CN208969732U - The screening system of eye fundus image lesion based on artificial intelligence - Google Patents
The screening system of eye fundus image lesion based on artificial intelligence Download PDFInfo
- Publication number
- CN208969732U CN208969732U CN201822203704.5U CN201822203704U CN208969732U CN 208969732 U CN208969732 U CN 208969732U CN 201822203704 U CN201822203704 U CN 201822203704U CN 208969732 U CN208969732 U CN 208969732U
- Authority
- CN
- China
- Prior art keywords
- eye fundus
- fundus image
- screening
- module
- information
- 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.)
- Active
Links
Landscapes
- Eye Examination Apparatus (AREA)
Abstract
The utility model provides a kind of screening system of eye fundus image lesion based on artificial intelligence comprising: memory module is deployed in Cloud Server；Module is obtained, is used to obtain two or more the eye fundus images from same person's difference eyes, and eye fundus image is uploaded to memory module；Recording module is used for typing examinee's information corresponding with eye fundus image and is associated with eye fundus image, and examinee's information is uploaded to memory module, and the recording module verifies examinee's information；And screening module, it is used to extract eye fundus image and examinee's information from memory module, and automatic interpretation is carried out to generate screening results with the presence or absence of lesion to the eye fundus image after association using the deep learning method based on convolutional neural networks.According to the utility model, the universal of eye fundus image lesion screening can be conducive to and promoted, and interpretation is carried out to generate screening results by deep learning method, can be improved screening efficiency.
Description
Technical field
The present invention relates to a kind of screening system, in particular to the screening system of a kind of eye fundus image lesion based on artificial intelligence
System.
Background technique
Diabetes are one group of metabolic diseases characterized by hyperglycemia.Hyperglycemia be then due to defect of insulin secretion or
Its biological effect is impaired.Long-standing hyperglycemia will lead to the chronic damage of various tissues especially eye, kidney, heart, blood vessel etc.
Harmful or dysfunction, especially to the damage of eyes.According to statistics, the ocular angiogenesis of the patient with diabetes many years can be damaged,
It may be led to fundus hemorrhage when serious and enable patient's vision fuzzy or complete Loss Of Vision.
Currently, diabetic retinopathy (referred to as " sugar net the disease ") patient in China is more than 30,000,000, about 80% with
On diabetic can occur sugar net disease, blindness risk it is 25 times higher than normal person.Sugared net disease is can be to avoid blindness
Eye disease, and the lesion in early stage retina is not obvious, if but can periodically carry out funduscopy at the initial stage of a disease, by sugar
Blindness risk caused by net disease can decline 90% or more.Therefore, early screening, early diagnosis, early treatment are that sugar net patient retains
The key of eyesight.
However, China's sugar net disease screening rate is less than 10%.And there is also many problems, examples for sugared net screening system
Such as, sugar nets sick intellectual analysis software and depends on computer hardware equipment, and for base's screening, carrying turnover is inconvenient, and each sieve
It looks into equipment or screening selects and needs separate configurations sugar net disease intellectual analysis software and computer hardware equipment, higher cost.Furthermore it is right
For huge diabetes community, the number wretched insufficiency of screening doctor, artificial interpretation method is time-consuming and laborious, and subjectivity is strong,
Also situations such as being easy to appear mistaken diagnosis.
Summary of the invention
The disclosure in view of the above-mentioned prior art situation and complete, its purpose is to provide one kind can convenient for promote and
The screening system of the eye fundus image lesion of implementation.
For this purpose, the screening system of present disclose provides a kind of eye fundus image lesion based on artificial intelligence comprising: storage
Module is deployed in Cloud Server；Module is obtained, is used to obtain two or more the eyes from the same different eyes
Base map picture, and the eye fundus image is uploaded to the memory module；Recording module is used for typing and the eye fundus image pair
It examinee's information for answering and is associated with the eye fundus image, and examinee's information is uploaded to the memory module, it is described
Recording module verifies examinee's information；And screening module, it is used for from the memory module described in extraction
Eye fundus image and examinee's information, and using the deep learning method based on convolutional neural networks to the eye after association
Base map picture carries out automatic interpretation with the presence or absence of lesion to generate screening results.
