, case history is used as by the key index (KPI, Key Performance Indicator) of various dimensions
The characteristic value of sample measures the similitude between the second case history sample and the first case history sample to be weighted analysis, wherein
KPIs (Key Performance Indicators) includes: patient's essential information, region, family's medical history, patient's illness, checks
One of report, medication history are a variety of.The similarity calculation of specific targets can pass through: Euclidean distance (Eucledian
Distance), cosine similarity (Cosine Similarity), Minkowski distance (Minkowski Distance),
Jaccard Similarity scheduling algorithm is calculated.
In an alternative embodiment of the invention, the similarity of the second case history sample and the first case history sample is obtained
Step includes:
Extract the characteristic value of the second case history sample and the first case history sample；
Obtain the similarity of the characteristic value of the second case history sample and the characteristic value of the first case history sample；
The similarity of the second case history sample and the first case history sample is obtained by following formula:
Wherein, S indicates the similarity of the second case history sample and the first case history sample；The index of i expression characteristic value；N is indicated
The quantity of characteristic value；The weighting parameters of k expression characteristic value；S indicates the characteristic value and the first case history sample of the second case history sample
The similarity of characteristic value.
Further, the similar of the characteristic value of the second case history sample and the characteristic value of the first case history sample is obtained by following formula
Degree:
Wherein, X indicates the characteristic value of the first case history sample, and Y indicates the characteristic value of the second case history sample, and s (X, Y) indicates X
With the similarity of Y, T indicates transposition, ∑-1Indicate the covariance matrix between X and Y.
In the present invention, input of the output of similar case history model as recommended models is treated by recommended models in output
When path, superiority and inferiority comparative analysis is carried out to the treatment path of multiple similar case histories of input, is exported corresponding to the second case history sample
Treatment path.When recommending treatment path according to the superiority and inferiority in treatment path, not only consider that the treatment path of similar case history obtains
Therapeutic effect etc., it is also contemplated that other auxiliary letters such as expense, the course for the treatment of, wound degree and relative risk involved in treatment path
The precondition of the reinforcings input such as theoretical knowledge or micro-judgment of breath and expert doctor, so that being recommended by recommended models
The diagnosis and treatment path obtained comprehensively considers the state of an illness and all preconditions of current goal patient, compares and is suitble to current target
Patient.Preferably, the step of exporting diagnosis and treatment path corresponding with the second case history sample include: each of acquire it is similar
The supplemental characteristic information of case history, wherein supplemental characteristic information is treated according to involved in the corresponding treatment path of the similar case history
Expense, the course for the treatment of, wound degree and relative risk and expert doctor, which extract the micro-judgment of the current state of an illness of patient, to be determined；It will be described
Supplemental characteristic information and similar case history input the trained recommended models；Pass through recommended models output and the second disease
Go through the corresponding diagnosis and treatment path of sample.
In an alternate embodiment of the present invention where, the step of recommended models output diagnosis and treatment path includes:
It obtains according to the following formula and recommends the corresponding probability for treating path of similar case history:
Wherein, PbIndicate the corresponding probability for treating path of b-th of similar case history of output；A indicates the second case history sample, b
Indicate the index of similar case history, Sa,bIndicate the similarity between b-th of similar case history sample and the second case history sample；RbIt indicates
The corresponding supplemental characteristic value of b-th of similar case history, B indicate the quantity of similar case history；
The corresponding treatment path of the maximum similar case history of probability value is chosen to export.
In the present invention, recommend optimal diagnosis and treatment path by recommended models, this diagnosis and treatment path is from various dimensions (including treatment side
Method, the course for the treatment of, curative effect track, may risk, complication and precautionary measures, cost control and detail etc.) each diagnosis and treatment side of detailed comparisons
The superiority and inferiority (including: foundation, knowledge mapping, Medical guidelines and the industry key cases of circular for confirmation) of case and visual core index
It compares (including individual difference, curative effect, cost etc.), last decision is done with adjuvant clinical doctor.For example, patient A is diagnosed as certain
Disease mid-term carries out the selection in treatment path, through the invention the diagnosis and treatment proposal recommending method, for patient A's
The current state of an illness carries out the matching of similar case history model and the recommendation in diagnosis and treatment path, to carry out the formulation of whole diagnosis and treatment scheme.It is logical
It crosses similar case history model and obtains similar case history from the history case history sample of magnanimity, and by recommended models recommend to obtain detailed
Diagnosis and treatment path includes: a variety of diagnosis and treatment schemes recommendations and circular for confirmation foundation；The superiority and inferiority of different diagnosis and treatment schemes is to when visual core
Index；Accurately similar case history and corresponding complete diagnosis and treatment scheme (including before examining, examine in and after examining)；Cure track and possibility
Risk and precautionary measures；Examine rear rehabilitation index and follow-up scheme.
Diagnosis and treatment proposal recommending method of the present invention is applied to electronic device, and electronic device can be television set, intelligent hand
The terminal devices such as machine, tablet computer, computer.The electronic device includes: processor；Memory, for storing diagnosis and treatment scheme
The step of recommended program, processor executes diagnosis and treatment scheme recommended program, realizes following diagnosis and treatment proposal recommending method:
The database of case history is constructed, the database of case history includes the first case history sample and mark label；Construct similar case history
Model and recommended models；The similar case history model and the recommended models are trained according to the data in the database of case history；
Obtain the second case history sample；The second case history sample is inputted into the trained similar case history model, output with it is described
The similar case history of the similar one or more of second case history sample, the similar case history input of obtained one or more is trained
The recommended models export diagnosis and treatment path corresponding with the second case history sample.
Memory includes the readable storage medium storing program for executing of at least one type, can be that flash memory, hard disk, CD etc. are non-volatile to be deposited
Storage media is also possible to plug-in type hard disk etc., and is not limited to this, can be in a manner of non-transitory store instruction or software with
And any associated data file and to processor provide instruction or software program so that the processor be able to carry out instruction or
Any device of software program.In the present invention, the software program of memory storage includes diagnosis and treatment scheme recommended program, and can be to
Processor provides the diagnosis and treatment scheme recommended program, so that processor can execute the diagnosis and treatment scheme recommended program, realizes diagnosis and treatment
Proposal recommending method.
Processor can be central processing unit, microprocessor or other data processing chips etc., can be in run memory
Storage program.
In the present invention, electronic device construct the database of case history the step of include: acquisition medical record data, the medical record data packet
Include: electronic health record, patient's audit report, medication record, historical therapeutic scheme, the data such as record of being hospitalized, may include all go through
History diagnosis and therapy recording；Structuring processing is carried out to the medical record data, obtains the first case history sample, and to the first case history sample
Mark label；The first case history sample and label are stored, the database of case history is formed.
Case history similitude is measured, similarity calculation etc. is carried out to case history, is built upon the medical record data base of structuring
On plinth, it is preferable that the step of electronic device carries out structuring processing to medical record data includes: to integrate to medical record data, shape
At complete data set；Data cleansing is carried out to the complete data set, to remove the data for not meeting specification；To the number after cleaning
Data analysis and extraction, the medical record data that garbled data is concentrated, to filter out satisfactory data are carried out according to collection；It will be through being sieved
It selects obtained medical record data to be converted, forms feature vector, medical record data feature is characterized with this feature vector；Store the spy
Vector is levied, the first case history sample is formed.
When constructing the database of case history, label for labelling is carried out to the first obtained case history sample, in order to according to case history number
Similar case history model and recommended models are trained according to the data in library.The label of mark includes corresponding with the first case history sample
Case treatment path, the treatment path based on the essential information of patient, symptom, check that data etc. determine, the treatment road
Diameter includes: medicining condition, percutaneous coronary intervention (for example, the treatment means such as drug, operation, radiation and chemotherapy), curative effect situation, wind
Danger, treatment cost and follow-up tracking etc., to form the treatment path of complete set.
In one embodiment of the present of invention, the step of electronic device training similar case history model, includes:
Training set is constructed, the training set includes multiple training samples, wherein training sample is selected from the database of case history, packet
Include the first case history sample and corresponding mark label；
Initialize the parameter of the similar case history model, the weighting parameters of the characteristic value including each training sample；
Input a training sample in training set；
Test label corresponding with the training sample is exported by similar case history model；
Loss model is constructed, test label is calculated by the loss model and marks the degree of loss between label；
The parameter of similar case history model is updated, the weighting parameters of the characteristic value of each training sample are carried out more according to the following formula
It is new:
K '=k+e
Wherein, k indicates the weighting parameters of the characteristic value of training sample before updating；The feature of training sample after k ' expression updates
The weighting parameters of value, e indicate error；
It according to updated similar case history model parameter, inputs next training sample and is trained, until all training
Sample all training, complete an iteration；
Judge whether similar case history model training meets termination condition, if meeting termination condition, exports described similar
Case history model continues to train if being unsatisfactory for termination condition, wherein the termination condition includes the first termination condition and/or the
Two termination conditions, first termination condition are the number of iterations that current iteration number is greater than setting, second termination condition
It is less than set target value for loss function value.
In the present invention, the number of iterations of setting can be according to the complexity (for example, GPU calculates power, calculating time etc.) of operation
It is determined with data sample (data volume, quality including data sample etc.), approaches 0% infinitely to lose curve as mesh
Mark.Loss function can be cross entropy loss function, quadratic loss function etc. in loss model.
Similarly, the training of recommended models is exported most by gradient descent algorithm using the smallest penalty values as recommended models
The standard in excellent treatment path.
After the similar case history model of training and recommended models, the medical record data of the characterization current goal patient state of an illness, packet are obtained
It includes: each dimension letter such as patient's essential information, household heredity factors, vital sign, audit report, history medication, state of an illness historical development
Breath carries out structuring processing to the medical record data of current goal patient, obtains the second case history sample.
The current state of an illness of intended patient is characterized using the second case history sample, by similar case history model, obtains likeness in form case history, this
In one alternative embodiment of invention, electronic device exports the similar case history of similar to the second case history sample one or more
The step of include: the similarity for obtaining the first case history sample in the second case history sample and the database of case history；According to similarity
Size, select the biggish one or more first case history samples of similarity as similar similar to the second case history sample
Case history；Export the similar case history.Wherein, by the first case history sample according to its similarity with the second case history sample from big to small
It is arranged successively, according to similarity size, therefrom chooses the biggish one or more first case history samples of similarity (for example, if choosing
A first case history sample is taken, then selects the corresponding first case history sample of maximum similarity, if the multiple first case history samples of selection,
Then according to similarity, successively selection is located at 10 the first case history samples before similarity arrangement list from big to small), form phase
Like case history list, the first case history sample in similar case history list is exported as similar case history, and is used for input and recommends
Model.
Feature by the key index (KPI, Key Performance Indicator) of various dimensions as case history sample
Value measures the similitude between the second case history sample and the first case history sample to be weighted analysis, wherein KPIs (Key
Performance Indicators) it include: patient's essential information, region, family's medical history, patient's illness, audit report, medication
One of history is a variety of.The similarity calculation of specific targets can pass through: Euclidean distance (Eucledian
Distance), cosine similarity (Cosine Similarity), Minkowski distance (Minkowski Distance),
Jaccard Similarity scheduling algorithm is calculated.
In an alternative embodiment of the invention, electronic device obtains the second case history sample and the first case history sample
The step of similarity includes: the characteristic value for extracting the second case history sample and the first case history sample；Obtain second case history
The similarity of the characteristic value of sample and the characteristic value of the first case history sample；By following formula obtain the second case history sample and
The similarity of the first case history sample:
Wherein, S indicates the similarity of the second case history sample and the first case history sample；The index of i expression characteristic value；N is indicated
The quantity of characteristic value；The weighting parameters of k expression characteristic value；S indicates the characteristic value and the first case history sample of the second case history sample
The similarity of characteristic value.
Further, the similar of the characteristic value of the second case history sample and the characteristic value of the first case history sample is obtained by following formula
Degree:
Wherein, X indicates the characteristic value of the first case history sample, and Y indicates the characteristic value of the second case history sample, and s (X, Y) indicates X
With the similarity of Y, T indicates transposition, ∑-1Indicate the covariance matrix between X and Y.
In the present invention, input of the output of similar case history model as recommended models is treated by recommended models in output
When path, superiority and inferiority comparative analysis is carried out to the treatment path of multiple similar case histories of input, is exported corresponding to the second case history sample
Treatment path.When recommending treatment path according to the superiority and inferiority in treatment path, not only consider that the treatment path of similar case history obtains
Therapeutic effect etc., it is also contemplated that other auxiliary letters such as expense, the course for the treatment of, wound degree and relative risk involved in treatment path
The precondition of the reinforcings input such as theoretical knowledge or micro-judgment of breath and expert doctor, so that being recommended by recommended models
The diagnosis and treatment path obtained comprehensively considers the state of an illness and all preconditions of current goal patient, compares and is suitble to current target
Patient.Preferably, the step of exporting diagnosis and treatment path corresponding with the second case history sample include: each of acquire it is similar
The supplemental characteristic information of case history, wherein supplemental characteristic information is treated according to involved in the corresponding treatment path of the similar case history
Expense, the course for the treatment of, wound degree and relative risk and expert doctor, which extract the micro-judgment of the current state of an illness of patient, to be determined；It will be described
Supplemental characteristic information and similar case history input the trained recommended models；Pass through recommended models output and the second disease
Go through the corresponding diagnosis and treatment path of sample.
In an alternate embodiment of the present invention where, the step of recommended models output diagnosis and treatment path includes:
It obtains according to the following formula and recommends the corresponding probability for treating path of similar case history:
Wherein, PbIndicate the corresponding probability for treating path of b-th of similar case history of output；A indicates the second case history sample, b
Indicate the index of similar case history, Sa,bIndicate the similarity between b-th of similar case history sample and the second case history sample；RbIt indicates
The corresponding supplemental characteristic value of b-th of similar case history, B indicate the quantity of similar case history；
The corresponding treatment path of the maximum similar case history of probability value is chosen to export.
In other embodiments, diagnosis and treatment scheme recommended program can also be divided into one or more module, one or
The multiple modules of person are stored in memory, and are executed by processor, to complete the present invention.The so-called module of the present invention refers to energy
Enough complete the series of computation machine program instruction section of specific function.The diagnosis and treatment scheme recommended program can be divided into: data
Library constructs module 1, model construction module 2, model training module 3, sample acquisition module 4, similar case history and obtains module 5 and output
Module 6.The functions or operations step that above-mentioned module is realized is similar as above, and and will not be described here in detail, illustratively, such as its
In:
Database sharing module 1, constructs the database of case history, and the database of case history includes the first case history sample and mark mark
Label；
Model construction module 2 constructs similar case history model and recommended models；
Model training module 3 according to the data training similar case history model in the database of case history and described pushes away
Recommend model；
Sample acquisition module 4 obtains the second case history sample；
Similar case history obtains module 5, and the second case history sample is inputted the trained similar case history model, defeated
The similar case history of one or more similar to the second case history sample out；
Output module 6, the similar case histories of one or more that will be obtained input the trained recommended models, output with
The corresponding diagnosis and treatment path of the second case history sample.
In one embodiment of the present of invention, computer readable storage medium, which can be, any includes or storage program or instruction
Tangible medium, program therein can be performed, and pass through the corresponding function of the relevant hardware realization of the program instruction of storage.Example
Such as, computer readable storage medium can be computer disk, hard disk, random access memory, read-only memory etc..The present invention
It is not limited to this, can be in a manner of non-transitory store instruction or software and any associated data files or data structure simultaneously
And processor is provided to so that processor executes any device of program therein or instruction.The computer-readable storage medium
It include that diagnosis and treatment scheme recommended program realizes above-mentioned diagnosis and treatment side when the diagnosis and treatment scheme recommended program is executed by processor in matter
Case recommended method, to avoid repeating, details are not described herein.
The specific embodiment of the computer readable storage medium of the present invention and above-mentioned diagnosis and treatment proposal recommending method, electronics fill
The specific embodiment set is roughly the same, and details are not described herein.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, device, article or the method that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, device, article or method institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, device of element, article or method.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.Pass through above embodiment party
The description of formula, it is required general that those skilled in the art can be understood that above-described embodiment method can add by software
The mode of hardware platform is realized, naturally it is also possible to which by hardware, but in many cases, the former is more preferably embodiment.It is based on
Such understanding, substantially the part that contributes to existing technology can be with software product in other words for technical solution of the present invention
Form embody, which is stored in a storage medium (such as ROM/RAM, magnetic disk, light as described above
Disk) in, including some instructions use is so that a terminal device (can be mobile phone, computer, server or the network equipment
Deng) execute method described in each embodiment of the present invention.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of diagnosis and treatment proposal recommending method, which comprises the following steps:
The database of case history is constructed, the database of case history includes the first case history sample and mark label；
Construct similar case history model and recommended models；
The similar case history model and the recommended models are trained according to the data in the database of case history；
Obtain the second case history sample；
The second case history sample is inputted into the trained similar case history model, output and the second case history sample phase
As one or more similar case histories；
The obtained similar case history of one or more is inputted into the trained recommended models, output and the second case history sample
This corresponding diagnosis and treatment path.
2. diagnosis and treatment proposal recommending method according to claim 1, which is characterized in that the step of the training similar case history model
Suddenly include:
Training set is constructed, the training set includes multiple training samples；
Initialize the parameter of the similar case history model, the weighting parameters of the characteristic value including each training sample；
Input a training sample in training set；
Test label corresponding with the training sample is exported by similar case history model；
Loss model is constructed, test label is calculated by the loss model and marks the degree of loss between label；
The parameter for updating similar case history model, is according to the following formula updated the weighting parameters of the characteristic value of each training sample:
K '=k+e
Wherein, k indicates the weighting parameters of the characteristic value of training sample before updating；The characteristic value of training sample after k ' expression updates
Weighting parameters, e indicate error；
It according to updated similar case history model parameter, inputs next training sample and is trained, until all training samples
An iteration is completed in all training；
Judge whether similar case history model training meets termination condition, if meeting termination condition, exports the similar case history
Model continues to train if being unsatisfactory for termination condition, wherein the termination condition includes the first termination condition and/or the second knot
Beam condition, first termination condition are the number of iterations that current iteration number is greater than setting, and second termination condition is damage
It loses functional value and is less than set target value.
3. diagnosis and treatment proposal recommending method according to claim 1, which is characterized in that output and the second case history sample phase
As one or more similar case histories the step of include:
Obtain the similarity of the first case history sample in the second case history sample and the database of case history；
According to the size of similarity, select the biggish one or more first case history samples of similarity as with second case history
The similar similar case history of sample；
Export the similar case history.
4. diagnosis and treatment proposal recommending method according to claim 3, which is characterized in that obtain the second case history sample and the
The step of similarity of one case history sample includes:
Extract the characteristic value of the second case history sample and the first case history sample；
Obtain the similarity of the characteristic value of the second case history sample and the characteristic value of the first case history sample；
The similarity of the second case history sample and the first case history sample is obtained by following formula:
Wherein, S indicates the similarity of the second case history sample and the first case history sample；The index of i expression characteristic value；N indicates feature
The quantity of value；The weighting parameters of k expression characteristic value；S indicates the characteristic value of the second case history sample and the feature of the first case history sample
The similarity of value.
5. diagnosis and treatment proposal recommending method according to claim 4, which is characterized in that the characteristic value of the second case history sample
It is obtained with the similarity of the characteristic value of the first case history sample by following formula:
Wherein, X indicates the characteristic value of the first case history sample, and Y indicates the characteristic value of the second case history sample, and s (X, Y) indicates X's and Y
Similarity, T indicate transposition, ∑-1Indicate the covariance matrix between X and Y.
6. diagnosis and treatment proposal recommending method according to claim 1, which is characterized in that output and the second case history sample pair
The step of diagnosis and treatment path answered includes:
Each of acquire the supplemental characteristic information of similar case history；
The supplemental characteristic information and the similar case history are inputted into the trained recommended models；
Diagnosis and treatment path corresponding with the second case history sample is exported by the recommended models.
7. diagnosis and treatment proposal recommending method according to claim 1, which is characterized in that the step of the building database of case history
Include:
Obtain medical record data；
Structuring processing is carried out to the medical record data, obtains the first case history sample, and mark and mark to the first case history sample
Label；
The first case history sample and label are stored, the database of case history is formed.
8. diagnosis and treatment proposal recommending method according to claim 7, which is characterized in that carry out structuring to the medical record data
The step of processing includes:
The medical record data is integrated, complete data set is formed；Data cleansing is carried out to the complete data set；To cleaning
Data set afterwards carries out data analysis and extraction, the medical record data that garbled data is concentrated；The medical record data that will be obtained by screening
It is converted, forms feature vector；Described eigenvector is stored, the first case history sample is formed.
9. a kind of electronic device, which is characterized in that the electronic device includes: memory and processor, includes in the memory
Diagnosis and treatment scheme recommended program is realized when the diagnosis and treatment scheme recommended program is executed by the processor as appointed in claim 1 to 8
The step of diagnosis and treatment proposal recommending method described in one.
10. a kind of computer readable storage medium, which is characterized in that include diagnosis and treatment scheme in the computer readable storage medium
Recommended program when the diagnosis and treatment scheme recommended program is executed by processor, is realized as described in any item of the claim 1 to 8
The step of diagnosis and treatment proposal recommending method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811592838.9A CN109785928B (en) | 2018-12-25 | 2018-12-25 | Diagnosis and treatment scheme recommendation method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811592838.9A CN109785928B (en) | 2018-12-25 | 2018-12-25 | Diagnosis and treatment scheme recommendation method, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109785928A true CN109785928A (en) | 2019-05-21 |
| CN109785928B CN109785928B (en) | 2024-06-25 |
Family
ID=66497691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811592838.9A Active CN109785928B (en) | 2018-12-25 | 2018-12-25 | Diagnosis and treatment scheme recommendation method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109785928B (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110297908A (en) * | 2019-07-01 | 2019-10-01 | 中国医学科学院医学信息研究所 | Diagnosis and treatment program prediction method and device |
| CN110569349A (en) * | 2019-07-30 | 2019-12-13 | 平安科技（深圳）有限公司 | Big data-based method, system, equipment and storage medium for pushing articles for education |
| CN110851725A (en) * | 2019-11-15 | 2020-02-28 | 北京盛诺一家医院管理咨询有限公司 | Medical resource recommendation method, system, computing device and storage medium |
| CN110911012A (en) * | 2019-12-05 | 2020-03-24 | 中国科学院自动化研究所 | Method and system for determining personalized diagnosis and treatment method based on utility model |
| CN110930705A (en) * | 2019-11-28 | 2020-03-27 | 北京易华录信息技术股份有限公司 | Intersection traffic decision system, method and equipment |
| CN110942232A (en) * | 2019-11-01 | 2020-03-31 | 支付宝(杭州)信息技术有限公司 | Model rapid disaster tolerance method and device and electronic equipment |
| CN110957046A (en) * | 2019-11-15 | 2020-04-03 | 合肥工业大学 | Medical health case knowledge matching method and system |
| CN110970109A (en) * | 2019-11-14 | 2020-04-07 | 厦门中翎易优创科技有限公司 | Deep learning-based orthopaedic postoperative rehabilitation scheme recommendation method |
| CN111091907A (en) * | 2019-11-15 | 2020-05-01 | 合肥工业大学 | Health medical knowledge retrieval method and system based on similar case library |
| CN111259111A (en) * | 2020-01-13 | 2020-06-09 | 安徽科大讯飞医疗信息技术有限公司 | Medical record-based decision-making assisting method and device, electronic equipment and storage medium |
| CN111714090A (en) * | 2020-06-19 | 2020-09-29 | 深圳市恬境科技有限公司 | Human body biological rhythm health management method and system |
| CN111755090A (en) * | 2020-06-24 | 2020-10-09 | 南京江北新区生物医药公共服务平台有限公司 | Medical record searching method, medical record searching device, storage medium and electronic equipment |
| CN111798973A (en) * | 2020-07-08 | 2020-10-20 | 广元量知汇科技有限公司 | Intelligent health terminal management system |
| CN112365988A (en) * | 2020-11-10 | 2021-02-12 | 杭州市肿瘤医院 | Prognosis prediction system |
| CN112394924A (en) * | 2019-08-16 | 2021-02-23 | 北京大数医达科技有限公司 | Method, apparatus, electronic device, and medium for generating question model |
| CN112397161A (en) * | 2019-08-16 | 2021-02-23 | 北京大数医达科技有限公司 | Method, apparatus, device and medium for generating an assay model |
| CN112509685A (en) * | 2020-12-31 | 2021-03-16 | 曜立科技(北京)有限公司 | Standardization method and system for screening operation modes |
| CN112562822A (en) * | 2020-12-17 | 2021-03-26 | 南京王师大数据有限公司 | Method and system for determining traditional Chinese medicine menstruation prescription |
| WO2021073277A1 (en) * | 2019-10-16 | 2021-04-22 | 平安科技（深圳）有限公司 | Personalized precise medication recommendation method and apparatus |
| CN113096795A (en) * | 2019-12-23 | 2021-07-09 | 四川医枢科技股份有限公司 | Multi-source data-aided clinical decision support system and method |
| WO2021232591A1 (en) * | 2020-05-22 | 2021-11-25 | 平安国际智慧城市科技股份有限公司 | User data classification method and apparatus based on internet of things data, computer device and storage medium |
| CN113782185A (en) * | 2021-08-25 | 2021-12-10 | 上海市儿童医院 | Clinical diagnosis and treatment intelligent decision analysis system and method |
| CN114491177A (en) * | 2022-02-15 | 2022-05-13 | 北京百度网讯科技有限公司 | Information determination method, model training method, model determination device and electronic equipment |
| CN114999599A (en) * | 2022-05-26 | 2022-09-02 | 中国科学技术信息研究所 | Data processing method and device, electronic equipment and computer readable storage medium |
| CN115148330A (en) * | 2022-05-24 | 2022-10-04 | 中国医学科学院北京协和医院 | POP treatment scheme forming method and system |
| CN115910323A (en) * | 2022-11-16 | 2023-04-04 | 山东仲雅信息技术有限公司 | Treatment scheme selection method, device, equipment and storage medium |
| CN116246749A (en) * | 2023-05-11 | 2023-06-09 | 西南医科大学附属医院 | Endocrine patient personalized health management system integrating electronic medical records |
| CN117059231A (en) * | 2023-10-10 | 2023-11-14 | 首都医科大学附属北京友谊医院 | Method for machine learning of traditional Chinese medicine cases and intelligent diagnosis and treatment system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016101351A1 (en) * | 2014-12-26 | 2016-06-30 | 深圳市前海安测信息技术有限公司 | Network hospital based general practitioner auxiliary diagnosis and treatment system and method |
| CN105809473A (en) * | 2016-02-29 | 2016-07-27 | 北京百度网讯科技有限公司 | Training method, service recommending method for coupling model parameters and associated apparatus |
| CN105912881A (en) * | 2016-06-14 | 2016-08-31 | 华南师范大学 | Individual-based treatment scheme recommendation system based on big data knowledge base |
| CN106778042A (en) * | 2017-01-26 | 2017-05-31 | 中电科软件信息服务有限公司 | Cardio-cerebral vascular disease patient similarity analysis method and system |
| CN107833629A (en) * | 2017-10-25 | 2018-03-23 | 厦门大学 | Aided diagnosis method and system based on deep learning |
| CN108417272A (en) * | 2018-02-08 | 2018-08-17 | 合肥工业大学 | Similar case with temporal constraint recommends method and device |
-
2018
- 2018-12-25 CN CN201811592838.9A patent/CN109785928B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016101351A1 (en) * | 2014-12-26 | 2016-06-30 | 深圳市前海安测信息技术有限公司 | Network hospital based general practitioner auxiliary diagnosis and treatment system and method |
| CN105809473A (en) * | 2016-02-29 | 2016-07-27 | 北京百度网讯科技有限公司 | Training method, service recommending method for coupling model parameters and associated apparatus |
| CN105912881A (en) * | 2016-06-14 | 2016-08-31 | 华南师范大学 | Individual-based treatment scheme recommendation system based on big data knowledge base |
| CN106778042A (en) * | 2017-01-26 | 2017-05-31 | 中电科软件信息服务有限公司 | Cardio-cerebral vascular disease patient similarity analysis method and system |
| CN107833629A (en) * | 2017-10-25 | 2018-03-23 | 厦门大学 | Aided diagnosis method and system based on deep learning |
| CN108417272A (en) * | 2018-02-08 | 2018-08-17 | 合肥工业大学 | Similar case with temporal constraint recommends method and device |
Non-Patent Citations (1)
| Title |
|---|
| 郑金龙;穆大力;解芳;史亚坤;宫辉;李昊;连九峥;王越;朱萧;党军红;: "电子病历临床决策系统的建立与应用", 中国病案, no. 10, 18 October 2017 (2017-10-18), pages 45 - 47 * |
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110297908A (en) * | 2019-07-01 | 2019-10-01 | 中国医学科学院医学信息研究所 | Diagnosis and treatment program prediction method and device |
| CN110569349B (en) * | 2019-07-30 | 2023-06-02 | 平安科技（深圳）有限公司 | Method, system, equipment and storage medium for pushing ill teaching article based on big data |
| CN110569349A (en) * | 2019-07-30 | 2019-12-13 | 平安科技（深圳）有限公司 | Big data-based method, system, equipment and storage medium for pushing articles for education |
| CN112397161A (en) * | 2019-08-16 | 2021-02-23 | 北京大数医达科技有限公司 | Method, apparatus, device and medium for generating an assay model |
| CN112394924A (en) * | 2019-08-16 | 2021-02-23 | 北京大数医达科技有限公司 | Method, apparatus, electronic device, and medium for generating question model |
| CN112394924B (en) * | 2019-08-16 | 2024-06-07 | 北京大数医达科技有限公司 | Method, device, electronic equipment and medium for generating questioning model |
| WO2021073277A1 (en) * | 2019-10-16 | 2021-04-22 | 平安科技（深圳）有限公司 | Personalized precise medication recommendation method and apparatus |
| CN110942232B (en) * | 2019-11-01 | 2022-07-19 | 支付宝(杭州)信息技术有限公司 | Model rapid disaster tolerance method and device and electronic equipment |
| CN110942232A (en) * | 2019-11-01 | 2020-03-31 | 支付宝(杭州)信息技术有限公司 | Model rapid disaster tolerance method and device and electronic equipment |
| CN110970109A (en) * | 2019-11-14 | 2020-04-07 | 厦门中翎易优创科技有限公司 | Deep learning-based orthopaedic postoperative rehabilitation scheme recommendation method |
| CN111091907A (en) * | 2019-11-15 | 2020-05-01 | 合肥工业大学 | Health medical knowledge retrieval method and system based on similar case library |
| CN110851725B (en) * | 2019-11-15 | 2023-10-27 | 北京盛诺一家医院管理咨询有限公司 | Medical resource recommendation method, system, computing device and storage medium |
| CN110957046B (en) * | 2019-11-15 | 2024-01-19 | 合肥工业大学 | Medical health case knowledge matching method and system |
| CN110957046A (en) * | 2019-11-15 | 2020-04-03 | 合肥工业大学 | Medical health case knowledge matching method and system |
| CN110851725A (en) * | 2019-11-15 | 2020-02-28 | 北京盛诺一家医院管理咨询有限公司 | Medical resource recommendation method, system, computing device and storage medium |
| CN110930705B (en) * | 2019-11-28 | 2020-10-27 | 北京易华录信息技术股份有限公司 | Intersection traffic decision system, method and equipment |
| CN110930705A (en) * | 2019-11-28 | 2020-03-27 | 北京易华录信息技术股份有限公司 | Intersection traffic decision system, method and equipment |
| CN110911012B (en) * | 2019-12-05 | 2022-10-14 | 中国科学院自动化研究所 | Method and system for determining personalized diagnosis and treatment method based on utility model |
| CN110911012A (en) * | 2019-12-05 | 2020-03-24 | 中国科学院自动化研究所 | Method and system for determining personalized diagnosis and treatment method based on utility model |
| CN113096795A (en) * | 2019-12-23 | 2021-07-09 | 四川医枢科技股份有限公司 | Multi-source data-aided clinical decision support system and method |
| CN113096795B (en) * | 2019-12-23 | 2023-10-03 | 四川医枢科技有限责任公司 | Multi-source data-aided clinical decision support system and method |
| CN111259111A (en) * | 2020-01-13 | 2020-06-09 | 安徽科大讯飞医疗信息技术有限公司 | Medical record-based decision-making assisting method and device, electronic equipment and storage medium |
| CN111259111B (en) * | 2020-01-13 | 2022-02-18 | 安徽科大讯飞医疗信息技术有限公司 | Medical record-based decision-making assisting method and device, electronic equipment and storage medium |
| WO2021232591A1 (en) * | 2020-05-22 | 2021-11-25 | 平安国际智慧城市科技股份有限公司 | User data classification method and apparatus based on internet of things data, computer device and storage medium |
| CN111714090B (en) * | 2020-06-19 | 2023-06-20 | 深圳市恬境健康管理有限公司 | Human biological rhythm health management method and system |
| CN111714090A (en) * | 2020-06-19 | 2020-09-29 | 深圳市恬境科技有限公司 | Human body biological rhythm health management method and system |
| CN111755090A (en) * | 2020-06-24 | 2020-10-09 | 南京江北新区生物医药公共服务平台有限公司 | Medical record searching method, medical record searching device, storage medium and electronic equipment |
| CN111798973A (en) * | 2020-07-08 | 2020-10-20 | 广元量知汇科技有限公司 | Intelligent health terminal management system |
| CN112365988A (en) * | 2020-11-10 | 2021-02-12 | 杭州市肿瘤医院 | Prognosis prediction system |
| CN112365988B (en) * | 2020-11-10 | 2023-08-04 | 杭州市肿瘤医院 | Prognosis prediction system |
| CN112562822A (en) * | 2020-12-17 | 2021-03-26 | 南京王师大数据有限公司 | Method and system for determining traditional Chinese medicine menstruation prescription |
| CN112509685A (en) * | 2020-12-31 | 2021-03-16 | 曜立科技(北京)有限公司 | Standardization method and system for screening operation modes |
| CN113782185B (en) * | 2021-08-25 | 2024-03-08 | 上海市儿童医院 | Intelligent decision analysis system and method for clinical diagnosis and treatment |
| CN113782185A (en) * | 2021-08-25 | 2021-12-10 | 上海市儿童医院 | Clinical diagnosis and treatment intelligent decision analysis system and method |
| CN114491177A (en) * | 2022-02-15 | 2022-05-13 | 北京百度网讯科技有限公司 | Information determination method, model training method, model determination device and electronic equipment |
| CN115148330B (en) * | 2022-05-24 | 2023-07-25 | 中国医学科学院北京协和医院 | POP treatment scheme forming method and system |
| CN115148330A (en) * | 2022-05-24 | 2022-10-04 | 中国医学科学院北京协和医院 | POP treatment scheme forming method and system |
| CN114999599A (en) * | 2022-05-26 | 2022-09-02 | 中国科学技术信息研究所 | Data processing method and device, electronic equipment and computer readable storage medium |
| CN115910323A (en) * | 2022-11-16 | 2023-04-04 | 山东仲雅信息技术有限公司 | Treatment scheme selection method, device, equipment and storage medium |
| CN115910323B (en) * | 2022-11-16 | 2023-09-29 | 山东仲雅信息技术有限公司 | Treatment scheme selection method, device, equipment and storage medium |
| CN116246749A (en) * | 2023-05-11 | 2023-06-09 | 西南医科大学附属医院 | Endocrine patient personalized health management system integrating electronic medical records |
| CN117059231A (en) * | 2023-10-10 | 2023-11-14 | 首都医科大学附属北京友谊医院 | Method for machine learning of traditional Chinese medicine cases and intelligent diagnosis and treatment system |
| CN117059231B (en) * | 2023-10-10 | 2023-12-22 | 首都医科大学附属北京友谊医院 | Method for machine learning of traditional Chinese medicine cases and intelligent diagnosis and treatment system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109785928B (en) | 2024-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785928A (en) | Diagnosis and treatment proposal recommending method, device and storage medium | |
| EP3292492B1 (en) | Predicting likelihoods of conditions being satisfied using recurrent neural networks | |
| CN107851462B (en) | Analyzing health events using a recurrent neural network | |
| Guan | A novel method of plant leaf disease detection based on deep learning and convolutional neural network | |
| CN109902222A (en) | Recommendation method and device | |
| CN107995992A (en) | Health event is analyzed using recurrent neural network | |
| Vistro et al. | The cricket winner prediction with application of machine learning and data analytics | |
| CN108648020A (en) | User behavior quantization method, system, equipment and storage medium | |
| Cho et al. | Using social network analysis and gradient boosting to develop a soccer win–lose prediction model | |
| CN111785366B (en) | Patient treatment scheme determination method and device and computer equipment | |
| CN109525384A (en) | The DPA attack method and system, terminal being fitted using neural network | |
| CN107958708A (en) | Risk trend appraisal procedure and system after institute | |
| CN109259784A (en) | AI prediction technique, device, equipment and the storage medium of cerebral infarction | |
| CN108038413A (en) | Cheat probability analysis method, apparatus and storage medium | |
| JP7191443B2 (en) | Target object attribute prediction method based on machine learning, related equipment and computer program | |
| CN106529110A (en) | Classification method and equipment of user data | |
| CN106096286A (en) | Clinical path formulating method and device | |
| Andersen et al. | The dreaming variational autoencoder for reinforcement learning environments | |
| CN113705092B (en) | Disease prediction method and device based on machine learning | |
| CN113990454A (en) | Malicious behavior identification method based on federal learning and feature extraction | |
| CN113345564B (en) | Early prediction method and device for patient hospitalization duration based on graph neural network | |
| CN110287938A (en) | Event recognition method, system, equipment and medium based on critical segment detection | |
| CN117034142B (en) | Unbalanced medical data missing value filling method and system | |
| CN116469534A (en) | Hospital number calling management system and method thereof | |
| CN109934352A (en) | The automatic evolvement method of model of mind |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |