Single corneal endothelial cell function evaluation method based on cellular morphology

By acquiring corneal endothelial cell images and calculating confidence levels, the problem that traditional detection methods cannot accurately reflect cell function has been solved, enabling more efficient functional assessment and visualization, and improving diagnostic accuracy.

CN118918087APending Publication Date: 2024-11-08EYE INST OF SHANDONG FIRST MEDICAL UNIV
View PDF 13 Cites 0 Cited by

Patent Information

Application Number
CN202411012268.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Traditional methods for detecting corneal endothelial cells cannot accurately reflect cell function, affecting the accuracy of clinical diagnosis. In particular, when the cell count is below 400-500 cells/mm2, it may lead to problems such as corneal edema and opacity.

Method used

By acquiring images of corneal endothelial cells, image recognition and processing are performed to calculate the node coordinates, area, and diagonal of each cell, calculate the confidence value, and visualize the data to assess cell function.

Benefits of technology

It improves the accuracy of corneal endothelial cell function assessment and the visualization of diagnosis, better reflects the functional status of each cell, and reduces the difficulty of diagnosis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918087A_ABST
    Figure CN118918087A_ABST
Patent Text Reader

Abstract

The invention provides a single corneal endothelial cell function evaluation method based on cellular morphology, and relates to the technical field of cell image recognition, and the method comprises the following steps: obtaining a target corneal endothelial cell image, and carrying out image recognition and processing on the target corneal endothelial cell image to obtain a corneal endothelial cell function evaluation result; obtaining node coordinates of each cell in the target corneal endothelial cell image; calculating the cell area, the longest diagonal line and the shortest diagonal line of each cell according to the node coordinates, and calculating the signal value of each cell according to the cell area, the longest diagonal line and the shortest diagonal line; and performing visual display according to the signal value and the node coordinate, and evaluating the function of each cell. The information value of each cell is calculated by combining the cell area and the diagonal line, and compared with parameters obtained by a traditional method, the parameters can better reflect the cell function.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of cell image recognition, and in particular to a method for evaluating the function of a single corneal endothelial cell based on cell morphology. Background Art

[0002] Corneal endothelial cells are a single layer of hexagonal cells on the innermost part of the cornea. They maintain normal corneal visual function by regulating corneal transparency and thickness. Corneal endothelial cells have limited proliferation capacity. After cell damage, they cannot regenerate and can only be repaired through the expansion and migration of surrounding cells. When the number of cells is less than 400-500 / mm 2 When corneal endothelial cells are detected, it can cause corneal edema, opacity, and even corneal blindness. Although the number and morphology of corneal endothelial cells are correlated with the function of the corneal endothelium, there is a clinical situation where traditional corneal endothelial cell number and morphology examinations cannot accurately reflect corneal thickness and transparency, seriously affecting clinicians' diagnosis of patients' conditions. Therefore, improving the accuracy of clinical examinations of corneal endothelial cell function and developing evaluation methods at the single-cell level of the corneal endothelium are of great significance to help prevent and treat clinical corneal endothelial lesions.

[0003] Normal corneal endothelial cells are regular hexagonal cells. According to the principle of minimum mechanical energy, the regular hexagon is the most stable morphological structure in physics, satisfying the maximum area and minimum perimeter. The diagonal is a line segment connecting two non-adjacent vertices in a polygon, and the relationship between different diagonals reflects the change in morphology. The traditional detection method uses the hexagon ratio as an indicator for the evaluation of corneal endothelial cells, but the overall cell heteromorphism cannot reflect the function of each cell. Therefore, it is necessary to use diagonals to assist in analyzing the degree of distortion of each cell, thereby indirectly reflecting the cell function. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a method for evaluating the function of a single corneal endothelial cell based on cell morphology.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for evaluating the function of individual corneal endothelial cells based on cell morphology, comprising:

[0007] Acquire a target corneal endothelial cell image, and perform image recognition and processing on the target corneal endothelial cell image to obtain node coordinates of each cell in the target corneal endothelial cell image;

[0008] Calculating the cell area, the longest diagonal and the shortest diagonal of each cell according to the node coordinates, and calculating the signal value of each cell according to the cell area, the longest diagonal and the shortest diagonal;

[0009] The signal value and the node coordinates are visualized and the function of each cell is evaluated.