In the disclosure, operator can be facilitated to be acquired and upload eyeground in different places by Cloud Server
Image, to be conducive to the universal of eye fundus image lesion screening and promote, moreover, passing through the depth based on convolutional neural networks
Learning method carries out interpretation to eye fundus image to generate screening results, to alleviate the burden of operator, improves screening efficiency.
In addition, further including for exporting optionally in the screening system of eye fundus image lesion involved in the disclosure
State the output module of screening results.Thereby, it is possible to facilitate doctor and user to obtain screening results.
In addition, optionally, examinee's information is also in the screening system of eye fundus image lesion involved in the disclosure
It include: the information such as name, gender, identity code (ID), medical history.Thereby, it is possible to the effectively letters such as comprehensive examinee's information
Breath, can more accurately judge whether eye fundus image occurs lesion.
In addition, optionally, the eye fundus image is at least in the screening system of eye fundus image lesion involved in the disclosure
Optic disk and macular area including eyeground.Thereby, it is possible to according to the effective information of eye fundus image to whether lesion occur judging.
In addition, optionally, the acquisition module is hand in the screening system of eye fundus image lesion involved in the disclosure
Hold formula fundus camera.Thereby, it is possible to facilitate operator or collector to carry.
In addition, optionally, the output module also wraps in the screening system of eye fundus image lesion involved in the disclosure
It includes and exports screening results at report form, and be uploaded to the memory module.Thereby, it is possible to easily depositing from Cloud Server
Store up inquiry screening report in module.
In addition, further including by screening results optionally in the screening system of eye fundus image lesion involved in the disclosure
It is uploaded to the memory module.System can inquire screening results from memory module as a result, and carry out to result further
Processing.
In addition, optionally, screening module is deployed in cloud in the screening system of eye fundus image lesion involved in the disclosure
In server.Thereby, it is possible to carry out automatic screening to eye fundus image by Cloud Server.
In addition, optionally, the acquisition module is also wrapped in the screening system of eye fundus image lesion involved in the disclosure
Include for judge the acquired eye fundus image whether He Ge judging unit.Thereby, it is possible to the eye fundus image of acquisition into
The preliminary judgement of row, facilitates operator to collect correct eye fundus image.
In addition, optionally, the eye fundus image includes in the screening system of eye fundus image lesion involved in the disclosure
Two left eye eye fundus images and two right eye eye fundus images from the same person.It in this case, can be more accurately
Judge whether eye fundus image occurs lesion.
In accordance with the invention it is possible to provide a kind of screening system that can be convenient for popularization and the eye fundus image lesion implemented.
Detailed description of the invention
Fig. 1 be show the screening system of eye fundus image lesion involved in embodiment of the present disclosure one kind it is exemplary
System block diagram.
Fig. 2 is to show another example of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
System block diagram.
Fig. 3 is to show the acquisition module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.
Fig. 4 is to show the recording module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.
Fig. 5 is to show the screening module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.
Fig. 6 is to show the schematic diagram of an example of eye fundus image involved in embodiment of the present disclosure.
Fig. 7 is to show the output result report of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
The schematic diagram for the example accused.
Specific embodiment
Hereinafter, explaining the preferred embodiment of the present invention in detail with reference to attached drawing.In the following description, for identical
Component assign identical symbol, the repetitive description thereof will be omitted.Scheme in addition, attached drawing is only schematical, the mutual ruler of component
Very little shape of ratio or component etc. can be with actual difference.
It should be noted that term " includes " and " having " and their any deformation in the present invention, such as wrapped
Include or the process, method, system, product or equipment of possessed a series of steps or units are not necessarily limited to be clearly listed that
A little step or units, but may include or with being not clearly listed or for these process, methods, product or equipment
Intrinsic other step or units.
In addition, the subhead etc. designed in description below the present invention be not meant to limit the present invention in perhaps model
It encloses, is merely possible to the suggesting effect read.Such subhead can neither be interpreted as also not answering for dividing article content
Content under subhead is limited only in the range of subhead.
Fig. 1 be show the screening system of eye fundus image lesion involved in embodiment of the present disclosure one kind it is exemplary
System block diagram.Fig. 2 is to show the another kind of the screening system of eye fundus image lesion involved in embodiment of the present disclosure to show
The system block diagram of example.
