Ophthalmologic robot optical coherence tomography eyeball fixation induction method and device

Abstract

The invention discloses an ophthalmology robot optical coherence tomography eyeball gaze induction method and device, and relates to the field of pupil tracking and eyeball gaze direction induction, and the method comprises the steps: calculating the current actual gaze angle of an eyeball based on a camera recognition result; calculating the average value of the actual fixation angles which are successfully calculated once or continuously multiple times; according to the average value of the successfully calculated actual fixation angles and the expected fixation angle, calculating the spatial position deviation of the eyeball actual fixation point and the expected fixation point on the objective lens plane; calculating the relative step length of movement of the fixation lamp according to the spatial position deviation; the fixation lamp moves from the current position according to the calculated relative step length, the actual fixation point of the eyeballs is induced to move towards the expected fixation point, and accurate induction of the fixation angle and the fixation position of the eyeballs is achieved. According to the method, accurate induction of the eyeball angle and the watching position in the OCT scanning process of free posture ophthalmology robot OCT imaging can be achieved, and the problems of blurred scanned images and the like caused by large-amplitude movement of eyeballs are solved.

Description

Technical Field

[0001] This application relates to the technical field of pupil tracking and eye induction. Specifically, it relates to an ophthalmic robot optical coherence tomography eye fixation induction method and device. Background Art

[0002] Ophthalmic OCT devices have been widely used in non-invasive ophthalmic imaging and play an important role in the diagnosis and treatment of ophthalmic diseases. The free-form ophthalmic robot OCT imaging solves the problem that traditional ophthalmic OCT devices rely heavily on patient cooperation and operator expertise, but it still cannot collect fundus information at specific angles of the human eye. At the same time, during the OCT scanning process after alignment, the eyeball moves significantly due to the lack of a fixation point, resulting in blurred scanned images. Summary of the Invention

[0003] To solve the deficiencies of the prior art, the present invention provides an ophthalmic robot optical coherence tomography eye fixation induction, and the solution is as follows:

[0004] The ophthalmic robot optical coherence tomography eye fixation induction includes:

[0005] S1: After the eye induction starts, turn on the fixation light, and calculate the current actual fixation angle of the eyeball based on the camera recognition result; during the above process, the calculation of the current actual fixation angle of the eyeball is carried out at any time during the eye induction process, and the calculation and change of the fixation light position are two independent threads;

[0006] S2: Take the current actual fixation angle of the eyeball that has been calculated successfully once or continuously for multiple times and obtain the average value of the actual fixation angles;

[0007] S3: If the deviation value between the actual fixation angle or the average value of the actual fixation angles and the expected fixation angle is greater than the deviation threshold, enter S4; otherwise, enter S2 for continuous induction; it should be noted that if only one actual fixation angle is taken in S2, then in S3, the deviation value is calculated between this actual fixation angle and the expected fixation angle;

[0008] S4: According to the average value of the actual fixation angles calculated successfully once or continuously for multiple times in S2 and the expected fixation angle, calculate the spatial position deviation between the actual fixation point and the expected fixation point of the eyeball on the objective lens plane;

[0009] S5: Calculate the relative step length that the fixation light needs to move according to the spatial position deviation;

[0010] S6: The fixation light moves from the current position according to the calculated relative step length, inducing the actual fixation point of the eyeball to move towards the expected fixation point, and realizing the precise induction of the eye fixation angle and fixation position.

[0011] Specifically, S1 is specifically as follows:

[0012] S101. After the robot completes the tracking and alignment of the eye pupil, utilize the reflection of the pupil center and the infrared LED array on the cornea to calculate the coordinates of the geometric center of the LED array reflection projected onto the imaging plane relative to the pupil center.

[0013] S102. According to the coordinates calculate the actual gaze angle (R x , R y ), and filter out the results of failed actual gaze angle calculations, where R x represents the rotation angle of the eye around the x-axis on the imaging plane, and R y represents the rotation angle of the eye around the y-axis on the imaging plane.

