Quantitative evaluation method, apparatus, and storage medium for conjunctival hyperemia

Abstract

The present invention discloses a method, an apparatus, and a storage medium for quantitatively evaluating conjunctival congestion. The method for quantitatively evaluating conjunctival congestion includes step S1 of acquiring a video of the movement of the eye from the inner side to the outer side or from the outer side to the inner side, step S2 of performing eye image segmentation on the video to obtain an eye conjunctival image and extracting the red channel value and the blue channel value of each pixel from the eye conjunctival image, and step S3 of determining the percentage of conjunctival congestion based on the proportional value between the red channel value and the blue channel value of each pixel. By the method for quantitatively evaluating conjunctival congestion, an accurate value of conjunctival congestion can be obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of eye detection, and particularly to a method, apparatus, and storage medium for quantitatively evaluating conjunctival congestion.

Background Art

[0002] The degree of conjunctival congestion is one of the important reference indicators when a doctor evaluates a patient's eye inflammation. Clinically, the doctor needs to judge whether the degree of conjunctival congestion is mild, moderate, or severe within 1 m from the patient by artificial experience. Due to the subjective consciousness of different doctors, there is no unified standard for judging mild, moderate, and severe degrees, which is likely to cause misdiagnosis and missed diagnosis.

[0003] There is a problem that the accuracy in manual measurement is affected by human factors. In the prior art, image recognition-related technical means depending on shape have been proposed. As described in Chinese Patent CN111212594A (Application No.: CN201880066273.2, Publication Date: May 29, 2020), an image including an eye photographed by a camera is acquired, one or more blood vessels in the image are identified, and the degree of conjunctival congestion is determined based on the size of the one or more identified blood vessels. Further, in the prior art, a method for examining the degree of conjunctival congestion by using the ratio of the red pixel value of the eye part is disclosed. In this method, specifically, a high-resolution eye image is acquired, the eye image is converted from RGB to gray, a Gaussian filter is used to find an area where the difference in gradients in the horizontal and vertical directions is large, and a filter heat map is used to utilize a threshold value to divide the congested part and the non-congested part.

[0004] However, eye redness is often conjunctival congestion, and the redness in this area is often diffuse. Therefore, conventional methods that rely on shape for identification have the following drawbacks: 1. Identification is inaccurate because it is easily affected by the intensity of the blood vessel color, making it impossible to accurately extract red pixel values. 2. Due to the characteristics of the human eyeball structure, there is shading at the intersection of the eyelid and the eyeball, which affects accuracy. 3. When the same light beam is shone on the eyeball, the light rays become uneven because the eyeball is spherical. [Overview of the project] [Problems that the invention aims to solve]

[0005] To address the problem of low accuracy in measuring the degree of conjunctival hyperemia using conventional techniques, the present invention provides a quantitative evaluation method for conjunctival hyperemia. [Means for solving the problem]

[0006] To achieve the above objective, the present invention employs the following technical means.

[0007] In one embodiment, the quantitative evaluation method for conjunctival hyperemia according to the present invention is: Step S1 involves obtaining a video of the movement of the eyeball from the inside to the outside or from the outside to the inside, Step S2 involves performing eye image segmentation on the aforementioned video to obtain an eye conjunctival image, and extracting the red channel value and blue channel value of each pixel from the eye conjunctival image. The process includes step S3, which determines the percentage of conjunctival hyperemia based on the proportional value between the red channel value and the blue channel value of each pixel.

[0008] Furthermore, in step S2, second eye position images are acquired when the eyeball has moved to its medial and lateral limits, respectively, and the second eye position image when the eyeball has moved to its medial limit and the second eye position image when the eyeball has moved to its lateral limit are combined to acquire the conjunctival image of the eye.

[0009] Furthermore, in step S3, it is determined that a pixel point where the proportional value between the red channel value and the blue channel value exceeds a threshold is a conjunctival hyperemia area, and the percentage of conjunctival hyperemia is the percentage of the conjunctival area of ​​the eye conjunctival image occupied by the area of ​​the conjunctival hyperemia pixel point.

[0010] Furthermore, the threshold is either a pre-set fixed value or a manually set value.

[0011] In one embodiment, the quantitative evaluation device for conjunctival hyperemia according to the present invention is: A video acquisition module that acquires video of eye movement from the inside to the outside or from the outside to the inside in response to a video acquisition command, A feature extraction module that performs eye image segmentation on the aforementioned video to obtain an eye conjunctival image, and extracts the red channel value and blue channel value of each pixel from the eye conjunctival image, It includes a calculation module that calculates the proportional value between the red channel value and the blue channel value of each pixel to determine the percentage of conjunctival hyperemia.

[0012] Furthermore, the feature extraction module further includes an image extraction module that extracts a second eye position image from the video showing the eyeball moving to its limit positions inward and outward.

[0013] In one embodiment, the computer-readable storage medium according to the present invention stores at least one program code that is loaded and executed by a processor to implement the quantitative evaluation method of conjunctival hyperemia in the first embodiment described above. [Effects of the Invention]

[0014] Compared to the prior art, the present invention has at least the following beneficial effects.

[0015] In this invention, an ocular conjunctival image is synthesized using a second eye position image in which the eyeball has moved to its limit inward and outward positions. Then, red channel values ​​and blue channel values ​​are extracted from each pixel of the image. Since the light reflected in the congested area is mainly red light, and the sclera itself has high reflectivity to both red and blue light, the red channel value is significantly higher than the blue channel value in areas with congestion, and the red and blue channel values ​​are equal in areas without congestion. The congested points can be accurately determined by the proportional value between the red and blue channel values ​​of each pixel, and the percentage of conjunctival congestion can be determined. [Brief explanation of the drawing]

[0016] [Figure 1] This is a flowchart of the quantitative evaluation method for conjunctival hyperemia according to the present invention. [Figure 2] This is a block diagram of the quantitative evaluation principle of conjunctival hyperemia according to the present invention. [Figure 3] This is a step diagram illustrating the quantitative evaluation of conjunctival hyperemia according to the present invention. [Figure 4] This is a schematic diagram of the quantitative evaluation device for conjunctival hyperemia according to the present invention. [Modes for carrying out the invention]

[0017] Specific embodiments of the present invention will be described in more detail below with reference to the drawings and examples. The following examples are for illustrative purposes only and do not limit the scope of the present invention.

[0018] As shown in Figures 1-3, the quantitative evaluation method for conjunctival hyperemia includes the following steps S1-S3.

[0019] In step S1, we obtain a video of the movement of the eyeball from the inside to the outside or from the outside to the inside. Specifically, in step S1, the eye and head of the user are fixed by an eye evaluation or measurement device, and a video of the eyeball moving horizontally from the inside to the outside or from the outside to the inside is captured and obtained by a camera provided in front of the user's eye. The movement of the user's eyeball may be instructed using an indicator lamp.

[0020] Videos of the necessary eye movements may be obtained by other eye devices or video collection devices.

[0021] In step S2, eye image segmentation is performed on the video to obtain an eye conjunctiva image, and the red channel value and blue channel value of each pixel are extracted from the eye conjunctiva image. Specifically, in step S2, an eye conjunctiva image with the eyeball removed can be obtained by performing eye image segmentation on the video. Specifically, second eye position images in which the eyeball moves to the inner and outer limit positions from the video are respectively extracted, and the second eye position image in which the extracted eyeball moves to the inner limit position and the second eye position image in which the eyeball moves to the outer limit position are combined to obtain an eye conjunctiva image, thereby obtaining an eye conjunctiva image unaffected by the eyeball image, and the red channel value and blue channel value of each pixel point are extracted from the eye conjunctiva image.

[0022] Note that the eye position refers to the position of the eyeball in an ophthalmological examination and is divided into the first eye position, the second eye position, and the third eye position. The first eye position refers to the eye position when both eyes are looking straight at an infinite point on the horizontal plane. The second eye position refers to the eye position when the eyeball rotates upward, downward, inward, and outward. The third eye position refers to the eye position when the eyeball rotates diagonally upward, downward, inward, and outward, that is, the eye position when it rotates upward, downward, superior temporally, and inferior temporally. Accordingly, the first, second, and third eye position images respectively refer to images captured when the eyeball is in each eye position.

[0023] In step S3, the percentage of conjunctival congestion is determined based on the proportional value of the red channel value and the blue channel value of each pixel.

[0024] In step S3, a pixel point where the proportional value between the red channel value and the blue channel value exceeds a threshold is determined to be a conjunctival hyperemia area, and the percentage of conjunctival hyperemia is the percentage of the conjunctival area of ​​the conjunctival image occupied by the area of ​​the conjunctival hyperemia pixel point. Since the light reflected in the conjunctival hyperemia area is mainly red light, and the sclera itself has a high reflectivity for both red and blue light, the red channel value is significantly higher than the blue channel value in areas with conjunctival hyperemia, and the red channel value and blue channel value are equal in areas without conjunctival hyperemia. Therefore, the percentage of conjunctival hyperemia is determined by the proportional value between the red channel value and the blue channel value of each pixel. Specifically, the threshold is either a preset fixed value or a manually set value. For example, a threshold of 3 may be pre-entered in the computer, and a pixel point where the proportional value between the red channel value and the blue channel value exceeds 3 is determined to be a congested area. If the RGB value of a certain pixel point is RGB(250, 245, 240), and the red channel value and the blue channel value are equal, and the ratio of the two is close to 1, then this pixel point is not a congested area but the white color of the sclera itself. If the RGB value of a certain pixel point is RGB(250, 20, 15), and the proportional value between the red channel value and the blue channel value exceeds the threshold of 3, then this pixel point is determined to be a congested area. In addition to pre-entering the threshold, the threshold may also be a value manually set by a technician such as a physician or operator based on their experience.

[0025] Figure 3 shows a step-by-step diagram of the quantitative evaluation method for conjunctival hyperemia in more detail. First, the user controls the left-right movement of the eyeball in front of the camera and records it on video. Next, the eye image is segmented, and during the segmentation, second eye position images are obtained when the eyeball has moved to its medial and lateral limits. The second eye position image when the eyeball has moved to its medial limit and the second eye position image when the eyeball has moved to its lateral limit are combined to obtain an eye conjunctival image. The red channel value and blue channel value of each pixel point in the eye conjunctival image are extracted, and the ratio of the red and blue channel pixel values ​​of each pixel point is obtained. Furthermore, the percentage of the hyperemia is obtained by determining whether each pixel point is a hyperemic area based on a preset threshold or a manually set threshold.

[0026] Figure 4 is a schematic diagram of a quantitative evaluation device 40 for conjunctival hyperemia according to an embodiment of the present invention. As shown in Figure 4, the device includes a video acquisition module 401, a feature extraction module 402, and a calculation module 403.

[0027] The video acquisition module 401 acquires video of eye movement from the inside to the outside or from the outside to the inside in response to a video acquisition command. In one possible embodiment, when the video acquisition module receives an acquisition command, it can guide the user's eye movements by controlling a camera or other imaging equipment to activate and by controlling audio equipment or lighting equipment to acquire video of the required eye movements.

[0028] The feature extraction module 402 performs eye image segmentation on the video to obtain an eye conjunctival image, and extracts the red channel value and blue channel value of each pixel from the eye conjunctival image.

[0029] The calculation module 403 calculates the proportional value between the red channel value and the blue channel value for each pixel to determine the percentage of conjunctival hyperemia.

[0030] Furthermore, the feature extraction module further includes an image extraction module that extracts a second eye position image from the video showing the eyeball moving to its limit positions inward and outward.

[0031] In an exemplary embodiment, a computer-readable storage medium is further provided, which includes a memory that stores at least one program code that is loaded and executed by a processor to implement the quantitative evaluation method of conjunctival hyperemia in the above embodiment. For example, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CDROM), a magnetic tape, a floppy disk, or an optical data storage device.

[0032] Those skilled in the art will understand that all or part of the steps of the above embodiment may be completed by hardware, or by hardware associated with at least one program code, and that the program may be stored in a computer-readable storage medium, the storage medium of which may be read-only memory, a magnetic disk, or an optical disk, etc.

[0033] The foregoing are merely preferred embodiments of the present invention and do not limit it. All modifications, equivalent substitutions, and improvements made within the conceptual framework and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for quantitatively evaluating conjunctival hyperemia performed by computer, Step S1 involves obtaining a video of the movement of the eyeball from the inside to the outside or from the outside to the inside, Step S2 involves performing eye image segmentation on the aforementioned video to obtain an eye conjunctival image, and extracting the red channel value and blue channel value of each pixel from the eye conjunctival image. Step S3 includes determining the percentage of conjunctival hyperemia based on the proportional value between the red channel value and the blue channel value of each pixel, In step S2, by performing eye image segmentation on the video, second eye position images are obtained, respectively, of the eyeball moving to its medial and lateral limits. The second eye position image of the eyeball moving to its medial limit and the second eye position image of the eyeball moving to its lateral limit are combined to obtain an eye conjunctival image with the eyeball removed. A quantitative evaluation method for conjunctival hyperemia, characterized by the following:

2. In step S3, a pixel point where the proportional value between the red channel value and the blue channel value exceeds a threshold is determined to be a conjunctival hyperemia area, and the percentage of conjunctival hyperemia is the percentage of the conjunctival area of ​​the ocular conjunctival image occupied by the area of ​​the conjunctival hyperemia pixel point. The method for quantitatively evaluating conjunctival hyperemia according to feature 1.

3. The threshold is either a pre-set fixed value or a manually set value. The quantitative evaluation method for conjunctival hyperemia according to feature 2.

4. A video acquisition module that acquires video of eye movement from the inside to the outside or from the outside to the inside in response to a video acquisition command, A feature extraction module that performs eye image segmentation on the aforementioned video to obtain an eye conjunctival image, and extracts the red channel value and blue channel value of each pixel from the eye conjunctival image, It includes a calculation module that calculates the proportional value between the red channel value and the blue channel value of each pixel to determine the percentage of conjunctival hyperemia, The aforementioned feature extraction module is This image processing submodule includes the following steps: performing eye image segmentation on the video extracts second eye position images where the eyeball has moved to its medial and lateral limits, and then combining the second eye position image where the eyeball has moved to its medial limit and the second eye position image where the eyeball has moved to its lateral limit to obtain the conjunctival image of the eye region with the eyeball removed. A quantitative evaluation device for conjunctival hyperemia, characterized by the following features.

5. A processor stores at least one program code that is loaded and performs the quantitative evaluation method of conjunctival hyperemia according to any one of claims 1 to 3. A computer-readable storage medium characterized by the following features.

Images