CN113361482A - Nuclear cataract identification method, device, electronic device and storage medium - Google Patents
Nuclear cataract identification method, device, electronic device and storage medium Download PDFInfo
- Publication number
- CN113361482A CN113361482A CN202110766487.4A CN202110766487A CN113361482A CN 113361482 A CN113361482 A CN 113361482A CN 202110766487 A CN202110766487 A CN 202110766487A CN 113361482 A CN113361482 A CN 113361482A
- Authority
- CN
- China
- Prior art keywords
- region
- nuclear
- eye image
- pixel
- sample eye
- 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
- 206010007759 Cataract nuclear Diseases 0.000 title claims abstract description 90
- 208000029552 nuclear cataract Diseases 0.000 title claims abstract description 90
- 238000000034 method Methods 0.000 title claims abstract description 34
- 210000000695 crystalline len Anatomy 0.000 claims abstract description 76
- 238000012545 processing Methods 0.000 claims description 14
- 230000009467 reduction Effects 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 6
- 230000000903 blocking effect Effects 0.000 claims description 4
- 238000003745 diagnosis Methods 0.000 abstract description 4
- 230000036285 pathological change Effects 0.000 abstract description 4
- 231100000915 pathological change Toxicity 0.000 abstract description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 208000002177 Cataract Diseases 0.000 description 11
- 238000012014 optical coherence tomography Methods 0.000 description 9
- 210000001653 corpus striatum Anatomy 0.000 description 8
- 230000003287 optical effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000011218 segmentation Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000001356 surgical procedure Methods 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 206010024214 Lenticular opacities Diseases 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000000875 corresponding effect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The invention discloses a nuclear cataract identification method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: extracting the region characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the region characteristics and the shape characteristics; and constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified. By the technical scheme, the pathological change characteristics of the crystalline lens nuclear region in the fundus image can be accurately acquired, so that the grade judgment of the nuclear cataract is more accurate, a new thought is provided for the automatic diagnosis of the nuclear cataract, and an ophthalmologist can be assisted to efficiently, objectively and accurately diagnose the nuclear cataract.
Description
Technical Field
The embodiment of the invention relates to the technical field of artificial intelligence, in particular to the technical field of image processing, and specifically relates to a nuclear cataract identification method and device, electronic equipment and a storage medium.
Background
Nuclear cataracts are a major blinding and vision-impairing ophthalmic disease with clinical symptoms manifested by clouding and progressive stiffening and darkening of the nuclear region of the lens. Early intervention and cataract surgery can effectively improve the vision and quality of life of cataract patients. The optical coherence tomography (AS-OCT) is an Optical Coherence Tomography (OCT). The AS-OCT image has the characteristics of non-contact, high detection sensitivity, quickness, high resolution, quickness in inspection, objective quantitative measurement and the like, and can acquire the whole lens structure.
In the clinical basic research, scholars have studied the correlation between the mean pixel value of the nuclear region of the crystalline lens and the severity of the nuclear cataract, and clinical statistics show that the mean pixel value of the nuclear region of the crystalline lens and the severity of the nuclear cataract have a good correlation. Under the Lens opacity Classification System (LOCS III), ophthalmologists typically classify nuclear cataracts into two grades: low and high; the low stage refers to that the nuclear region of the crystalline lens of cataract patients has turbid symptoms but is not obvious (the nuclear cataract grades are 1 grade and 2 grade), and the patients can use drip medicine to relieve the development process of cataract; advanced stage refers to the appearance of significant opacity in the nuclear region of the lens of cataract patients, with grade 3 and above, who need to undergo clinical follow-up or cataract surgery. Clinical studies also found that the mean of pixels in the lower part of the nuclear region of the lens correlated less strongly with the severity of the cataract than in the whole nuclear region of the lens, while clinical studies also analyzed the relationship between the diameter and thickness of the nuclear region and the thickness of the cataract. These clinical studies provide clinical support for automated nuclear cataract classification based on anterior segment OCT images, but less automated nuclear cataract classification based on anterior segment OCT images is currently in progress.
Disclosure of Invention
The invention provides a nuclear cataract identification method, a nuclear cataract identification device, electronic equipment and a storage medium, which are used for realizing automatic nuclear cataract classification based on an anterior segment OCT image.
In a first aspect, an embodiment of the present invention provides a nuclear cataract identification method, including:
extracting the region characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the region characteristics and the shape characteristics;
and constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified.
