CN106021920A - Automatic diagnosis oriented medical data processing system and device based on neural network - Google Patents
Automatic diagnosis oriented medical data processing system and device based on neural network Download PDFInfo
- Publication number
- CN106021920A CN106021920A CN201610330801.3A CN201610330801A CN106021920A CN 106021920 A CN106021920 A CN 106021920A CN 201610330801 A CN201610330801 A CN 201610330801A CN 106021920 A CN106021920 A CN 106021920A
- Authority
- CN
- China
- Prior art keywords
- data
- health check
- index
- value
- numerical value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses an automatic diagnosis oriented medical data processing system and device based on a neural network, and relates to a medical data processing system and device. The system comprises physical examination data acquisition equipment and a data server, wherein the physical examination data acquisition equipment is used for obtaining first physical examination data used for covering each body part of a patient to be diagnosed, and sending the first physical examination data to the data server; the data server is used for utilizing the neural network to operate the first physical examination data to obtain a diagnosis result corresponding to the first physical examination data; and the neural network is obtained through training according to second physical examination data which covers each body part of the diagnosed patient and diagnosis data corresponding to the second physical examination data. The medical data processing system and device provided by the invention does not need to consume excessive time to carry out each body examination in a diagnosis result generation process, does not need to consume time to carry out detailed research on each examination result by a doctor, can operate according to the neural network to obtain a diagnosis result after the first physical examination data sent from the physical examination data acquisition equipment is received, and improves the generation efficiency of a diagnosis result.
Description
Technical field
The present invention relates to technical field of data processing, particularly relate to based on neutral net towards diagnosis automatically
Medical treatment data processing system and device.
Background technology
Along with the development of computer technology, hospital utilize computer to record the medical data of patient, and will doctor
Treat data and store server.Every day, substantial amounts of patient went to see a doctor, and each patient was being gone to a doctor by doctor
The medical data storage produced in journey is in server.
But, after by medical data storage to server, substantial amounts of medical data needs to take more
Memory space, and these are not fully utilized.When patient goes to see a doctor simultaneously, need to expend
Plenty of time carries out various types of physical examination, and doctor needs the consuming time to meet various inspections to carry out in detail
Thin research, could make a definite diagnosis patient thus obtain confirmed result, the formation efficiency of the confirmed result of patient
Relatively low.
Summary of the invention
Based on this, it is necessary to for the problem that the formation efficiency of the confirmed result of patient is relatively low, it is provided that a kind of doctor
Treat data handling system and device.
A kind of medical treatment data processing system, it is characterised in that described system include health check-up data acquisition equipment and
Data server;
Described health check-up data acquisition equipment treats the first health check-up number of patient diagnosed's body parts for obtaining covering
According to, and described first health check-up data are sent to described data server;
Described data server is used for utilizing neutral net that described first health check-up data carry out computing and obtains described
The confirmed result that first health check-up data are corresponding;Described neutral net is according to covering each portion of patient diagnosed's health
Point the make a definite diagnosis data training with the second health check-up data corresponding of the second health check-up data obtain.
Wherein in an embodiment, described health check-up data acquisition equipment is additionally operable to obtain described second health check-up number
According to, and it is sent to described data server through described second health check-up data;
Described data server is additionally operable to corresponding for described second health check-up data patient diagnosed the most is made a definite diagnosis data
It is associated storing in medical data base;The second health check-up data and second are extracted from described medical data base
What health check-up data were corresponding makes a definite diagnosis data;Corresponding with the second health check-up data according to the second health check-up data extracted
Make a definite diagnosis data training and obtain neutral net.
Wherein in an embodiment, described data server is additionally operable to determine exception according to described confirmed result
Health check-up index;At the numerical value that health check-up index abnormal described in described first health check-up extracting data is corresponding;According to institute
State numerical value corresponding to abnormal health check-up index and described confirmed result generates and makes a definite diagnosis report.
Wherein in an embodiment, described data server is additionally operable to extract in described first health check-up data respectively
The numerical value of health check-up index, the numerical value to described each health check-up index carries out the coding described each health check-up index of generation respectively
Corresponding coded data, described coded data includes index mark and index value;By defeated for described coded data
Enter described neutral net, make described neutral net described coded data be input to and described coded data middle finger
The node of mark mark correspondence carries out computing and obtains operation result;Described first body is determined according to described operation result
The disease name that inspection data are corresponding.
Wherein in an embodiment, described data server is additionally operable to extract in described first health check-up data respectively
The numerical value of health check-up index, is standardized process and obtains described each health check-up and refer to the numerical value of described each health check-up index
Target standardized value;The standardized value of described each health check-up index is inputted described neutral net, makes described nerve
The standardized value of described each health check-up index is inputted corresponding node by network, and according to the power between described node
Weight values carries out computing and obtains described first corresponding the making a definite diagnosis of health check-up data the standardized value of described each health check-up index
Result.
Above-mentioned medical treatment data processing system, data server receive health check-up data acquisition equipment collection wait make a definite diagnosis
The first health check-up data of patient, the first health check-up data cover treats patient diagnosed's body parts, can be the most anti-
Reflect the health status of patient body.Utilize according to the second health check-up data covering patient diagnosed's body parts
To covering, the neutral net that training obtains treats that the first health check-up data of patient diagnosed's body parts carry out computing,
The confirmed result that the first health check-up data are corresponding is obtained by the computing of neutral net.Generation in confirmed result
Cheng Zhong, it is not necessary to expending the too much time carries out every physical examination, expends every inspection result without doctor
Time is studied in detail, only need to after receiving the first health check-up data that health check-up data acquisition equipment sends,
Computing can be carried out according to neutral net and obtain confirmed result, improve the formation efficiency of confirmed result, and then
Auxiliary doctor more quickly, accurately, reasonably makes diagnosis.
A kind of medical data processing means, it is characterised in that described device includes:
First data acquisition module, covers, for obtaining, treat that patient diagnosed's body parts treats patient diagnosed the
One health check-up data;
Confirmed result obtains module, is used for utilizing neutral net that described first health check-up data are carried out computing and obtains
The confirmed result that described first health check-up data are corresponding;Described neutral net is according to covering patient diagnosed's health
What the second health check-up data of each several part were corresponding with the second health check-up data makes a definite diagnosis what data training obtained.
Wherein in an embodiment, described device also includes:
Second data acquisition module, is used for obtaining described second health check-up data;
Data association memory module, for making a definite diagnosis number by corresponding for described second health check-up data patient diagnosed the most
Store in medical data base according to being associated;
Associated data extraction module, for extracting the second health check-up data and the second body from described medical data base
What inspection data were corresponding makes a definite diagnosis data;
Neural metwork training module, for corresponding with the second health check-up data according to the second health check-up data extracted
Make a definite diagnosis data training obtain neutral net.
Wherein in an embodiment, described device also includes:
Abnormal index determines module, for determining abnormal health check-up index according to confirmed result;
Abnormal numerical value extraction module, in health check-up index pair abnormal described in described first health check-up extracting data
The numerical value answered;
Make a definite diagnosis report generation module, for according to numerical value corresponding to described abnormal health check-up index with described make a definite diagnosis knot
Fruit generates and makes a definite diagnosis report.
Wherein in an embodiment, described confirmed result obtains module and includes:
Health check-up numerical value extraction module, for extracting the numerical value of each health check-up index in described first health check-up data;
Health check-up numeric coding module, carries out coding for numerical value to described each health check-up index respectively and generates described
The coded data that each health check-up index is corresponding, described coded data includes index mark and index value;
Coded data computing module, for described coded data is inputted described neutral net, makes described nerve
Described coded data is input to carry out computing with node corresponding to index mark in described coded data and obtains by network
To operation result;
Disease name determines module, for determining that described first health check-up data are corresponding according to described operation result
Disease name.
Wherein in an embodiment, described confirmed result obtains module and includes:
Index value extraction module, for extracting the numerical value of each health check-up index in described first health check-up data;
Index value processing module, obtains institute for the numerical value of described each health check-up index is standardized process
State the standardized value of each health check-up index;
Standardized value computing module, for the standardized value of described each health check-up index is inputted described neutral net,
Make described neutral net that the standardized value of described each health check-up index is inputted corresponding node, and according to described god
Weighted value between nodes carries out computing and obtains described the standardized value of described each health check-up index
The confirmed result that one health check-up data are corresponding.
Above-mentioned medical data processing means, obtains the first health check-up data, wherein the first health check-up treating patient diagnosed
Data cover treats patient diagnosed's body parts, it is possible to reflect the health status of patient body comprehensively.Utilize
According to the neutral net that the second health check-up data training covering patient diagnosed's body parts obtains, covering is treated
First health check-up data of patient diagnosed's body parts carry out computing, obtain first by the computing of neutral net
The confirmed result that health check-up data are corresponding.During the generation of confirmed result, it is not necessary to expending the too much time is carried out respectively
Physical examination, without doctor to checking that result expends the time and studies in detail, only need to get the
After one health check-up data, according to neutral net, the first health check-up data can be carried out computing and obtain confirmed result, carry
The high formation efficiency of confirmed result, and then auxiliary doctor more quickly, accurately, reasonably makes diagnosis.
Accompanying drawing explanation
Fig. 1 is the applied environment figure of medical treatment data processing system in an embodiment;
Fig. 2 is the structured flowchart of medical treatment data processing equipment in an embodiment;
Fig. 3 is the structured flowchart of medical treatment data processing equipment in another embodiment;
Fig. 4 is the structured flowchart of medical treatment data processing equipment in further embodiment;
Fig. 5 is the structured flowchart of confirmed result acquisition module in an embodiment;
Fig. 6 is the structured flowchart of confirmed result acquisition module in another embodiment;
Fig. 7 is the schematic flow sheet of medical treatment data processing method in an embodiment.
Detailed description of the invention
In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and reality
Execute example, the present invention is further elaborated.Only should be appreciated that specific embodiment described herein
Only in order to explain the present invention, it is not intended to limit the present invention.
Fig. 1 is the applied environment figure of medical treatment data processing system, Analysis of Medical Treatment Data system in an embodiment
Including health check-up data acquisition equipment 110 and data server 120, health check-up data acquisition equipment 110 and data clothes
Business device 120 is connected by network.Health check-up collecting device 110 includes control terminal for data acquisition 112, sweep unit 114
With acquisition platform 116, control terminal for data acquisition 112 gathers by controlling sweep unit 114 and acquisition platform 116
It is positioned at the health check-up data of patient on acquisition platform 116.Sweep unit 114 can be specifically B ultrasonic (B-mode
Ultrasonography) ultrasound scanner in equipment, it is also possible to be CT (x-ray computer tomography,
Computed Tomography) X-ray tube and detector in equipment, it is also possible to be MRI (nuclear magnetic resonance,
Magnetic Resonance Imaging) magnetic field generator, radiowave generator and spy in imaging device
Survey device.
In one embodiment, health check-up data acquisition equipment 110 is used for obtaining covering and treats that patient diagnosed's health is each
First health check-up data of part, and the first health check-up data are sent to described data server 120.
Specifically, health check-up data acquisition equipment 110 is treated patient diagnosed's parts of body and is incited somebody to action inspection, by right
That treats patient diagnosed's body parts checks the first health check-up data generating patient, and the first health check-up data are for covering
Treat that the health check-up data of patient diagnosed's body parts, the first health check-up data include that treating patient diagnosed's each several part enters
Row checks the data obtained, and the first health check-up data specifically can include the whole body health check-up data treating patient diagnosed.
First health check-up data, after getting the first health check-up data, are sent out by health check-up data acquisition equipment 110 by network
Give data server 120.
In one embodiment, health check-up data acquisition equipment 110 can be medical scanning equipment, medical treatment scanning
Equipment is scanned obtaining the first health check-up data treating patient diagnosed for treating patient diagnosed's body part.
Data server 120 is used for utilizing neutral net that the first health check-up data are carried out computing and obtains the first health check-up
The confirmed result that data are corresponding;Neutral net is according to the second health check-up covering patient diagnosed's body parts
What data were corresponding with the second health check-up data makes a definite diagnosis what data training obtained.
Specifically, neutral net, also known as artificial neural network, (Artificial Neural Networks, is abbreviated as
ANNs) it is a kind of imitation animal nerve network behavior feature, carries out the algorithm number of distributed parallel information processing
Learn model.Neutral net relies on the complexity of system, is connected with each other between internal great deal of nodes by adjusting
Relation, thus reach the purpose of process information, and there is self study and adaptive ability.
The second health check-up data that data server 120 obtains patient diagnosed are corresponding with the second health check-up data
Make a definite diagnosis data, using the second health check-up data of getting corresponding with the second health check-up data make a definite diagnosis data as training
Sample, using the second health check-up data as the input of neutral net, and the second health check-up data corresponding make a definite diagnosis trouble
Person makes a definite diagnosis the data output as neutral net, obtains neutral net by training.At data server 120
Being input to train in the neutral net obtained by the first health check-up data got, neutral net is to the first health check-up
Data carry out computing and obtain the confirmed result that the first health check-up data are corresponding, wherein confirmed result can with disease name,
It can also be the coding that disease name is corresponding.
