Intelligent corneal topograph

Through the combination of the guidance system and the focus system, using deep learning algorithms and convolutional neural network models, the automatic alignment of the corneal topography instrument and high-precision image acquisition are realized, and the problems of operation difficulties and large errors in the existing technology are solved. Improved accuracy and consistency of corneal topographic maps.

CN115137292BActive Publication Date: 2024-11-08CHONGMINGNIAO (SUZHOU) INTELLIGENT TECH CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202210836671.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-15
Publication Date
2024-11-08
Estimated Expiration
2042-07-15

AI Technical Summary

Technical Problem

The existing corneal topographic instruments are difficult to operate, time consuming and poor centering accuracy, resulting in large errors in corneal topographic maps.

Method used

Adopting a guide system and a focus system, including a convolutional neural network eye tracking motion guidance algorithm based on deep learning and a Placido ring center tracking motion guidance algorithm, combined with a thick focus and a fine focus system, the automatic pairing of the main imaging optical components is achieved through the motion control system. It is accurate and strengthens the training database through the convolutional neural network model segmentation sampling algorithm to improve the segmentation accuracy of corneal images.

Benefits of technology

Real-time detection of eye pupil position and fast and accurate centering of corneal centers are achieved, which reduces the difficulty of operation, reduces the error of corneal topographic map, and obtains high-definition and high-quality corneal images, providing reliable for subsequent reconstruction. Data foundation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115137292B_ABST
    Figure CN115137292B_ABST
Patent Text Reader

Abstract

The present invention discloses an intelligent corneal topographer, which relates to the field of ophthalmic medical devices. Most of the existing corneal topographers use manual operation to center the optical center of the Placido disk with the corneal observation center, which is difficult and time-consuming to operate and has poor centering accuracy, resulting in large errors in the reconstructed corneal topography. The following scheme is proposed, including a guidance system and a focusing system, wherein the guidance system includes a guidance camera component, a main imaging optical component, a motion control system component, and an image guidance detection algorithm; the image guidance detection algorithm includes a first sub-algorithm and a second sub-algorithm, wherein the first sub-algorithm is an eye tracking motion guidance algorithm based on a deep learning convolutional neural network. The present invention proposes an intelligent corneal topographer, which can detect the position of the pupil in real time, improve the centering accuracy, reduce the difficulty of operation in clinical applications, and reduce the error of the corneal topography.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of ophthalmic medical instruments, and in particular to an intelligent corneal topograph. Background Art

[0002] Corneal topograph is a new device that presents the image of corneal surface curvature with the aid of computer. Computer-assisted corneal topography is characterized by its ability to accurately analyze the changes in the morphology and curvature of the entire corneal surface, making it possible to systematically, objectively and accurately analyze corneal properties. Corneal topograph consists of three parts: 1. Placido's disc projection system, 2. real-time image monitoring system, and 3. computer image processing system.

[0003] However, most of the existing corneal topographers use manual operation to align the optical center of the Placido disk with the corneal observation center, which is difficult and time-consuming to operate and has poor alignment accuracy, resulting in large errors in the reconstructed corneal topography. Summary of the invention

[0004] (I) Purpose of the invention

[0005] In order to solve the technical problems existing in the background technology, the present invention proposes an intelligent corneal topography device, which can detect the position of the pupil in real time, improve the centering accuracy, reduce the operating difficulty in clinical applications, and reduce the error of corneal topography.