In a second aspect, an embodiment of the present invention further provides a nuclear cataract identification device, including:
the target feature determination module is used for extracting the region feature of the crystalline lens nuclear region in the sample eye image, extracting the shape feature of the crystalline lens nuclear region in the sample eye image, and determining the target feature of the sample eye image according to the region feature and the shape feature;
and the identification model building module is used for building a nuclear cataract identification model based on the target characteristics of the sample eye image and determining the nuclear cataract grade of the eye image to be identified.
In a third aspect, an embodiment of the present invention further provides an electronic device, including:
one or more processors;
a memory for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement a method of nuclear cataract identification as provided by any of the embodiments of the present invention.
In a fourth aspect, embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements a nuclear cataract identification method as provided in any of the embodiments of the present invention.
The method comprises the steps of extracting the regional characteristics of a crystalline lens nuclear region in a sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the regional characteristics and the shape characteristics; and then constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, wherein the nuclear cataract identification model is used for determining the nuclear cataract grade of the eye image to be identified. By the technical scheme, the pathological change characteristics of the crystalline lens nuclear region in the fundus image can be accurately acquired, so that the grade judgment of the nuclear cataract is more accurate, a new thought is provided for the automatic diagnosis of the nuclear cataract, and an ophthalmologist can be assisted to efficiently, objectively and accurately diagnose the nuclear cataract.
Drawings
Fig. 1 is a flowchart of a nuclear cataract identification method according to an embodiment of the present invention;
fig. 2 is a flowchart of a nuclear cataract identification method according to a second embodiment of the present invention;
fig. 3 is a schematic structural diagram of a nuclear cataract identification device according to a third embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a nuclear cataract identification method according to an embodiment of the present invention; the embodiment is applicable to the case of nuclear cataract identification, and the method can be executed by a nuclear cataract identification device, which can be implemented by software/hardware and can be integrated in an electronic device carrying the nuclear cataract identification function, for example, a server device.
As shown in fig. 1, the method may specifically include:
s110, extracting the region characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the region characteristics and the shape characteristics.
In the present embodiment, the sample fundus image refers to a fundus image of an ophthalmic clinic collected in advance, and may be, for example, an AS-OCT image of an eye portion including a normal eye portion image and a nuclear cataract lesion eye portion image. The lenticular nucleus region refers to the region of the eye image where the lens is located.
The regional characteristics refer to pixel characteristics of a crystalline lens nuclear region in an eye image, and may include at least one of the following regional characteristic dimensions: pixel mean, pixel variance, maximum pixel value, minimum pixel value, pixel range, pixel standard deviation, pixel median, energy, root mean square, pixel gray scale value within a preset range, pixel four-quadrant spacing, pixel average absolute deviation, pixel skewness, pixel kurtosis, energy based on image gray histogram, entropy, and consistency.
The pixel mean value is the mean value of all pixels in the nuclear region of the crystalline lens in the eye image, and reflects the average reflection intensity of the opacity of the nuclear cataract in the eye image.
The pixel variance is used to measure the degree of dispersion (uniformity) of the pixels in the lenticular region in the eye image, and specifically, the pixel variance may be determined by calculating a square value of a difference value between each pixel value of the lenticular region in the eye image and a pixel mean value, and calculating a mean value of all the square values, and taking the mean value as the pixel variance.
The maximum pixel value is a maximum pixel value of a lenticular region in an eye image.
The minimum pixel value is a minimum pixel value of a lenticular region in an eye image.
The pixel range is a difference between a maximum pixel value of the lenticular region in the eye image and a minimum pixel value of the lenticular region in the eye image.
The pixel standard deviation is the square root of the variance of pixel values of the lenticular region in the eye image.
The pixel median is a value in which the gray level value of the pixel in the lenticular region in the eye image is in the middle level.
The energy is an average value of the sum of squares of all pixel values of the lenticular region in the eye image.
The root mean square refers to the square root of the energy characteristics of the pixels of the lenticular nuclear region in the eye image.
The pixel gray scale value in the preset range refers to the pixel gray scale value within the preset range after all the pixel gray scale values of the lenticular nucleus region in the eye image are sorted from small to large, wherein the preset range is set by a person skilled in the art according to actual situations, and may be, for example, the pixel gray scale value located 10% in front or the pixel gray scale value located 10% in back.
The quarter-bit distance of the pixels is the difference between the pixel gray value ranked 75% and the pixel gray value ranked 25% after all the pixel gray values of the lenticular and nuclear regions in the eye image are sorted from small to large.
The mean absolute deviation of the pixels refers to the mean value of the sum of absolute values of differences between all pixel gray-scale values of the lenticular region in the eye image and the pixel mean value, and is used for measuring the degree of deviation between all pixel gray-scale values and the pixel mean value. The larger the average absolute deviation value is, the pixel gray value distribution is discrete; and conversely, the distribution of the pixel gray values is concentrated.
The skewness of the pixels is used for measuring the skewness direction and the skewness of all pixel gray value distribution of the lens nucleus region in the eye image and is used for counting the digital characteristics of the asymmetry degree of the pixel distribution.
The kurtosis of a pixel is used to measure the peak value (flatness) of the distribution of the gray-scale values of the pixel; the higher the kurtosis, the more concentrated the pixel distribution is at the tail rather than at the mean; the lower the kurtosis, the more concentrated the pixel distribution is at the mean.
The energy, entropy and consistency characteristics based on the image gray level histogram refer to characteristics obtained by performing statistical analysis on the histogram of the crystalline nucleus region in the eye image, wherein the entropy is used for measuring the uncertainty (randomness) of the histogram and the average information content contained in the measured image; coherence is used to measure histogram randomness. Specifically, the pixel grayscale segmentation interval of the histogram is set to a fixed value (e.g., 25), and then the pixel range is between 0 and 255, and there are 11 pixel segmentation intervals; calculating the square sum of the pixel quantity of all gray level segments in the histogram, and taking the obtained result as the energy based on the image gray level histogram; calculating the entropy of each gray level segment in the histogram, summing the entropies of the gray level segments, and taking the summed result as the entropy based on the image gray level histogram; and taking the sum of squares of the probability of each gray scale segment in the histogram as a consistency characteristic based on the image gray scale histogram, wherein the probability of the gray scale segment refers to the ratio of the number of pixels of each gray scale segment to the total number of pixels of the phaco-nuclear region.
The shape feature is the thickness and length of the lenticular nucleus region in the eye image.
In this embodiment, the lenticular nucleus region may be cut out from the sample eye image based on the depth segmentation network; extracting 18-dimensional region features of a crystalline lens nuclear region in the sample eye image, wherein the features are respectively as follows: pixel mean, pixel variance, maximum pixel value, minimum pixel value, pixel range, pixel standard deviation, pixel median, energy, root mean square, pixel gray scale value at the first 10%, pixel gray scale value at the last 10%, quarter-bit distance of a pixel, mean absolute deviation of a pixel, skewness of a pixel, kurtosis of a pixel, energy based on image gray histogram, entropy, and consistency; and extracting 2-dimensional shape features of the crystalline lens nuclear region in the sample eye image, wherein the shape features are the thickness and the length of the crystalline lens nuclear region respectively. And then splicing the 18-dimensional region features and the 2-dimensional shape features in sequence to obtain 20-dimensional vector features, and taking the vector features as target features of the sample eye image.
In order to fully represent the characteristics of the lenticular nucleus region in the eye image, as an optional way of the embodiment of the present invention, the extracting of the region characteristics of the lenticular nucleus region in the sample eye image may be performed by performing a blocking process on the lenticular nucleus region to obtain at least two sub-images of the lenticular nucleus region; and extracting the area characteristics of the sub-images of the crystalline lens nuclear area, and determining the area characteristics of the crystalline lens nuclear area in the sample eye image according to the area characteristics of the sub-image blocks of the crystalline lens nuclear area.
Specifically, dividing a crystalline lens nuclear region in a sample eye image into an upper part and a lower part to obtain two crystalline lens nuclear region sub-images; respectively extracting 18-dimensional region features of the lens nuclear region sub-images, splicing the 18-dimensional region features of each lens nuclear region sub-image to obtain 36-dimensional vector features, and taking the vector features as the region features of the lens nuclear region in the sample eye image. And then, the 36-dimensional region feature and the 2-dimensional shape feature are spliced to obtain a 38-dimensional vector feature, and the vector feature is used as a target feature of the sample eye image.
Further, the 18-dimensional region features of the lens nuclear region sub-images, the 18-dimensional region features of the lens nuclear region, and the 2-dimensional shape features of the lens nuclear region may be concatenated to obtain 56-dimensional vector features, which may be used as target features of the sample eye image.
It should be noted that, in the embodiment of the present invention, only an example that the crystalline lens nuclear region in the sample eye image is divided into two parts is given, the crystalline lens nuclear region in the sample eye image may also be divided into other numbers of crystalline lens nuclear region sub-images, and according to the above optional embodiment, the region feature of each crystalline lens nuclear region sub-image is determined, and then the target feature of the sample eye image is determined according to the region feature of each crystalline lens nuclear region sub-image and the shape feature of the crystalline lens nuclear region.
120. And constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified.
The eye image to be identified is an eye image which needs to be subjected to nuclear cataract grade judgment in clinic. Nuclear cataracts are classified into low grade and high grade.
In this embodiment, a supervised classification recognition algorithm in machine learning may be adopted, a nuclear cataract recognition model may be constructed based on the target features of the sample eye images and the nuclear cataract levels of the sample eye images, and then the core cataract level of the eye image to be recognized may be determined based on the nuclear cataract recognition model.
Optionally, an unsupervised classification recognition algorithm in machine learning may be adopted, a nuclear cataract recognition model is constructed based on the target features of the sample eye images, and then the grade of the core cataract of the eye image to be recognized is determined based on the nuclear cataract recognition model.
According to the technical scheme of the embodiment of the invention, the target characteristics of the sample eye image are determined by extracting the regional characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image and determining the target characteristics of the sample eye image according to the regional characteristics and the shape characteristics; and then constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, wherein the nuclear cataract identification model is used for determining the nuclear cataract grade of the eye image to be identified. By the technical scheme, the pathological change characteristics of the crystalline lens nuclear region in the fundus image can be accurately acquired, so that the grade judgment of the nuclear cataract is more accurate, a new thought is provided for the automatic diagnosis of the nuclear cataract, and an ophthalmologist can be assisted to efficiently, objectively and accurately diagnose the nuclear cataract.
Example two
Fig. 2 is a flowchart of a nuclear cataract identification method according to a second embodiment of the present invention; on the basis of the above embodiment, an optional implementation scheme is provided by further optimizing "determining the target feature of the sample eye image according to the region feature and the shape feature".
As shown in fig. 2, the method may specifically include:
s210, extracting the regional characteristics of the crystalline lens nuclear region in the sample eye image, and extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image.
And S220, determining candidate characteristics of the sample eye image according to the region characteristics and the shape characteristics.
In this embodiment, the region feature and the shape feature of the crystalline lens nuclear region in the sample eye image are used as candidate features of the sample eye image. For example, the 56-dimensional vector feature obtained in the above embodiment may be used as a candidate feature of the sample eye image.
And S230, performing dimension reduction processing on the candidate features, and determining the target features of the sample eye images.
In this embodiment, because some candidate features in the sample eye image have features that do not have a great influence on the subsequent nuclear cataract level identification, and in order to improve the speed of the subsequent nuclear cataract identification, the candidate features are subjected to dimension reduction processing, and the target features of the sample eye image are determined. Specifically, Principal Component Analysis (PCA) may be used to perform dimensionality reduction on the candidate features, and the target features of the sample eye image may be determined according to the result of dimensionality reduction.
Optionally, the candidate features may be analyzed based on an importance analysis model to determine scores of the candidate features; and taking the candidate features with the scores larger than the set value as the target features of the sample eye images. Wherein, the importance analysis model is determined by the technicians in the field based on a logistic regression method and a characteristic iterative deletion method; the set value is set by a person skilled in the art according to actual conditions.
Specifically, candidate features of the sample eye image are input into the importance analysis model to obtain scores of the candidate features, and if the scores are larger than or equal to a set value, the candidate features are reserved; and if the score is smaller than the set value, deleting the candidate feature. For example, the candidate feature of 56 dimensions obtained in the above embodiment is input into the importance analysis model, where the pixel range, the pixel minimum value, and the score of the pixel gray scale value located in the first 10% are all smaller than the set value, that is, the three features have small influence on the subsequent nuclear cataract level determination, and therefore, the three features are deleted to obtain a vector feature of 53 dimensions, and the vector feature is used as the target feature of the sample eye image.
240. And constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified.
According to the technical scheme of the embodiment of the invention, the candidate characteristics of the sample eye image are determined according to the region characteristics and the shape characteristics, and then the candidate characteristics are subjected to dimension reduction processing to determine the target characteristics of the sample eye image. According to the technical scheme, the regional characteristics and the shape characteristics are screened, so that the target characteristics of the eye image are more accurate, the nuclear cataract grade judgment is more accurate, meanwhile, the regional characteristics and the shape characteristics are screened, the dimensionality of the target characteristics is reduced, and the speed of subsequently identifying the nuclear cataract grade is improved.
EXAMPLE III
Fig. 3 is a schematic structural diagram of a nuclear cataract identification device according to a third embodiment of the present invention; the embodiment can be applied to the nuclear cataract identification, and the device can be realized by software/hardware and can be integrated in electronic equipment bearing the nuclear cataract identification function, such as server equipment.
As shown in fig. 3, the apparatus may include a target feature determination module 310 and a recognition model construction module 320, wherein,
the target feature determination module 310 is configured to extract a region feature of a crystalline lens nuclear region in the sample eye image, extract a shape feature of the crystalline lens nuclear region in the sample eye image, and determine a target feature of the sample eye image according to the region feature and the shape feature;
and the identification model building module 320 is used for building a nuclear cataract identification model based on the target characteristics of the sample eye image, and is used for determining the nuclear cataract grade of the eye image to be identified.
According to the technical scheme of the embodiment of the invention, the target characteristics of the sample eye image are determined by extracting the regional characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image and determining the target characteristics of the sample eye image according to the regional characteristics and the shape characteristics; and then constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, wherein the nuclear cataract identification model is used for determining the nuclear cataract grade of the eye image to be identified. By the technical scheme, the pathological change characteristics of the crystalline lens nuclear region in the fundus image can be accurately acquired, so that the grade judgment of the nuclear cataract is more accurate, a new thought is provided for the automatic diagnosis of the nuclear cataract, and an ophthalmologist can be assisted to efficiently, objectively and accurately diagnose the nuclear cataract.
Further, the regional features include at least one of the following regional feature dimensions: pixel mean, pixel variance, maximum pixel value, minimum pixel value, pixel range, pixel standard deviation, pixel median, energy, root mean square, pixel gray scale value within a preset range, pixel four-quadrant spacing, pixel average absolute deviation, pixel skewness, pixel kurtosis, energy based on image gray histogram, entropy, and consistency.
Further, the target feature determination module 310 comprises a sub-image determination unit and a region feature determination unit, wherein,
the subimage determining unit is used for carrying out blocking processing on the crystalline lens nuclear region to obtain at least two crystalline lens nuclear region subimages;
and the region characteristic determining unit is used for extracting the region characteristics of the sub-images of the crystalline lens nuclear region and determining the region characteristics of the crystalline lens nuclear region in the sample eye image according to the region characteristics of the sub-image blocks of the crystalline lens nuclear region.
Further, the target feature determination module 310 further comprises a candidate feature determination unit and a target feature determination unit, wherein,
a candidate feature determination unit configured to determine a candidate feature of the sample eye image according to the region feature and the shape feature;
and the target characteristic determining unit is used for performing dimension reduction processing on the candidate characteristics and determining the target characteristics of the sample eye image.
Further, the target feature determination unit comprises a score determination subunit and a target feature determination subunit, wherein,
the score determining subunit is used for analyzing the candidate characteristics based on the importance analysis model and determining the scores of the candidate characteristics;
and the target characteristic determining subunit is used for taking the candidate characteristic with the score larger than the set value as the target characteristic of the sample eye image.
The nuclear cataract identification device can execute the nuclear cataract identification method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
Example four
Fig. 4 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention, and fig. 4 shows a block diagram of an exemplary device suitable for implementing the embodiment of the present invention. The device shown in fig. 4 is only an example and should not bring any limitation to the function and the scope of use of the embodiments of the present invention.
As shown in FIG. 4, electronic device 12 is embodied in the form of a general purpose computing device. The components of electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including the system memory 28 and the processing unit 16.
Electronic device 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.
The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)30 and/or cache memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 4, and commonly referred to as a "hard drive"). Although not shown in FIG. 4, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.
A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 42 generally carry out the functions and/or methodologies of embodiments described herein.
Electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), with one or more devices that enable a user to interact with electronic device 12, and/or with any devices (e.g., network card, modem, etc.) that enable electronic device 12 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via the network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via the bus 18. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
The processing unit 16 executes programs stored in the system memory 28 to perform various functional applications and data processing, such as implementing the nuclear cataract identification method provided by the embodiments of the present invention.
EXAMPLE five
Fifth, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program (or referred to as computer-executable instructions) is stored, where the computer program is used to, when executed by a processor, perform the nuclear cataract identification method provided by the fifth embodiment of the present invention, where the method includes:
extracting the region characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the region characteristics and the shape characteristics;
and constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified.
Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for embodiments of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the embodiments of the present invention have been described in more detail through the above embodiments, the embodiments of the present invention are not limited to the above embodiments, and many other equivalent embodiments may be included without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (10)
1. A method of identifying nuclear cataracts, comprising:
extracting the region characteristics of the crystalline lens nuclear region in the sample eye image, extracting the shape characteristics of the crystalline lens nuclear region in the sample eye image, and determining the target characteristics of the sample eye image according to the region characteristics and the shape characteristics;
and constructing a nuclear cataract identification model based on the target characteristics of the sample eye image, and determining the nuclear cataract grade of the eye image to be identified.
2. The method of claim 1, wherein the regional characteristics comprise at least one regional characteristic dimension of: pixel mean, pixel variance, maximum pixel value, minimum pixel value, pixel range, pixel standard deviation, pixel median, energy, root mean square, pixel gray scale value within a preset range, pixel four-quadrant spacing, pixel average absolute deviation, pixel skewness, pixel kurtosis, energy based on image gray histogram, entropy, and consistency.
3. The method of claim 1, wherein extracting the regional features of the lenticular nuclear region in the sample eye image comprises:
the lens nucleus region is subjected to blocking processing to obtain at least two lens nucleus region sub-images;
and extracting the area characteristics of the sub-images of the crystalline lens nuclear area, and determining the area characteristics of the crystalline lens nuclear area in the sample eye image according to the area characteristics of the sub-image blocks of the crystalline lens nuclear area.
4. The method of claim 1, wherein determining the target feature of the sample eye image from the region feature and the shape feature comprises:
determining candidate features of the sample eye image according to the region features and the shape features;
and performing dimension reduction processing on the candidate features to determine target features of the sample eye images.
5. The method of claim 4, wherein performing dimension reduction on the candidate features to determine the target feature of the sample eye image comprises:
analyzing the candidate characteristics based on an importance analysis model to determine scores of the candidate characteristics;
and taking the candidate features with the scores larger than a set value as target features of the sample eye image.
6. A nuclear cataract identification device, comprising:
the target feature determination module is used for extracting the region feature of the crystalline lens nuclear region in the sample eye image, extracting the shape feature of the crystalline lens nuclear region in the sample eye image, and determining the target feature of the sample eye image according to the region feature and the shape feature;
and the identification model building module is used for building a nuclear cataract identification model based on the target characteristics of the sample eye image and determining the nuclear cataract grade of the eye image to be identified.
7. The apparatus of claim 6, wherein the regional characteristics comprise at least one regional characteristic dimension: pixel mean, pixel variance, maximum pixel value, minimum pixel value, pixel range, pixel standard deviation, pixel median, energy, root mean square, pixel gray scale value within a preset range, average absolute deviation of pixels, skewness of pixels, kurtosis of pixels, energy based on image gray histogram, entropy, and consistency.
8. The apparatus of claim 6, wherein the target feature determination module comprises:
the subimage determining unit is used for carrying out blocking processing on the crystalline lens nuclear region to obtain at least two crystalline lens nuclear region subimages;
and the region feature determining unit is used for extracting the region features of the sub-images of the crystalline lens nuclear region and determining the region features of the crystalline lens nuclear region in the sample eye image according to the region features of the sub-image blocks of the crystalline lens nuclear region.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the nuclear cataract identification method of any one of claims 1-5.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out a method for nuclear cataract identification as claimed in any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110766487.4A CN113361482A (en) | 2021-07-07 | 2021-07-07 | Nuclear cataract identification method, device, electronic device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110766487.4A CN113361482A (en) | 2021-07-07 | 2021-07-07 | Nuclear cataract identification method, device, electronic device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113361482A true CN113361482A (en) | 2021-09-07 |
Family
ID=77538717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110766487.4A Pending CN113361482A (en) | 2021-07-07 | 2021-07-07 | Nuclear cataract identification method, device, electronic device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113361482A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116309584A (en) * | 2023-05-22 | 2023-06-23 | 泰安光明爱尔眼科医院有限公司 | Image processing system for cataract area identification |
| CN116612339A (en) * | 2023-07-21 | 2023-08-18 | 中国科学院宁波材料技术与工程研究所 | Construction device and grading device of nuclear cataract image grading model |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4191176A (en) * | 1976-02-24 | 1980-03-04 | Novo Laboratories, Inc. | Intralenticular cataract surgery |
| US20030186294A1 (en) * | 2000-08-25 | 2003-10-02 | Xiangyin Kong | Method of diagnosing and treating lesion of crystalline lens using human crygs gene and coding product thereof |
| CN102202557A (en) * | 2008-05-20 | 2011-09-28 | 科学、技术与研究机构 | An automatic opacity detection system for cortical cataract diagnosis |
| US20120155726A1 (en) * | 2009-08-24 | 2012-06-21 | Huiqi Li | method and system of determining a grade of nuclear cataract |
| JP2012239639A (en) * | 2011-05-19 | 2012-12-10 | Konan Medical Inc | Crystalline lens image analysis instrument |
| CN109376782A (en) * | 2018-10-26 | 2019-02-22 | 北京邮电大学 | Support vector machines cataract stage division and device based on eye image feature |
| CN109872322A (en) * | 2019-02-27 | 2019-06-11 | 电子科技大学 | A kind of nuclear cataract lesion region localization method based on cascade detection model |
| CN109919932A (en) * | 2019-03-08 | 2019-06-21 | 广州视源电子科技股份有限公司 | The recognition methods of target object and device |
| CN110738527A (en) * | 2019-10-17 | 2020-01-31 | 中国建设银行股份有限公司 | feature importance ranking method, device, equipment and storage medium |
| CN110909005A (en) * | 2019-11-29 | 2020-03-24 | 广州市百果园信息技术有限公司 | Model feature analysis method, device, equipment and medium |
| CN112006651A (en) * | 2020-09-10 | 2020-12-01 | 孙礼华 | Cataract surgery auxiliary diagnosis system and method thereof |
-
2021
- 2021-07-07 CN CN202110766487.4A patent/CN113361482A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4191176A (en) * | 1976-02-24 | 1980-03-04 | Novo Laboratories, Inc. | Intralenticular cataract surgery |
| US20030186294A1 (en) * | 2000-08-25 | 2003-10-02 | Xiangyin Kong | Method of diagnosing and treating lesion of crystalline lens using human crygs gene and coding product thereof |
| CN102202557A (en) * | 2008-05-20 | 2011-09-28 | 科学、技术与研究机构 | An automatic opacity detection system for cortical cataract diagnosis |
| US20120155726A1 (en) * | 2009-08-24 | 2012-06-21 | Huiqi Li | method and system of determining a grade of nuclear cataract |
| CN102984997A (en) * | 2009-08-24 | 2013-03-20 | 新加坡保健服务集团有限公司 | A Method and system of determining a grade of nuclear cataract |
| JP2012239639A (en) * | 2011-05-19 | 2012-12-10 | Konan Medical Inc | Crystalline lens image analysis instrument |
| CN109376782A (en) * | 2018-10-26 | 2019-02-22 | 北京邮电大学 | Support vector machines cataract stage division and device based on eye image feature |
| CN109872322A (en) * | 2019-02-27 | 2019-06-11 | 电子科技大学 | A kind of nuclear cataract lesion region localization method based on cascade detection model |
| CN109919932A (en) * | 2019-03-08 | 2019-06-21 | 广州视源电子科技股份有限公司 | The recognition methods of target object and device |
| CN110738527A (en) * | 2019-10-17 | 2020-01-31 | 中国建设银行股份有限公司 | feature importance ranking method, device, equipment and storage medium |
| CN110909005A (en) * | 2019-11-29 | 2020-03-24 | 广州市百果园信息技术有限公司 | Model feature analysis method, device, equipment and medium |
| CN112006651A (en) * | 2020-09-10 | 2020-12-01 | 孙礼华 | Cataract surgery auxiliary diagnosis system and method thereof |
Non-Patent Citations (6)
| Title |
|---|
| A L WONG 等: "Quantitative assessment of lens opacities with anterior segment optical coherence tomography", BRITISH JOURNAL OF OPHTHALMOLOGY, 6 October 2008 (2008-10-06), pages 61 - 65 * |
| JULIE-ANNE LITTLE等: "In-vivo anterior segment OCT imaging provides unique insight into cerulean blue-dot opacities and cataracts in Down syndrome", NATURE, 22 June 2020 (2020-06-22), pages 1 - 7 * |
| MARIA A HENRIQUEZ 等: "Correlation between lens thickness and lens density in patients with mild to moderate cataracts", BRITISH JOURNAL OF OPHTHALMOLOGY, vol. 104, no. 10, 19 December 2019 (2019-12-19), pages 1 - 8 * |
| XIAOQING ZHANG 等: "Nuclear cataract classification in anterior segment OCT based on clinical global–local features", COMPLEX & INTELLIGENT SYSTEMS, no. 9, 16 September 2022 (2022-09-16), pages 1479 - 1493 * |
| 唐由之, 马文新, 王萌建, 黄晓燕, 何进, 王吉龙: "白内障的诊断与晶状体图像计算机分析系统的研究", 中国中医眼科杂志, no. 03, 11 August 1998 (1998-08-11), pages 139 - 144 * |
| 杨平 等: "影响高度近视白内障患者疗效的相关因素分析", 《国际眼科杂志》, vol. 17, no. 7, 31 July 2017 (2017-07-31), pages 1374 - 1377 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116309584A (en) * | 2023-05-22 | 2023-06-23 | 泰安光明爱尔眼科医院有限公司 | Image processing system for cataract area identification |
| CN116612339A (en) * | 2023-07-21 | 2023-08-18 | 中国科学院宁波材料技术与工程研究所 | Construction device and grading device of nuclear cataract image grading model |
| CN116612339B (en) * | 2023-07-21 | 2023-11-14 | 中国科学院宁波材料技术与工程研究所 | Construction device and grading device of nuclear cataract image grading model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bilal et al. | Diabetic retinopathy detection and classification using mixed models for a disease grading database | |
| CN108464840B (en) | Automatic detection method and system for breast lumps | |
| CN110222759B (en) | Automatic identification system for vulnerable plaque of coronary artery | |
| CN110245657B (en) | Pathological image similarity detection method and detection device | |
| CN109191451B (en) | Abnormality detection method, apparatus, device, and medium | |
| CN111598867B (en) | Method, apparatus, and computer-readable storage medium for detecting specific facial syndrome | |
| WO2021114817A1 (en) | Oct image lesion detection method and apparatus based on neural network, and medium | |
| KR102155381B1 (en) | Method, apparatus and software program for cervical cancer decision using image analysis of artificial intelligence based technology | |
| SV | Computer-aided diagnosis of anterior segment eye abnormalities using visible wavelength image analysis based machine learning | |
| Xiao et al. | Major automatic diabetic retinopathy screening systems and related core algorithms: a review | |
| CN113361482A (en) | Nuclear cataract identification method, device, electronic device and storage medium | |
| CN115731203A (en) | Cataract image identification method and device, computer equipment and readable storage medium | |
| CN115578783A (en) | Device and method for identifying eye diseases based on eye images and related products | |
| Adorno III et al. | Advancing eosinophilic esophagitis diagnosis and phenotype assessment with deep learning computer vision | |
| CN116681923A (en) | Automatic ophthalmic disease classification method and system based on artificial intelligence | |
| Akut | FILM: finding the location of microaneurysms on the retina | |
| Zhang et al. | Artificial intelligence-assisted diagnosis of ocular surface diseases | |
| CN113485555B (en) | Medical image film reading method, electronic equipment and storage medium | |
| CN114005541A (en) | Dynamic dry eye early warning method and system based on artificial intelligence | |
| Ríos et al. | A deep learning model for classification of diabetic retinopathy in eye fundus images based on retinal lesion detection | |
| KR20210033902A (en) | Method, apparatus and software program for cervical cancer diagnosis using image analysis of artificial intelligence based technology | |
| CN115526882A (en) | Medical image classification method, device, equipment and storage medium | |
| EP3907696A1 (en) | Method and system for identifying abnormal images in a set of medical images | |
| Islam et al. | Severity grading of diabetic retinopathy using deep convolutional neural network | |
| US11798163B2 (en) | Systems and methods for quantitative phenotyping of fibrosis |
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 |