In the present embodiment, data server receive health check-up data acquisition equipment collection treat the first of patient diagnosed
Health check-up data, the first health check-up data cover is treated patient diagnosed's body parts, can be reflected patient body comprehensively
Health status.The the second health check-up data covering patient diagnosed's body parts are utilized to train the nerve obtained
To covering, network treats that the first health check-up data of patient diagnosed's body parts carry out computing, by neutral net
Computing obtains the confirmed result that the first health check-up data are corresponding.During the generation of confirmed result, it is not necessary to expend
Too much the time carries out every physical examination, without doctor, every inspection result is expended the time and grinds in detail
Study carefully, only need to be after receiving the first health check-up data that health check-up data acquisition equipment sends, can be according to nerve net
Network carries out computing and obtains confirmed result, improves the formation efficiency of confirmed result, and then auxiliary doctor more accelerates
Speed, accurately, reasonably make diagnosis.
In one embodiment, health check-up data acquisition equipment 110 is additionally operable to obtain the second health check-up data, and will
Second health check-up data are sent to data server.
Specifically, health check-up data acquisition equipment 110 is medical scanning equipment, and medical scanning equipment is for
Patient diagnosed's body part is scanned obtaining the second health check-up data of patient diagnosed.Health check-up data acquisition
Equipment 110 carries out covering body parts and is scanned obtaining the doctor of patient diagnosed patient diagnosed's health
Treating image, the characteristic area extracted in medical image is analyzed obtaining each portion of covering health of patient diagnosed
The the second health check-up data divided.Medical image can be specifically CT (x-ray computer tomography, Computed
Tomography) image, B ultrasonic (B-mode ultrasonography) image and MRI (nuclear magnetic resonance,
Magnetic Resonance Imaging) at least one in image.Health check-up data acquisition equipment 110 is by second
Health check-up data are sent to data server 120.
Data server is additionally operable to be associated the data of making a definite diagnosis of corresponding for the second health check-up data patient diagnosed
Store in medical data base;The second health check-up data are extracted corresponding with the second health check-up data from medical data base
Make a definite diagnosis data;Train according to the data of making a definite diagnosis that the second health check-up data extracted are corresponding with the second health check-up data
Obtain neutral net.
Specifically, data server 120, after receiving the second health check-up data, extracts in the second health check-up data
The Patient identification of patient diagnosed, then from make a definite diagnosis data base extract Patient identification corresponding make a definite diagnosis data.
Data server 120 makes a definite diagnosis data and the second health check-up data are associated also by corresponding for same Patient identification
Store in medical data base.
In one embodiment, the same Patient identification of data server 120 make a definite diagnosis data and the second health check-up
Data are associated storage, specifically corresponding for same Patient identification can be made a definite diagnosis data and the second health check-up number
It is stored in medical data base according to correspondence.By same Patient identification make a definite diagnosis data and the second health check-up data are carried out
Association storage, it is also possible to extract the spy made a definite diagnosis in data and the second health check-up data of same Patient identification respectively
Levy data, set up the mapping table of characteristic, by the mapping table of characteristic by same trouble
What person identified correspondence makes a definite diagnosis data and the second health check-up data association, and will make a definite diagnosis data and the second body after association
Inspection data store in medical data base.By same Patient identification make a definite diagnosis data and the second health check-up data are carried out
Association, it is also possible to corresponding for same Patient identification is made a definite diagnosis data and the second health check-up data by Patient identification
It is associated, by Patient identification, makes a definite diagnosis that data are corresponding with the second health check-up data to be stored in medical data base.
Data server 120 extracts that the Patient identification of multiple patient diagnosed is corresponding from medical data base
Data that what two health check-up data were corresponding with the second health check-up data make a definite diagnosis, and according to the second health check-up data extracted and
What the second health check-up data were corresponding makes a definite diagnosis data training neutral net.
In the present embodiment, by patient diagnosed make a definite diagnosis data and the second health check-up data are stored in medical data base
In, make to store in medical data base magnanimity patient diagnosed makes a definite diagnosis data and the second health check-up data, and
What same Patient identification was corresponding make a definite diagnosis data and health check-up data association is stored in medical data base, for training god
Big data supporting is provided, by the data of magnanimity in medical data base patient diagnosed to nerve net through network
Network is trained, and improves the accuracy being generated confirmed result by neutral net.
In one embodiment, data server 120 is additionally operable to determine abnormal health check-up index according to confirmed result;
At the numerical value that the first health check-up extracting data exception health check-up index is corresponding;According to the number that abnormal health check-up index is corresponding
Value and confirmed result generate makes a definite diagnosis report.
Specifically, health check-up index is the detection project that can reflect person health status in health check-up data,
Such as heart rate, blood pressure and uric acid etc., numerical value corresponding to each health check-up index has a normal range, according to
Whether the numerical value of health check-up index may determine that the health status of the person within normal range.The number of health check-up index
Value exceeds the normal range that this health check-up index is corresponding, then this health check-up index is abnormal health check-up index.
Data server 120 is after generating confirmed result, corresponding with abnormal health check-up index according to confirmed result
Relation determines the health check-up index mark of abnormal health check-up index, identifies in the first health check-up according to the health check-up index determined
Index is searched the numerical value of abnormal health check-up index and extracts, according to the numerical value of the abnormal health check-up index extracted and
Confirmed result generates and treats that patient diagnosed's makes a definite diagnosis report accordingly.Wherein confirmed result and abnormal health check-up index
Corresponding relation can take device 120 using the part or all of data of storage in medical data base as sample by data
Originally it is analyzed obtaining.
In the present embodiment, generating after confirmed result, according to confirmed result extract in the first health check-up data different
The numerical value of Chang Tijian index, makes a definite diagnosis report according to the numerical generation of confirmed result and abnormal health check-up index, makes a definite diagnosis
Report the most detailed reflection can treat the health status of patient diagnosed.
In one embodiment, data server is additionally operable to extract the number of each health check-up index in the first health check-up data
Value, the numerical value to each health check-up index carries out the coded data that each health check-up index of coding generation is corresponding, coding respectively
Data include index mark and index value;By coded data input neural network, neutral net is made to encode
Data are input to carry out computing with node corresponding to index mark in coded data and obtain operation result;According to fortune
Calculate result and determine the disease name that the first health check-up data are corresponding.
