CN114898172B - Multi-feature DAG network-based diabetic retinopathy classification modeling method - Google Patents
Multi-feature DAG network-based diabetic retinopathy classification modeling method Download PDFInfo
- Publication number
- CN114898172B CN114898172B CN202210362909.6A CN202210362909A CN114898172B CN 114898172 B CN114898172 B CN 114898172B CN 202210362909 A CN202210362909 A CN 202210362909A CN 114898172 B CN114898172 B CN 114898172B
- Authority
- CN
- China
- Prior art keywords
- image
- diabetic retinopathy
- characteristic
- features
- retinal
- 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
- 206010012689 Diabetic retinopathy Diseases 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 title claims abstract description 13
- 238000012549 training Methods 0.000 claims abstract description 23
- 230000002792 vascular Effects 0.000 claims abstract description 17
- 230000002207 retinal effect Effects 0.000 claims abstract description 16
- 206010029113 Neovascularisation Diseases 0.000 claims abstract description 11
- 230000004927 fusion Effects 0.000 claims abstract description 8
- 210000001525 retina Anatomy 0.000 claims description 23
- 208000037111 Retinal Hemorrhage Diseases 0.000 claims description 13
- 210000004204 blood vessel Anatomy 0.000 claims description 9
- 230000004044 response Effects 0.000 claims description 6
- 230000004256 retinal image Effects 0.000 claims description 5
- 208000017442 Retinal disease Diseases 0.000 claims description 4
- 206010038923 Retinopathy Diseases 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 230000036285 pathological change Effects 0.000 claims description 4
- 231100000915 pathological change Toxicity 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000000740 bleeding effect Effects 0.000 abstract description 3
- 238000013145 classification model Methods 0.000 abstract description 3
- 230000003902 lesion Effects 0.000 abstract description 3
- 206010012601 diabetes mellitus Diseases 0.000 description 8
- 238000012360 testing method Methods 0.000 description 3
- WQZGKKKJIJFFOK-GASJEMHNSA-N Glucose Natural products OC[C@H]1OC(O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-GASJEMHNSA-N 0.000 description 2
- 208000032843 Hemorrhage Diseases 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000008103 glucose Substances 0.000 description 2
- 230000002062 proliferating effect Effects 0.000 description 2
- 206010060891 General symptom Diseases 0.000 description 1
- 206010018429 Glucose tolerance impaired Diseases 0.000 description 1
- 206010020710 Hyperphagia Diseases 0.000 description 1
- 208000009857 Microaneurysm Diseases 0.000 description 1
- 208000034698 Vitreous haemorrhage Diseases 0.000 description 1
- 208000034158 bleeding Diseases 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 210000000416 exudates and transudate Anatomy 0.000 description 1
- 208000030533 eye disease Diseases 0.000 description 1
- 210000002747 omentum Anatomy 0.000 description 1
- 238000011369 optimal treatment Methods 0.000 description 1
- 206010036067 polydipsia Diseases 0.000 description 1
- 208000022530 polyphagia Diseases 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 210000002700 urine Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
- A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
- A61B3/12—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions for looking at the eye fundus, e.g. ophthalmoscopes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention discloses a multi-feature DAG network-based diabetic retinopathy classification modeling method, which comprises the steps of firstly extracting index features of diabetic retinopathy by using different methods, wherein the index features comprise bleeding spot features, fundus neovascularization features and retinal vascular curvature Zhang Tezheng; secondly, constructing an optimized DAG network, continuously changing a training scheme to train the network, realizing multi-feature fusion of the extracted features, and forming complex local or global features through the local features so as to restore the object; finally, normal and lesion classification is performed by a softmax classifier. The invention uses the DIARETDB1 data set and the third people hospital data in Dalian city to evaluate the performance of the classification model, and the result shows that the classification accuracy of the DIARETDB1 data set image and the hospital data is 98.7 percent and 98.5 percent respectively.
Description
Technical Field
The invention relates to the field of medical image processing, in particular to a diabetic retinopathy classification modeling method based on a multi-feature DAG network.
Background
Diabetic retinopathy is a now more severe blinding eye disease. Diabetic retinopathy is divided into stages VI, the first three stages being non-proliferative and the second three stages being proliferative. In addition to the general symptoms of diabetes, the diabetic patients have the symptoms of polydipsia, polyphagia, urine glucose and blood glucose elevation, and have eyeground changes with punctate bleeding, neovascularization and vascular varicose as main characteristics, so the characteristics of the diabetic retina in each period have great significance for diagnosis and estimation prognosis of diabetes. In the past, ophthalmologists manually evaluate fundus images according to the characteristics of diabetic retinopathy to detect whether the retina is diseased or not, but because part of patients are in the pre-diabetic retinopathy stage and the characteristics are not obvious, the problem that the optimal treatment time is missed due to inaccurate evaluation easily occurs. There is a need for a rapid and accurate automatic identification and classification system for diabetic retinopathy images that also identifies those images that are not distinct.
At present, classification of retinal images is realized by adopting a characteristic extraction mode, such as microaneurysm characteristic, exudate characteristic, vitreous hemorrhage characteristic and the like, and is not carried out on diabetic retinal hemorrhage spot characteristic, fundus neovascularization characteristic and retinal vascular varicose characteristic. Meanwhile, the existing research only extracts one or two kinds of characteristic information, so that the classification model cannot learn carefully and comprehensively when performing characteristic learning, and the classification accuracy is low.
Disclosure of Invention
The invention aims to solve the technical problems in the prior art and provides a diabetic retinopathy classification modeling method based on a multi-feature DAG network.
The technical scheme of the invention is as follows: a diabetes retina pathological change classification modeling method based on multi-feature DAG network is carried out according to the following steps:
step 1: preprocessing each retina image in the training set to obtain a characteristic image training set
Step 1.1 obtaining a characteristic image of the retinal hemorrhages
Extracting a green channel image in an RGB color mode of a retina image, graying the green channel image and dividing the green channel image into a plurality of sub-blocks, counting a cumulative distribution histogram of each sub-block, and setting a limited threshold T in the histogram c :
T c =max(1,T d ×h×w/S)
Wherein T is d Is an iterative adaptive soft threshold, S is the total pixels of the image, h and w are the length and width of the image;
the gray value in the histogram is combined with a set finite threshold T c Comparing, to compare the histogram with the threshold value T c Uniformly distributing gray value areas under the histogram, ensuring the total area of the histogram to be unchanged, and finally optimizing by using a linear interpolation method to make the characteristics of the retinal hemorrhages prominent, so as to obtain a retinal hemorrhages characteristic image;
step 2.2 obtaining fundus neovascularization characteristic images
First, 8 masks M are used for each pixel point in the retina image q Q=1, 2,..7 convolutionally derivative was performedEach mask M q Maximum response is made for 8 specific edge directions, and the maximum value G of the maximum response is taken as the output of the edge amplitude image:
G=max{|M 0 |,|M 1 |,|M 2 |,|M 3 |,|M 4 |,|M 5 |,|M 6 |,|M 7 |}
finally, binarizing the image according to the self-adaptive soft threshold value to make the characteristics of the new blood vessel stand out, namely obtaining a fundus new blood vessel characteristic image;
step 3.3 obtaining a retinal vascular curvature Zhang Tezheng image
Dividing the retina image into N sub-blocks, and iteratively calculating the clustering center C of the retina blood vessels in the retinopathy image v And corresponding membership degree D pv :
Wherein N is the total number of sub-blocks, C is the number of clusters, x p P=1, 2,..n represents the p-th sub-block, || denotes a measure of arbitrary distance, m e 1, ++) belongs to a weighted index; c (C) k A cluster center representing a kth class; stopping iteration when the membership degree meets the following iteration termination condition, and calculating a local optimal value J m Making retinal vascular curvature Zhang Tezheng stand out to obtain a retinal vascular curvature Zhang Tezheng image;
wherein, l is the iteration step number, epsilon is the error threshold;
step 3.4, taking the acquired retina characteristic image set as a characteristic image training set;
step 2: inputting the characteristic images in the characteristic image training set into a DAG network for training
Step 2.1, an optimized DAG network is established, the optimized DAG network is composed of one trunk, two branches, an add layer, a pooling layer avpool and a Full connection layer Full connection, the trunk is divided into five groups, each group is composed of a convolution layer Conv, a normalization layer BN and an activation function layer relu, the two branches are convolution layers skip Conv, and the trunk and the branches are simultaneously linked with the add layer;
step 2.2, inputting the images of the feature image training set into an optimized DAG network to realize multi-feature fusion and learning, and outputting a multi-feature fusion result F i add :
F i add =(X i +Y i +Z i )*K=X i *K+Y i *K+Z i *K
Where K represents the convolution layer convolution kernel, X represents the convolution, X i Characteristic of retinal hemorrhages, Y i Representing the characteristics of ocular fundus neovascularization; z is Z i Represents retinal vascular curvature Zhang Tezheng;
step 3: fusing the multiple features to the result F i add Sending the diabetic retinopathy to a softmax classifier, and calculating the prediction probability of the diabetic retinopathy or the diabetic retinopathy according to the following formula, so as to realize effective classification of the diabetic retinopathy and the diabetic retinopathy;
wherein R is a predictive probability, i=1, 2,..m represents an i-th image, M is the total number of retinal images, and e is a parameter;
and when Y epsilon (0, 0.5) judges that the image is normal, and when Y epsilon [0.5, 1) judges that the image is pathological change.
The invention firstly uses different methods to extract the index characteristics of diabetic retinopathy, including the characteristics of bleeding spots, ocular fundus neovascularization and retinal vascular curvature Zhang Tezheng; secondly, constructing an optimized DAG network, continuously changing a training scheme to train the network, realizing multi-feature fusion of the extracted features, and forming complex local or global features through the local features so as to restore the object; finally, normal and lesion classification is performed by a softmax classifier. The invention uses the DIARETDB1 data set and the third people hospital data in Dalian city to evaluate the performance of the classification model, and the result shows that the classification accuracy of the DIARETDB1 data set image and the hospital data is 98.7 percent and 98.5 percent respectively.
Drawings
FIG. 1 is a diagram showing the result of extracting omentum plaque characteristics according to an embodiment of the present invention.
Fig. 2 is a graph showing the result of extracting fundus neovascularization characteristics in the embodiment of the present invention.
Fig. 3 is a graph showing the result of extracting retinal vascular varicose features according to an embodiment of the present invention.
FIG. 4 is a block diagram of a DAG network optimized in accordance with an embodiment of the present invention.
Fig. 5 is an overall flow chart of an embodiment of the present invention.
Detailed Description
The invention discloses a multi-feature DAG network-based diabetic retinopathy classification modeling method, which is carried out according to the following steps as shown in figure 5:
step 1: taking DIARETDB1 data set and Dalian third people hospital data set (hospital data for short) to divide into training set and test set, preprocessing each diabetic retina image in the training set to obtain characteristic image training set
Step 1.1 obtaining a characteristic image of the retinal hemorrhages
Extracting a green channel image in an RGB color mode of a retina image, graying the green channel image and dividing the green channel image into a plurality of sub-blocks, counting a cumulative distribution histogram of each sub-block, and setting a limited threshold T in the histogram c :
T c =max(1,T d ×h×w/S)
Wherein T is d Is an iterative adaptive soft threshold, S is the total image pixel, h and w areThe length and width of the image;
the gray value in the histogram is combined with a set finite threshold T c Comparing, to compare the histogram with the threshold value T c Uniformly distributing gray value regions under the histogram and ensuring that the total area of the histogram is unchanged, and finally optimizing each sub-block transition problem by using a linear interpolation method to make the characteristics of the retinal hemorrhages prominent, namely obtaining a retinal hemorrhages characteristic image, as shown in figure 1;
step 2.2 obtaining fundus neovascularization characteristic images
First, 8 masks M are used for each pixel point in the retina image q Q=1, 2,..7 convolved to derivative, each mask M q Maximum response is made for 8 specific edge directions, and the maximum value G of the maximum response is taken as the output of the edge amplitude image:
G=max{|M 0 |,|M 1 |,|M 2 |,|M 3 |,|M 4 |,|M 5 |,|M 6 |,|M 7 |}
finally, binarizing the image according to the self-adaptive soft threshold value to make the characteristics of the new blood vessels stand out, namely obtaining a fundus new blood vessel characteristic image, as shown in figure 2;
step 3.3 obtaining a retinal vascular curvature Zhang Tezheng image
Dividing the retina image into N sub-blocks, and iteratively calculating the clustering center C of the retina blood vessels in the retinopathy image v And corresponding membership degree D pv :
Wherein N is the total number of sub-blocks, C is the number of clusters, x p P=1, 2,..n represents the p-th sub-block, the metric of arbitrary distance is represented, m e 1, ++) belongs to a weighted index; c (C) k Represents the kthA cluster center of the class; stopping iteration when the membership degree meets the following iteration termination condition, and calculating a local optimal value J m The retinal vascular curvature Zhang Tezheng is highlighted and a retinal vascular curvature Zhang Tezheng image is obtained as shown in fig. 3;
wherein, l is the iteration step number, epsilon is the error threshold;
step 3.4, taking the acquired retina characteristic image set as a characteristic image training set;
step 2: inputting the characteristic images in the characteristic image training set into a DAG network for training
Step 2.1, an optimized DAG network is established, wherein the optimized DAG network is composed of a trunk, two branches, an add layer, a pooling layer avpool and a Full connection layer Full connection (fc) as shown in fig. 4, the trunk is divided into five groups, each group is composed of a convolution layer Conv, a normalization layer BN and an activation function layer relu, the five groups are respectively composed of a convolution layer Conv1-5, a normalization layer BN1-5 and an activation function layer relu1-5, the two branches are respectively a convolution layer skip Conv-1 and a skip Conv-2, and the trunk and the two branches are simultaneously linked with the add layer;
step 2.2, inputting the images of the feature image training set into a DAG network to realize multi-feature fusion and learning, and outputting a multi-feature fusion result F i add :
F i add =(X i +Y i +Z i )*K=X i *K+Y i *K+Z i *K
Where K represents the convolution layer convolution kernel, X represents the convolution, X i Characteristic of retinal hemorrhages, Y i Representing the characteristics of ocular fundus neovascularization; z is Z i Represents retinal vascular curvature Zhang Tezheng;
the training parameters and training parameter values are shown in Table 1.
TABLE 1
Step 3: fusing the multiple features to the result F i add Sending the diabetic retinopathy to a softmax classifier, and calculating the prediction probability of the diabetic retinopathy or the diabetic retinopathy according to the following formula, so as to realize effective classification of the diabetic retinopathy and the diabetic retinopathy;
wherein R is a predictive probability, i=1, 2,..m represents an i-th image, M is the total number of diabetic retinal images, and e is a parameter;
and when Y epsilon (0, 0.5) judges that the image is normal, and when Y epsilon [0.5, 1) judges that the image is pathological change.
Experiment:
the test set images in the DIARETDB1 data set and the Dalian third people hospital data set (hospital data set for short) are input into the model established by the embodiment of the invention, and the model identifies the retina images of the data set and classifies the retina images into two major categories, namely a normal fundus image and a lesion fundus image. Aiming at the retinal image classification problem, the evaluation indexes are classification accuracy (accuracy), precision (Precision), recall (Recall), specificity (Specificity) and F1-score obtained after the data test set image enters the model to jointly evaluate the performance of the model, so that the model is more stable and reliable. The correlation formula is as follows:
where TP represents the number of positive samples correctly classified; TN represents the number of correctly classified negative samples; FP represents the number of negative samples that are falsely marked as positive; FN represents the number of positive samples that are falsely marked as negative.
In order to prove the importance of the extracted features, the invention uses hospital data to respectively carry out experimental result comparison by using self algorithms for extracting only one or two features of the non-extracted features. The comparison results are shown in Table 2.
TABLE 2
Meanwhile, in order to verify the performance of the model of the present invention, the same dataset DIARETDB1 was selected for performance comparison between the conventional model and the model of the present invention, and the comparison results are shown in Table 3.
TABLE 3 Table 3
Note that: 'to' represents a deficiency.
Claims (1)
1. A multi-feature DAG network-based diabetic retinopathy classification modeling method is characterized by comprising the following steps:
step 1: preprocessing each retina image in the training set to obtain a characteristic image training set
Step 1.1 obtaining a characteristic image of the retinal hemorrhages
Extracting a green channel image in an RGB color mode of a retina image, graying the green channel image and dividing the green channel image into a plurality of sub-blocks, counting a cumulative distribution histogram of each sub-block, and setting a limited threshold T in the histogram c :
T c =max(1,T d ×h×w/S)
Wherein T is d Is an iterative adaptive soft threshold, S is the total pixels of the image, h and w are the length and width of the image;
the gray value in the histogram is combined with a set finite threshold T c Comparing, to compare the histogram with the threshold value T c Uniformly distributing gray value areas under the histogram, ensuring the total area of the histogram to be unchanged, and finally optimizing by using a linear interpolation method to make the characteristics of the retinal hemorrhages prominent, so as to obtain a retinal hemorrhages characteristic image;
step 2.2 obtaining fundus neovascularization characteristic images
First, 8 masks M are used for each pixel point in the retina image q Q=1, 2,..7 convolved to derivative, each mask M q Maximum response is made for 8 specific edge directions, and the maximum value G of the maximum response is taken as the output of the edge amplitude image:
G=max{|M 0 |,|M 1 |,|M 2 |,|M 3 |,|M 4 |,|M 5 |,|M 6 |,|M 7 |}
finally, binarizing the image according to the self-adaptive soft threshold value to make the characteristics of the new blood vessel stand out, namely obtaining a fundus new blood vessel characteristic image;
step 3.3 obtaining a retinal vascular curvature Zhang Tezheng image
Dividing the retina image into N sub-blocks, and iteratively calculating the clustering center C of the retina blood vessels in the retinopathy image v And corresponding membership degree D pv :
Wherein N is the total number of sub-blocks, C is the number of clusters, x p P=1, 2,..n represents the p-th sub-block, || denotes a measure of arbitrary distance, m e 1, ++) belongs to a weighted index; c (C) k A cluster center representing a kth class; stopping iteration when the membership degree meets the following iteration termination condition, and calculating a local optimal value J m Making retinal vascular curvature Zhang Tezheng stand out to obtain a retinal vascular curvature Zhang Tezheng image;
wherein, l is the iteration step number, epsilon is the error threshold;
step 3.4, taking the acquired retina characteristic image set as a characteristic image training set;
step 2: inputting the characteristic images in the characteristic image training set into a DAG network for training
Step 2.1, an optimized DAG network is established, the optimized DAG network is composed of one trunk, two branches, an add layer, a pooling layer avpool and a Full connection layer Full connection, the trunk is divided into five groups, each group is composed of a convolution layer Conv, a normalization layer BN and an activation function layer relu, the two branches are convolution layers skip Conv, and the trunk and the branches are simultaneously linked with the add layer;
step 2.2, inputting the images of the feature image training set into an optimized DAG network to realize multi-feature fusion and learning, and outputting a multi-feature fusion result F i add :
F i add =(X i +Y i +Z i )*K=X i *K+Y i *K+Z i *K
Where K represents the convolution layer convolution kernel, X represents the convolution, X i Characteristic of retinal hemorrhages, Y i Representing the characteristics of ocular fundus neovascularization; z is Z i Represents retinal vascular curvature Zhang Tezheng;
step 3: fusing the multiple features to the result F i add Sending the diabetic retinopathy to a softmax classifier, and calculating the prediction probability of the diabetic retinopathy or the diabetic retinopathy according to the following formula, so as to realize effective classification of the diabetic retinopathy and the diabetic retinopathy;
wherein R is a predictive probability, i=1, 2,..m represents an i-th image, M is the total number of retinal images, and e is a parameter;
and when Y epsilon (0, 0.5) judges that the image is normal, and when Y epsilon [0.5, 1) judges that the image is pathological change.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210362909.6A CN114898172B (en) | 2022-04-08 | 2022-04-08 | Multi-feature DAG network-based diabetic retinopathy classification modeling method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210362909.6A CN114898172B (en) | 2022-04-08 | 2022-04-08 | Multi-feature DAG network-based diabetic retinopathy classification modeling method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114898172A CN114898172A (en) | 2022-08-12 |
CN114898172B true CN114898172B (en) | 2024-04-02 |
Family
ID=82715928
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210362909.6A Active CN114898172B (en) | 2022-04-08 | 2022-04-08 | Multi-feature DAG network-based diabetic retinopathy classification modeling method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114898172B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108021916A (en) * | 2017-12-31 | 2018-05-11 | 南京航空航天大学 | Deep learning diabetic retinopathy sorting technique based on notice mechanism |
CN110210570A (en) * | 2019-06-10 | 2019-09-06 | 上海延华大数据科技有限公司 | The more classification methods of diabetic retinopathy image based on deep learning |
WO2019196268A1 (en) * | 2018-04-13 | 2019-10-17 | 博众精工科技股份有限公司 | Diabetic retina image classification method and system based on deep learning |
-
2022
- 2022-04-08 CN CN202210362909.6A patent/CN114898172B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108021916A (en) * | 2017-12-31 | 2018-05-11 | 南京航空航天大学 | Deep learning diabetic retinopathy sorting technique based on notice mechanism |
WO2019196268A1 (en) * | 2018-04-13 | 2019-10-17 | 博众精工科技股份有限公司 | Diabetic retina image classification method and system based on deep learning |
CN110210570A (en) * | 2019-06-10 | 2019-09-06 | 上海延华大数据科技有限公司 | The more classification methods of diabetic retinopathy image based on deep learning |
Non-Patent Citations (1)
Title |
---|
李琼 ; 柏正尧 ; 刘莹芳 ; .糖尿病性视网膜图像的深度学习分类方法.中国图象图形学报.2018,(第10期),第1594-1603 页. * |
Also Published As
Publication number | Publication date |
---|---|
CN114898172A (en) | 2022-08-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Asiri et al. | Deep learning based computer-aided diagnosis systems for diabetic retinopathy: A survey | |
CN109325942B (en) | Fundus image structure segmentation method based on full convolution neural network | |
Gayathri et al. | Diabetic retinopathy classification based on multipath CNN and machine learning classifiers | |
JP2021531098A (en) | Systems and methods for determining eye condition using AI | |
CN109635862B (en) | Sorting method for retinopathy of prematurity plus lesion | |
Tian et al. | Multi-path convolutional neural network in fundus segmentation of blood vessels | |
Lu et al. | Automatic optic disc detection from retinal images by a line operator | |
Biswas et al. | Dilated deep neural network for segmentation of retinal blood vessels in fundus images | |
CN112465772B (en) | Fundus colour photographic image blood vessel evaluation method, device, computer equipment and medium | |
Pathan et al. | Automated segmentation and classifcation of retinal features for glaucoma diagnosis | |
Ghosh et al. | Grid color moment features in glaucoma classification | |
CN110751636B (en) | Fundus image retinal arteriosclerosis detection method based on improved coding and decoding network | |
CN113243887B (en) | Intelligent diagnosis and treatment instrument for macular degeneration of old people | |
CN104545792A (en) | Arteriovenous retinal vessel optic disk positioning method of eye fundus image | |
Sharma et al. | Machine learning approach for detection of diabetic retinopathy with improved pre-processing | |
Bilal et al. | Diabetic retinopathy detection using weighted filters and classification using CNN | |
CN115661066A (en) | Diabetic retinopathy detection method based on segmentation and classification fusion | |
Elsawy et al. | A novel network with parallel resolution encoders for the diagnosis of corneal diseases | |
KR102537470B1 (en) | Diagnosis method for Graves’orbitopathy using orbital computed tomography (CT) based on neural network-based algorithm | |
CN112102234B (en) | Ear sclerosis focus detection and diagnosis system based on target detection neural network | |
Lima et al. | A semiautomatic segmentation approach to corneal lesions | |
CN114898172B (en) | Multi-feature DAG network-based diabetic retinopathy classification modeling method | |
Vasudeva et al. | Classifying Skin Cancer and Acne using CNN | |
Wang et al. | Retinal OCT layer segmentation via joint motion correction and graph-assisted 3D neural network | |
Alves et al. | The challenges of applying deep learning for hemangioma lesion segmentation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |