CN112446876A - anti-VEGF indication distinguishing method and device based on image and electronic equipment - Google Patents
anti-VEGF indication distinguishing method and device based on image and electronic equipment Download PDFInfo
- Publication number
- CN112446876A CN112446876A CN202011464274.8A CN202011464274A CN112446876A CN 112446876 A CN112446876 A CN 112446876A CN 202011464274 A CN202011464274 A CN 202011464274A CN 112446876 A CN112446876 A CN 112446876A
- Authority
- CN
- China
- Prior art keywords
- image
- discrimination
- vegf
- optical coherence
- coherence tomography
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012014 optical coherence tomography Methods 0.000 claims abstract description 59
- 238000012360 testing method Methods 0.000 claims abstract description 50
- 238000012545 processing Methods 0.000 claims abstract description 19
- 238000001514 detection method Methods 0.000 claims description 27
- 238000004422 calculation algorithm Methods 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 14
- 238000007635 classification algorithm Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004821 distillation Methods 0.000 claims description 6
- 238000012850 discrimination method Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 10
- 230000003902 lesion Effects 0.000 description 9
- 208000000208 Wet Macular Degeneration Diseases 0.000 description 8
- 238000002560 therapeutic procedure Methods 0.000 description 8
- 208000005590 Choroidal Neovascularization Diseases 0.000 description 5
- 206010060823 Choroidal neovascularisation Diseases 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 3
- 239000000243 solution Substances 0.000 description 3
- 206010038848 Retinal detachment Diseases 0.000 description 2
- 102000005789 Vascular Endothelial Growth Factors Human genes 0.000 description 2
- 108010019530 Vascular Endothelial Growth Factors Proteins 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 239000012530 fluid Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 210000002301 subretinal fluid Anatomy 0.000 description 2
- 108010041308 Endothelial Growth Factors Proteins 0.000 description 1
- 230000002137 anti-vascular effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000001771 impaired effect Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 150000002632 lipids Chemical class 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 208000015122 neurodegenerative disease Diseases 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 210000003583 retinal pigment epithelium Anatomy 0.000 description 1
- 230000037390 scarring Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/60—Rotation of whole images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
- G06T5/92—Dynamic range modification of images or parts thereof based on global image properties
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10101—Optical tomography; Optical coherence tomography [OCT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application provides an anti-VEGF indication distinguishing method and device based on an image and an electronic device, wherein the distinguishing method comprises the following steps: acquiring an optical coherence tomography image to be detected; carrying out data processing on the optical coherence tomography image to obtain one or more test images; inputting one or more test images into the trained one or more discrimination models to obtain the discrimination result of each test image by the one or more discrimination models; according to the discrimination result of one or more discrimination models for each test image, the anti-VEGF indication discrimination result of the optical coherence tomography image is determined by a voting method, so that a diagnosis suggestion can be automatically given, and the discrimination accuracy is improved.
Description
Technical Field
The invention relates to the field of image processing, in particular to an anti-VEGF indication distinguishing method and device based on an image, electronic equipment and a computer storage medium.
Background
Wet Age-related Macular Degeneration (wAMD) is an Age-related degenerative disease that occurs in the Macular region with rapid progression, if not timely treated, resulting in severely impaired vision, due to the formation of Choroidal Neovascularization (CNV), which can cause sub-Macular fluid and lipid leakage, fibrous scarring. In addition, the disease is complicated, difficult to diagnose and treat and has higher requirements on the professional level of ophthalmologists.
Intraocular injection of anti-Vascular Endothelial Growth Factor (VEGF) drugs is currently the first line treatment for wAMD, and indications for anti-VEGF therapy are at the discretion of the physician who analyzes a collection of Optical Coherence Tomography (OCT) images of the patient's macular region, and if any characteristic focus (representing choroidal neovascularization activity) is present in any of the OCT images in the collection, then it is deemed necessary to perform anti-VEGF therapy.
The interpretation of the OCT images by different ophthalmologists is different, which causes the treatment scheme to be non-uniform, thereby affecting the treatment effect. Conventional VEGF therapy requires a visit to a hospital registry for treatment by a physician. If the machine can be used for replacing a doctor to identify whether the patient needs anti-VEGF treatment, the process can be carried out in a primary hospital, and the time and cost for the patient to see a doctor are greatly saved. Therefore, an automatic judgment system for anti-VEGF indications is urgently needed to provide diagnosis suggestions and improve judgment accuracy.
Disclosure of Invention
The embodiment of the application provides an anti-VEGF indication distinguishing method based on an image, which is used for automatically giving a diagnosis suggestion and improving the distinguishing accuracy.
The embodiment of the application provides an anti-VEGF indication distinguishing method based on an image, which comprises the following steps:
acquiring an optical coherence tomography image to be detected;
carrying out data processing on the optical coherence tomography image to obtain one or more test images;
inputting the one or more test images into one or more trained discrimination models to obtain a discrimination result of each test image by the one or more discrimination models;
and determining the anti-VEGF indication discrimination result of the optical coherence tomography image by a voting method according to the discrimination result of the discrimination model or the discrimination models on each test image.
In an embodiment, the performing data processing on the optical coherence tomography image to obtain one or more test images includes:
and carrying out left-right turning, rotation, picture scaling, contrast change, Gaussian noise change, image cutting or brightness change on the optical coherence tomography image.
In one embodiment, prior to said inputting the one or more test images into the trained one or more discriminative models, the method further comprises:
acquiring a sample image marked with a characteristic focus;
and taking the sample image as the input of a target detection algorithm, and adjusting the parameters of the target detection algorithm to ensure that the focus detection result output by the target detection algorithm is the same as the marked characteristic focus to obtain the discrimination model.
In one embodiment, prior to said inputting the one or more test images into the trained one or more discriminative models, the method further comprises:
acquiring a sample image marked as wAMD positive and an optical coherence tomography sample image marked as having an anti-VEGF indication;
and taking the wAMD positive sample image and the optical coherence tomography sample image with the anti-VEGF indication as input of a classification algorithm, adjusting parameters of the classification algorithm, and enabling a classification result output by the classification algorithm to be the same as a marked result to obtain the discrimination model.
In an embodiment, after obtaining the discriminant model, the method further comprises:
and carrying out model distillation on the discrimination model to obtain a new discrimination model.
The embodiment of the application also provides a device for discriminating an anti-VEGF indication based on an image, which comprises:
the image acquisition module is used for acquiring an optical coherence tomography image to be detected;
the image processing module is used for carrying out data processing on the optical coherence tomography image to obtain one or more test images;
the image distinguishing module is used for inputting the one or more test images into one or more trained distinguishing models to obtain the distinguishing result of each test image by the one or more distinguishing models;
and the result determining module is used for determining the anti-VEGF indication distinguishing result of the optical coherence tomography image by a voting method according to the distinguishing result of the one or more distinguishing models to each test image.
In an embodiment, the image enhancement module is to: and carrying out left-right turning, rotation, picture scaling, contrast change, Gaussian noise change and brightness change on the optical coherence tomography image to obtain at least two test images.
In one embodiment, the apparatus further comprises:
the sample acquisition module is used for acquiring a sample image marked with a characteristic focus;
a model training module for taking the sample image as the input of the target detection algorithm, adjusting the parameters of the target detection algorithm to make the focus detection result output by the target detection algorithm the same as the marked characteristic focus, and obtaining the discrimination model
An embodiment of the present application further provides an electronic device, including:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to perform the image-based anti-VEGF indication discrimination method.
The embodiment of the application also provides a computer readable storage medium, wherein the storage medium stores a computer program, and the computer program can be executed by a processor to complete the anti-VEGF indication distinguishing method of the image.
According to the technical scheme provided by the embodiment of the application, one or more test images are obtained by processing the optical coherence tomography image, the one or more test images are input into one or more trained discrimination models, and the anti-VEGF indication discrimination result of the optical coherence tomography image is determined by a voting method based on the discrimination result of the one or more discrimination models for each test image, so that a diagnosis suggestion can be automatically given to improve the discrimination accuracy.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly described below.