A kind of screening system 1 of the eye fundus image lesion based on artificial intelligence involved in embodiment of the present disclosure can be with
Include: memory module 10, obtain module 20, recording module 30 and screening module 40 (referring to Fig. 1).The eye involved in the disclosure
Base map as lesion screening system 1 in, memory module 10 is deployed in Cloud Server 2；Module 20 is obtained, is used to obtain
From two or more eye fundus images of same person's difference eyes, and eye fundus image is uploaded to memory module 10；Typing
Module 30 is used for typing examinee's information corresponding with eye fundus image and is associated with eye fundus image, and will be in examinee's information
Memory module 10 is reached, recording module 30 verifies examinee's information；And screening module 40, it is used for from storage
Eye fundus image and examinee's information (such as patient information) are extracted in module 10, and utilize the depth based on convolutional neural networks
Learning method carries out automatic interpretation with the presence or absence of lesion to the eye fundus image after association to generate screening results.
When using the screening system of eye fundus image lesion involved in the disclosure, eye fundus image is can be used in collector
Equipment such as fundus camera is acquired, memory module 10 is uploaded to after having acquired eye fundus image, in the collection process of eye fundus image
In can simultaneously typing examinee's information, typing examinee's information can also be shifted to an earlier date, examinee's information is uploaded to and is arranged in cloud
After memory module 10 in server 2, using the deep learning method based on convolutional neural networks screening module 40 to its into
Row automatic interpretation, and generate screening results.Thereby, it is possible to be conducive to the universal of eye fundus image lesion screening and promote, moreover, logical
It crosses the deep learning method based on convolutional neural networks and interpretation is carried out to eye fundus image to generate screening results, to alleviate behaviour
Make the burden of personnel, improves screening efficiency.
In addition, in some instances, the screening system 1 of eye fundus image lesion can also include for exporting the screening knot
The output module 50 (referring to fig. 2) of fruit.
(memory module)
In some instances, memory module 10 can be nonvolatile memory.In some instances, memory module 10 can
To be flash memory (Flash Memory).But the present disclosure is not limited thereto, and memory module 10 can also be for example: ferroelectric random
Memory (FeRAM), magnetic RAM (MRAM), phase-change random access memory (PRAM) or resistive random access memory
(RRAM).Thereby, it is possible to reduce a possibility that causing loss of data because of sudden power-off.
In other example, memory module 10 can also be other types of readable storage medium storing program for executing, such as: it is read-only to deposit
Reservoir (Read-Only Memory, ROM), may be programmed read-only deposit at random access memory (Random Access Memory, RAM)
Reservoir (Programmable Read-only Memory, PROM), Erasable Programmable Read Only Memory EPROM (Erasable
Programmable Read Only Memory, EPROM), disposable programmable read-only memory (One-time
Programmable Read-Only Memory, OTPROM), the electronics formula of erasing can make carbon copies read-only memory
(Electrically-Erasable Programmable Read-Only Memory, EEPROM), CD-ROM (Compact
Disc Read-Only Memory, CD-ROM) or other disc memories, magnetic disk storage, magnetic tape storage or can
For carrying or any other computer-readable medium of storing data.Thereby, it is possible to suitable with selecting according to different situations
Memory.
As described above, memory module 10 can be set in Cloud Server 2.Wherein, Cloud Server 2 can be rental.
Thus, it is possible to reduce the maintenance cost of server.In other examples, Cloud Server 2 is also possible to what oneself was built.At this
Kind in the case of, memory module 10 can be arranged in the server oneself built, it is ensured that the confidentiality of data, prevent client or
The leaking data of patient.
(obtaining module)
Fig. 3 is to show the acquisition module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.
In the disclosure, obtaining module 20 can be used for obtaining two or more the eyes from same person's difference eyes
Base map picture.
In some instances, the available unit 21 of module 20 and judging unit 22 are obtained (referring to Fig. 3).In some examples
In, whether acquiring unit 21 includes obtaining the acquiring unit 21 of eye fundus image and closing to the eye fundus image obtained by acquiring unit 21
The judging unit 22 that lattice are judged.