Specifically, data server 120 extracts the numerical value of each health check-up index in the first health check-up data, search with
The index mark of health check-up index name coupling, index mark is used for distinguishing different health check-up index, can be specifically
At least one in character and numeral.Index mark is added in the numerical value of health check-up index and forms coded data,
Coded data includes index mark and index value.Coded data is input to nerve by data server 120
In network, neutral net reads the index mark in coded data, and searches the node that index mark is corresponding,
Extract the index value in coded data, by node corresponding for index value input pointer mark.Neutral net
According to input index value computing obtain operation result, operation result can be in character or numerical value at least
One, searches the disease name corresponding with operation result, and using the disease name found output as treating really
Examine the confirmed result of patient.
In the present embodiment, carry out coding by each health check-up index and generate coded data, according to coded data middle finger
Index value in coded data is input to corresponding neural network node, by health check-up index by mark mark
Carry out encoding the input accuracy that ensure that neural network node, so that operation result is more accurate.
In one embodiment, data server 120 is additionally operable to extract each health check-up index in the first health check-up data
Numerical value, be standardized the numerical value of each health check-up index processing the standardized value obtaining each health check-up index;Will
The standardized value input neural network of each health check-up index, makes neutral net by defeated for the standardized value of each health check-up index
Enter corresponding node, and according to the weighted value between node, the standardized value of each health check-up index is carried out computing and obtain
To the confirmed result that the first health check-up data are corresponding.
Specifically, data server 120 extracts the numerical value of each health check-up index in the first health check-up data, by mark
Quasi-ization processes incites somebody to action each by the numerical value of health check-up index each in the first health check-up data according to the mapping relations of each health check-up index
The numerical value of health check-up index is mapped to the numerical value in specific numerical intervals, and wherein specific numerical intervals specifically may be used
To be [-1,0], [0,1] or [-1,1].
In one embodiment, data server 120 utilizes the numerical value of health check-up index to deduct this health check-up index
Minima obtains difference x1, recycle the maximum of this health check-up index and deduct minima and obtain difference x2, utilize x1Remove
With x2Obtain the numerical value of the health check-up index standardized value in numerical intervals [0,1].
In one embodiment, data server 120 is additionally operable to extract each health check-up index in the first health check-up data
Corresponding value, searches the health check-up index that the value extracted is not numerical value, the health check-up found being referred to, target value turns
Chemical conversion numerical value, with numerical value corresponding to each health check-up index as standardized value.Illustrate, health check-up index " urine Portugal
Grape sugar " corresponding value is " negative ", value corresponding to this health check-up index is not numerical value, " negative " is converted into
" 1 ", if " positive ", is then converted into " 0 " by " positive ".
In one embodiment, the numeric format that in the first health check-up index, each health check-up index is corresponding is different, data
The form of the numerical value that server 120 is corresponding to each health check-up index is adjusted, and makes the number that each health check-up index is corresponding
Value form is consistent, using the numerical value after Format adjusting as standardized value corresponding to each health check-up index.
Data server 120 obtains each health check-up index after being standardized the numerical value of each health check-up index processing
Standardized value, by the standardized value input neural network of each body index, neutral net is by each health check-up index
Standardized value inputs node corresponding to each health check-up index, neutral net according to the weighted value between each node to defeated
The standardized value of each health check-up index entered computes weighted, and obtains final computing by multi-time weighted computing
As a result, search corresponding disease name according to operation result, and using disease name as the first health check-up data pair
The confirmed result answered.
In the present embodiment, data take server 120 and are standardized the first health check-up data processing, and obtain the
The standardized value of each health check-up index of one health check-up data, utilizes the neutral net standardized value to each health check-up index
Carry out computing, obtain confirmed result.By standardized value is calculated, save the generation of confirmed result
During calculation resources, improve operation efficiency, also improve accuracy and the generation of confirmed result simultaneously
Efficiency.
As in figure 2 it is shown, in one embodiment, it is provided that a kind of medical data processing means 200, this device bag
Include: the first data acquisition module 202 and confirmed result obtain module 204.
For obtaining covering, first data acquisition module 202, treats that patient diagnosed's body parts treats patient diagnosed's
First health check-up data.
Specifically, treat patient diagnosed's parts of body will check, by treating patient diagnosed's body parts
Check generate patient the first health check-up data, the first health check-up data are that patient diagnosed's body parts is treated in covering
Health check-up data, the first health check-up data include that treating patient diagnosed's each several part carries out checking the data obtained,
One health check-up data specifically include the whole body health check-up data treating patient diagnosed.Specifically can treat patient diagnosed to carry out
The scanning covering body parts obtains the first health check-up data treating patient diagnosed.
Confirmed result obtains module 204, is used for utilizing neutral net that the first health check-up data carry out computing and obtains the
The confirmed result that one health check-up data are corresponding;Neutral net is according to covering the of patient diagnosed's body parts
What two health check-up data were corresponding with the second health check-up data makes a definite diagnosis what data training obtained.
Specifically, neutral net, also known as artificial neural network, (Artificial Neural Networks, is abbreviated as
ANNs) it is a kind of imitation animal nerve network behavior feature, carries out the algorithm number of distributed parallel information processing
Learn model.Neutral net relies on the complexity of system, is connected with each other between internal great deal of nodes by adjusting
Relation, thus reach the purpose of process information, and there is self study and adaptive ability.
Wherein neutral net is according to the second health check-up data of the patient diagnosed got and the second health check-up data
Corresponding data of making a definite diagnosis are as training sample, using the second health check-up data as the input of neutral net, and second
The patient diagnosed that health check-up data are corresponding makes a definite diagnosis the data output as neutral net, is obtained by training.
Being input to train in the neutral net obtained by the first health check-up data got, neutral net is to the first health check-up
Data carry out computing and obtain the confirmed result that the first health check-up data are corresponding, wherein confirmed result can with disease name,
It can also be the coding that disease name is corresponding.
In the present embodiment, obtaining the first health check-up data treating patient diagnosed, the first health check-up data cover is waited to make a definite diagnosis
Patient body each several part, can reflect the health status of patient body comprehensively.Utilization covers patient diagnosed's body
The neutral net that second health check-up data training of body each several part obtains treats patient diagnosed's body parts to covering
First health check-up data carry out computing, by the computing of neutral net obtain the first health check-up data corresponding make a definite diagnosis knot
Really.During the generation of confirmed result, it is not necessary to expending the too much time carries out every physical examination, without
Every inspection result is expended the time and studies in detail by doctor, only need to receive health check-up data acquisition equipment
After the first health check-up data sent, computing can be carried out according to neutral net and obtain confirmed result, improve really
Examine the formation efficiency of result, and then auxiliary doctor more quickly, accurately, reasonably makes diagnosis.