[0006] (II) Technical solution

[0007] The present invention provides an intelligent corneal topographer, including a guidance system and a focusing system, wherein the guidance system includes a guidance camera component, a main imaging optical component, a motion control system component, and an image guidance detection algorithm;

[0008] The image-guided detection algorithm includes a first sub-algorithm and a second sub-algorithm, the first sub-algorithm is an eye-tracking motion guidance algorithm based on a deep learning convolutional neural network, and the second sub-algorithm is a Placido ring center tracking motion guidance algorithm based on a deep learning convolutional neural network;

[0009] The guiding camera component automatically determines the left and right eyeballs of the user through the eyeball tracking motion guiding algorithm, and feeds back the real-time position of the currently detected eyeball to the motion control system component, and the motion control system component controls the main imaging optical component to move toward the real-time position of the eyeball according to the current real-time position of the eyeball;

[0010] The main imaging optical component detects the center of the Placido ring in real time through the Placido ring center tracking motion guidance algorithm, and feeds back the real-time position of the current Placido ring center to the motion control system component, and the motion control system component controls the main imaging optical component to move toward the real-time position of the Placido ring center according to the real-time position of the current Placido ring center;

[0011] The focusing system includes a coarse focusing system, a fine focusing system and an image clarity evaluation algorithm. The image clarity evaluation algorithm is used to evaluate the image clarity value in real time and accumulate the fitted image clarity curve in real time. The coarse focusing system and the fine focusing system control the movement of the main imaging optical component through the image clarity evaluation algorithm.

[0012] As a further optimization of the present invention: the guidance system also includes a first strategy and a second strategy. The first strategy includes: when the main imaging optical component is about to reach the target position, instantly judging whether the main imaging optical component has reached it through an eye-tracking motion guidance algorithm; if the main imaging optical component has reached the target position, the eye-tracking motion guidance algorithm feeds back an arrival signal to the motion control system component to stop it from moving; if the main imaging optical component has not reached the target position, continuing to control the main imaging optical component to move toward the real-time position of the eyeball through the motion control system component.

[0013] As a further optimization of the present invention: the second strategy includes: when the main imaging optical component is about to reach the target position, instantly judging whether the main imaging optical component has reached it through the Placido ring center tracking motion guidance algorithm; if the main imaging optical component has reached the target position, the Placido ring center tracking motion guidance algorithm feeds back an arrival signal to the motion control system component to stop it from moving; if the main imaging optical component has not reached the target position, continuing to control the main imaging optical component to move toward the real-time position of the eyeball through the motion control system component.

[0014] As a further optimization of the present invention: the motion control system component includes a motion controller and an X-axis motion component, a Y-axis motion component and a Z-axis motion component electrically connected to the motion controller respectively, and the motion controller controls the movement of the guide camera component and the main imaging optical component through the X-axis motion component, the Y-axis motion component and the Z-axis motion component.

[0015] As a further optimization of the present invention: it also includes a forehead rest lifting mechanism, which is arranged on one side of the main imaging optical component, and a forehead rest is arranged on the forehead rest lifting mechanism, and the forehead rest lifting mechanism is used to control the lifting of the forehead rest.

[0016] As a further optimization of the present invention: it also includes a screen assembly, and the screen assembly is arranged on a side of the main imaging optical assembly away from the forehead support lifting mechanism.

[0017] As a further optimization of the present invention: it also includes a fixation lamp, which is arranged between the main imaging optical component and the forehead support lifting component. The fixation lamp includes a lamp panel, a white light lamp, an infrared lamp and a blue light lamp. The white light lamp, the blue light lamp and the infrared lamp are distributed in a ring along the midpoint of the lamp panel.

[0018] As a further optimization of the present invention: the coarse focusing system includes a third strategy, and the third strategy is that during the Z-axis motion component, when the image clarity evaluation algorithm determines that an obvious peak of the curve appears, the motion control system component controls the main imaging optical component to move back to the physical position where the clarity peak is located.

[0019] As a further optimization of the present invention: the fine focusing system includes a fourth strategy, and the fourth strategy is that after completing the coarse focusing, the motion control system component starts the voice coil motor, and drives the main imaging image sensor to move back and forth at the back focus position through the voice coil motor, and when it is determined by the image clarity evaluation algorithm that when an obvious peak of the curve appears, the motion control system component controls the main imaging optical component to move back to the physical position where the clarity peak is located.

