CN114445666A - Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images - Google Patents
Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images Download PDFInfo
- Publication number
- CN114445666A CN114445666A CN202210100811.3A CN202210100811A CN114445666A CN 114445666 A CN114445666 A CN 114445666A CN 202210100811 A CN202210100811 A CN 202210100811A CN 114445666 A CN114445666 A CN 114445666A
- Authority
- CN
- China
- Prior art keywords
- image
- optic disc
- positions
- classifying
- yellow spots
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- A61B3/0016—Operational features thereof
- A61B3/0025—Operational features thereof characterised by electronic signal processing, e.g. eye models
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses a classification method for left and right eyes and visual field positions of eyeground images based on deep learning, which is characterized by comprising the following steps of: step 1, preprocessing an eye fundus image to obtain a preprocessed image; step 2, building a convolutional neural network model based on the preprocessed image, and detecting yellow spots and optic discs to obtain a detection result; step 3, obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection result, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image; step 4, classifying left and right eyes according to the relative position of the yellow spots and the optic disc, wherein the optic disc judges the right eye image on the right side of the yellow spots, and otherwise, the left eye image is obtained; and 5, classifying the positions of the visual fields according to the positions of the yellow spots and the optic discs in the image, judging the center of a connecting line of the optic discs and the yellow spots in the image as a single-visual-field image, and otherwise, judging the image as a double-visual-field image.
Description
Technical Field
The invention belongs to the technical field of computer vision and medical image processing, and particularly relates to a method and a system for classifying left and right eye and visual field positions of eyeground images based on deep learning.
Background
Medical imaging technology has rapidly developed and become an indispensable technology in medical diagnosis. Since the digital image age, the generation of massive data provides more possibilities for the future development of medical images. Therefore, how to further analyze and mine the medical image big data, how to extract valuable information from the medical image high-dimensional data, and how to closely combine the development of modern medical images with precise medical treatment become important topics for the future development of medical images.
In recent years, with the enhancement of computing power and the explosive increase of data, artificial intelligence techniques represented by deep learning have advanced sufficiently, and have begun to be applied to various fields in production and life. The deep learning algorithm can automatically extract features, and complex processing of high-dimensional medical image data is avoided. Under the common promotion of more and more public medical image data resources, open-source artificial intelligence algorithm resources and open high-performance computing resources, the deep learning algorithm is further developed rapidly in the field of medical images.
Eye health is an important component of national health, visual impairment including blindness seriously affects the physical health and life quality of people, increases the burden of families and society, threatens social and economic production activities, and is a major public health problem and social problem related to livelihood. With the rapid development of economic society, the accelerated aging process of population and the continuous improvement of the demand of people on eye health, the prevention and treatment of eye diseases in China still has a difficult task. China is still one of the countries with the largest number of blind and vision-impaired patients in the world, the prevalence rate of age-related eye diseases is increased, the problems of ametropia of teenagers and the like are increasingly prominent, and the problem of cataract blindness of rural poverty-stricken population is not completely solved; the problems of insufficient total quantity, low quality and uneven distribution of ophthalmic medical resources still exist, and the basic eye health care work still needs to be strengthened; the healthy life concept of people for caring eyes and eyes needs to be continuously strengthened.
The fundus color photography can be used as a rapid screening tool for common fundus diseases, and a simple, convenient and easy observation and detection means is provided for the prevention and treatment of basic ophthalmopathy. Fundus photography examined morphological changes of the entire retina. The principle is to record the scene seen under the ophthalmoscope with a specially made camera. The fundus photography can observe the forms of retina, optic disc, macular area, retinal blood vessel, and the change of retinal hemorrhage, exudation, hemangioma, retinal degeneration area, retinal hole, new blood vessel, atrophic spot, pigment disorder, etc.
The fundus photography has two shooting positions, namely a single-view shooting method and a double-view shooting method. The single-vision shooting method takes the midpoint of the connecting line of the macula lutea and the optic disc as the center of the shooting vision, and the imaging at least covers 60 percent of the retina area. In the double-vision shooting method, the vision 1 takes the central fovea of the macula lutea as the center of the shooting vision, the imaging at least covers a 45-degree retina area, the vision 2 takes the optic disc as the center of the shooting vision, and the imaging at least covers a 45-degree retina area.
Before artificial intelligence aided diagnosis of common fundus diseases is performed by using color fundus images, original images need to be processed, and it is important to obtain position information of left and right eyes and visual fields of the fundus images.
Disclosure of Invention
The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a method and a system for classifying left and right eye and visual field positions based on a deep learning fundus image.
The invention provides a classification method for left and right eyes and visual field positions of eyeground images based on deep learning, which is characterized by comprising the following steps: step 1, preprocessing an eye fundus image to obtain a preprocessed image; step 2, building a convolutional neural network model based on the preprocessed image, and detecting yellow spots and optic discs to obtain a detection result; step 3, obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection result, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image; step 4, classifying left and right eyes according to the relative position of the yellow spots and the optic disc, wherein the optic disc judges the right eye image on the right side of the yellow spots, and otherwise, the left eye image is obtained; and 5, classifying the positions of the visual fields according to the positions of the macula lutea and the optic disc in the image, judging the center of a connecting line of the optic disc and the macula lutea as a single-visual-field image at the central position of the image, and otherwise, judging the image as a double-visual-field image.
The method for classifying the left and right eye and visual field positions based on the deep learning fundus image according to the present invention may further include the following features: in step 1, the fundus image is a color fundus image.
The method for classifying the left and right eye and visual field positions based on the deep learning fundus image according to the present invention may further include the following features: wherein, in step 1, the pretreatment comprises: the method comprises the following steps of data enhancement, image cropping, image turning and rotation and image transformation, wherein the image cropping comprises random cropping and center cropping, the image turning and rotation comprises horizontal or vertical turning and random rotation, and the image transformation comprises standardization, brightness adjustment, contrast adjustment and saturation adjustment.
The method for classifying the left and right eye and visual field positions based on the deep learning fundus image according to the present invention may further include the following features: wherein, step 2 includes the following steps: step 2-1, constructing a convolutional neural network model, wherein model parameters contained in the convolutional neural network model are randomly set; 2-2, sequentially inputting each training image in the training set into a convolutional neural network model and carrying out one iteration; step 2-3, after iteration, calculating loss errors by adopting the model parameters of the last layer, and then reversely propagating the loss errors so as to update the model parameters; and 2-4, repeating the step 2-2 to the step 2-3 until a training completion condition is reached, and obtaining the trained convolutional neural network model.
The method for classifying the left and right eye and visual field positions based on the deep learning fundus image according to the present invention may further include the following features: in step 5, the dual-view images may be classified, and if the center of the optic disc is closer to the center of the image, the dual-view image with the optic disc as the center is obtained, otherwise, the dual-view image with the macula lutea as the center is obtained.
The invention provides a classification system for left and right eyes and visual field positions based on deep learning eyeground images, which has the following characteristics: the preprocessing part is used for preprocessing the fundus image to obtain a preprocessed image; the target detection part is used for building a convolution neural network model, detecting the yellow spots and the optic disc of the preprocessed image and obtaining the position information of the yellow spots and the optic disc of the fundus image through detection, wherein the position information comprises the relative position of the yellow spots and the optic disc and the position of the yellow spots and the optic disc in the fundus image; and the classification predicting part is used for classifying left and right eyes according to the relative positions of the yellow spots and the optic disc, the optic disc judges the right eye image on the right side of the yellow spots, otherwise, the right eye image is a left eye image, the vision field positions are classified according to the positions of the yellow spots and the optic disc in the image, and the connecting line center of the optic disc and the yellow spots judges the image as a single-vision field image in the image center position, otherwise, the image is a double-vision field image.
Action and Effect of the invention
According to the classification method for left and right eye and visual field positions based on the deep learning fundus image of the present invention, the classification method comprises: step 1, preprocessing an eye fundus image to obtain a preprocessed image; step 2, building a convolutional neural network model based on the preprocessed image, and detecting yellow spots and optic discs to obtain a detection result; step 3, obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection result, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image; step 4, classifying left and right eyes according to the relative position of the yellow spots and the optic disc, wherein the optic disc judges the right eye image on the right side of the yellow spots, and otherwise, the left eye image is obtained; and 5, classifying the positions of the visual fields according to the positions of the yellow spots and the optic discs in the image, judging the center of a connecting line of the optic discs and the yellow spots in the image as a single-visual-field image, and otherwise, judging the image as a double-visual-field image.
Therefore, according to the fundus image left and right eye and visual field position classification method based on the deep learning of the present invention, since the left and right eye and visual field position classification is performed using the positioning of the macula lutea and the optic disc, the classification result of the model is more accurate than that of directly classifying the entire image.
In addition, the method is simple to implement, can be used for various conventional convolutional neural network models, and is simple, convenient and fast. Under the condition of ensuring high accuracy, the method occupies resources as little as possible, is easy to deploy, has strong platform compatibility, and is suitable for production environment.
Drawings
FIG. 1 is a flowchart of a method for classifying left and right eye and visual field positions of fundus images based on deep learning according to an embodiment of the present invention;
FIG. 2 is a block diagram of a convolutional neural network model employed in an embodiment of the present invention;
fig. 3 is a schematic diagram of a classification method of positions of left and right eyes and visual field of fundus images based on deep learning in an embodiment of the present invention.
Detailed Description
In order to make the technical means, creation features, achievement purposes and effects of the present invention easy to understand, the following embodiments specifically describe a method and a system for classifying left and right eye and visual field positions of an eyeground image based on deep learning according to the present invention with reference to the accompanying drawings.
In the embodiment, a method and a system for classifying left and right eye and visual field positions of a fundus image based on deep learning are provided.
The data set used in this embodiment is a community-based data set collected by centers for preventing and treating eye diseases in Shanghai city, which is not disclosed, and contains fundus images of 2000 patients and labels of positions of optic discs and macula lutea.
In addition, the hardware platform implemented in this embodiment needs an NVIDIA TITANX graphics card (GPU acceleration).
Fig. 1 is a flowchart of a method for classifying positions of left and right eyes and a visual field of a fundus image based on deep learning in the present embodiment.
As shown in fig. 1, a method for classifying positions of left and right eyes and visual field of a fundus image based on deep learning according to the present embodiment includes the steps of:
step S1, the fundus image is preprocessed to obtain a preprocessed image.
In this embodiment, the image to be measured is a color fundus image, which may be a left eye image or a right eye image of the patient, and is captured by a dual-view photographing method or a single-view photographing method. Performing data enhancement on the image, and horizontally turning to realize data expansion; adjusting brightness, contrast and saturation; the size is cropped, normalized to 224x224 (i.e., 224x224 pixels), resulting in a pre-processed image.
In the above process of this embodiment, the horizontal flipping is to increase the number of acquired images, and implement data expansion, so that the amount of data acquired from the image to be measured is richer, and then the epoch of iteration is increased. The brightness, contrast and saturation are adjusted for image enhancement.
Step S2, building a convolution neural network model based on the preprocessed image, detecting the macula lutea and the optic disc to obtain a detection result, comprising the following steps:
and step S2-1, constructing a convolutional neural network model, wherein the model parameters are randomly set.
And step S2-2, sequentially inputting each training image in the training set into the convolutional neural network model and carrying out one iteration.
And step S2-3, after iteration, calculating loss errors by using the model parameters of the last layer, and then reversely propagating the loss errors so as to update the model parameters.
And S2-4, repeating the step S2-2 to the step S2-3 until a training completion condition is reached, and obtaining the trained convolutional neural network model.
Fig. 2 is a schematic structural diagram of the convolutional neural network model of the present embodiment.
As shown in fig. 2, the YOLO network is mainly composed of three main components: the Backbone aggregates and forms a convolutional neural network of image features over different image fine granularities. The hack series mixes and combines network layers of image features and passes the image features to a prediction layer. Head predicts image features, generates bounding boxes and predicts classes.
In step S3, position information of the macula lutea and the optic disc of the fundus image including the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image is obtained from the detection result.
In the embodiment, the central positions of the macula lutea and the optic disc are calculated by obtaining the coordinates of the regression frames of the optic disc and the macula lutea.
And step S4, classifying the left and right eyes according to the relative position of the macula lutea and the optic disc, wherein the optic disc is judged as a right eye image on the right side of the macula lutea, and is judged as a left eye image if not.
And step S5, carrying out visual field position classification according to the positions of the macula lutea and the optic disc in the image, and judging that the central position of the connecting line of the optic disc and the macula lutea in the image is a single-visual-field image, otherwise, the central position is a double-visual-field image.
And if the center of the optic disc is closer to the center of the image, the two-view image with the optic disc as the center is obtained, and otherwise, the two-view image with the macula lutea as the center is obtained.
The embodiment adopts a community-based data set collected by the centers for preventing and treating eye diseases in Shanghai city, and is used for classifying the positions of left and right eyes and visual fields of the eyeground images as a test set.
Fig. 3 is a schematic diagram of a method and a system for classifying left and right eye and visual field positions of a fundus image based on deep learning according to an embodiment of the present invention.
As shown in fig. 3, the model structure of this embodiment is Yolov5, and the model training is fast.
In this embodiment, the accuracy of the trained convolutional neural network model to the left and right eye and visual field positions of the test set is 92%.
In the present embodiment, there is provided a fundus image left and right eye and visual field position classification system based on deep learning, including:
and a preprocessing section for performing preprocessing by the method of step S1 in the present embodiment.
The target detection unit performs target detection by the method of step S2 to step S3 in this embodiment.
The classification prediction unit performs classification prediction by the method of step S4 to step S5 in this embodiment.
Effects and effects of the embodiments
According to the method and the system for classifying the left eye, the right eye and the visual field of the fundus image based on the deep learning, because the step 1, the fundus image is preprocessed to obtain a preprocessed image; step 2, building a convolutional neural network model based on the preprocessed image, and detecting yellow spots and optic discs to obtain a detection result; step 3, obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection result, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image; step 4, classifying left and right eyes according to the relative position of the yellow spots and the optic disc, wherein the optic disc judges the right eye image on the right side of the yellow spots, and otherwise, the left eye image is obtained; and 5, classifying the positions of the visual fields according to the positions of the yellow spots and the optic discs in the image, judging the center of a connecting line of the optic discs and the yellow spots in the image as a single-visual-field image, and otherwise, judging the image as a double-visual-field image.
Therefore, according to the method and system for classifying positions of left and right eyes and visual field of the fundus image based on the deep learning provided by the embodiment, the positions of the left and right eyes and the visual field are classified by using the positions of the macula lutea and the optic disc, so that the classification result of the model is more accurate than that of directly classifying the whole image.
In addition, the method in the embodiment is simple to implement, can be used for various conventional convolutional neural network models, and is simple, convenient and fast. Under the condition of ensuring high accuracy, the method occupies resources as little as possible, is easy to deploy, has strong platform compatibility, and is suitable for production environment.
The above embodiments are preferred examples of the present invention, and are not intended to limit the scope of the present invention.
In the preprocessing process in the above embodiment, the image is horizontally flipped, in other embodiments, the image may not be horizontally flipped, or other data expansion methods in the prior art (for example, vertical flipping, combination of horizontal flipping and vertical flipping, etc.) are adopted, or other image transformations (for example, linear affine transformation, probability-based transformation into gray-scale map, etc.) may be performed without adjusting brightness, contrast, and saturation, so as to achieve the object of the present invention.
Claims (6)
1. A method for classifying the positions of left and right eyes and visual field of an eyeground image based on deep learning, which is characterized in that the positions of the left and right eyes and visual field of the eyeground image are classified by using the positions of detected yellow spots and a detected optic disc, and the method comprises the following steps:
step 1, preprocessing the fundus image to obtain a preprocessed image;
step 2, building a convolutional neural network model based on the preprocessed image, and detecting yellow spots and optic discs to obtain a detection result;
step 3, obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection result, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image;
step 4, classifying left and right eyes according to the relative position of the yellow spots and the optic disc, and judging the optic disc as a right eye image on the right side of the yellow spots, otherwise, judging the optic disc as a left eye image;
and 5, classifying the positions of the visual fields according to the positions of the yellow spots and the optic discs in the image, judging the center of a connecting line of the optic discs and the yellow spots as a single-visual-field image in the central position of the image, and otherwise, judging the image as a double-visual-field image.
2. The method for classifying positions of left and right eyes and visual field of an eyeground image based on deep learning according to claim 1, wherein:
in step 1, the fundus image is a color fundus image.
3. The method for classifying positions of left and right eyes and visual field of an eyeground image based on deep learning according to claim 1, wherein:
wherein, in step 1, the pretreatment comprises: data enhancement, image cropping, image flipping and rotation, and image transformation,
the image cropping includes a random cropping and a center cropping,
the image flipping and rotation includes horizontal or vertical flipping, random rotation,
the image transformation includes normalization, adjusting brightness, contrast, and saturation.
4. The method for classifying positions of left and right eyes and visual field of an eyeground image based on deep learning according to claim 1, wherein:
wherein, step 2 includes the following steps:
step 2-1, constructing a convolutional neural network model, wherein model parameters contained in the convolutional neural network model are randomly set;
2-2, sequentially inputting each training image in the training set into the convolutional neural network model and carrying out one iteration;
step 2-3, after iteration, calculating loss errors by using model parameters of the last layer, and then reversely propagating the loss errors so as to update the model parameters;
and 2-4, repeating the step 2-2 to the step 2-3 until a training completion condition is reached, and obtaining the trained convolutional neural network model.
5. The method for classifying positions of left and right eyes and a visual field based on a deep learning fundus image according to claim 1, wherein:
in step 5, the dual-view images may be classified, and if the center of the optic disc is closer to the center of the image, the dual-view images are the dual-view images with the optic disc as the center, otherwise, the dual-view images with the macula lutea as the center.
6. A system for classifying the position of left and right eyes and visual field of an eyeground image based on deep learning, which classifies the position of left and right eyes and visual field of an eyeground image using the positions of detected macula lutea and optic disc, comprising:
a preprocessing unit that preprocesses the fundus image to obtain a preprocessed image;
the target detection part is used for building a convolution neural network model, detecting the macula lutea and the optic disc of the preprocessed image, and obtaining the position information of the macula lutea and the optic disc of the fundus image through the detection, wherein the position information comprises the relative position of the macula lutea and the optic disc and the position of the macula lutea and the optic disc in the fundus image;
and the classification predicting part is used for classifying left and right eyes according to the relative positions of the yellow spots and the optic disc, the optic disc judges the right eye image on the right side of the yellow spots, otherwise, the right eye image is a left eye image, the vision position classification is carried out according to the positions of the yellow spots and the optic disc in the image, the center of a connecting line of the optic disc and the yellow spots judges the image as a single-vision image in the image center position, and otherwise, the image is a double-vision image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210100811.3A CN114445666A (en) | 2022-01-27 | 2022-01-27 | Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210100811.3A CN114445666A (en) | 2022-01-27 | 2022-01-27 | Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114445666A true CN114445666A (en) | 2022-05-06 |
Family
ID=81369079
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210100811.3A Pending CN114445666A (en) | 2022-01-27 | 2022-01-27 | Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114445666A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115170503A (en) * | 2022-07-01 | 2022-10-11 | 上海市第一人民医院 | Eye fundus image visual field classification method and device based on decision rule and deep neural network |
| CN115457306A (en) * | 2022-08-09 | 2022-12-09 | 夏晓波 | Universal method for automatically splitting left and right eye photos |
-
2022
- 2022-01-27 CN CN202210100811.3A patent/CN114445666A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115170503A (en) * | 2022-07-01 | 2022-10-11 | 上海市第一人民医院 | Eye fundus image visual field classification method and device based on decision rule and deep neural network |
| CN115170503B (en) * | 2022-07-01 | 2023-12-19 | 上海市第一人民医院 | Fundus image visual field classification method and device based on decision rule and deep neural network |
| CN115457306A (en) * | 2022-08-09 | 2022-12-09 | 夏晓波 | Universal method for automatically splitting left and right eye photos |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108876775B (en) | Method for rapidly detecting diabetic retinopathy | |
| CN109726743B (en) | Retina OCT image classification method based on three-dimensional convolutional neural network | |
| Kauppi | Eye fundus image analysis for automatic detection of diabetic retinopathy | |
| CN114445666A (en) | Deep learning-based method and system for classifying left eye, right eye and visual field positions of fundus images | |
| CN110428421A (en) | Macula lutea image region segmentation method and apparatus | |
| CN110516685A (en) | Lenticular opacities degree detecting method based on convolutional neural networks | |
| CN110013216B (en) | Artificial intelligence cataract analysis system | |
| CN112233087A (en) | Artificial intelligence-based ophthalmic ultrasonic disease diagnosis method and system | |
| CN109344763A (en) | A kind of strabismus detection method based on convolutional neural networks | |
| Maqsood et al. | Detection of macula and recognition of aged-related macular degeneration in retinal fundus images | |
| Ovreiu et al. | An exploratory study for glaucoma detection using densely connected neural networks | |
| CN111179258A (en) | Artificial intelligence method and system for identifying retinal hemorrhage image | |
| Giancardo | Automated fundus images analysis techniques to screen retinal diseases in diabetic patients | |
| CN106446805A (en) | Segmentation method and system for optic cup in eye ground photo | |
| Consejo et al. | Detection of subclinical keratoconus with a validated alternative method to corneal densitometry | |
| Aurangzeb et al. | Systematic development of AI-Enabled diagnostic systems for glaucoma and diabetic retinopathy | |
| CN116452571A (en) | Image recognition method based on deep neural network | |
| Thanh et al. | A Real-Time Classification Of Glaucoma from Retinal Fundus Images Using AI Technology | |
| CN112991289B (en) | Processing method and device for standard section of image | |
| Purwanithami et al. | Hemorrhage diabetic retinopathy detection based on fundus image using neural network and FCM segmentation | |
| Fu et al. | A GAN-based deep enhancer for quality enhancement of retinal images photographed by a handheld fundus camera | |
| Gandhimathi et al. | Detection of neovascularization in proliferative diabetic retinopathy fundus images. | |
| Arvind et al. | Deep learning regression-based retinal layer segmentation process for early diagnosis of retinal anamolies and secure data transmission through thingspeak | |
| Fang et al. | Lens structure segmentation from AS-OCT images via shape-based learning | |
| CN114140371A (en) | Eye disease diagnosis device and method based on deep neural network by utilizing eye fundus image |
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 |