CN113850762A - Eye disease identification method, device, equipment and storage medium based on anterior segment image - Google Patents
Eye disease identification method, device, equipment and storage medium based on anterior segment image Download PDFInfo
- Publication number
- CN113850762A CN113850762A CN202111024233.1A CN202111024233A CN113850762A CN 113850762 A CN113850762 A CN 113850762A CN 202111024233 A CN202111024233 A CN 202111024233A CN 113850762 A CN113850762 A CN 113850762A
- Authority
- CN
- China
- Prior art keywords
- segment image
- anterior
- anterior segment
- image
- eye disease
- 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 62
- 208000030533 eye disease Diseases 0.000 title claims abstract description 59
- 230000001575 pathological effect Effects 0.000 claims abstract description 94
- 238000004458 analytical method Methods 0.000 claims abstract description 59
- 238000012545 processing Methods 0.000 claims abstract description 25
- 230000011218 segmentation Effects 0.000 claims description 21
- 238000007689 inspection Methods 0.000 claims description 7
- 238000002372 labelling Methods 0.000 claims description 7
- 230000007246 mechanism Effects 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 5
- 230000037430 deletion Effects 0.000 claims description 5
- 238000013473 artificial intelligence Methods 0.000 claims description 4
- 230000002146 bilateral effect Effects 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 2
- 208000022873 Ocular disease Diseases 0.000 claims 3
- 201000010099 disease Diseases 0.000 abstract description 40
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 abstract description 40
- 230000008569 process Effects 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 230000003935 attention Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 210000005036 nerve Anatomy 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000010332 selective attention Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012163 sequencing technique Methods 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
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30041—Eye; Retina; Ophthalmic
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Public Health (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Pathology (AREA)
- Data Mining & Analysis (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method, a device, equipment and a storage medium for identifying eye diseases based on anterior segment images, and belongs to the technical field of image processing. The eye disease identification method comprises the steps of obtaining an anterior segment image; standardizing the anterior segment image to obtain a standard anterior segment image; the standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images; extracting pathological features in the local anterior segment image; performing characteristic importance analysis on each pathological characteristic to determine a target pathological characteristic; and identifying the blinding eye disease according to the target pathological characteristics and the trained AI analysis model. The eye disease identification method can realize simultaneous identification of multiple anterior segment diseases, and has high identification efficiency.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to a method, an apparatus, a device, and a storage medium for eye disease identification based on an anterior segment image.
Background
At present, when identifying anterior ocular segment diseases, a single anterior ocular segment image is often used to identify one anterior ocular segment disease, and multiple anterior ocular segment diseases cannot be identified at the same time, so that the problem of low identification efficiency exists.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. Therefore, the invention provides an eye disease identification method based on the anterior segment image, which can realize simultaneous identification of various anterior segment diseases and has higher identification efficiency.
The invention also provides an eye disease identification device based on the anterior segment image.
The invention also provides eye disease identification equipment based on the anterior segment image.
The invention also provides a computer readable storage medium.
According to the embodiment of the first aspect of the invention, the eye disease identification method based on the anterior segment image comprises the following steps:
acquiring an anterior segment image;
standardizing the anterior segment image to obtain a standard anterior segment image;
performing segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
extracting pathological features in the local anterior segment image;
performing characteristic importance analysis on each pathological feature to determine a target pathological feature;
and identifying the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
The eye disease identification method based on the anterior segment image, provided by the embodiment of the invention, has the following beneficial effects: the eye disease identification method obtains the standard anterior segment image by acquiring the anterior segment image and standardizing the anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images, pathological features in the local anterior ocular segment images are extracted, and the pathological features of various anterior ocular segment diseases can be conveniently obtained. And then, feature importance analysis is carried out on each pathological feature, a target pathological feature is determined, blinding eye diseases are identified according to the target pathological feature and the trained AI analysis model, a plurality of anterior segment diseases can be identified simultaneously, and the identification efficiency is high.
According to some embodiments of the invention, the acquiring an anterior segment image comprises:
and acquiring an anterior ocular segment image and a corresponding flow modulation parameter of the anterior ocular segment image.
According to some embodiments of the invention, the normalizing the anterior ocular segment image to obtain a standard anterior ocular segment image includes:
screening the anterior segment image to obtain a screened anterior segment image;
and labeling the screened anterior segment image according to a cross inspection method to obtain a standard anterior segment image.
According to some embodiments of the present invention, the segmenting the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images includes:
acquiring three-dimensional structure parameters of the standard anterior ocular segment image;
and carrying out segmentation processing on the standard anterior ocular segment image according to the three-dimensional structure parameters and the trained multi-target segmentation model to obtain a plurality of local anterior ocular segment images.
According to some embodiments of the invention, the performing a feature importance analysis on each of the pathological features to determine a target pathological feature includes:
and performing characteristic importance analysis on each pathological characteristic according to at least one of a Pearson correlation coefficient method, a bilateral multivariable method and a characteristic iterative deletion method to determine the target pathological characteristic.
According to some embodiments of the invention, the identifying the blinding eye disease according to the target pathological characteristics and the trained AI analysis model comprises:
constructing a multi-view convolution module according to the graph convolution and the target pathological features;
obtaining an AI analysis model according to the multi-view convolution module and the multi-region selection attention mechanism module;
and identifying the blinding eye disease according to the trained AI analysis model.
According to some embodiments of the invention, the eye disease identification method further comprises:
identifying the anterior segment image according to an artificial intelligence algorithm to obtain identification data;
configuring a corresponding reference file according to the identification data;
and optimizing the reference file according to the obtained feedback data to obtain a final reference file.
An eye disease recognition apparatus based on an anterior segment image according to a second aspect of the present invention includes:
the anterior segment image acquisition module is used for acquiring an anterior segment image;
the standardization processing module is used for carrying out standardization processing on the anterior segment image to obtain a standard anterior segment image;
the image segmentation module is used for carrying out segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
the characteristic extraction module is used for extracting pathological characteristics in the local anterior segment image;
the characteristic analysis module is used for carrying out characteristic importance analysis on each pathological characteristic to determine a target pathological characteristic;
and the recognition module is used for recognizing the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
The eye disease recognition device based on the anterior segment image has the following beneficial effects: the eye disease recognition device obtains the anterior segment image through the anterior segment image obtaining module, and the standardization processing module carries out standardization processing on the anterior segment image to obtain the standard anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The image segmentation module performs segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images, and the feature extraction module extracts pathological features in the local anterior ocular segment images, so that pathological features of various anterior ocular segment diseases can be conveniently obtained. And the feature analysis module performs feature importance analysis on each pathological feature to determine a target pathological feature, and the recognition module recognizes the blinding eye disease according to the target pathological feature and the trained AI analysis model, so that various anterior segment diseases can be recognized simultaneously, and the recognition efficiency is high.
An eye disease recognition apparatus based on an anterior ocular segment image according to an embodiment of the third aspect of the present invention includes:
at least one processor, and,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions for execution by the at least one processor to cause the at least one processor, when executing the instructions, to implement a method for eye condition identification as described in embodiments of the first aspect.
The eye disease identification device based on the anterior segment image has the following beneficial effects: the electronic equipment adopts the eye disease identification method based on the anterior segment image, and obtains the standard anterior segment image by acquiring the anterior segment image and carrying out standardization processing on the anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images, pathological features in the local anterior ocular segment images are extracted, and the pathological features of various anterior ocular segment diseases can be conveniently obtained. And then, feature importance analysis is carried out on each pathological feature, a target pathological feature is determined, blinding eye diseases are identified according to the target pathological feature and the trained AI analysis model, a plurality of anterior segment diseases can be identified simultaneously, and the identification efficiency is high.
According to a fourth aspect of the present invention, there is provided a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the eye disease recognition method based on an anterior segment image according to the first aspect.
The computer-readable storage medium according to the embodiment of the invention has at least the following advantages: the computer-readable storage medium executes the above eye disease recognition method to obtain a standard anterior segment image by acquiring the anterior segment image and normalizing the anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images, pathological features in the local anterior ocular segment images are extracted, and the pathological features of various anterior ocular segment diseases can be conveniently obtained. And then, feature importance analysis is carried out on each pathological feature, a target pathological feature is determined, blinding eye diseases are identified according to the target pathological feature and the trained AI analysis model, a plurality of anterior segment diseases can be identified simultaneously, and the identification efficiency is high.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The invention is further described with reference to the following figures and examples, in which:
fig. 1 is a flowchart of an eye disease recognition method based on an anterior segment image according to an embodiment of the present invention;
FIG. 2 is a flowchart of step S200 in FIG. 1;
FIG. 3 is a flowchart of step S300 in FIG. 1;
FIG. 4 is a flowchart of step S600 in FIG. 1;
FIG. 5 is another flow chart of the eye condition identification method based on the anterior segment image of FIG. 1;
fig. 6 is a schematic structural diagram of an eye disease recognition apparatus based on an anterior segment image according to an embodiment of the present invention.
Reference numerals: 610. an anterior ocular segment image acquisition module; 620. a standardization processing module; 630. an image segmentation module; 640. a feature extraction module; 650. a feature analysis module; 660. and identifying the module.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
In the description of the present invention, it should be understood that the orientation or positional relationship referred to in the description of the orientation, such as the upper, lower, front, rear, left, right, etc., is based on the orientation or positional relationship shown in the drawings, and is only for convenience of description and simplification of description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention.
In the description of the present invention, the meaning of a plurality is one or more, the meaning of a plurality is two or more, and the above, below, exceeding, etc. are understood as excluding the present numbers, and the above, below, within, etc. are understood as including the present numbers. If the first and second are described for the purpose of distinguishing technical features, they are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated or implicitly indicating the precedence of the technical features indicated.
In the description of the present invention, unless otherwise explicitly limited, terms such as arrangement, installation, connection and the like should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above terms in the present invention in combination with the specific contents of the technical solutions.
In the description of the present invention, reference to the description of the terms "one embodiment," "some embodiments," "an illustrative embodiment," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
In a first aspect, referring to fig. 1, a method for identifying an eye disease based on an anterior segment image according to an embodiment of the present invention includes:
s100, acquiring an anterior segment image;
s200, standardizing the anterior segment image to obtain a standard anterior segment image;
s300, segmenting the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
s400, extracting pathological features in the local anterior segment image;
s500, performing feature importance analysis on each pathological feature to determine a target pathological feature;
s600, identifying the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
In the process of identifying the anterior segment disease, an anterior segment image is first acquired. It should be noted that these anterior segment images are an AS-OCT three-dimensional image. The three-dimensional image has the characteristics of three-dimensional imaging non-contact, high detection sensitivity, high resolution, quick inspection, objective quantitative measurement and the like, and the corresponding whole anterior segment structure information can be obtained through the acquired anterior segment image. Such anterior segment images may be acquired by the AS-OCT apparatus and acquired in a scanning mode at multiple angles, or in other manners, without being limited thereto. Further, the anterior segment image is normalized to obtain a standard anterior segment image. Specifically, the standard anterior segment image can be obtained by screening the acquired anterior segment image, eliminating the abnormal image, and labeling the screened anterior segment image through a cross inspection or the like. For the identification of a plurality of anterior ocular segment diseases, different parts of the anterior ocular segment structure need to be analyzed, so that the standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images. And analyzing the anterior segment diseases according to the local anterior segment images, extracting pathological features in the local anterior segment images, performing feature importance analysis and feature correlation analysis on different types of pathological features, and determining target pathological features. And identifying the blinding eye disease movement according to the target pathological characteristics and the trained AI analysis model, wherein the trained A analysis model is based on a mixed model of two-dimensional convolution nerves and a three-dimensional deep neural network, and simultaneously comprises a multi-view-angle graph convolution module and a multi-region selection attention mechanism module, so that the comprehensiveness and the accuracy of the ophthalmic disease identification can be improved. The identification method can realize simultaneous identification of multiple anterior segment diseases, and has high identification efficiency.
In some embodiments, step S100 includes:
and acquiring the anterior ocular segment image and the corresponding flow modulation parameter of the anterior ocular segment image.
In the process of identifying the anterior segment disease, an anterior segment image needs to be acquired. These anterior segment images are an AS-OCT three-dimensional image. And obtaining corresponding whole anterior segment structure information through the obtained anterior segment image. Such anterior segment images can be acquired by an AS-OCT device and acquired at multiple angles in a scanning mode. In order to improve the identification accuracy, the flow modulation parameters corresponding to each anterior segment image need to be acquired, wherein the flow modulation parameters include age, gender, naked eye vision and the like, and the flow modulation parameters can help to analyze and identify different ophthalmic diseases, so that the analysis accuracy is improved.
Referring to fig. 2, in some embodiments, step S200 includes:
s210, screening the anterior segment image to obtain a screened anterior segment image;
and S220, labeling the screened anterior segment image according to a cross-checking method to obtain a standard anterior segment image.
In the process of identifying the anterior segment disease, an anterior segment image is first acquired. After the anterior ocular segment image is acquired, in order to avoid interference of abnormal data, the anterior ocular segment image needs to be screened and the abnormal anterior ocular segment image needs to be removed. In order to improve the image labeling quality, the labeled images of the pre-ocular segment after screening are further required to be labeled according to a cross inspection method, so that a standard pre-ocular segment image is obtained. The cross-over inspection method here is a method of performing labeling processing on each region of a selected anterior segment image a plurality of times, and then integrating a plurality of labels in the region for each region to form a final label. By the marking mode, the marking quality of the image can be improved, and the data accuracy can be improved.
Referring to fig. 3, in some embodiments, step S300 includes:
s310, acquiring three-dimensional structure parameters of the standard anterior segment image;
and S320, segmenting the standard anterior ocular segment image according to the three-dimensional structure parameters and the trained multi-target segmentation model to obtain a plurality of local anterior ocular segment images.
After the obtained anterior ocular segment images are screened and labeled to obtain a standard anterior ocular segment image, in order to improve the accuracy of anterior ocular segment structure segmentation, three-dimensional structure parameters of the standard anterior ocular segment image need to be obtained, and the standard anterior ocular segment image is segmented and detected according to the three-dimensional structure parameters and a trained multi-target segmentation model to obtain a plurality of local anterior ocular segment images. Specifically, the multi-target segmentation model is mainly an end-to-end multi-target segmentation model based on multi-information interaction, and the convolution layer used by the segmentation model can be expressed as follows:
wherein HcijA value, W, representing the position of c in the feature map i in the hidden layer generated by the input sample jk,m,iRepresenting the weight values at k positions for the convolution kernel of the feature map i in the hidden layer for the m channels of the input samples,
the pixel value corresponding to the data of the m channel representing the input sample j at the d ° c + k position. d represents the step vector of the downsampling convolution, determining how many pixels to sample, and degree represents the element-to-element product.
The cross entropy loss function used by the multi-target segmentation model in the segmentation of the standard anterior ocular segment image is
Wherein n is the number of categories, y represents the true label,
indicating the segmentation prediction result.
The MES loss function used when the multi-target segmentation model detects the standard anterior ocular segment image is
Wherein, yiAnd
the label value and the predicted value of the model are represented, and n represents the number of regression coordinate points.
In some embodiments, step S500 includes:
and performing feature importance analysis on each pathological feature according to at least one of a Pearson correlation coefficient method, a bilateral multivariate method and a feature iterative deletion method to determine the target pathological feature.
In order to fully utilize the correlation between different regions and different features related to various diseases of the anterior ocular segment image, different types of pathological features can be extracted by using an image omics method and the like after the standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images. And analyzing the feature importance of each pathological feature according to at least one of a Pearson correlation coefficient method, a bilateral multivariate method and a feature iterative deletion method, analyzing and sequencing the feature importance, and analyzing each pathological feature so as to determine the target pathological feature. It should be noted that, when the pearson correlation coefficient method is used to analyze the feature importance of each pathological feature, the pearson correlation coefficient method can be expressed as
Wherein, E [ (X-mu)X)(Y-μY)]Represents the covariance, σ, of X and YX、σYStandard deviations for X and Y, respectively.
The Pearson correlation coefficient method is high in calculation speed, is suitable for analyzing a large amount of data, and can improve the identification efficiency.
In addition, because the pearson coefficient is only sensitive to the linear relation, for the nonlinear model, a characteristic iterative deletion method can be adopted for characteristic analysis. And training a sample through a model, then scoring each pathological feature, sorting the pathological features with the smallest feature score, then re-training the model by using the remaining pathological features, performing the next iteration, finally selecting the required pathological features, and taking the pathological features as target pathological features.
Referring to fig. 4, in some embodiments, step S600 includes:
s610, constructing a multi-view convolution module according to the graph convolution and the target pathological characteristics;
s620, obtaining an AI analysis model according to the multi-view convolution module and the multi-region selection attention mechanism module;
and S630, identifying the blinding eye disease according to the trained AI analysis model.