[0020] As a further optimization of the present invention, it also includes a convolutional neural network model segmentation sampling algorithm, which uses a large amount of corneal image data with different clinical morphologies as training data, and strengthens the training database through an effective data enhancement algorithm. The convolutional neural network model segmentation sampling algorithm includes the following steps:

[0021] S1: Data annotation, wherein a large number of Placido ring corneal images of human eyes are collected for data annotation, and the segmentation and annotation of Placido black and white rings are performed by combining traditional segmentation algorithms with manual annotation;

[0022] S2: Data enhancement, which is to first perform geometric transformation enhancement on the basic data, i.e., flip, rotate, crop, deform, and scale; and then perform color transformation enhancement, i.e., noise, blur, color transformation, smearing, and filling;

[0023] S3: Model training, wherein the model training adopts a cross-validation training method, divides the data into multiple hold-out training batches with different test samples, trains a model respectively, and selects the model with the best training result;

[0024] S4: extracting the effective contour of the white ring, wherein the effective contour of the white ring is extracted by edge detection and geometric condition filtering to obtain the effective white ring contour;

[0025] S5: sorting the rings, wherein the sorting of the rings is used to detect that the image point where the fixation light is located in the image is the center point of the Placido ring, calculate each distance from the center point, and sort the ring contours in order from the smallest to the largest distance;

[0026] S6: Acquire sampling points. The sampling points are acquired with the center point of the Placido ring as the origin, and the polar equation is established with a certain step angle. The intersection points of all polar lines and contours in the circumferential direction are calculated, which are the sampling points of each contour and serve as the input for the subsequent reconstruction of corneal physical values.

[0027] Compared with the prior art, the above technical solution of the present invention has the following beneficial technical effects:

[0028] The present invention can detect the position of the pupil in real time and at high speed on an embedded hardware platform by setting a guidance system and a focusing system, and interact with the motion control system component for position information, so that the center of the Placido ring of the guidance mechanism is quickly and accurately aligned with the center of the cornea, and at the same time, a high-definition, high-quality corneal image is obtained by combining coarse focusing and fine focusing, providing a reliable data basis for the reconstruction of subsequent corneal graphics, and strengthening the training database through an effective data enhancement algorithm to train a segmentation model with high robustness, high adaptability and high accuracy, which can adapt to complex ambient light conditions and complex and diverse corneal morphologies, and has high segmentation accuracy and good consistency. The present invention proposes an intelligent corneal topograph, which can detect the position of the pupil in real time, improve the centering accuracy, reduce the difficulty of operation in clinical applications, and reduce the error of corneal topography. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the first three-dimensional structure of the present invention;

[0030] Figure 2 It is a schematic diagram of the second three-dimensional structure of the present invention;

[0031] Figure 3 This is a schematic diagram of the structure of the present invention from a first viewing angle;

[0032] Figure 4 for Figure 3 Schematic diagram of the cross-sectional structure of the CC;

[0033] Figure 5 It is a schematic diagram of the structure of the second viewing angle of the present invention;

[0034] Figure 6 This is a schematic diagram of the structure of the present invention from a third viewing angle;

[0035] Figure 7 It is a schematic diagram of the structure from the fourth viewing angle of the present invention.

[0036] Figure numerals: 1. Main imaging optical component; 2. Guide camera component; 3. Fixation lamp; 31. White light lamp; 32. Blue light lamp; 33. Infrared lamp; 4. Forehead rest lifting mechanism; 41. Forehead rest; 5. Motion control system component; 51. Motion controller; 52. X-axis motion component; 53. Y-axis motion component; 54. Z-axis motion component; 6. Screen assembly; 7. Input and output ports; 8. Network port; 9. Power input port. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0038] pass Figure 1-7 As shown, the intelligent corneal topographer proposed by the present invention includes a guidance system and a focusing system. The guidance system is divided into two stages, namely, long-stroke guidance and short-stroke guidance. The guidance system includes a guidance camera component 2, a main imaging optical component 1, a motion control system component 5, and an image guidance detection algorithm;

[0039] The image-guided detection algorithm includes a first sub-algorithm and a second sub-algorithm, the first sub-algorithm is an eye-tracking motion guidance algorithm based on a deep learning convolutional neural network, and the second sub-algorithm is a Placido ring center tracking motion guidance algorithm based on a deep learning convolutional neural network;

[0040] Long-stroke guidance: The guiding camera component 2 automatically determines the left and right eyeballs of the user through the eye-tracking motion guidance algorithm, and feeds back the real-time position of the currently detected eyeball to the motion control system component 5. The motion control system component 5 controls the main imaging optical component 1 to move toward the real-time position of the eyeball according to the current real-time position of the eyeball; the long-stroke guidance also includes a first strategy, which includes instantly judging whether the main imaging optical component 1 has reached the target position through the eye-tracking motion guidance algorithm when the main imaging optical component 1 is about to reach the target position. If the main imaging optical component 1 has reached the target position, the eye-tracking motion guidance algorithm feeds back the arrival signal to the motion control system component 5 to stop the movement; if the main imaging optical component 1 has not reached the target position, the motion control system component 5 continues to control the main imaging optical component 1 to move toward the real-time position of the eyeball. The guiding camera component 2 is composed of a wide-band miniature camera and a wide-angle lens;

[0041] Small-stroke guidance: the main imaging optical component 1 detects the Placido ring center in real time through the Placido ring center tracking motion guidance algorithm, and feeds back the current real-time position of the Placido ring center to the motion control system component 5. The motion control system component 5 controls the main imaging optical component 1 to move toward the real-time position of the Placido ring center according to the current real-time position of the Placido ring center; the small-stroke guidance also includes a second strategy, which includes when the main imaging optical component 1 is about to reach the target position, using the Placido ring center tracking motion guidance algorithm to instantly determine whether the main imaging optical component 1 has arrived. If the main imaging optical component 1 has reached the target position, the Placido ring center tracking motion guidance algorithm feeds back the arrival signal to the motion control system component 5 to stop it from moving; if the main imaging optical component 1 has not reached the target position, the motion control system component 5 continues to control the main imaging optical component 1 to move toward the real-time position of the eyeball. The main imaging optical component 1 is composed of a wide-band camera and a telecentric lens.

[0042] The focusing system includes a coarse focusing system, a fine focusing system and an image clarity evaluation algorithm. The image clarity evaluation algorithm is used to evaluate the image clarity value in real time and accumulate the fitted image clarity curve in real time. The coarse focusing system and the fine focusing system control the movement of the main imaging optical component 1 through the image clarity evaluation algorithm. The coarse focusing system is composed of a Z-axis motion component 54, a main imaging camera, and a telecentric lens, while the fine focusing system is composed of a voice coil motor, a main imaging image sensor, and a telecentric lens.

[0043] The coarse focusing system includes a third strategy, which is that during the Z-axis motion component 54, the image clarity evaluation algorithm determines that when an obvious peak of the curve appears, the motion control system component 5 controls the main imaging optical component 1 to move back to the physical position where the clarity peak is located; the fine focusing system includes a fourth strategy, which is that after completing the coarse focusing, the motion control system component 5 starts the voice coil motor, drives the main imaging image sensor to move back and forth at the back focus position through the voice coil motor, and determines that when an obvious peak of the curve appears through the image clarity evaluation algorithm, the motion control system component 5 controls the main imaging optical component 1 to move back to the physical position where the clarity peak is located. Since the voice coil motor has a high movement speed and high position accuracy, ultra-high-speed focusing can be achieved.

[0044] The motion control system component 5 includes a motion controller 51 and an X-axis motion component 52, a Y-axis motion component 53 and a Z-axis motion component 54 which are electrically connected to the motion controller 51 respectively. The motion controller 51 controls and guides the movement of the camera component 2 and the main imaging optical component 1 through the X-axis motion component 52, the Y-axis motion component 53 and the Z-axis motion component 54.

[0045] The corneal topograph also includes a frontal support 41 lifting mechanism 4, which is arranged on one side of the main imaging optical component 1, and a frontal support 41 is arranged on the frontal support 41 lifting mechanism 4, and the frontal support 41 lifting mechanism 4 is used to control the lifting and lowering of the frontal support 41; it also includes a screen assembly 6, and the screen assembly 6 is arranged on the side of the main imaging optical component 1 away from the frontal support 41 lifting mechanism 4; it also includes a fixation lamp 3, and the fixation lamp 3 is arranged between the main imaging optical component 1 and the frontal support 41 lifting assembly. The fixation lamp 3 includes a lamp panel, a white light lamp 31, an infrared lamp 33 and a blue light lamp 32, and the white light lamp 31, the blue light lamp 32 and the infrared lamp 33 are distributed in a ring along the midpoint of the lamp panel, and also includes a power input port 9 connected to the power supply, as well as an input / output port 7 and a network port 8, and the corneal topograph is connected to the matching equipment for use through the input / output port 7 and the network port 8.

[0046] The corneal topographer also includes a convolutional neural network model segmentation sampling algorithm. The convolutional neural network model segmentation sampling algorithm uses a large amount of corneal image data with different clinical morphologies as training data, and strengthens the training database through an effective data enhancement algorithm to train a highly robust, highly adaptable, and highly accurate segmentation model that can adapt to complex ambient light conditions and complex and diverse corneal morphologies, and has high segmentation accuracy and good consistency. The convolutional neural network model segmentation sampling algorithm includes the following steps:

[0047] S1: Data annotation: A large number of Placido ring corneal images of human eyes are collected for data annotation. The traditional segmentation algorithm is combined with manual annotation to perform segmentation and annotation of Placido black and white rings.

[0048] S2: Data enhancement: Data enhancement is performed by first performing geometric transformation enhancement on the basic data, i.e. flipping, rotating, cropping, deforming, and scaling; and then performing color transformation enhancement, i.e. noise, blurring, color transformation, smearing, and filling;

[0049] S3: Model training. Model training adopts cross-validation training method, which divides the data into multiple hold-out training batches with different test samples, trains a model for each batch, and selects the model with the best training result.

[0050] S4: Extraction of effective white ring contours. The effective white ring contour extraction uses edge detection and geometric condition filtering to obtain the effective white ring contour;

[0051] S5: sorting the rings. Sorting the rings is used to detect that the image point where the fixation lamp 3 is located in the image is the center point of the Placido ring, and the distance from each center point is calculated, and the ring contours are sorted in order from the smallest to the largest distance;

[0052] S6: Obtain sampling points. Take the center point of the Placido ring as the origin, establish the polar equation with a certain step angle, calculate the intersection points of all polar lines and contours in the circumferential direction, which are the sampling points of each contour and serve as the input for the subsequent reconstruction of corneal physical values.

[0053] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0054] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0055] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent corneal topographer, comprising a guidance system and a focusing system, characterized in that: The guidance system comprises a guidance camera component (2), a main imaging optical component (1), a motion control system component (5) and an image guidance detection algorithm; The image-guided detection algorithm includes a first sub-algorithm and a second sub-algorithm, the first sub-algorithm is an eye-tracking motion guidance algorithm based on a deep learning convolutional neural network, and the second sub-algorithm is a Placido ring center tracking motion guidance algorithm based on a deep learning convolutional neural network; The guiding camera component (2) automatically determines the left and right eyeballs of the user through the eyeball tracking motion guiding algorithm, and feeds back the real-time position of the currently detected eyeball to the motion control system component (5), and the motion control system component (5) controls the main imaging optical component (1) to move toward the real-time position of the eyeball according to the current real-time position of the eyeball; The main imaging optical component (1) detects the center of the Placido ring in real time through a Placido ring center tracking motion guidance algorithm, and feeds back the current real-time position of the Placido ring center to the motion control system component (5). The motion control system component (5) controls the main imaging optical component (1) to move toward the real-time position of the Placido ring center according to the current real-time position of the Placido ring center. The focusing system comprises a coarse focusing system, a fine focusing system and an image clarity evaluation algorithm, wherein the image clarity evaluation algorithm is used to evaluate the image clarity value in real time and accumulate a fitted image clarity curve in real time, and the coarse focusing system and the fine focusing system control the movement of the main imaging optical component (1) through the image clarity evaluation algorithm; The guidance system further comprises a first strategy and a second strategy. The first strategy comprises: when the main imaging optical component (1) is about to reach the target position, the eye tracking motion guidance algorithm is used to instantly determine whether the main imaging optical component (1) has reached the target position; if the main imaging optical component (1) has reached the target position, the eye tracking motion guidance algorithm feeds back an arrival signal to the motion control system component (5) to stop the movement; if the main imaging optical component (1) has not reached the target position, the motion control system component (5) continues to control the main imaging optical component (1) to move toward the real-time position of the eyeball; The second strategy includes: when the main imaging optical component (1) is about to reach the target position, the Placido ring center tracking motion guidance algorithm is used to instantly determine whether the main imaging optical component (1) has reached the target position; if the main imaging optical component (1) has reached the target position, the Placido ring center tracking motion guidance algorithm feeds back an arrival signal to the motion control system component (5) to stop the movement; if the main imaging optical component (1) has not reached the target position, the motion control system component (5) continues to control the main imaging optical component (1) to move toward the real-time position of the eyeball.

2. The intelligent corneal topographer according to claim 1, characterized in that: The motion control system component (5) comprises a motion controller (51) and an X-axis motion component (52), a Y-axis motion component (53) and a Z-axis motion component (54) which are electrically connected to the motion controller (51) respectively. The motion controller (51) controls the movement of the guide camera component (2) and the main imaging optical component (1) through the X-axis motion component (52), the Y-axis motion component (53) and the Z-axis motion component (54).

3. The intelligent corneal topographer according to claim 1, characterized in that: It also comprises a forehead rest (41) lifting mechanism (4), wherein the forehead rest (41) lifting mechanism (4) is arranged on one side of the main imaging optical assembly (1), a forehead rest (41) is arranged on the forehead rest (41) lifting mechanism (4), and the forehead rest (41) lifting mechanism (4) is used to control the lifting of the forehead rest (41).

4. The intelligent corneal topographer according to claim 3, characterized in that: It also comprises a screen assembly (6), wherein the screen assembly (6) is arranged on a side of the main imaging optical assembly (1) away from the forehead support (41) lifting mechanism (4).

5. The intelligent corneal topographer according to claim 3, characterized in that: The device also comprises a fixation lamp (3), the fixation lamp (3) being arranged between the main imaging optical component (1) and the forehead support (41) lifting component, the fixation lamp (3) comprising a lamp panel, a white light lamp (31), an infrared lamp (33) and a blue light lamp (32), the white light lamp (31), the blue light lamp (32) and the infrared lamp (33) being distributed in a ring shape along the midpoint of the lamp panel.

6. The intelligent corneal topographer according to claim 2, characterized in that: The coarse focusing system includes a third strategy, wherein during the Z-axis motion component (54), when the image clarity evaluation algorithm determines that an obvious peak of the curve appears, the motion control system component (5) controls the main imaging optical component (1) to move back to the physical position where the clarity peak is located.

7. The intelligent corneal topographer according to claim 2, characterized in that: The fine focusing system includes a fourth strategy, wherein after completing the coarse focusing, the motion control system component (5) starts the voice coil motor, and drives the main imaging image sensor to move forward and backward at the back focus position through the voice coil motor, and when it is determined by the image clarity evaluation algorithm that an obvious peak of the curve appears, the motion control system component (5) controls the main imaging optical component (1) to move back to the physical position where the clarity peak is located.

Citation Information

Patent Citations

  • Corneal topographic map discrimination method and system based on deep learning

    CN110517219A

  • Optical coherence tomography device and control method thereof

    CN112155512A

  • Calibration-free eye movement tracking system and application

    CN112732071A

  • Corneal topography reconstruction method and device

    CN113822928A

  • Cornea image generation method and device

    CN113827181A