Fig. 1 is a schematic view of an application scenario of an image-based anti-VEGF indication determination method according to an embodiment of the present application;
fig. 2 is a schematic diagram of an electronic device provided in an embodiment of the present application;
FIG. 3 is a schematic view of a complete flow chart of an image-based anti-VEGF indication determination method provided in the embodiment of the present application;
FIG. 4 is a schematic flow chart of a target detection algorithm model of the image-based anti-VEGF indication discrimination method provided in the embodiment of the present application;
fig. 5 is a block diagram of a determination apparatus of an image-based anti-VEGF indication determination method according to an embodiment of the present application;
fig. 6 is a block diagram of a device for discriminating an anti-VEGF indication method based on an image according to another embodiment of the present application based on the corresponding embodiment of fig. 5;
fig. 7 is a block diagram of a device for discriminating an anti-VEGF indication based on an image according to another embodiment of the present application based on the corresponding embodiment of fig. 6.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
In the description of the present application, suffixes such as "module", "means", or "unit" used to denote elements are used in the description only for the convenience of description of the present invention, and have no specific meaning in themselves.
Fig. 1 is a schematic view of an application scenario of an image-based anti-VEGF indication determination method according to an embodiment of the present application. As shown in fig. 1, the application scenario includes a client 110 and a server 120, where the client 110 may send an optical coherence tomography image to be processed to the server 120, and the server 120 may perform data processing on the optical coherence tomography image to obtain one or more test images. After the anti-VEGF indication discriminant model based on the image is trained, the server 120 may input one or more test images into the trained one or more discriminant models to obtain a result of the one or more discriminant models for each test image; and determining the anti-VEGF indication discrimination result of the optical coherence tomography image by a voting method according to the discrimination result of one or more discrimination models on each test image.
Fig. 2 is a schematic view of an electronic device provided in an embodiment of the present application. The electronic device 210 can be used as the server 120, and the electronic device 210 includes: a processor 230; a memory 220 for storing instructions executable by processor 230; wherein, the processor 230 is configured to execute the image-based anti-VEGF indication distinguishing method provided by the embodiment of the present application. A communication interface 240 for communication between the electronic device 210 and an external device; communication bus 250 provides for communication among memory 220, processor 230, and communication interface 240.
The Memory 220 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk or optical disk. The memory is also provided with a plurality of modules which are respectively executed by the processor so as to complete the steps of the anti-VEGF indication distinguishing method based on the image.
Embodiments of the present application also provide a computer-readable storage medium storing a computer program executable by processor 230 to perform the following image-based anti-VEGF indication determination method.
Fig. 3 is a complete flow chart of the image-based anti-VEGF indication determination method according to the embodiment of the present application. As shown in fig. 3, the determination method may include the following steps S310 to S340.
Step S310: and acquiring an optical coherence tomography image to be detected.
For the discrimination, the optical coherence tomography image for which it is unknown whether anti-VEGF therapy is required or not may be referred to as an optical coherence tomography image to be detected, and may also be referred to as an original optical coherence tomography image. Wherein, the optical coherence tomography image to be detected, namely the original optical coherence tomography image is used for the prediction process; an optical coherence tomography image, which is known to be in need of anti-VEGF therapy or not, may be referred to as a sample image, where the sample image is used for the training process.
Step S320: and carrying out data processing on the optical coherence tomography image to obtain one or more test images.
The data processing may be flipping the optical coherence tomography image left or right, rotating (e.g., 15 left or right), picture scaling, contrast change (e.g., [0.8,1.2]), gaussian noise change, image cropping, or brightness change. The optical coherence tomography image is transformed through a plurality of enhancing modes, and a plurality of test images can be obtained. Without distinction, the optical coherence tomography image is an image obtained by image processing, and is referred to herein as a test image.
Step S330: and inputting the one or more test images into the trained one or more discriminant models to obtain the discrimination result of each test image by the one or more discriminant models.
In one embodiment, one or more discriminant models may be trained in advance. The discriminant model training process can be seen in the following detailed discussion. The same test image can be input into a plurality of discrimination models to obtain discrimination results of the discrimination models on the test image. The discrimination is used to indicate whether there is an indication of anti-VEGF. Having an indication of anti-VEGF may alert the physician to anti-VEGF therapy.
Step S340: and determining the anti-VEGF indication discrimination result of the optical coherence tomography image by a voting method according to the discrimination result of the discrimination model or the discrimination models on each test image.
For example, the voting method may be that when two or more discrimination models among the plurality of discrimination models discriminate that the same optical coherence tomography image needs anti-VEGF therapy, the image is finally determined to need anti-VEGF therapy. In the discrimination results of a plurality of test images of the same optical coherence tomography image, if at least two test images are considered to have the anti-VEGF indication, the optical coherence tomography image can be considered to have the anti-VEGF indication. The result of the output of the discriminant model on the test image may be referred to as a discriminant result. The result of determining whether the optical coherence tomography image finally obtained by the voting method has an anti-VEGF indication may be referred to as an anti-VEGF indication determination result.
Fig. 4 is a schematic flow chart of the above discriminant model according to an embodiment of the present disclosure. In one embodiment, the discriminant model may be trained by a target detection algorithm. As shown in fig. 4, the training process of the discriminant model includes:
step S410: and acquiring a sample image marked with the characteristic focus.
Since the criterion for the evaluation of anti-VEGF indication is whether there are 4 characteristic lesions SRF (subretinal fluid), IRF (intraretinal fluid), PED (retinal pigment epithelium detachment) and SRHM (subretinal hyperreflective material) representing CNV activity, any OCT image showing any of the 4 characteristic lesions is considered to have an anti-VEGF indication. So here the sample image may be an optical coherence tomography image with a characteristic lesion location marked.
Step S420: and taking the sample image as the input of a target detection algorithm, and adjusting the parameters of the target detection algorithm to ensure that the focus detection result output by the target detection algorithm is the same as the marked characteristic focus to obtain the discrimination model.
The target detection algorithm may output a lesion location representative of CNV activity for the sample image. The training goal is to make the output lesion location as identical as possible to the already labeled lesion location.
Fig. 5 is a schematic flow chart of the above discriminant model according to another embodiment of the present disclosure. In one embodiment, the discriminant model may be trained by a classification algorithm. As shown in fig. 5, the training process of the discriminant model includes:
step S510: sample images marked as wAMD positive and optical coherence tomography sample images marked as having an indication of anti-VEGF were acquired.
The sample image labeled as wmda positive is an image in the absence of wet age-related macular degeneration.
Step S520: and taking the wAMD positive sample image and the optical coherence tomography sample image with the anti-VEGF indication as input of a classification algorithm, adjusting parameters of the classification algorithm, and enabling a classification result output by the classification algorithm to be the same as a marked result to obtain the discrimination model.
The classification algorithm can classify sample images as having an anti-VEGF indication, and as not having an anti-VEGF indication. The training objective is to make the classification result (e.g., anti-VEGF indication or no anti-VEGF indication) of the sample image as identical as possible to the actual labeling result.
In an embodiment, the discriminant model can also be implemented by using an example segmentation algorithm or a semantic segmentation algorithm, and whether the specific anti-VEGF treatment indication is obtained according to the segmentation result.
In one embodiment, the sample image may be pre-processed, and the data pre-processing includes two parts, namely data cleaning and data enhancement. And carrying out model training and testing on the preprocessed image.
The data cleaning refers to screening out the image with the wrong label. The data enhancement can perform left-right turning, rotation, picture scaling, contrast change, Gaussian noise change, image cropping or brightness change on the optical coherence tomography image, wherein the data enhancement is used for improving the generalization capability of the model.
In one embodiment, a cross-validation approach may be used to train multiple discriminant models. The Backbone network adopts Resnet series and ResNeXt series, model parameters are initialized by using pre-training model parameters of ImageNet, and because semantic information difference between pictures and medical pictures in ImageNet is large, only the model parameter of the first stage is frozen during training, and other parameters are optimized on the current data set. And finally, selecting the first 3 discrimination models with the highest accuracy rate in the cross validation process for prediction. Inputting a predicted image into three discrimination models, outputting 3 results (whether anti-VEGF treatment is needed or not), and finally determining a final result by adopting a voting method (when two or more than two models consider that anti-VEGF treatment is needed, the final result is that anti-VEGF treatment is needed). The backbone in the embodiment of the application adopts Resnet101 with the highest accuracy.
After the discriminant model is trained, model distillation can be performed on the discriminant model to obtain a new discriminant model for deployment.
The model distillation is to train a complex model (teacher) with high accuracy, then retrain a simple model, and use the simple model (student) to fit the output of the complex model (teacher) in the training process. In one embodiment, the known result and the image to be detected can be combined into a new data set to be placed in another simple discriminant model for learning by using the output result of the discriminant model trained in advance as the known result.
In the stage of model deployment, the discriminant model needs to be deployed on hardware equipment, at this time, factors such as calculation complexity, model size and speed need to be further considered, and in order to obtain better trade-off of model size and accuracy, the embodiment of the present application performs model distillation (Resnet50 distillation Resnet101) on the original integrated model, so that the accuracy is slightly reduced compared with the original accuracy, the hardware requirement of model deployment is reduced, and the prediction speed is increased.
Fig. 6 is a block diagram of an apparatus for discriminating an anti-VEGF indication based on an image according to an embodiment of the present application. As shown in fig. 6, the apparatus includes: an image acquisition module 610, an image processing module 620, an image discrimination module 630, and a result determination module 640.
An image acquisition module 610, configured to acquire an optical coherence tomography image to be detected;
an image processing module 620, configured to perform data processing on the optical coherence tomography image to obtain one or more test images;
an image discrimination module 630, configured to input the one or more test images into one or more trained discrimination models, so as to obtain a discrimination result of each test image by the one or more discrimination models;
and a result determining module 640, configured to determine, according to a determination result of the one or more determination models for each test image, an anti-VEGF indication determination result of the optical coherence tomography image by a voting method.
In one embodiment, the data processing method includes: and carrying out left-right turning, rotation, picture scaling, contrast change, Gaussian noise change, image cutting or brightness change on the optical coherence tomography image.
Fig. 7 is a block diagram of a device for discriminating an anti-VEGF indication based on an image according to another embodiment of the present application based on the corresponding embodiment of fig. 6. As shown in fig. 7, the apparatus includes: a sample acquisition module 710 and a model training module 720.
A sample acquiring module 710, configured to acquire a sample image with a characteristic lesion marked;
and the model training module 720 is configured to use the sample image as an input of a target detection algorithm, adjust parameters of the target detection algorithm, and obtain the discrimination model by using a lesion detection result output by the target detection algorithm to be the same as the labeled characteristic lesion.
The implementation processes of the functions and actions of the modules in the device are specifically described in the implementation processes of the corresponding steps in the image-based anti-VEGF indication distinguishing method, and are not described herein again.
According to the technical scheme, the diagnosis suggestion can be automatically given, the judgment accuracy is improved, the patient treatment time can be saved after practical application, and the patient treatment experience is improved. The scheme can replace a doctor to perform image recognition through a machine, namely whether a patient corresponding to the image needs anti-VEGF treatment or not is recognized from the image, so that the time and cost for the patient to see a doctor are greatly saved, and the scheme belongs to domestic initiatives.
In the embodiments provided in the present application, the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
Claims (10)
1. An image-based anti-VEGF indication discrimination method, comprising:
acquiring an optical coherence tomography image to be detected;
carrying out data processing on the optical coherence tomography image to obtain one or more test images;
inputting the one or more test images into one or more trained discrimination models to obtain a discrimination result of each test image by the one or more discrimination models;
and determining the anti-VEGF indication discrimination result of the optical coherence tomography image by a voting method according to the discrimination result of the discrimination model or the discrimination models on each test image.
2. The method of claim 1, wherein said data processing the optical coherence tomography image to obtain one or more test images comprises:
and carrying out left-right turning, rotation, picture scaling, contrast change, Gaussian noise change, image cutting or brightness change on the optical coherence tomography image.
3. The method of claim 1, wherein prior to said inputting the one or more test images into the trained one or more discriminant models, the method further comprises:
acquiring a sample image marked with a characteristic focus;
and taking the sample image as the input of a target detection algorithm, and adjusting the parameters of the target detection algorithm to ensure that the focus detection result output by the target detection algorithm is the same as the marked characteristic focus to obtain the discrimination model.
4. The method of claim 1, wherein prior to said inputting the one or more test images into the trained one or more discriminant models, the method further comprises:
acquiring a sample image marked as wAMD positive and an optical coherence tomography sample image marked as having an anti-VEGF indication;
and taking the wAMD positive sample image and the optical coherence tomography sample image with the anti-VEGF indication as input of a classification algorithm, adjusting parameters of the classification algorithm, and enabling a classification result output by the classification algorithm to be the same as a marked result to obtain the discrimination model.
5. The method of claim 3 or 4, wherein after obtaining the discriminative model, the method further comprises:
and carrying out model distillation on the discrimination model to obtain a new discrimination model.
6. An apparatus for image-based anti-VEGF indication discrimination, comprising:
the image acquisition module is used for acquiring an optical coherence tomography image to be detected;
the image processing module is used for carrying out data processing on the optical coherence tomography image to obtain one or more test images;
the image distinguishing module is used for inputting the one or more test images into one or more trained distinguishing models to obtain the distinguishing result of each test image by the one or more distinguishing models;
and the result determining module is used for determining the anti-VEGF indication distinguishing result of the optical coherence tomography image by a voting method according to the distinguishing result of the one or more distinguishing models to each test image.
7. The apparatus of claim 6, wherein the data processing means comprises: and carrying out left-right turning, rotation, picture scaling, contrast change, Gaussian noise change, image cutting or brightness change on the optical coherence tomography image.
8. The apparatus of claim 6, further comprising:
the sample acquisition module is used for acquiring a sample image marked with a characteristic focus;
and the model training module is used for taking the sample image as the input of a target detection algorithm, adjusting the parameters of the target detection algorithm, enabling the focus detection result output by the target detection algorithm to be the same as the marked characteristic focus, and obtaining the discrimination model.
9. An electronic device, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to perform a method of anti-VEGF indication discrimination of an image according to any one of claims 1 to 7.
10. A computer-readable storage medium, wherein the storage medium stores a computer program, the computer program being executable by a processor to perform a method for anti-VEGF indication determination of an image according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011464274.8A CN112446876A (en) | 2020-12-11 | 2020-12-11 | anti-VEGF indication distinguishing method and device based on image and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011464274.8A CN112446876A (en) | 2020-12-11 | 2020-12-11 | anti-VEGF indication distinguishing method and device based on image and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112446876A true CN112446876A (en) | 2021-03-05 |
Family
ID=74739605
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011464274.8A Pending CN112446876A (en) | 2020-12-11 | 2020-12-11 | anti-VEGF indication distinguishing method and device based on image and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112446876A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113768452A (en) * | 2021-09-16 | 2021-12-10 | 重庆金山医疗技术研究院有限公司 | Intelligent timing method and device for electronic endoscope |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180132725A1 (en) * | 2015-05-05 | 2018-05-17 | Medizinische Universität Wien | Computerized Device and Method for Processing Image Data |
| CN108346145A (en) * | 2018-01-31 | 2018-07-31 | 浙江大学 | The recognition methods of unconventional cell in a kind of pathological section |
| CN110010219A (en) * | 2019-03-13 | 2019-07-12 | 杭州电子科技大学 | Optical coherence tomography image retinopathy intelligent checking system and detection method |
| CN110021022A (en) * | 2019-02-21 | 2019-07-16 | 哈尔滨理工大学 | A kind of thyroid gland nuclear medical image diagnostic method based on deep learning |
| CN110363226A (en) * | 2019-06-21 | 2019-10-22 | 平安科技（深圳）有限公司 | Ophthalmology disease classifying identification method, device and medium based on random forest |
| JP2020093076A (en) * | 2018-08-14 | 2020-06-18 | キヤノン株式会社 | Medical image processing device, learned model, medical image processing method and program |
-
2020
- 2020-12-11 CN CN202011464274.8A patent/CN112446876A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180132725A1 (en) * | 2015-05-05 | 2018-05-17 | Medizinische Universität Wien | Computerized Device and Method for Processing Image Data |
| CN108346145A (en) * | 2018-01-31 | 2018-07-31 | 浙江大学 | The recognition methods of unconventional cell in a kind of pathological section |
| JP2020093076A (en) * | 2018-08-14 | 2020-06-18 | キヤノン株式会社 | Medical image processing device, learned model, medical image processing method and program |
| CN110021022A (en) * | 2019-02-21 | 2019-07-16 | 哈尔滨理工大学 | A kind of thyroid gland nuclear medical image diagnostic method based on deep learning |
| CN110010219A (en) * | 2019-03-13 | 2019-07-12 | 杭州电子科技大学 | Optical coherence tomography image retinopathy intelligent checking system and detection method |
| CN110363226A (en) * | 2019-06-21 | 2019-10-22 | 平安科技（深圳）有限公司 | Ophthalmology disease classifying identification method, device and medium based on random forest |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113768452A (en) * | 2021-09-16 | 2021-12-10 | 重庆金山医疗技术研究院有限公司 | Intelligent timing method and device for electronic endoscope |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kauppi et al. | The diaretdb1 diabetic retinopathy database and evaluation protocol. | |
| KR20200005409A (en) | Fundus image management device and method for determining suitability of fundus image | |
| Deepak et al. | Automatic assessment of macular edema from color retinal images | |
| Tennakoon et al. | Image quality classification for DR screening using convolutional neural networks | |
| KR102313143B1 (en) | Diabetic retinopathy detection and severity classification apparatus Based on Deep Learning and method thereof | |
| KR20200023029A (en) | Diagnosis assistance system and control method thereof | |
| Sousa et al. | Automatic segmentation of retinal layers in OCT images with intermediate age-related macular degeneration using U-Net and DexiNed | |
| Rahim et al. | Automatic screening and classification of diabetic retinopathy fundus images | |
| US10058243B2 (en) | Clinic evaluation via outer retinal layer assessment | |
| CN116848588A (en) | Automatic labeling of health features in medical images | |
| KR102569285B1 (en) | Method and system for training machine learning model for detecting abnormal region in pathological slide image | |
| CN112446876A (en) | anti-VEGF indication distinguishing method and device based on image and electronic equipment | |
| Giancardo | Automated fundus images analysis techniques to screen retinal diseases in diabetic patients | |
| Aurangzeb et al. | Systematic development of AI-Enabled diagnostic systems for glaucoma and diabetic retinopathy | |
| Mohammedhasan et al. | A new deeply convolutional neural network architecture for retinal blood vessel segmentation | |
| US20220036558A1 (en) | Method and system for predicting expression of biomarker from medical image | |
| Niemeijer | Automatic detection of diabetic retinopathy in digital fundus photographs | |
| Azeroual et al. | Convolutional Neural Network for Segmentation and Classification of Glaucoma. | |
| Sutradhar et al. | Blind-spot network for image anomaly detection: A new approach to diabetic retinopathy screening. | |
| WO2019082203A1 (en) | A system and method for detection and classification of retinal disease | |
| Civit Masot et al. | Multi-dataset training for medical image segmentation as a service | |
| Moorthi et al. | Automatic detection of new vessels in retinal photographs using machine learning | |
| Aris et al. | Diabetic Retinopathy Detection Using Grey Level Co-Occurrence Matrix | |
| Silva | Automatic detection of cataract in fundus images | |
| Amin et al. | A Retinal Blood Vessel Segmentation in Matlab |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |