CN109872812A - A kind of fititious doctor diagnostic system and method based on convolutional neural networks - Google Patents
A kind of fititious doctor diagnostic system and method based on convolutional neural networks Download PDFInfo
- Publication number
- CN109872812A CN109872812A CN201711269723.1A CN201711269723A CN109872812A CN 109872812 A CN109872812 A CN 109872812A CN 201711269723 A CN201711269723 A CN 201711269723A CN 109872812 A CN109872812 A CN 109872812A
- Authority
- CN
- China
- Prior art keywords
- doctor
- patient
- fititious
- diagnostic
- convolutional neural
- 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
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The present invention relates to technical field of medical equipment, in particular to a kind of fititious doctor diagnostic system and method based on convolutional neural networks.This method are as follows: obtain the registration information of patient, select hospital, the information of fititious doctor individual and the condition symptoms information of patient's input to be checked, and using the fititious doctor individual of patient's selection, tentative diagnosis is carried out according to condition symptoms, obtain tentative diagnosis result, wherein, hospital to be checked is used to carry out physical examination to patient；When receiving patient body inspection data, using the fititious doctor individual of patient's selection, secondary diagnosis is carried out according to the physical examination data of the patient received, obtains secondary diagnostic result；Corresponding electronic prescription and/or electronic health record are generated according to tentative diagnosis result and/or secondary diagnostic result.Using the above method, human simulation diagnosis process is realized, provides consulting services by online mode for patient, medical patient is alleviated and registers difficulty, is lined up the problem of consultation time goes out.
Description
Technical field
The present invention relates to technical field of medical equipment, in particular to a kind of fititious doctor diagnosis based on convolutional neural networks
System and method.
Background technique
Since medical resource is unevenly distributed, the Medical Devices of infirmary, medical environment and medical technology etc., with large hospital
Between gap it is larger, therefore, mostly to go large hospital to stand in the queue to register medical for selection after sick by people, in this way, resulting in curing greatly
Institute's congestion, patient register difficulty, and queuing of patients's consultation time is long.
Currently, also there are some for detecting human health backmans as medical expert's progress medical diagnosis on disease
The mycin system of tool, by the thought process of simulative medicine expert diagnosis disease, by the diagnostic experiences of medical expert and
Knowledge is stored in computer in the form of rules, and establishes corresponding rule base, and the mode of symbolization reasoning carries out medical treatment and examines
It is disconnected.
However, the limitation of the Knowledge Representation Schemes due to symbol, the complexity of some difficult and complicated illness be difficult with rule come
It is described, or even is difficult to be carried out with simple language to express, and rule-based mycin system, as rule base is advised
The continuous increase of mould, it is also more and more by the combination between each rule, due to containing a large amount of nothings during cycle of inference
The rule combination of effect is attempted, then, Reasoning Efficiency will reduce.That is, existing mycin system cannot complete generation
Medical diagnosis on disease is completed for doctor, cannot still alleviate medical patient and register the long problem of difficulty and medical queuing of patients's consultation time.
In view of this, needing to design a kind of fititious doctor diagnostic system and method based on convolutional neural networks, to make up
Defect and shortcoming existing in the prior art.
Summary of the invention
The purpose of the embodiment of the present invention is that a kind of fititious doctor diagnostic system and method based on convolutional neural networks is provided,
It registers the long problem of difficulty and medical queuing of patients's consultation time to solve medical patient existing in the prior art.
The specific technical solution provided in the embodiment of the present invention is as follows:
A kind of fititious doctor diagnostic system based on convolutional neural networks, the fititious doctor diagnostic system, which includes at least, to be used
Receive module, fititious doctor module and electronic prescription and electronic health record generation module in family, wherein
The user receives module, for completing the registration of patient, patient hospital to be checked, fititious doctor
The input of the condition symptoms of the selection of body and the patient, wherein the hospital to be checked is used to carry out body inspection to patient
It looks into；
The fititious doctor module, it is built-up by convolutional neural networks, it include the fititious doctor that multiple training are completed
Individual carries out tentative diagnosis according to the condition symptoms for the fititious doctor individual using patient selection, obtains preliminary
Diagnostic result, and using the fititious doctor individual, secondary examine is carried out according to the physical examination data of the patient received
It is disconnected, obtain secondary diagnostic result；
The electron process and electronic health record generation module, for according to the tentative diagnosis result and/or secondary diagnosis
As a result corresponding electronic prescription and/or electronic health record are generated.
Preferably, using the patient selection fititious doctor individual, according to the condition symptoms to the patient into
Before row tentative diagnosis, the fititious doctor module is further used for:
Each doctor's historical diagnostic process sample data is obtained respectively；
Sample training is carried out to the convolutional neural networks respectively according to each doctor's historical diagnostic process sample data,
Form corresponding each fititious doctor individual.
Preferably, carrying out sample to the convolutional neural networks respectively according to each doctor's historical diagnostic process sample data
This training, when forming corresponding each fititious doctor individual, the fititious doctor module is specifically used for:
Place is trained in the historical diagnostic process sample data input convolutional neural networks for the doctor that will acquire
Reason, obtains real processing results；
If it is determined that when the difference between real processing results and ideal process result is not less than given threshold, according to minimization
Error approach reversely adjusts weight matrix in convolutional neural networks, until the difference between real processing results and ideal process result
Not little Yu given threshold, it is individual to form corresponding with one doctor fititious doctor.
Preferably, the convolutional neural networks include input layer, and at least one convolutional layer, at least one sample level, at least
One full articulamentum and classifier.
Preferably, the doctor that will acquire historical diagnostic process sample data input convolutional neural networks in into
Row training managing, when obtaining real processing results, the fititious doctor module is specifically used for:
Process of convolution at least once is carried out using historical diagnostic process sample data of the convolution Boltzmann machine to input, is mentioned
Corresponding characteristic value is taken, and carries out pondization operation and sparse regularization operation for the characteristic value obtained after each process of convolution,
Obtain corresponding diagnostic result.
A kind of fititious doctor diagnostic method based on convolutional neural networks, comprising:
The registration information for obtaining patient selects hospital to be checked, the information of fititious doctor individual and the patient defeated
The condition symptoms information entered, and using the fititious doctor individual of patient selection, it is tentatively examined according to the condition symptoms
It is disconnected, obtain tentative diagnosis result, wherein the hospital to be checked is used to carry out physical examination to patient；
When receiving the patient body and checking data, using the fititious doctor individual of patient selection, according to connecing
The physical examination data of the patient received carry out secondary diagnosis, obtain secondary diagnostic result；
Corresponding electronic prescription and/or electronic health record are generated according to the tentative diagnosis result and/or secondary diagnostic result.
Preferably, when receiving the illness symptom of patient's input, using the fititious doctor individual of patient selection, root
Tentative diagnosis is carried out in equity quilt to the patient according to the condition symptoms, further comprises:
Each doctor's historical diagnostic process sample data is obtained respectively；
Sample training is carried out to convolutional neural networks respectively according to each doctor's historical diagnostic process sample data, is formed
Corresponding each fititious doctor individual.
Preferably, carrying out sample instruction to convolutional neural networks respectively according to each doctor's historical diagnostic process sample data
Practice, form corresponding each fititious doctor individual, specifically include:
Place is trained in the historical diagnostic process sample data input convolutional neural networks for the doctor that will acquire
Reason, obtains real processing results；
If it is determined that when the difference between real processing results and ideal process result is not less than given threshold, according to minimization
Error approach reversely adjusts weight matrix in convolutional neural networks, until the difference between real processing results and ideal process result
Not little Yu given threshold, it is individual to form corresponding with one doctor fititious doctor.
Preferably, the convolutional neural networks include input layer, and at least one convolutional layer, at least one sample level, at least
One full articulamentum and classifier.
Preferably, being carried out in the historical diagnostic process sample data input convolutional neural networks for the doctor that will acquire
Training managing obtains real processing results, specifically includes:
Process of convolution at least once is carried out using historical diagnostic process sample data of the convolution Boltzmann machine to input, is mentioned
Corresponding characteristic value is taken, and carries out pondization operation and sparse regularization operation for the characteristic value obtained after each process of convolution,
Obtain corresponding diagnostic result.
The present invention has the beneficial effect that:
In conclusion during being diagnosed using fititious doctor diagnostic system, obtaining and suffering from the embodiment of the present invention
The registration information of person selects hospital, the information of fititious doctor individual and the condition symptoms letter of patient input to be checked
Breath, and using the fititious doctor individual of patient selection, tentative diagnosis is carried out according to the condition symptoms, obtains tentative diagnosis
As a result, wherein the hospital to be checked is used to carry out physical examination to patient；Data are checked receiving the patient body
When, using the fititious doctor individual of patient selection, carried out according to the physical examination data of the patient received secondary
Diagnosis, obtains secondary diagnostic result；Corresponding electronic prescription is generated according to the tentative diagnosis result and/or secondary diagnostic result
And/or electronic health record.
Using the above method, sample training is carried out to convolutional neural networks by doctor's history diagnosis and treatment process sample data,
Human simulation diagnosis process is realized, diagnosis and treatment flow services is provided by online mode for patient, is effectively relieved
Medical patient registers difficulty at present, is lined up the problem of consultation time goes out.
Detailed description of the invention
Fig. 1 is a kind of fititious doctor diagnostic system structural schematic diagram based on convolutional neural networks in the embodiment of the present invention；
Fig. 2 is the detailed stream of the progress data processing of convolutional neural networks in fititious doctor module in the embodiment of the present invention
Cheng Tu；
Fig. 3 is a kind of detailed process of the fititious doctor diagnostic method based on convolutional neural networks in the embodiment of the present invention
Figure；
Fig. 4 is in the embodiment of the present invention, and the diagnostic process of the fititious doctor diagnostic system based on convolutional neural networks is illustrated
Figure.
Specific embodiment
Register that difficulty and medical queuing of patients's consultation time are long to ask to solve medical patient existing in the prior art
It inscribes, a kind of new fititious doctor diagnostic system and method based on convolutional neural networks, the party is provided in the embodiment of the present invention
Method are as follows: the registration information for obtaining patient selects hospital to be checked, the information of fititious doctor individual and patient input
Condition symptoms information, and using the fititious doctor individual of patient selection, tentative diagnosis is carried out according to the condition symptoms, is obtained
To tentative diagnosis result, wherein the hospital to be checked is used to carry out physical examination to patient；Receiving the patient body
When checking data, using the fititious doctor individual of patient selection, according to the physical examination data of the patient received
Secondary diagnosis is carried out, secondary diagnostic result is obtained；It is generated according to the tentative diagnosis result and/or secondary diagnostic result corresponding
Electronic prescription and/or electronic health record.
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, is not whole embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
As shown in fig.1, the fititious doctor diagnostic system based on convolutional neural networks includes at least in the embodiment of the present invention
User receives module, fititious doctor module and electronic prescription and electronic health record generation module, wherein
Above-mentioned user receives module, for completing the registration of patient, above-mentioned patient hospital to be checked, fititious doctor
The input of the condition symptoms of the selection of body and above-mentioned patient, wherein above-mentioned hospital to be checked is used to carry out body inspection to patient
It looks into；
Above-mentioned fititious doctor module, it is built-up by convolutional neural networks, it include multiple virtual doctors for having trained and having completed
Raw individual carries out tentative diagnosis according to above-mentioned condition symptoms, obtains just for the fititious doctor individual using above-mentioned patient selection
Diagnostic result is walked, and using above-mentioned fititious doctor individual, is carried out according to the physical examination data of the above-mentioned patient received secondary
Diagnosis, obtains secondary diagnostic result；
Above-mentioned electron process and electronic health record generation module, for according to above-mentioned tentative diagnosis result and/or secondary diagnosis
As a result corresponding electronic prescription and/or electronic health record are generated.
In practical application, before carrying out medical diagnosis on disease using the fititious doctor diagnostic system based on convolutional neural networks,
It needs to carry out the operation of fititious doctor individual training simulation for the fititious doctor module in the fititious doctor diagnostic system.
In the embodiment of the present invention, fititious doctor individual is carried out for the fititious doctor module in the fititious doctor diagnostic system
The process of training simulation operation are as follows: obtain each doctor's historical diagnostic process sample data respectively, and according to each doctor's history
Diagnosis process sample data carries out sample training to the convolutional neural networks respectively, forms corresponding each fititious doctor individual.
Specifically, the historical diagnostic process sample data for each doctor executes following operation respectively: by a doctor
Historical diagnostic process sample data as training sample, by the convolutional Neural in a training sample input fititious doctor module
In network, simulated training operation is carried out, actual processing result is obtained, wherein the real processing results are fititious doctor module
According to the condition symptoms of the patient in said one training sample, the state of an illness information of the patient judged, and by actual treatment knot
Fruit is compared with the ideal process result for characterizing the true state of an illness information of patient, judges real processing results and ideal process
Whether the difference between as a result is not less than given threshold, if it is determined that the difference between real processing results and ideal process result is not
When less than given threshold, weight matrix in convolutional neural networks is reversely adjusted according to minimization error approach, then using adjustment
The convolutional neural networks of weight matrix carry out simulated training to subsequent training sample, until at real processing results and ideal
The difference managed between result is less than given threshold, after largely training, completes virtual doctor corresponding with said one doctor
The simulated training of raw individual.
It is possible to which different virtual doctors is respectively trained out according to the historical diagnostic process sample data of different doctors
Raw individual, i.e., different patients can select to be suitble to oneself according to their needs after receiving module by user and completing registration
The fititious doctor individual of condition symptoms.
In the embodiment of the present invention, fititious doctor module is built-up by improved convolutional neural networks, the convolutional Neural net
Network includes input layer, at least one convolutional layer, at least one sample level, at least one full articulamentum and classifier.Wherein, training
Sample inputs convolutional neural networks by input layer, by the convolution Boltzmann machine in convolutional layer to the training sample of input network
This progress characteristics extraction, and pondization operation and sparse regularization operation are carried out to the characteristic value extracted, to obtain corresponding
Diagnostic result.
For example, as shown in fig.2, in fititious doctor module convolutional neural networks flow chart of data processing figure.By illness disease
Shape and/or physical examination data input convolutional neural networks, by the input layer in convolutional neural networks, convolutional layer, and sample level,
After full articulamentum and classifier processing, diagnostic result is exported.Specifically, by condition symptoms and/or physical examination data through pulleying
Product Boltzmann machine carries out characteristics extraction, and in C1 layers of generation Feature Mapping, characteristic value is set through summation, weighted value, biasing, passed through
S2 layers of characteristic value are obtained after the processing of Sigmoid function, characteristics extraction is being carried out by convolution Boltzmann machine, is obtaining C3 layers of spy
Value indicative, then set by characteristic value summation, weighted value, biasing, S4 layers of characteristic value are obtained after handling by Sigmoid function, again
After convolution Boltzmann machine carries out characteristics extraction, C5 layers of characteristic value are obtained, through dot product, biasing, the processing of Sigmoid function
Afterwards, F6 layers of characteristic value are obtained, after handling finally by European radial basis function, obtain diagnostic result.
The solution of the present invention will be described in detail by specific embodiment below, certainly, the present invention is not limited to
Lower embodiment.
As shown in fig.3, in the embodiment of the present invention, a kind of fititious doctor diagnostic method based on convolutional neural networks it is detailed
Thread journey is as follows:
Step 300: obtain the registration information of patient, select hospital to be checked, fititious doctor individual information and on
State patient input condition symptoms information, and using above-mentioned patient selection fititious doctor individual, according to above-mentioned condition symptoms into
Row tentative diagnosis obtains tentative diagnosis result, wherein above-mentioned hospital to be checked is used to carry out physical examination to patient.
In practical application, user completes the registration of medical patient information by logging in fititious doctor diagnostic system, and
Selection can carry out hospital to be checked (e.g., the doctor nearest from medical patient location of physical examination to medical patient accordingly
Institute), and select to be suitble to go to a doctor from the fititious doctor module of fititious doctor diagnostic system according to the condition symptoms of medical patient
The fititious doctor individual of patient condition's symptom is completed to register, and hospital to be checked can lead to after the selection of fititious doctor individual
The condition symptoms for inputting medical patient are crossed, so that fititious doctor individual is according to the condition symptoms of the medical patient of the input to patient
Tentative diagnosis is carried out, and obtains corresponding tentative diagnosis result.
For example, condition symptoms are subjected to characteristics extraction by convolution Boltzmann machine, it is special in C1 layers of generation Feature Mapping
Value indicative is set through summation, weighted value, biasing, obtains S2 layers of characteristic value after handling by Sigmoid function, is passing through convolution Bohr hereby
Graceful machine carries out characteristics extraction, obtains C3 layers of characteristic value, then set by characteristic value summation, weighted value, biasing, passes through Sigmoid
S4 layers of characteristic value are obtained after function processing, after again passing by convolution Boltzmann machine progress characteristics extraction, obtain C5 layers of feature
Value obtains F6 layers of characteristic value after dot product, biasing, the processing of Sigmoid function, after being handled finally by European radial basis function,
Obtain tentative diagnosis result.
Step 310: when receiving above-mentioned patient body inspection data, using the fititious doctor of above-mentioned patient selection
Body carries out secondary diagnosis according to the physical examination data of the above-mentioned patient received, obtains secondary diagnostic result.
In practical application, the patient that goes to a doctor completes online registration and hospital to be checked and fititious doctor individual
After selection and the input of condition symptoms information, medical patient needs hospital to be detected to carry out corresponding physical examination, to
After completing physical examination, physical examination data can be transmitted to fititious doctor diagnostic system, fititious doctor by physical examination equipment
Fititious doctor module in diagnostic system is in the physical examination data for receiving medical patient, using the void of medical patient selection
Quasi- doctor's individual carries out secondary diagnosis to medical patient according to the physical examination data received, obtains final diagnostic result.
For example, physical examination data are carried out characteristics extraction by convolution Boltzmann machine, reflected in C1 layers of generation feature
It penetrates, characteristic value is set through summation, weighted value, biasing, obtains S2 layers of characteristic value after handling by Sigmoid function, is passing through convolution
Boltzmann machine carries out characteristics extraction, obtains C3 layers of characteristic value, then set by characteristic value summation, weighted value, biasing, passes through
S4 layers of characteristic value are obtained after the processing of Sigmoid function, after again passing by convolution Boltzmann machine progress characteristics extraction, obtain C5
Layer characteristic value obtains F6 layers of characteristic value, finally by European radial basis function after dot product, biasing, the processing of Sigmoid function
After processing, last diagnostic result is obtained.
Step 320: according to above-mentioned tentative diagnosis result and/or secondary diagnostic result generate corresponding electronic prescription and/or
Electronic health record.
Specifically, diagnostic result is sent in fititious doctor diagnostic system by fititious doctor individual after completing diagnosis
Electronic prescription and electronic health record generation module in, examined with triggering electronic prescription with electronic health record generation module according to what is received
Break as a result, generating corresponding electronic prescription and/or generating electronic health record according to treatment process.Patient can use according to the electronic prescription
Medicine etc. is taken to operate in subsequent, patient can check the electronic health record, and when next time is medical, can directly transfer the electronic health record.
Above-described embodiment is described in further detail using specific application scenarios below, as shown in fig.4, of the invention
In embodiment, the diagnostic process schematic diagram of the fititious doctor diagnostic system based on convolutional neural networks.
Specifically, firstly, fititious doctor module previously according to doctor's diagnosis and treatment process sample data, to convolutional neural networks into
(pondization operation and sparse regularization operation is added, to improve training effectiveness) in row sample training during sample training, formed more
A fititious doctor individual；
Then, the patient that goes to a doctor by user receives module progress online registration registration, selects hospital to be checked, fititious doctor
Individual describes condition symptoms, and the fititious doctor individual in the fititious doctor module of patient's selection of going to a doctor is according to the illness disease of description
Shape carries out the online tentative diagnosis of network to patient, obtains tentative diagnosis result；
Then, medical patient goes to the hospital to be checked of selection to carry out related physical inspection, and physical examination data are transmitted to
Fititious doctor diagnostic system, the fititious doctor individual in fititious doctor module that medical patient selects are gone to a doctor according to what is received
The physical examination data of patient carry out network secondary diagnosis online to medical patient, obtain last diagnostic result；
Finally, the electronic prescription and electronic health record generation module in fititious doctor diagnostic system are according to initial diagnostic result
And/or last diagnostic result generates corresponding electronic prescription and/or electronic health record.
In conclusion during being diagnosed using fititious doctor diagnostic system, obtaining and suffering from the embodiment of the present invention
The registration information of person selects hospital, the information of fititious doctor individual and the condition symptoms letter of patient input to be checked
Breath, and using the fititious doctor individual of patient selection, tentative diagnosis is carried out according to the condition symptoms, obtains tentative diagnosis
As a result, wherein the hospital to be checked is used to carry out physical examination to patient；Data are checked receiving the patient body
When, using the fititious doctor individual of patient selection, carried out according to the physical examination data of the patient received secondary
Diagnosis, obtains secondary diagnostic result；Corresponding electronic prescription is generated according to the tentative diagnosis result and/or secondary diagnostic result
And/or electronic health record.
Using the above method, sample training is carried out to convolutional neural networks by doctor's history diagnosis and treatment process sample data,
Human simulation diagnosis process is realized, diagnosis and treatment flow services is provided by online mode for patient, is effectively relieved
Medical patient registers difficulty at present, is lined up the problem of consultation time goes out.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, those skilled in the art can carry out various modification and variations without departing from this hair to the embodiment of the present invention
The spirit and scope of bright embodiment.In this way, if these modifications and variations of the embodiment of the present invention belong to the claims in the present invention
And its within the scope of equivalent technologies, then the present invention is also intended to include these modifications and variations.
Claims (10)
1. a kind of fititious doctor diagnostic system based on convolutional neural networks, which is characterized in that the fititious doctor diagnostic system
Module, fititious doctor module and electronic prescription and electronic health record generation module are received including at least user, wherein
The user receives module, for completing the registration of patient, patient hospital to be checked, fititious doctor individual
The input of the condition symptoms of selection and the patient, wherein the hospital to be checked is used to carry out physical examination to patient；
The fititious doctor module, it is built-up by convolutional neural networks, it include the fititious doctor individual that multiple training are completed,
For the fititious doctor individual using patient selection, tentative diagnosis is carried out according to the condition symptoms, obtains tentative diagnosis
As a result, and being obtained using the fititious doctor individual according to the secondary diagnosis of the physical examination data of the patient received progress
To secondary diagnostic result；
The electron process and electronic health record generation module, for according to the tentative diagnosis result and/or secondary diagnostic result
Generate corresponding electronic prescription and/or electronic health record.
2. fititious doctor diagnostic system as described in claim 1, which is characterized in that in the virtual doctor using patient selection
Raw individual, before carrying out tentative diagnosis to the patient according to the condition symptoms, the fititious doctor module is further used for:
Each doctor's historical diagnostic process sample data is obtained respectively；
Sample training is carried out to the convolutional neural networks respectively according to each doctor's historical diagnostic process sample data, is formed
Corresponding each fititious doctor individual.
3. fititious doctor diagnostic system as claimed in claim 2, which is characterized in that according to each doctor's historical diagnostic process
Sample data carries out sample training to the convolutional neural networks respectively, when forming corresponding each fititious doctor individual, the void
Quasi- client module is specifically used for:
It is trained processing in the historical diagnostic process sample data input convolutional neural networks for the doctor that will acquire, is obtained
To real processing results；
If it is determined that when the difference between real processing results and ideal process result is not less than given threshold, according to minimization error
Method reversely adjusts weight matrix in convolutional neural networks, until the difference between real processing results and ideal process result is small
In given threshold, fititious doctor individual corresponding with one doctor is formed.
4. fititious doctor diagnostic system as claimed in claim 3, which is characterized in that the convolutional neural networks include input
Layer, at least one convolutional layer, at least one sample level, at least one full articulamentum and classifier.
5. fititious doctor diagnostic system as claimed in claim 4, which is characterized in that in the history for the doctor that will acquire
It is trained processing in diagnosis process sample data input convolutional neural networks, when obtaining real processing results, the virtual doctor
Raw module is specifically used for:
Process of convolution at least once is carried out using historical diagnostic process sample data of the convolution Boltzmann machine to input, extracts phase
The characteristic value answered, and pondization operation and sparse regularization operation are carried out for the characteristic value obtained after each process of convolution, it obtains
Corresponding diagnostic result.
6. a kind of fititious doctor diagnostic method based on convolutional neural networks is applied to as described in any one in claim 1-5
Fititious doctor diagnostic system characterized by comprising
The registration information for obtaining patient selects hospital to be checked, the information of fititious doctor individual and the patient to input
Condition symptoms information, and using the fititious doctor individual of patient selection, tentative diagnosis is carried out according to the condition symptoms, is obtained
To tentative diagnosis result, wherein the hospital to be checked is used to carry out physical examination to patient；
When receiving the patient body and checking data, using the fititious doctor individual of patient selection, according to receiving
The physical examination data of the patient carry out secondary diagnosis, obtain secondary diagnostic result；
Corresponding electronic prescription and/or electronic health record are generated according to the tentative diagnosis result and/or secondary diagnostic result.
7. fititious doctor diagnostic method as claimed in claim 6, which is characterized in that in the illness symptom for receiving patient's input
When, using the fititious doctor individual of patient selection, tentative diagnosis is carried out in stock to the patient according to the condition symptoms
Quilt is weighed, further comprises:
Each doctor's historical diagnostic process sample data is obtained respectively；
Sample training is carried out to convolutional neural networks respectively according to each doctor's historical diagnostic process sample data, is formed corresponding
Each fititious doctor individual.
8. fititious doctor diagnostic method as claimed in claim 7, which is characterized in that according to each doctor's historical diagnostic process
Sample data carries out sample training to convolutional neural networks respectively, forms corresponding each fititious doctor individual, specifically includes:
It is trained processing in the historical diagnostic process sample data input convolutional neural networks for the doctor that will acquire, is obtained
To real processing results；
If it is determined that when the difference between real processing results and ideal process result is not less than given threshold, according to minimization error
Method reversely adjusts weight matrix in convolutional neural networks, until the difference between real processing results and ideal process result is small
In given threshold, fititious doctor individual corresponding with one doctor is formed.
9. fititious doctor diagnostic method as claimed in claim 8, which is characterized in that the convolutional neural networks include input
Layer, at least one convolutional layer, at least one sample level, at least one full articulamentum and classifier.
10. fititious doctor diagnostic method as claimed in claim 9, which is characterized in that the history for the doctor that will acquire
It is trained processing in diagnosis process sample data input convolutional neural networks, real processing results is obtained, specifically includes:
Process of convolution at least once is carried out using historical diagnostic process sample data of the convolution Boltzmann machine to input, extracts phase
The characteristic value answered, and pondization operation and sparse regularization operation are carried out for the characteristic value obtained after each process of convolution, it obtains
Corresponding diagnostic result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711269723.1A CN109872812A (en) | 2017-12-05 | 2017-12-05 | A kind of fititious doctor diagnostic system and method based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711269723.1A CN109872812A (en) | 2017-12-05 | 2017-12-05 | A kind of fititious doctor diagnostic system and method based on convolutional neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109872812A true CN109872812A (en) | 2019-06-11 |
Family
ID=66915155
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711269723.1A Pending CN109872812A (en) | 2017-12-05 | 2017-12-05 | A kind of fititious doctor diagnostic system and method based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109872812A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110379507A (en) * | 2019-06-27 | 2019-10-25 | 南京市卫生信息中心 | A kind of aided diagnosis method based on patient's vector-valued image |
| CN110584618A (en) * | 2019-08-15 | 2019-12-20 | 济南市疾病预防控制中心 | Infectious disease machine recognition system based on artificial intelligence |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101681455A (en) * | 2007-04-23 | 2010-03-24 | 三星电子株式会社 | Remote medical-diagnosis system and method |
| US20150223759A1 (en) * | 2010-03-15 | 2015-08-13 | Singapore Health Services Pte Ltd. | Predicting acute cardiopulmonary events and survivability of a patient |
| CN106126917A (en) * | 2016-06-22 | 2016-11-16 | 扬州立兴科技发展合伙企业（有限合伙） | A kind of remote diagnosis system based on virtual reality technology |
| CN106372390A (en) * | 2016-08-25 | 2017-02-01 | 姹ゅ钩 | Deep convolutional neural network-based lung cancer preventing self-service health cloud service system |
| CN106709254A (en) * | 2016-12-29 | 2017-05-24 | 天津中科智能识别产业技术研究院有限公司 | Medical diagnostic robot system |
-
2017
- 2017-12-05 CN CN201711269723.1A patent/CN109872812A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101681455A (en) * | 2007-04-23 | 2010-03-24 | 三星电子株式会社 | Remote medical-diagnosis system and method |
| US20150223759A1 (en) * | 2010-03-15 | 2015-08-13 | Singapore Health Services Pte Ltd. | Predicting acute cardiopulmonary events and survivability of a patient |
| CN106126917A (en) * | 2016-06-22 | 2016-11-16 | 扬州立兴科技发展合伙企业（有限合伙） | A kind of remote diagnosis system based on virtual reality technology |
| CN106372390A (en) * | 2016-08-25 | 2017-02-01 | 姹ゅ钩 | Deep convolutional neural network-based lung cancer preventing self-service health cloud service system |
| CN106709254A (en) * | 2016-12-29 | 2017-05-24 | 天津中科智能识别产业技术研究院有限公司 | Medical diagnostic robot system |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110379507A (en) * | 2019-06-27 | 2019-10-25 | 南京市卫生信息中心 | A kind of aided diagnosis method based on patient's vector-valued image |
| CN110379507B (en) * | 2019-06-27 | 2023-03-14 | 南京市卫生信息中心 | Auxiliary diagnosis method based on patient vector image |
| CN110584618A (en) * | 2019-08-15 | 2019-12-20 | 济南市疾病预防控制中心 | Infectious disease machine recognition system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hussain et al. | Notice of retraction: AI techniques for COVID-19 | |
| JP6463572B1 (en) | Support system to estimate the internal state of the target system | |
| Shah et al. | Deep GRU-CNN model for COVID-19 detection from chest X-rays data | |
| US20170147777A1 (en) | Method and apparatus for predicting health data value through generation of health data pattern | |
| Ghadiri Hedeshi et al. | Coronary Artery Disease Detection Using a Fuzzy‐Boosting PSO Approach | |
| JP7304594B2 (en) | Diagnostic imaging device and method using deep learning model | |
| WO2021180244A1 (en) | Disease risk prediction system, method and apparatus, device and medium | |
| WO2021179630A1 (en) | Complications risk prediction system, method, apparatus, and device, and medium | |
| CN112259246A (en) | Disease prediction method fusing medical concept hierarchical structure and related equipment | |
| CN111524598A (en) | Perioperative complication prediction method and system | |
| Soundrapandiyan et al. | AI-based wavelet and stacked deep learning architecture for detecting coronavirus (COVID-19) from chest X-ray images | |
| Bajpai et al. | Early prediction of cardiac arrest using hybrid machine learning models | |
| Nugroho et al. | Performance of root-mean-square propagation and adaptive gradient optimization algorithms on covid-19 pneumonia classification | |
| CN109872812A (en) | A kind of fititious doctor diagnostic system and method based on convolutional neural networks | |
| CN112365962A (en) | Medical resource allocation method and device based on lung CT (computed tomography) image and clinical information | |
| CN108595432A (en) | Medical document error correction method | |
| JP7173482B2 (en) | Health care data analysis system, health care data analysis method and health care data analysis program | |
| Shaikh et al. | Automated classification of pneumonia from chest x-ray images using deep transfer learning efficientnet-b0 model | |
| Domyati et al. | Robust detection of cardiac disease using machine learning algorithms: Robust detection | |
| US20230395228A1 (en) | Physical therapy imaging and prediction | |
| Yilmaz et al. | Use of Autopsy to Determine Live or Stillbirth: New Approaches in Decision‐support Systems | |
| Khan et al. | Dental caries and non caries detection using deep learning | |
| Chowdary et al. | Detecting retinopathy of prematurity disease based on Fundus Image Dataset | |
| Nandini et al. | Smart IoT Solutions for Precise COVID-19 Lung and Heart Disease Identification via Bayesian Neural Networks and Hybrid Optimization | |
| Indira et al. | Prediction and analysis of Covid-19 using the Deep Learning Models |
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 |
Application publication date: 20190611 |
|
| RJ01 | Rejection of invention patent application after publication |