In order to fully utilize the pathological features of various diseases contained in the anterior segment image and the spatial structure information of the image, a mixed model of two-dimensional convolution nerve and a three-dimensional deep neural network can be used as a basic network framework of an AI analysis model. And processing the target pathological features by using graph convolution to construct the correlation among the anterior segment images at different angles to form a multi-view convolution module. Meanwhile, in order to characterize the difference between different regions in the anterior segment image, a multi-region selective attention mechanism module can be constructed. And (3) an AI analysis model is obtained by integrating the multi-view convolution module and the multi-region selection attention mechanism module into the mixed model. The multi-view multi-disease AI analysis model can conveniently identify blinding eye diseases, and improves the identification comprehensiveness and the analysis accuracy of ophthalmic diseases.
Referring to fig. 5, in some embodiments, the eye disease recognition method further includes:
s700, identifying the anterior segment image according to an artificial intelligence algorithm to obtain identification data;
s800, configuring a corresponding reference file according to the identification data;
and S900, optimizing the reference file according to the obtained feedback data to obtain the final reference file.
In the process of identifying the anterior segment diseases, the anterior segment images can be identified according to an artificial intelligence algorithm and identification data can be obtained. Specifically, analysis algorithms and calculation service functions of different diseases can be called according to different received commands, and the recognition efficiency can be improved. By acquiring the identification data, the corresponding reference file can be inquired and configured in the resource library according to the identification data. Such as disease analysis data and treatment protocols, etc., stored in a repository. In order to provide a more matched scheme for different anterior segment diseases, feedback data can be obtained through an interactive feedback system, and a reference file is optimized according to the feedback data to generate a final reference file. It should be noted that the interactive feedback system can control the output of the feedback data through the principles of rule driving, data driving and the like, and this way can improve the accuracy of the feedback data, so that the generated reference file can assist in analyzing the corresponding anterior segment diseases more accurately, and improve the accuracy of disease identification and analysis.
Second aspect, referring to fig. 6, an eye disease recognition apparatus based on an anterior ocular segment image according to an embodiment of the present invention includes:
an anterior segment image obtaining module 610, configured to obtain an anterior segment image;
the standardization processing module 620 is configured to standardize the anterior segment image to obtain a standard anterior segment image;
an image segmentation module 630, configured to perform segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
the feature extraction module 640 is used for extracting pathological features in the local anterior segment image;
the characteristic analysis module 650 is used for performing characteristic importance analysis on each pathological characteristic to determine a target pathological characteristic;
and the recognition module 660 is used for recognizing the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
In the process of identifying the anterior ocular segment disease, the anterior ocular segment image acquisition module 610 first acquires an anterior ocular segment image. And obtaining corresponding whole anterior segment structure information through the obtained anterior segment image. Further, the normalization processing module 620 performs normalization processing on the anterior segment image to obtain a standard anterior segment image. Specifically, the standard anterior segment image can be obtained by screening the acquired anterior segment image, eliminating the abnormal image, and labeling the screened anterior segment image through a cross inspection or the like. Since different parts of the anterior ocular segment structure need to be analyzed for identifying various anterior ocular segment diseases, the image segmentation module 630 performs segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images. According to the local anterior segment images, the disease of the anterior segment is analyzed, the characteristic extraction module 640 extracts pathological characteristics in the local anterior segment images, and the characteristic analysis module 650 performs characteristic importance analysis and characteristic correlation analysis on different types of pathological characteristics to determine target pathological characteristics. The identification module 660 identifies the blinding eye disease movement according to the target pathological characteristics and the trained AI analysis model, wherein the trained A analysis model is based on a mixed model of two-dimensional convolution nerves and a three-dimensional deep neural network, and simultaneously comprises a multi-view-map convolution module and a multi-region selective attention mechanism module, so that the comprehensiveness and accuracy of ophthalmic disease identification can be improved. The identification method can realize simultaneous identification of multiple anterior segment diseases, and has high identification efficiency.
In a third aspect, an eye disease recognition device based on an anterior segment image according to an embodiment of the present invention includes at least one processor, and a memory communicatively coupled to the at least one processor; the memory stores instructions that are executed by the at least one processor, so that the at least one processor, when executing the instructions, implements the eye condition recognition method according to the first aspect.
According to the eye disease identification device provided by the embodiment of the invention, at least the following beneficial effects are achieved: the electronic equipment adopts the eye disease identification method, and obtains the standard anterior segment image by acquiring the anterior segment image and carrying out standardization processing on the anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images, pathological features in the local anterior ocular segment images are extracted, and the pathological features of various anterior ocular segment diseases can be conveniently obtained. And then, feature importance analysis is carried out on each pathological feature, a target pathological feature is determined, blinding eye diseases are identified according to the target pathological feature and the trained AI analysis model, a plurality of anterior segment diseases can be identified simultaneously, and the identification efficiency is high.
In a fourth aspect, the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions for causing a computer to execute the eye disease identification method based on the anterior segment image as an embodiment of the first aspect.
The computer-readable storage medium according to the embodiment of the invention has at least the following advantages: such a computer-readable storage medium executes the above eye disease recognition method, and obtains a standard anterior segment image by obtaining an anterior segment image and normalizing the anterior segment image. The whole anterior segment structure information can be conveniently acquired by acquiring the anterior segment image. The standard anterior ocular segment image is segmented to obtain a plurality of local anterior ocular segment images, pathological features in the local anterior ocular segment images are extracted, and the pathological features of various anterior ocular segment diseases can be conveniently obtained. And then, feature importance analysis is carried out on each pathological feature, a target pathological feature is determined, blinding eye diseases are identified according to the target pathological feature and the trained AI analysis model, a plurality of anterior segment diseases can be identified simultaneously, and the identification efficiency is high.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention. Furthermore, the embodiments of the present invention and the features of the embodiments may be combined with each other without conflict.
Claims (10)
1. An eye disease identification method based on an anterior segment image is characterized by comprising the following steps:
acquiring an anterior segment image;
standardizing the anterior segment image to obtain a standard anterior segment image;
performing segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
extracting pathological features in the local anterior segment image;
performing characteristic importance analysis on each pathological feature to determine a target pathological feature;
and identifying the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
2. The eye disease recognition method of claim 1, wherein the acquiring an anterior segment image comprises:
and acquiring an anterior ocular segment image and a corresponding flow modulation parameter of the anterior ocular segment image.
3. The method for identifying an eye disease according to claim 1, wherein the normalizing the anterior ocular segment image to obtain a standard anterior ocular segment image comprises:
screening the anterior segment image to obtain a screened anterior segment image;
and labeling the screened anterior segment image according to a cross inspection method to obtain a standard anterior segment image.
4. The method for identifying an eye disease according to claim 1, wherein the step of segmenting the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images comprises:
acquiring three-dimensional structure parameters of the standard anterior ocular segment image;
and carrying out segmentation processing on the standard anterior ocular segment image according to the three-dimensional structure parameters and the trained multi-target segmentation model to obtain a plurality of local anterior ocular segment images.
5. The method for identifying an ocular disease as recited in claim 1, wherein said analyzing feature importance of each of said pathological features to determine a target pathological feature comprises:
and performing characteristic importance analysis on each pathological characteristic according to at least one of a Pearson correlation coefficient method, a bilateral multivariable method and a characteristic iterative deletion method to determine the target pathological characteristic.
6. The method for identifying an ocular disease as recited in claim 1, wherein the identifying a blinding ocular disease based on the target pathological features and the trained AI analysis model comprises:
constructing a multi-view convolution module according to the graph convolution and the target pathological features;
obtaining an AI analysis model according to the multi-view convolution module and the multi-region selection attention mechanism module;
and identifying the blinding eye disease according to the trained AI analysis model.
7. The eye disease identification method according to any one of claims 1 to 6, further comprising:
identifying the anterior segment image according to an artificial intelligence algorithm to obtain identification data;
configuring a corresponding reference file according to the identification data;
and optimizing the reference file according to the obtained feedback data to obtain a final reference file.
8. Eye disease recognition device based on anterior segment image, characterized by including:
the anterior segment image acquisition module is used for acquiring an anterior segment image;
the standardization processing module is used for carrying out standardization processing on the anterior segment image to obtain a standard anterior segment image;
the image segmentation module is used for carrying out segmentation processing on the standard anterior ocular segment image to obtain a plurality of local anterior ocular segment images;
the characteristic extraction module is used for extracting pathological characteristics in the local anterior segment image;
the characteristic analysis module is used for carrying out characteristic importance analysis on each pathological characteristic to determine a target pathological characteristic;
and the recognition module is used for recognizing the blinding eye disease according to the target pathological characteristics and the trained AI analysis model.
9. Eye disease recognition equipment based on anterior segment images, comprising:
at least one processor, and,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions for execution by the at least one processor to cause the at least one processor, when executing the instructions, to implement a method for eye condition identification based on an anterior ocular segment image as recited in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions for causing a computer to execute the eye disease identification method based on an anterior segment image according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111024233.1A CN113850762A (en) | 2021-09-02 | 2021-09-02 | Eye disease identification method, device, equipment and storage medium based on anterior segment image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111024233.1A CN113850762A (en) | 2021-09-02 | 2021-09-02 | Eye disease identification method, device, equipment and storage medium based on anterior segment image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113850762A true CN113850762A (en) | 2021-12-28 |
Family
ID=78976787
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111024233.1A Pending CN113850762A (en) | 2021-09-02 | 2021-09-02 | Eye disease identification method, device, equipment and storage medium based on anterior segment image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113850762A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114463319A (en) * | 2022-02-15 | 2022-05-10 | 北京百度网讯科技有限公司 | Data prediction method and device, electronic equipment and storage medium |
-
2021
- 2021-09-02 CN CN202111024233.1A patent/CN113850762A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114463319A (en) * | 2022-02-15 | 2022-05-10 | 北京百度网讯科技有限公司 | Data prediction method and device, electronic equipment and storage medium |
| CN114463319B (en) * | 2022-02-15 | 2024-01-02 | 北京百度网讯科技有限公司 | Data prediction method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pantazi et al. | Automated leaf disease detection in different crop species through image features analysis and One Class Classifiers | |
| US11842556B2 (en) | Image analysis method, apparatus, program, and learned deep learning algorithm | |
| US20210326624A1 (en) | Method, system and device for difference automatic calibration in cross modal target detection | |
| US20190355113A1 (en) | Multi-sample Whole Slide Image Processing in Digital Pathology via Multi-resolution Registration and Machine Learning | |
| WO2018121690A1 (en) | Object attribute detection method and device, neural network training method and device, and regional detection method and device | |
| CN110033007B (en) | Pedestrian clothing attribute identification method based on depth attitude estimation and multi-feature fusion | |
| CN112633297B (en) | Target object identification method and device, storage medium and electronic device | |
| CN111274926B (en) | Image data screening method, device, computer equipment and storage medium | |
| EP3006551B1 (en) | Image processing device, image processing method, program, and storage medium | |
| CN104680185B (en) | Hyperspectral image classification method based on boundary point reclassification | |
| CN112464983A (en) | Small sample learning method for apple tree leaf disease image classification | |
| CN116012291A (en) | Industrial part image defect detection method and system, electronic equipment and storage medium | |
| CN113052295B (en) | Training method of neural network, object detection method, device and equipment | |
| Wang et al. | Local defect detection and print quality assessment | |
| CN113011528B (en) | Remote sensing image small target detection method based on context and cascade structure | |
| CN108805181B (en) | Image classification device and method based on multi-classification model | |
| CN113850762A (en) | Eye disease identification method, device, equipment and storage medium based on anterior segment image | |
| CN116468690B (en) | Subtype analysis system of invasive non-mucous lung adenocarcinoma based on deep learning | |
| Montalvo et al. | A novel threshold to identify plant textures in agricultural images by Otsu and Principal Component Analysis | |
| Ramamoorthy et al. | Reliable and accurate plant leaf disease detection with treatment suggestions using enhanced deep learning techniques | |
| CN114080644A (en) | System and method for diagnosing small bowel cleanliness | |
| CN116630268A (en) | Road disease detection method, system, equipment and medium | |
| CN116245855A (en) | Crop variety identification method, device, equipment and storage medium | |
| CN109558771B (en) | Behavior state identification method, device and equipment of marine ship and storage medium | |
| CN111815613B (en) | Liver cirrhosis disease stage identification method based on envelope line morphological feature analysis |
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 |