[0014] Specifically, S101 is specifically as follows: Illuminate the eye through the annular near-infrared LED array around the objective lens to form a circular reflection bright spot on the cornea. Segment the circular bright spot through the image threshold method, and perform ellipse fitting on the segmented bright spot. The obtained ellipse center (x g , y g ) is the geometric center of the LED array reflection; taking the pupil center (x p , y p ) as the coordinate origin, according to the resolution factor r and the corneal curvature radius R eye , convert the coordinates on the image to physical space coordinates, and calculate the coordinates of the geometric center of the LED array reflection projected onto the imaging plane relative to the pupil center The calculation method is as follows:

[0015]

[0016] If θ = 0, then:

[0017]

[0018] If θ ≠ 0, then:

[0019]

[0020] Specifically, in S102, utilize the coordinates of the geometric center of the LED array reflection projected onto the imaging plane relative to the pupil center calculated in S101 to calculate the actual gaze angle (R x , R y ), and the calculation formula is:

[0021]

[0022] The result of filtering out the failed angle calculation is: if the number of light spots required for ellipse fitting is insufficient due to the occlusion of eyelids and eyelashes, causing the angle calculation to fail, the angle that failed in this calculation is assigned as the expected gaze angle and filtered out, and is not used as a basis for changing the position of the fixation lamp to induce the eyeball.

[0023] Specifically, the S2 is:

[0024] Eye gaze angle R x The change of will cause the change of the y value of the gaze point on the objective lens plane. Similarly, the eye gaze angle R y The change of will cause the change of the x value of the fixation point on the objective lens plane. Considering the relationship between the reaction time of the human eye and the frequency of the fixation light change, the actual fixation angle (R x ,R y ), when the number of recorded data reaches n, calculate the mean of n data Determine the deviation from the expected gaze angle. If the deviation is no more than 1.5°, it is considered that the current gaze angle is close to the expected angle, the position of the fixation light remains unchanged, and the process returns to S1 to continue calculating and monitoring the change of the current gaze angle. If the deviation is greater than 1.5°, the process proceeds to S3 for subsequent steps.

[0025]

[0026] Specifically, the S4 includes:

[0027] S401. According to the actual gaze angle average value obtained in S2, the spatial position of the actual gaze point on the objective lens plane is calculated using the triangular relationship formed by the line between the pupil center and the objective lens center, the line between the pupil center and the actual gaze point, and the line between the actual gaze point and the objective lens center. The calculation method is:

[0028]

[0029] in, is the spatial position of the actual eye gaze point on the objective lens plane, Represent the actual gaze angle (R x ,R y )’s mean;

[0030] S402. According to the desired gaze angle, the spatial position of the desired gaze point on the objective lens plane is calculated using the triangular relationship formed by the line between the pupil center and the objective lens center, the line between the pupil center and the desired gaze point, and the line between the desired gaze point and the objective lens center:

[0031]

[0032] Among them, D is the working distance during OCT scanning after the robot pupil tracking and alignment are completed, is the expected fixation angle, (x exp , y exp ) is the spatial position of the expected fixation point on the objective lens plane;

[0033] S403. Calculate the spatial position deviation between the actual fixation point of the eyeball and the expected fixation point on the objective lens plane. The calculation method is as follows:

[0034]

[0035] Specifically, the calculation method of S4 is specifically as follows:

[0036]

[0037] Among them, ΔL x’ and ΔL y’ are the relative step lengths that the fixation lamp needs to move in the x-axis and y-axis directions respectively, D M is the diameter of the objective lens, x' and y' are the x-axis and y-axis on the induced screen coordinate system respectively, N x‘ and N y’ are the total number of pixels of the induced screen in the x' and y' directions respectively.

[0038] Specifically, in S5, the fixation lamp moves based on the relative step length calculated in S4 with the current position as the reference, inducing the actual fixation point of the eyeball to move towards the expected fixation point.

[0039]

[0040] Among them, and are the horizontal and vertical coordinates of the current coordinate position of the fixation lamp respectively, L x‘ and L y’ are the horizontal and vertical coordinates of the coordinate position of the fixation lamp after movement.

[0041] Preferably, in S2, the current actual fixation angles of the eyeball that are successfully calculated continuously are taken, and the number of multiple values in the obtained average value of the actual fixation angles is 15.

[0042] Preferably, in S3, the deviation threshold is 1.5°.

[0043] The present invention calculates the step size that the fixation lamp needs to move according to the deviation between the actual fixation point spatial position and the desired fixation point spatial position, and adjusts the position of the fixation lamp in real time, realizing the precise induction of the eye angle and fixation position during the scanning process of the free-body ophthalmic robot OCT imaging. By effectively controlling the eye movement, problems such as blurred scanning images caused by large eye movements are overcome, thereby improving the clarity and accuracy of the scanning images. In the specific implementation of the present invention, the above technical solution was tested on the human eye.

[0044] The present invention is not limited to a specific angle, can achieve eye induction at any angle, and can modify the desired induction angle at any time according to requirements.

[0045] Meanwhile, the present invention discloses a real-time eye fixation induction device for an ophthalmic robot optical coherence tomography imaging. The device is used to induce the eye fixation direction in real time during the optical coherence tomography imaging. The induction method adopts the above-mentioned real-time eye fixation direction induction method for an ophthalmic robot optical coherence tomography imaging by obtaining and calculating the eye fixation angle in real time.

[0046] The device is used to induce the eye fixation direction in real time during the ophthalmic optical coherence tomography imaging. The specific induction method adopts the above-mentioned eye fixation induction method for an ophthalmic robot optical coherence tomography imaging. The device includes: an infrared LED array, a side camera, an objective lens, a short-wave pass dichroic mirror, a long-wave pass dichroic mirror, a central camera, an induction screen, and an OCT scanning arm optical path. The objective lens, the short-wave pass dichroic mirror, the long-wave pass dichroic mirror, the central camera, and the induction screen are arranged in sequence. The infrared LED array is located around the objective lens. There are 2 side cameras, which are respectively arranged on both sides of the objective lens. The OCT scanning arm optical path is arranged below the short-wave pass dichroic mirror and includes a galvanometer and an optical fiber collimator arranged in sequence.

[0047] Furthermore, the device further includes an intermediate optical path correction module one, an intermediate optical path correction module two, and an OCT optical path correction module. The intermediate optical path correction module one is located between the short-wave pass dichroic mirror and the long-wave pass dichroic mirror. The intermediate optical path correction module two is located between the long-wave pass dichroic mirror and the central camera. The OCT optical path correction module is located between the short-wave pass dichroic mirror and the galvanometer.

[0048] The beneficial effects of the present invention are: The real-time eye induction method and device for an ophthalmic robot optical coherence tomography imaging proposed by the present invention can achieve the precise induction of the eye angle and fixation position during the scanning process of the free-body ophthalmic robot OCT imaging, and overcome problems such as blurred scanning images caused by large eye movements, thereby improving the clarity and accuracy of the scanning images. Description of the Drawings

[0049] To more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required for the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation of the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0050] Figure 1 Is the main flowchart of the present invention;

[0051] Figure 2 Is the structural schematic diagram of the free-form ophthalmic robot OCT based on the present invention;

[0052] Figure 3 Is Figure 2 The optical path diagram corresponding to the robot of;

[0053] Figure 4 Is the schematic diagram of the angle calculation principle of the present invention;

[0054] Figure 5 Is one of the schematic diagrams of the spatial position calculation of the present invention, showing the fixation point O of the eye gaze direction on the objective plane;

[0055] Figure 6 Is one of the schematic diagrams of the spatial position calculation of the present invention, showing the fixation point O on the y-direction projection of the objective plane y ;

[0056] Figure 7 Is one of the schematic diagrams of the spatial position calculation of the present invention, showing the fixation point O on the x-direction projection of the objective plane x ;

[0057] Figure 8 Is the expected fixation angle during the eye gaze induction process of the present invention When the corresponding angle Rlx of the fixation light display position and the average value of the actual fixation angle Curve comparison diagram;

[0058] Figure 9 Is the expected fixation angle during the eye gaze induction process of the present invention When the corresponding angle Rly of the fixation light display position and the average value of the actual fixation angle Curve comparison diagram;

[0059] In the figure, the marks are: 1 - infrared LED array, 2 - side camera, 3 - objective lens, 4 - short-wave dichroic mirror, 5 - intermediate optical path correction module 1, 6 - long-wave dichroic mirror, 7 - intermediate optical path correction module 2, 8 - central camera, 9 - induction screen, 10 - OCT optical path correction module, 11 - galvanometer, 12 - fiber collimator, 13 - eyeball. Detailed implementation mode