In some instances, obtaining module 20 can be hand-held fundus camera.Thereby, it is possible to facilitate operator each
Examinee's eye fundus image is obtained in the case of kind.In some instances, obtaining module 20 can also be desk-top fundus camera.
In addition, as shown in figure 5, in some instances, eye fundus image may include the optic disk and macular area on eyeground.Wherein,
Optic disk namely discus nervi optici.In other examples, eye fundus image can also include the artery and vein in each region in eyeground.As a result, can
It is enough to make accurate judgement to whether eyeground occurs lesion by observing eye fundus image.
In some instances, the acquiring unit 21 for obtaining module 20 can be by obtaining from same person's difference eyes
Two or more eye fundus images.As a result, to guarantee the reliability of eye fundus image obtained, it can reduce and screening mistake occur
The case where probability.
In other examples, eye fundus image can also include two left eye eye fundus images and two from the same person
Right eye eye fundus image.In this case, can one in two left-eye images be used as Main Analysis target eye fundus image, separately
One is used as with reference to eye fundus image, thereby, it is possible to simulate the diagnosis process of doctor, with reference to other eyeground figures from the same person
As judging target image, to be conducive to improve the judging nicety rate to eye fundus image lesion, can it is cross-referenced and
By calculation amount control in minimum zone, to improve the efficiency of screening.In addition, above-mentioned two left-eye images (two right sides
Eye image) it can be same eye fundus image.At this point, same eye fundus image had not only been used as Main Analysis target eye fundus image, but also
As reference eye fundus image.
In other examples, obtaining the eye fundus image obtained of module 20 can also include multiple from the same person
(two or more) eye fundus image.In this case, can using any one width therein as Main Analysis target eye fundus image,
Other images are as reference picture, thus, it is possible to judge whether eye fundus image occurs lesion more accurately.Further, may be used
To use the eye fundus image from left and right eyes of equal amount.
In some instances, it obtains eye fundus image acquired in module 20 to be not particularly limited, can be color image example
Such as RGB image, it is also possible to gray level image etc..
In some instances, obtaining the eye fundus image obtained of module 20 can be uploaded in memory module 10, that is, eye
Base map picture can be uploaded in Cloud Server 2.As a result, obtaining the image obtained of module 20 can be stored in time
Or analysis, while but also acquisition can carry out whenever and wherever possible, become more flexible.
In the disclosure, judging unit 22 can be used for judging whether eye fundus image obtained is qualified.In such case
Under, acquired eye fundus image can judge whether qualification by judging unit 21 in advance.For example, acquired eye fundus image
Whether clear, whether angle meets the requirements.Thereby, it is possible to carry out preliminary judgement to the eye fundus image of acquisition, facilitate operation
Personnel acquire correct effective eye fundus image.
(recording module)
Fig. 4 is to show the recording module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.
In the disclosure, recording module 30 can be used for typing examinee's information corresponding with eye fundus image and with eyeground figure
As association.
In some instances, recording module 30 may include typing unit 31, verification unit 32 and associative cell 33.One
In a little examples, typing unit 31 can be with typing examinee's information corresponding with the eye fundus image.Verification unit 32 can be to quilt
Inspection person's information is verified, such as the ID card information of input mismatches, input bit miscounts mistake etc..Associative cell 33 can be to quilt
It inspection person's information and is associated with eye fundus image, it is possible thereby to generate with the associated image ID number of the eye fundus image etc..
In some instances, recording module 30 can be cell phone application.Thereby, it is possible to conveniently and efficiently be convenient for whenever and wherever possible
Testing staff or examinee's typing information.But present embodiment is without being limited thereto, and recording module 30 can also be computer, touch-screen equipment
Or other input equipments.In some instances, examinee's information may include name, gender, identity code (ID), previously disease
The information such as history.Screening system can be based on the essential information of examinee, such as medical history, to more accurately judge as a result,
Whether eye fundus image occurs lesion.
In addition, in some instances, examinee's information of typing can be by examinee oneself for example by shifting to an earlier date on the net
The information of input.In such a case, it is possible to reduce the time that testing staff spends in input information, and input letter in advance
System can reserve row number after breath for examinee, and testing staff can be directly associated with the image of acquisition by input information, greatly
It improves work efficiency greatly.In other examples, the information of typing, which can also be, to be transferred from other databases, such as is cured
The database of institute or the database of public security.Thereby, it is possible to ensure the accuracy and reliability of typing information.
In addition, in some instances, recording module 30 can also verify examinee by the data in other databases
Whether the data of information are correct.The case where avoiding wrong record as a result, and accidentally recording.
In some instances, it by associative cell 33, by examinee's information of 30 typings of recording module and can obtain
Eye fundus image acquired in module 20 is associated.In such a case, it is possible to pass through the basic information and eye fundus image of examinee
It combines, relatively reliable judging result is obtained with this.
In some instances, examinee's information of 30 typings of recording module can be uploaded in memory module 10, that is,
It is uploaded in Cloud Server 2.As a result, system can be associated with the eye fundus image from same people beyond the clouds.Some
In example, examinee's information of 30 typings of recording module common again after associated with eye fundus image collected can be uploaded
Into memory module 10.Subtracted as a result, system need beyond the clouds by the eye fundus image of acquisition it is associated with examinee's information this
One step improves the operational efficiency of system.
(screening module)
Fig. 5 is to show the screening module of the screening system of eye fundus image lesion involved in embodiment of the present disclosure
Block diagram.Fig. 6 is to show the schematic diagram of an example of eye fundus image involved in embodiment of the present disclosure.Fig. 7 is to show
One example of the output result report of the screening system of eye fundus image lesion involved in embodiment of the present disclosure shows
It is intended to.
In the disclosure, screening module 40 can be used for extracting eye fundus image and examinee's information from memory module 10,
And automatic interpretation is carried out with the presence or absence of lesion to the eye fundus image after association to generate screening results based on deep learning method
In some instances, screening module 40 may include pretreatment unit 41 and screening unit 42 (referring to Fig. 5).One
In a little examples, pretreatment unit 41 can be pre-processed eye fundus image such as gray proces, normalization.Screening unit 42
It can use the deep learning method based on convolutional neural networks and automatic interpretation carried out to pretreated eye fundus image.
In some instances, the deep learning method based on convolutional neural networks can also be utilized to target eye fundus image and
It is operated respectively with reference to eye fundus image, to obtain the feature of target eye fundus image and with reference to the feature of eye fundus image.Above-mentioned
In deep learning method, such as the eye fundus image respectively from eyes can be obtained by convolutional neural networks, it will wherein one
Eye fundus image is opened as target eye fundus image, and other eye fundus images, which are used as, refers to eye fundus image.Thus, it is possible to be based on target eye
Base map picture and better interpretation result is obtained with reference to eye fundus image.In addition, since convolutional neural networks are conducive to have part
The advantage that receptive field and weight are shared, and be conducive to extract the advanced features of eye fundus image, therefore can be improved operation efficiency,
Save hardware spending.
In some instances, screening module 40 can be the convolutional neural networks based on deep learning method.Specifically,
Artificial neural network based on deep learning method can also in training to eye fundus image to (including target eye fundus image and ginseng
Examine eye fundus image) carry out data amplification, with improve training data sample amount, thus improve eyeground pathological changes are judged it is accurate
Rate.Thereby, it is possible to reduce the influence of human factor, the accuracy rate to eye fundus image screening is improved, and improve efficiency.
In some instances, pretreatment unit 41 can carry out eyeground region detection, image cutting-out, size to eye fundus image
The pretreatments such as adjustment, normalization.Thereby, it is possible to facilitate artificial neural network to identify and judge eye fundus image.
In some instances, screening module 40 can be by the characteristics of image of the eye fundus image acquired in identifying, with judgement
Whether eyeground occurs lesion.In other examples, eyeground pathological changes type that screening module 40 may determine that can be with are as follows: 1. view
Film angiosis；2. acquired maculopathy related diseases；3. inflammation venereal disease, including non-infectious systemic disease and it is infectious (virus,
Bacterium, fungi) sick and other diseases；4. eyeground dystrophia, including retina dystrophia, vitreoretinopathy,
And choroid dystrophia etc.；5. detachment of retina；6. tumour；7. acquired optic nerve disease；8. birth defect etc..On
Clue can be shown on eye fundus image by stating lesion, such as late discovery, early treatment, it would be possible to cause blindness to patient
Risk.For example, diabetic retinopathy is one of the complication of diabetes as common eyeground pathological changes, it has also become work
One of the main reason for adult's blinding of age bracket.It is estimated that falling ill within existing diabetic 92,400,000,5 years in China
Rate is 43%, blind rate 10%.It can early find that early treatment reduces the risk of blindness with this by screening as a result,.
In some instances, screening module 40 can the screening results to eye fundus image can be divided into it is ill and disease-free.Into
One step, in some instances, screening module 40 can be further divided into the disease of eyeground pathological changes to the screening results of eye fundus image
Become grade and disease-free.Further, in other examples, screening module 40 can also divide the screening results of eye fundus image
For the corresponding type of eyeground pathological changes and type and disease-free.In this case, eye can be immediately arrived at by screening results
The reason of bottom lesion and type had both reduced patient and had needed instead to greatly reduce the subsequent examination process of testing staff
The trouble looked into is rechecked, and facilitates medical staff to the lasting tracking of patient and checks.
In some instances, obtaining screening results by screening module 40 can directly be uploaded in memory module 10.As a result,
System can call screening results from memory module 10, and result is further processed.
In some instances, screening module 40 can be deployed in Cloud Server 2.In this case, screening module 40
Screening directly can be carried out to the image stored in Cloud Server 2, reduce the data time to be expended of transmission.
As shown in fig. 7, in some instances, the result that screening module 40 is obtained can be exported by output module 50 at
Report form.Thereby, it is possible to facilitate doctor and user query screening results.In some instances, output module 50 can also be
After screening results are exported into report form, report is uploaded to memory module 10.User can inquire sieve from storage as a result,
Look into report.
In addition, in some instances, output module 50 can obtain screening results data from memory module 10.
Although being illustrated in conjunction with the accompanying drawings and embodiments to the present invention above, it will be appreciated that above description
The invention is not limited in any way.Those skilled in the art without departing from the true spirit and scope of the present invention may be used
To deform and change to the present invention as needed, these deformations and variation are within the scope of the present invention.
Claims (10)
1. a kind of screening system of the eye fundus image lesion based on artificial intelligence, which is characterized in that
Include:
Memory module is deployed in Cloud Server；
Module is obtained, is used to obtain two or more the eye fundus images from same person's difference eyes, and by the eye
Base map picture is uploaded to the memory module；
Recording module is used for typing examinee's information corresponding with the eye fundus image and is associated with the eye fundus image, and
Examinee's information is uploaded to the memory module, the recording module verifies examinee's information；And
Screening module is used to from the memory module extract the eye fundus image and examinee's information, and utilizes base
In convolutional neural networks deep learning method to the eye fundus image after association with the presence or absence of lesion carry out automatic interpretation with
Generate screening results.
2. screening system according to claim 1, which is characterized in that
It further include the output module for exporting the screening results.
3. screening system according to claim 1, which is characterized in that
Examinee's information further include: the information such as name, gender, identity code (ID), medical history.
4. screening system according to claim 1, which is characterized in that
The eye fundus image includes at least the optic disk and macular area on eyeground.
5. screening system according to claim 1, which is characterized in that
The acquisition module is hand-held fundus camera.
6. screening system according to claim 2, which is characterized in that
The output module further includes exporting the screening results at report form, and be uploaded to the memory module.
7. screening system according to claim 1, which is characterized in that
The screening results are also uploaded to the memory module by the screening module.
8. screening system according to claim 1, which is characterized in that
The screening module is deployed in the Cloud Server.
9. screening system according to claim 1, which is characterized in that
It is described obtain module further include for judge the acquired eye fundus image whether He Ge judging unit.
10. screening system according to claim 5, which is characterized in that
The eye fundus image includes two left eye eye fundus images and two right eye eye fundus images from the same person.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201822203704.5U CN208969732U (en) | 2018-12-26 | 2018-12-26 | The screening system of eye fundus image lesion based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201822203704.5U CN208969732U (en) | 2018-12-26 | 2018-12-26 | The screening system of eye fundus image lesion based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN208969732U true CN208969732U (en) | 2019-06-11 |
Family
ID=66766581
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201822203704.5U Active CN208969732U (en) | 2018-12-26 | 2018-12-26 | The screening system of eye fundus image lesion based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN208969732U (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110974151A (en) * | 2019-11-06 | 2020-04-10 | 中山大学中山眼科中心 | Artificial intelligence system and method for identifying retinal detachment |
| CN114451860A (en) * | 2022-01-27 | 2022-05-10 | 广东康软科技股份有限公司 | Fundus oculi lesion diagnosis method, system and device based on deep learning |
-
2018
- 2018-12-26 CN CN201822203704.5U patent/CN208969732U/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110974151A (en) * | 2019-11-06 | 2020-04-10 | 中山大学中山眼科中心 | Artificial intelligence system and method for identifying retinal detachment |
| CN114451860A (en) * | 2022-01-27 | 2022-05-10 | 广东康软科技股份有限公司 | Fundus oculi lesion diagnosis method, system and device based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109411086A (en) | The screening system of eye fundus image lesion | |
| CN107423571B (en) | Diabetic retinopathy identifying system based on eye fundus image | |
| CN107358605B (en) | The deep neural network apparatus and system of diabetic retinopathy for identification | |
| WO2020200087A1 (en) | Image-based detection of ophthalmic and systemic diseases | |
| CN108230296A (en) | The recognition methods of characteristics of image and device, storage medium, electronic device | |
| EP3373798B1 (en) | Method and system for classifying optic nerve head | |
| CN109528155B (en) | Intelligent screening system suitable for high myopia complicated with open angle glaucoma and establishment method thereof | |
| CN107680683A (en) | A kind of AI eye healths appraisal procedure | |
| CN112101424B (en) | Method, device and equipment for generating retinopathy identification model | |
| Raja et al. | An automatic detection of blood vessel in retinal images using convolution neural network for diabetic retinopathy detection | |
| CN109464120A (en) | A kind of screening for diabetic retinopathy method, apparatus and storage medium | |
| CN208969732U (en) | The screening system of eye fundus image lesion based on artificial intelligence | |
| CN110101361A (en) | Based on big data on-line intelligence diagnostic platform and its operation method and storage medium | |
| CN108416371A (en) | A kind of diabetic retinopathy automatic testing method | |
| Ludwig et al. | Automatic identification of referral-warranted diabetic retinopathy using deep learning on mobile phone images | |
| Nayak et al. | Automatic identification of diabetic maculopathy stages using fundus images | |
| CN109994173A (en) | A kind of retinopathy monitoring system | |
| CN111292286B (en) | Analysis method and system for change of characteristic data of sugar mesh bottom and storage device | |
| CN112869697A (en) | Judgment method for simultaneously identifying stage and pathological change characteristics of diabetic retinopathy | |
| Suedumrong et al. | Application of deep convolutional neural networks vgg-16 and googlenet for level diabetic retinopathy detection | |
| CN204813807U (en) | Box -like traditional chinese medical science electron equipment of diagnosing of multiunit | |
| US11450433B2 (en) | System and method for remote diagnosis of disease progression | |
| CN113010722B (en) | Slow clinical research queue query system and method for fusing fundus images | |
| Kashmoola | Diagnosis Retinal Disease by using Deep Learning Models | |
| Shylaja | Algorithmic approach for prediction and early detection of diseases using retinal images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190902 Address after: 518000 District C, 3rd Building, 4th Floor, Tingwei Industrial Park, 6 Liufang Road, Xin'an Street, Baoan District, Shenzhen City, Guangdong Province Patentee after: Shenzhen silicon base Intelligent Technology Co., Ltd. Address before: 518000 District C, 3rd Building, 4th Floor, Tingwei Industrial Park, 6 Liufang Road, Xin'an Street, Baoan District, Shenzhen City, Guangdong Province Co-patentee before: SHENZHEN SIBIONICS CO., LTD. Patentee before: Shenzhen silicon base Intelligent Technology Co., Ltd. |
|
| TR01 | Transfer of patent right |