[0010] Preferably, performing image recognition and processing on the target corneal endothelial cell image to obtain the node coordinates of each cell in the target corneal endothelial cell image includes:

[0011] Using a histogram equalization technique to enhance the contrast and details of the target corneal endothelial cell image, and converting the enhanced image into a grayscale image;

[0012] Applying filtering technology to denoise the grayscale image to obtain a denoised image, and calculating a segmentation threshold;

[0013] The denoised image is segmented into two regions, a cell and a background, by binarization according to the segmentation threshold, and a hole filling algorithm and an erosion operation are used to process the cell region after the threshold segmentation to obtain a target region image;

[0014] Calculate the center point position of the cell according to the target area image;

[0015] Generate a Voronoi diagram using the original information of the center point position and the target corneal endothelial cell image, and save the Voronoi diagram coordinate information as a json file;

[0016] The json file is exported to obtain the node coordinates of each cell.

[0017] Preferably, the calculation formula of the confidence value is: Among them, XV is the signal value, S is the cell area, S th is the critical cell area, D max is the longest diagonal of the cell, D min The shortest diagonal of the cell.

[0018] Preferably, performing visual display according to the signal value and the node coordinates includes:

[0019] Reading the node coordinates of the cells derived after identification, and establishing a three-dimensional cell model according to the node coordinates; the signal value of each cell in the three-dimensional cell model corresponds to a preset cell color scale; the value and color on the cell correspond to the color scale;

[0020] The three-dimensional cell model is visualized.

[0021] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0022] The present invention provides a method for evaluating the function of a single corneal endothelial cell based on cell morphology, comprising: acquiring a target corneal endothelial cell image, and performing image recognition and processing on the target corneal endothelial cell image to obtain the node coordinates of each cell in the target corneal endothelial cell image; calculating the cell area, the longest diagonal and the shortest diagonal of each cell according to the node coordinates, and calculating the signal value of each cell according to the cell area, the longest diagonal and the shortest diagonal; performing visualization according to the signal value and the node coordinates, and evaluating the function of each cell. The present invention combines the cell area and diagonal calculation to obtain the signal value of each cell, which can better reflect the cell function than the parameters obtained by the traditional method. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0024] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0025] Figure 2 A schematic diagram of cell identification provided by an embodiment of the present invention;

[0026] Figure 3 A schematic diagram of visualization of cells and signal values ​​provided in an embodiment of the present invention;

[0027] Figure 4 A schematic diagram of three-dimensional visualization of cells and signal values ​​provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a method for evaluating the function of a single corneal endothelial cell based on cell morphology, comprising:

[0031] Step 100: acquiring a target corneal endothelial cell image, and performing image recognition and processing on the target corneal endothelial cell image to obtain the node coordinates of each cell in the target corneal endothelial cell image;

[0032] Step 200: Calculate the cell area, the longest diagonal and the shortest diagonal of each cell according to the node coordinates, and calculate the signal value of each cell according to the cell area, the longest diagonal and the shortest diagonal;

[0033] Step 300: Perform visualization according to the signal value and the node coordinates, and evaluate the function of each cell.

[0034] Specifically, the detailed steps of step 100 in this embodiment are as follows:

[0035] Using a semi-supervised image recognition algorithm, corneal endothelial cells were identified. Figure 2 , the input corneal endothelial cell image is converted into a grayscale image after enhancing the contrast and details by using histogram equalization technology, and the image is denoised by using filtering technology and then the appropriate threshold is calculated. After the image is divided into two regions, cells and background, by binarization, the image after threshold segmentation is further processed by using hole filling algorithm and corrosion operation. The corrosion operation helps to refine the cell edge and eliminate the noise and artifacts introduced by threshold segmentation and hole filling. The center point position of the cell is calculated according to the segmented image to accurately locate and mark the cell. For the images with automatic identification errors, manual adjustments are made by adding and deleting the center point. Finally, the Voronoi diagram is generated using the center point data and the information of the original image, and the coordinate information is saved as a json file. The results are exported to the coordinates of the cell nodes for subsequent signal value calculation and visualization. Among them, the original image information is the size of the image (length and width), because a Voronoi diagram of the same size as the original image is to be generated.

[0036] Furthermore, step 200 of this embodiment is specifically as follows:

[0037] Calculate the signal value of each cell. The signal value calculation method is:

[0038]

[0039] XV is the signal value, S is the cell area, S th is the critical cell area, D max is the longest diagonal of the cell, D min The shortest diagonal of the cell; the known quantity is S th , the unknown quantities are S and D max and D minThe area of ​​each cell, the longest diagonal, and the shortest diagonal were calculated based on the node coordinates of image recognition. The critical cell area of ​​the decompensation period was based on the maximum value of the average cell area in the postoperative compensation period image of Qingdao Eye Hospital affiliated to Shandong First Medical University, which was a constant of 2854.41 μm 2 .

[0040] The Shoelace formula is used to calculate the area of ​​a polygon using the vertex coordinates. This formula is particularly effective for a simple polygon with vertices arranged in sequence. Suppose the vertices of the polygon are (x1, y1), (x2, y2), ..., (x n ,y n ) can be calculated using the following formula:

[0041]

[0042] Traverse all nodes and calculate the distance between the current node and other nodes (points on the same edge are not calculated) using the Euclidean distance formula. Given the coordinates of two points (x1, y1) and (x2, y2), the distance formula is:

[0043]

[0044] Calculate the longest diagonal D of each polygon max and the shortest diagonal D min .

[0045] Calculate the signal value for each cell.

[0046] Furthermore, step 300 of this embodiment is specifically as follows:

[0047] Visualize each cell and X, see Figure 3 . Read the cell node coordinates derived after identification, build a three-dimensional cell model, and assign a value to each cell. The value is the XV corresponding to the cell, see Figure 4 .like Figure 4 As shown in the figure, it is divided into two parts: 3D cell model and color scale. The 3D cell model is set to different colors, with a number in the middle of the cell, which is the cell XV; the upper limit of the color scale is the critical value of endothelial cell decompensation, and the lower limit is the minimum XV value of normal endothelial cells, which is obtained from research. The value and color on the cell correspond to the color scale. Since the outermost circle of the cell is not fully photographed, it is set to gray and does not participate in the evaluation.

[0048] Endothelial cell function is evaluated based on XV. The higher the proportion of cells whose XV exceeds the critical value of decompensation, the worse the cell function. When the proportion of cells whose XV exceeds the critical value of decompensation exceeds 10%, the corneal endothelial function is considered to be abnormal and further diagnosis is required.

[0049] The beneficial effects of the present invention are as follows:

[0050] (1) The semi-supervised learning image recognition method of the present invention can more accurately obtain the node information of each cell, thereby increasing the credibility of subsequent data;

[0051] (2) The present invention combines cell area and diagonal calculation to obtain the signal value of each cell, which can better reflect the cell function than the parameters obtained by traditional methods;

[0052] (3) The present invention displays the state of each cell through three-dimensional visualization. Compared with the traditional method, the cell model can be rotated and scaled, and the cells can be observed from multiple angles. The function of the corneal endothelium can be more intuitively displayed, thereby improving the accuracy of diagnosis and reducing the difficulty of diagnosis.

[0053] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0054] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for evaluating the function of a single corneal endothelial cell based on cell morphology, characterized in that: include: Acquire a target corneal endothelial cell image, and perform image recognition and processing on the target corneal endothelial cell image to obtain node coordinates of each cell in the target corneal endothelial cell image; Calculating the cell area, the longest diagonal and the shortest diagonal of each cell according to the node coordinates, and calculating the signal value of each cell according to the cell area, the longest diagonal and the shortest diagonal; The signal value and the node coordinates are visualized and the function of each cell is evaluated.

2. The method for evaluating the function of a single corneal endothelial cell based on cell morphology according to claim 1, characterized in that: Performing image recognition and processing on the target corneal endothelial cell image to obtain the node coordinates of each cell in the target corneal endothelial cell image includes: Using a histogram equalization technique to enhance the contrast and details of the target corneal endothelial cell image, and converting the enhanced image into a grayscale image; Applying filtering technology to denoise the grayscale image to obtain a denoised image, and calculating a segmentation threshold; The denoised image is segmented into two regions, a cell and a background, by binarization according to the segmentation threshold, and a hole filling algorithm and an erosion operation are used to process the cell region after the threshold segmentation to obtain a target region image; Calculate the center point position of the cell according to the target area image; Generate a Voronoi diagram using the original information of the center point position and the target corneal endothelial cell image, and save the Voronoi diagram coordinate information as a json file; The json file is exported to obtain the node coordinates of each cell.

3. The method for evaluating the function of a single corneal endothelial cell based on cell morphology according to claim 1, characterized in that: The calculation formula of the letter value is: Among them, XV is the signal value, S is the cell area, S th is the critical cell area, D max is the longest diagonal of the cell, D min The shortest diagonal of the cell.

4. The method for evaluating the function of a single corneal endothelial cell based on cell morphology according to claim 1, characterized in that: Visual display is performed according to the signal value and the node coordinates, including: Reading the node coordinates of the cells derived after identification, and establishing a three-dimensional cell model according to the node coordinates; the signal value of each cell in the three-dimensional cell model corresponds to a preset cell color scale; the value and color on the cell correspond to the color scale; The three-dimensional cell model is visualized.

Citation Information

Patent Citations

  • Process of statistic validation of corneal endothelium cells analysed samples

    CN101203185A

  • Devices and methods for assessing changes in corneal health

    CN102884551A

  • Cone-cell density calculation method based on image identification

    CN103745257A

  • Accurate quantification method and device for cell basic indexes in cornea confocal image

    CN113592783A

  • Analysis method of corneal endothelial cell active factors based on deep learning

    CN114693646A