As it is shown on figure 3, in one embodiment, this medical data processing means 200 also includes: the second number
Instruct according to acquisition module 206, data association memory module 208, associated data extraction module 210 and neutral net
Practice module 212.
Second data acquisition module 206, for obtaining the second health check-up data.
Specifically, medical scanning equipment is utilized to be scanned obtaining to patient diagnosed's body part and make a definite diagnosis
The second health check-up data of patient.Can be specifically patient diagnosed's health to carry out cover body parts carry out
Scanning obtains the medical image of patient diagnosed, and the characteristic area extracted in medical image is analyzed obtaining
The second health check-up data covering body parts of patient diagnosed.Medical image can be specifically CT (x-ray meter
Calculation machine tomography, Computed Tomography) image, B ultrasonic (B-mode ultrasonography)
At least one in image and MRI (nuclear magnetic resonance, Magnetic Resonance Imaging) image.
Second health check-up data are sent to data server 120 by health check-up data acquisition equipment 110.
Data association memory module 208, for making a definite diagnosis data by corresponding for the second health check-up data patient diagnosed the most
It is associated storing in medical data base.
Specifically, after getting the second health check-up data, extract in the second health check-up data patient diagnosed's
Patient identification, then from make a definite diagnosis data base extract Patient identification corresponding make a definite diagnosis data.Same patient is marked
Know and corresponding make a definite diagnosis data and the second health check-up data are associated and store in medical data base.
In one embodiment, the same Patient identification of data server 120 make a definite diagnosis data and the second health check-up
Data are associated storage, specifically corresponding for same Patient identification can be made a definite diagnosis data and the second health check-up number
It is stored in medical data base according to correspondence.By same Patient identification make a definite diagnosis data and the second health check-up data are carried out
Association storage, it is also possible to extract the spy made a definite diagnosis in data and the second health check-up data of same Patient identification respectively
Levy data, set up the mapping table of characteristic, by the mapping table of characteristic by same trouble
What person identified correspondence makes a definite diagnosis data and the second health check-up data association, and will make a definite diagnosis data and the second body after association
Inspection data store in medical data base.By same Patient identification make a definite diagnosis data and the second health check-up data are carried out
Association, it is also possible to corresponding for same Patient identification is made a definite diagnosis data and the second health check-up data by Patient identification
It is associated, by Patient identification, makes a definite diagnosis that data are corresponding with the second health check-up data to be stored in medical data base.
Associated data extraction module 210, for extracting the second health check-up data and the second health check-up from medical data base
What data were corresponding makes a definite diagnosis data.
Neural metwork training module 212, for according to the second health check-up data extracted and the second health check-up data pair
The data training of making a definite diagnosis answered obtains neutral net.
Specifically, from medical data base, extract the second health check-up that the Patient identification of multiple patient diagnosed is corresponding
What data were corresponding with the second health check-up data makes a definite diagnosis data, obtains the second health check-up data and of patient diagnosed
What two health check-up data were corresponding makes a definite diagnosis data, corresponding with the second health check-up data with the second health check-up data got
Make a definite diagnosis data as training sample, using the second health check-up data as the input of neutral net, and the second health check-up number
Make a definite diagnosis the data output as neutral net according to corresponding patient diagnosed, obtain neutral net by training.
In the present embodiment, by patient diagnosed make a definite diagnosis data and the second health check-up data are stored in medical data base
In, make to store in medical data base magnanimity patient diagnosed makes a definite diagnosis data and the second health check-up data, and
What same Patient identification was corresponding make a definite diagnosis data and health check-up data association is stored in medical data base, for training god
Big data supporting is provided, by the data of magnanimity in medical data base patient diagnosed to nerve net through network
Network is trained, and improves the accuracy being generated confirmed result by neutral net.
As shown in Figure 4, in one embodiment, this medical data processing means 200 also includes: extremely refer to
Mark determines module 214, abnormal numerical value extraction module 216 and makes a definite diagnosis report generation module 218.
Abnormal index determines module 214, for determining abnormal health check-up index according to confirmed result.
Abnormal numerical value extraction module 216, for the number corresponding in the first health check-up extracting data exception health check-up index
Value.
Make a definite diagnosis report generation module 218, generate for the numerical value corresponding according to abnormal health check-up index and confirmed result
Make a definite diagnosis report.
Specifically, health check-up index is the detection project that can reflect person health status in health check-up data,
Such as heart rate, blood pressure and uric acid etc., numerical value corresponding to each health check-up index has a normal range, according to
Whether the numerical value of health check-up index may determine that the health status of the person within normal range.The number of health check-up index
Value exceeds the normal range that this health check-up index is corresponding, then this health check-up index is abnormal health check-up index.
After generating confirmed result, determine anomalous body according to the corresponding relation of confirmed result with abnormal health check-up index
The health check-up index mark of inspection index, the health check-up index mark according to determining is searched abnormal in the first health check-up index
The numerical value of health check-up index also extracts, and numerical value and confirmed result according to the abnormal health check-up index extracted generate phase
That answers treats that patient diagnosed's makes a definite diagnosis report.Wherein confirmed result can be led to the corresponding relation of abnormal health check-up index
Cross data to take device 120 and be analyzed obtaining as sample using the part or all of data of storage in medical data base,
It can also be the mapping table of confirmed result and the abnormal health check-up index pre-set.
In the present embodiment, generating after confirmed result, according to confirmed result extract in the first health check-up data different
The numerical value of Chang Tijian index, makes a definite diagnosis report according to the numerical generation of confirmed result and abnormal health check-up index, makes a definite diagnosis
Report the most detailed reflection can treat the health status of patient diagnosed.
As it is shown in figure 5, in one embodiment, confirmed result obtains module 204 and includes: health check-up numerical value carries
Delivery block 204a, health check-up numeric coding module 204b, coded data computing module 204c and disease name determine
Module 204d.
Health check-up numerical value extraction module 204a, for extracting the numerical value of each health check-up index in the first health check-up data;
Health check-up numeric coding module 204b, carries out coding for numerical value to each health check-up index respectively and generates each body
The coded data that inspection index is corresponding, coded data includes index mark and index value;
Coded data computing module 204c, for by coded data input neural network, makes neutral net to compile
Code data are input to carry out computing with node corresponding to index mark in coded data and obtain operation result;
Disease name determines module 204d, for determining, according to operation result, the disease that the first health check-up data are corresponding
Title.
Specifically, health check-up data extraction module extracts the numerical value of each health check-up index in the first health check-up data, searches
The index mark mated with health check-up index name, index mark, for distinguishing different health check-up indexs, specifically may be used
To be at least one in character and numeral.Index mark is added in the numerical value of health check-up index and form coding
Data, coded data includes index mark and index value, and coded data can be specifically binary number.
Coded data is input in neutral net by data server 120, and neutral net reads the finger in coded data
Mark mark, and search the node that index mark is corresponding, extract the index value in coded data, by index number
The node that value input pointer mark is corresponding.Neutral net obtains operation result according to the index value computing of input,
Operation result can be at least one in character or numerical value, searches the disease name corresponding with operation result,
And the disease name found is exported as the confirmed result treating patient diagnosed.
In the present embodiment, in the present embodiment, carry out coding by each health check-up index and generate coded data, according to
In coded data, the index value in coded data is input to corresponding neural network node by index mark, logical
Cross and health check-up index is encoded the input accuracy that ensure that neural network node, so that operation result
More accurate, and then assist doctor more quickly, accurately, reasonably to make diagnosis.
As shown in Figure 6, in one embodiment, confirmed result acquisition module 204 includes: index value carries
Delivery block 204e, index value processing module 204f and standardized value computing module 204g.
Index value extraction module 204e, for extracting the numerical value of each health check-up index in the first health check-up data.
Index value processing module 204f, obtains respectively for the numerical value of each health check-up index is standardized process
The standardized value of health check-up index.
Specifically, the numerical value of each health check-up index in the first health check-up data is extracted, by standardization by first
In health check-up data the numerical value of each health check-up index according to the mapping relations of each health check-up index by the numerical value of each health check-up index
Being mapped to the numerical value in specific interval, wherein specific interval can be specifically [-1,0], [0,1] or [-1,1].
In one embodiment, the numerical value utilizing health check-up index deducts the minima of this health check-up index and obtains difference x1,
Recycle the maximum of this health check-up index to deduct minima and obtain difference x2, utilize x1Divided by x2Obtain health check-up index
Numerical value standardized value in interval [0,1].
In one embodiment, data server 120 is additionally operable to extract each health check-up index in the first health check-up data
Corresponding value, searches the health check-up index that the value extracted is not numerical value, the health check-up found being referred to, target value turns
Chemical conversion numerical value, with numerical value corresponding to each health check-up index as standardized value.Illustrate, health check-up index " urine Portugal
Grape sugar " corresponding value is " negative ", value corresponding to this health check-up index is not numerical value, " negative " is converted into
" 1 ", if " positive ", is then converted into " 0 " by " positive ".
In one embodiment, the numeric format that in the first health check-up index, each health check-up index is corresponding is different, data
The form of the numerical value that server 120 is corresponding to each health check-up index is adjusted, and makes the number that each health check-up index is corresponding
Value form is consistent, using the numerical value after Format adjusting as standardized value corresponding to each health check-up index.
Standardized value computing module 204g, for by the standardized value input neural network of each health check-up index, making
The standardized value of each health check-up index is inputted corresponding node by neutral net, and according to neutral net interior joint it
Between weighted value the standardized value of each health check-up index is carried out computing obtain the first health check-up data corresponding make a definite diagnosis knot
Really.
Specifically, after being standardized processing, obtain the standard of each health check-up index to the numerical value of each health check-up index
Change value, by the standardized value input neural network of each body index, neutral net is by the standardization of each health check-up index
Value inputs the node that each health check-up index is corresponding, and neutral net is each to input according to the weighted value between each node
The standardized value of health check-up index computes weighted, and obtains final operation result by multi-time weighted computing,
According to the disease name that operation result lookup is corresponding, and using disease name as the first health check-up data correspondence really
Examine result.
In the present embodiment, data take server 120 and are standardized the first health check-up data processing, and obtain the
The standardized value of each health check-up index of one health check-up data, utilizes the neutral net standardized value to each health check-up index
Carry out computing, obtain confirmed result.By standardized value is calculated, save the generation of confirmed result
During calculation resources, improve operation efficiency, also improve accuracy and the generation of confirmed result simultaneously
Efficiency.
As it is shown in fig. 7, in one embodiment, it is provided that a kind of medical data processing method, the method is concrete
Comprise the following steps:
Step 702, obtains to cover and treats that patient diagnosed's body parts treats the first health check-up data of patient diagnosed.
Step 704, utilizing neutral net that the first health check-up data carry out computing, to obtain the first health check-up data corresponding
Confirmed result;Neutral net is according to the second health check-up data and second covering patient diagnosed's body parts
What health check-up data were corresponding makes a definite diagnosis what data training obtained.
In the present embodiment, obtaining the first health check-up data treating patient diagnosed, the first health check-up data cover is waited to make a definite diagnosis
Patient body each several part, can reflect the health status of patient body comprehensively.Utilization covers patient diagnosed's body
The neutral net that second health check-up data training of body each several part obtains treats patient diagnosed's body parts to covering
First health check-up data carry out computing, by the computing of neutral net obtain the first health check-up data corresponding make a definite diagnosis knot
Really.During the generation of confirmed result, it is not necessary to expending the too much time carries out every physical examination, without
Every inspection result is expended the time and studies in detail by doctor, only need to receive health check-up data acquisition equipment
After the first health check-up data sent, computing can be carried out according to neutral net and obtain confirmed result, improve really
Examine the formation efficiency of result.
In one embodiment, also including the step training neutral net before step 702, concrete steps are such as
Under: obtain the second health check-up data;The data of making a definite diagnosis of corresponding for the second health check-up data patient diagnosed are closed
Connection stores in medical data base;The second health check-up data and second health check-up of association is extracted from medical data base
What data were corresponding makes a definite diagnosis data;According to the second health check-up data extracted and the second corresponding making a definite diagnosis of health check-up data
Data training obtains neutral net.
In the present embodiment, by patient diagnosed make a definite diagnosis data and the second health check-up data are stored in medical data base
In, make to store in medical data base magnanimity patient diagnosed makes a definite diagnosis data and the second health check-up data, and
What same Patient identification was corresponding make a definite diagnosis data and health check-up data association is stored in medical data base, for training god
Big data supporting is provided, by the data of magnanimity in medical data base patient diagnosed to nerve net through network
Network is trained, and improves the accuracy being generated confirmed result by neutral net.
In one embodiment, the generation step making a definite diagnosis report, concrete steps bag are also included after step 704
Include herein below: determine abnormal health check-up index according to confirmed result;At the first health check-up extracting data anomalous body
The numerical value that inspection index is corresponding;The numerical value corresponding according to abnormal health check-up index and confirmed result generate makes a definite diagnosis report.
In the present embodiment, generating after confirmed result, according to confirmed result extract in the first health check-up data different
The numerical value of Chang Tijian index, makes a definite diagnosis report according to the numerical generation of confirmed result and abnormal health check-up index, makes a definite diagnosis
Report the most detailed reflection can treat the health status of patient diagnosed, and then auxiliary doctor is quicker, accurate
Really, diagnosis is reasonably made.
In one embodiment, step 704 specifically also includes the coding step of index value, specifically include with
Lower step: extract the numerical value of each health check-up index in the first health check-up data;Respectively the numerical value of each health check-up index is entered
Row coding generates the coded data that each health check-up index is corresponding, and coded data includes index mark and index value;
By coded data input neural network, neutral net is made coded data to be input to and index mark in coded data
The node of knowledge correspondence carries out computing and obtains operation result;Determine that the first health check-up data are corresponding according to operation result
Disease name.
In the present embodiment, in the present embodiment, carry out coding by each health check-up index and generate coded data, according to
In coded data, the index value in coded data is input to corresponding neural network node by index mark, logical
Cross and health check-up index is encoded the input accuracy that ensure that neural network node, so that operation result
More accurate.
In one embodiment, step 704 specifically also includes normalizing steps, specifically includes following step
Rapid: to extract the numerical value of each health check-up index in the first health check-up data;Index value processing module, for each body
The numerical value of inspection index is standardized processing the standardized value obtaining each health check-up index;Standardized value computing module,
For the standardized value input neural network by each health check-up index, make neutral net by the standard of each health check-up index
Change value inputs corresponding node, and according to the mark to each health check-up index of the weighted value between neutral net interior joint
Quasi-ization value carries out computing and obtains the confirmed result that the first health check-up data are corresponding.
In the present embodiment, data take server 120 and are standardized the first health check-up data processing, and obtain the
The standardized value of each health check-up index of one health check-up data, utilizes the neutral net standardized value to each health check-up index
Carry out computing, obtain confirmed result.By standardized value is calculated, save the generation of confirmed result
During calculation resources, improve operation efficiency, also improve accuracy and the generation of confirmed result simultaneously
Efficiency, and then assist doctor more quickly, accurately, reasonably to make diagnosis.
Each technical characteristic of above example can combine arbitrarily, for making description succinct, not to above-mentioned
The all possible combination of each technical characteristic in embodiment is all described, but, as long as these technology are special
There is not contradiction in the combination levied, is all considered to be the scope that this specification is recorded.
Above example only have expressed the several embodiments of the present invention, and it describes more concrete and detailed, but
Can not therefore be construed as limiting the scope of the patent.It should be pointed out that, general for this area
For logical technical staff, without departing from the inventive concept of the premise, it is also possible to make some deformation and improvement,
These broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be wanted with appended right
Ask and be as the criterion.
Claims (10)
1. a medical treatment data processing system, it is characterised in that described system includes health check-up data acquisition equipment
And data server;
Described health check-up data acquisition equipment treats the first health check-up number of patient diagnosed's body parts for obtaining covering
According to, and described first health check-up data are sent to described data server;
Described data server is used for utilizing neutral net that described first health check-up data carry out computing and obtains described
The confirmed result that first health check-up data are corresponding;Described neutral net is according to covering each portion of patient diagnosed's health
Point the make a definite diagnosis data training with the second health check-up data corresponding of the second health check-up data obtain.
System the most according to claim 1, it is characterised in that described health check-up data acquisition equipment is also used
In obtaining described second health check-up data, and described second health check-up data are sent to described data server;
Described data server is additionally operable to corresponding for described second health check-up data patient diagnosed the most is made a definite diagnosis data
It is associated storing in medical data base;The second health check-up data and second are extracted from described medical data base
What health check-up data were corresponding makes a definite diagnosis data;Corresponding with the second health check-up data according to the second health check-up data extracted
Make a definite diagnosis data training and obtain neutral net.
System the most according to claim 1, it is characterised in that described data server is additionally operable to basis
Described confirmed result determines abnormal health check-up index;Refer in health check-up abnormal described in described first health check-up extracting data
The numerical value that mark is corresponding;The numerical value corresponding according to described abnormal health check-up index and described confirmed result generate makes a definite diagnosis report
Accuse.
System the most according to claim 1, it is characterised in that described data server is additionally operable to extract
The numerical value of each health check-up index in described first health check-up data, compiles the numerical value of described each health check-up index respectively
Code generates the coded data that described each health check-up index is corresponding, and described coded data includes index mark and index number
Value;Described coded data is inputted described neutral net, makes described neutral net described coded data be inputted
Carry out computing to the node corresponding with index mark in described coded data and obtain operation result;According to described fortune
Calculate result and determine the disease name that described first health check-up data are corresponding.
System the most according to claim 1, it is characterised in that described data server is additionally operable to extract
The numerical value of each health check-up index in described first health check-up data, is standardized the numerical value of described each health check-up index
Process the standardized value obtaining described each health check-up index;The standardized value of described each health check-up index is inputted described
Neutral net, makes described neutral net that the standardized value of described each health check-up index is inputted corresponding node, and
According to the weighted value between described node, the standardized value of described each health check-up index is carried out computing and obtain described
The confirmed result that one health check-up data are corresponding.
6. a medical data processing means, it is characterised in that described device includes:
First data acquisition module, covers, for obtaining, treat that patient diagnosed's body parts treats patient diagnosed the
One health check-up data;
Confirmed result obtains module, is used for utilizing neutral net that described first health check-up data are carried out computing and obtains
The confirmed result that described first health check-up data are corresponding;Described neutral net is according to covering patient diagnosed's health
What the second health check-up data of each several part were corresponding with the second health check-up data makes a definite diagnosis what data training obtained.
Device the most according to claim 6, it is characterised in that described device also includes:
Second data acquisition module, is used for obtaining described second health check-up data;
Data association memory module, for making a definite diagnosis number by corresponding for described second health check-up data patient diagnosed the most
Store in medical data base according to being associated;
Associated data extraction module, for extracting the second health check-up data and the second body from described medical data base
What inspection data were corresponding makes a definite diagnosis data;
Neural metwork training module, for corresponding with the second health check-up data according to the second health check-up data extracted
Make a definite diagnosis data training obtain neutral net.
Device the most according to claim 6, it is characterised in that described device also includes:
Abnormal index determines module, for determining abnormal health check-up index according to confirmed result;
Abnormal numerical value extraction module, in health check-up index pair abnormal described in described first health check-up extracting data
The numerical value answered;
Make a definite diagnosis report generation module, for according to numerical value corresponding to described abnormal health check-up index with described make a definite diagnosis knot
Fruit generates and makes a definite diagnosis report.
Device the most according to claim 6, it is characterised in that described confirmed result obtains module and includes:
Health check-up numerical value extraction module, for extracting the numerical value of each health check-up index in described first health check-up data;
Health check-up numeric coding module, carries out coding for numerical value to described each health check-up index respectively and generates described
The coded data that each health check-up index is corresponding, described coded data includes index mark and index value;
Coded data computing module, for described coded data is inputted described neutral net, makes described nerve
Described coded data is input to carry out computing with node corresponding to index mark in described coded data and obtains by network
To operation result;
Disease name determines module, for determining that described first health check-up data are corresponding according to described operation result
Disease name.
Device the most according to claim 6, it is characterised in that described confirmed result obtains module bag
Include:
Index value extraction module, for extracting the numerical value of each health check-up index in described first health check-up data;
Index value processing module, obtains institute for the numerical value of described each health check-up index is standardized process
State the standardized value of each health check-up index;
Standardized value computing module, for the standardized value of described each health check-up index is inputted described neutral net,
Make described neutral net that the standardized value of described each health check-up index is inputted corresponding node, and according to described god
Weighted value between nodes carries out computing and obtains described the standardized value of described each health check-up index
The confirmed result that one health check-up data are corresponding.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610330801.3A CN106021920A (en) | 2016-05-17 | 2016-05-17 | Automatic diagnosis oriented medical data processing system and device based on neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610330801.3A CN106021920A (en) | 2016-05-17 | 2016-05-17 | Automatic diagnosis oriented medical data processing system and device based on neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106021920A true CN106021920A (en) | 2016-10-12 |
Family
ID=57098476
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610330801.3A Pending CN106021920A (en) | 2016-05-17 | 2016-05-17 | Automatic diagnosis oriented medical data processing system and device based on neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106021920A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108182389A (en) * | 2017-12-14 | 2018-06-19 | 华南师范大学 | User data processing method, robot system based on big data and deep learning |
CN111986762A (en) * | 2020-09-03 | 2020-11-24 | 平安国际智慧城市科技股份有限公司 | Abnormal medical index determination method and device, electronic equipment and storage medium |
WO2021146941A1 (en) * | 2020-01-21 | 2021-07-29 | 京东方科技集团股份有限公司 | Disease location acquisition method, apparatus, device and computer readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102446243A (en) * | 2010-10-08 | 2012-05-09 | 深圳Tcl教育科技有限责任公司 | Medical diagnosis system and diagnostic method |
CN103116707A (en) * | 2013-03-04 | 2013-05-22 | 北京工业大学 | Heart disease intelligent diagnostic method based on case reasoning |
CN104657621A (en) * | 2015-03-09 | 2015-05-27 | 上海万达全程健康门诊部有限公司 | System and method for evaluating body health states based on multi-point collection |
-
2016
- 2016-05-17 CN CN201610330801.3A patent/CN106021920A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102446243A (en) * | 2010-10-08 | 2012-05-09 | 深圳Tcl教育科技有限责任公司 | Medical diagnosis system and diagnostic method |
CN103116707A (en) * | 2013-03-04 | 2013-05-22 | 北京工业大学 | Heart disease intelligent diagnostic method based on case reasoning |
CN104657621A (en) * | 2015-03-09 | 2015-05-27 | 上海万达全程健康门诊部有限公司 | System and method for evaluating body health states based on multi-point collection |
Non-Patent Citations (1)
Title |
---|
赵卫东等: "基于神经网络的案例推理医疗诊断", 《东南大学学报(自然科学版》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108182389A (en) * | 2017-12-14 | 2018-06-19 | 华南师范大学 | User data processing method, robot system based on big data and deep learning |
CN108182389B (en) * | 2017-12-14 | 2021-07-30 | 华南师范大学 | User data processing method based on big data and deep learning and robot system |
WO2021146941A1 (en) * | 2020-01-21 | 2021-07-29 | 京东方科技集团股份有限公司 | Disease location acquisition method, apparatus, device and computer readable storage medium |
CN111986762A (en) * | 2020-09-03 | 2020-11-24 | 平安国际智慧城市科技股份有限公司 | Abnormal medical index determination method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP4131160A1 (en) | Image obtaining method and system, image quality determination method and system, and medical image acquisition method and system | |
CN106021915B (en) | Big data-oriented automatic diagnosis and treatment-oriented medical data analysis system and device | |
CN111985584B (en) | Disease auxiliary detection equipment, method, device and medium based on multi-mode data | |
JP7218215B2 (en) | Image diagnosis device, image processing method and program | |
CN107103182A (en) | A kind of heart disease Warning System and method based on deep learning algorithm | |
CN106021920A (en) | Automatic diagnosis oriented medical data processing system and device based on neural network | |
KR20170047423A (en) | Automatic diagnosis system of digital xray based on cad | |
CN110731773A (en) | abnormal electrocardiogram screening method based on fusion of global and local depth features of electrocardiogram | |
CN109919912A (en) | A kind of quality evaluating method and device of medical image | |
CN110532926A (en) | Pig neurogenic disease intelligence Forecasting Method based on deep learning | |
Zeng et al. | Machine Learning-Based Medical Imaging Detection and Diagnostic Assistance | |
CN110223772A (en) | A kind for the treatment of endocrine diseases diagnostic device and data information processing method | |
KR102245219B1 (en) | Method for discriminating suspicious lesion in medical image, method for interpreting medical image, and computing device implementing the methods | |
CN111667457B (en) | Automatic identification method, system, terminal and storage medium for vertebral body information based on medical image | |
CN109979554A (en) | A kind of clinical case data collection system and acquisition method | |
CN115050456A (en) | Artificial intelligence medical image automatic diagnosis system and method | |
CN111588359B (en) | New coronaries pneumonia early warning system | |
CN111242930B (en) | Method and device for acquiring pulmonary tuberculosis recognition model, storage medium and processor | |
CN115089112A (en) | Method and device for establishing post-stroke cognitive disorder risk assessment model and electronic equipment | |
CN113793662A (en) | Correlation identification method, correlation identification device, electronic equipment and computer-readable storage medium | |
CN113990487A (en) | Malignant tumor identification system based on blood routine, biochemical and urine routine test data | |
CN110147830B (en) | Method for training image data generation network, image data classification method and device | |
CN113554640A (en) | AI model training method, use method, computer device and storage medium | |
CN116763268B (en) | Human body multi-index detection method and device | |
CN114898892B (en) | Infection risk control method for basic medical community medical institution based on Internet of things system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161012 |