[0060] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0061] As Figure 1 shown, an ophthalmic robot optical coherence tomography real-time eye fixation induction method, which is based on a free-body ophthalmic robot OCT imaging vision-assisted pupil tracking device Figure 2 shows a schematic structural diagram of the robot of the present invention. The structure mainly consists of a mechanical support part, three near-infrared cameras and an OCT scanning arm optical path. Specifically, it includes an infrared LED array 1, a side camera 2, an objective lens 3, a short-wave pass dichroic mirror 4, an intermediate optical path correction module 5, a long-wave pass dichroic mirror 6, an intermediate optical path correction module 7, a central camera 8, an induction screen 9, an OCT optical path correction module 10, a galvanometer 11 and an optical fiber collimator 12; the infrared LED array 1 is located around the objective lens 3, the side cameras 2 are symmetrically arranged on both sides of the objective lens 3, and the OCT scanning arm optical path includes an OCT optical path correction module 10, a galvanometer 11 and an optical fiber collimator 12; the induction optical path for realizing the eye induction mentioned in this patent includes an objective lens 3, a short-wave pass dichroic mirror 4, a long-wave pass dichroic mirror 6, and an induction screen 9. The induction screen 9 is located below the long-wave pass dichroic mirror 6, and the optical fiber collimator 12 and the galvanometer 11 are sequentially located below the short-wave pass dichroic mirror 4. The device is also provided with a support mechanism, which includes a horizontal part and a vertical part. The vertical part is used to support the optical fiber collimator 12, the galvanometer 11 and the OCT optical path correction module 10, and the horizontal part is used to support the side camera 2, the objective lens 3, the short-wave pass dichroic mirror 4, the intermediate optical path correction module 5, the long-wave pass dichroic mirror 6, the intermediate optical path correction module 7 and the central camera 8. It should be understood that the support mechanism is used to support other components. The specific connection method between the support mechanism and other components can be conventionally selected by those skilled in the art, such as a fixed connection method or a detachable connection method. How the specific support mechanism supports is prior art rather than the improvement point and protection point of the present invention. Figure 3 shows the optical path diagram of the present invention. The fixation light is the yellow cross cursor displayed on the induction screen, and the position movement is the movement of the fixation light, while the position of the induction screen remains unchanged. After the pupil tracking alignment is completed, the fixation light on the induction screen 9 and the infrared LED array 1, the long-wave pass dichroic mirror 6, and the objective lens 3 jointly complete the induction of the eye fixation angle and fixation direction. The specific principle is prior art and will not be described and explained in detail here.

[0062] After the pupil tracking and alignment are completed, the eye gaze angle is obtained based on the camera recognition result. The position coordinates of the actual fixation point on the incident plane are calculated through the angle, and compared with the position coordinates at the incident plane corresponding to the desired gaze angle. The relative step length that the fixation lamp needs to move is calculated through the deviation of the two fixation points in the spatial position. The fixation lamp moves based on the current position, inducing the actual fixation point of the eye to move towards the desired fixation point, so as to achieve precise induction of the eye gaze angle and position during the OCT scanning process.

[0063] Based on the above principle, the method of this specific implementation includes the following steps:

[0064] S1: After the eye induction starts, turn on the fixation lamp, and calculate the current actual gaze angle of the eye based on the camera recognition result; specifically, S1 is:

[0065] S101. After the robot completes the tracking and alignment of the eye pupil, use the reflection of the pupil center and the infrared LED array on the cornea to calculate the coordinates of the geometric center of the LED array reflection relative to the projection of the pupil center on the imaging plane; illuminate the eye through the annular near-infrared LED array around the objective lens, form a circular reflection bright spot on the cornea, segment the circular bright spot through the image threshold method, and perform elliptical fitting on the segmented bright spot. The obtained ellipse center (x g , y g ) is the geometric center of the LED array reflection; taking the pupil center (x p , y p ) as the coordinate origin, according to the resolution factor r and the corneal curvature radius R eye , convert the coordinates on the image to the physical space coordinates, and calculate the coordinates of the geometric center of the LED array reflection relative to the projection of the pupil center on the imaging plane The calculation method is as follows:

[0066]

[0067] If θ = 0, then:

[0068]

[0069] If θ ≠ 0, then:

[0070]

[0071] S102. According to the coordinates of the geometric center of the LED array reflection relative to the projection of the pupil center on the imaging plane, calculate the actual gaze angle (R x , R y ) of the eye at this time, where R x represents the rotation angle of the eye with the x-axis on the imaging plane as the axis, and R yRepresents the rotation angle of the eyeball with the y-axis on the imaging plane as the axis, and the calculation method is as follows:

[0072]

[0073] The result of filtering the failed calculation of the actual fixation angle is: If the number of light spots required for ellipse fitting is insufficient due to the occlusion of the eyelids and eyelashes, resulting in a failed angle calculation, then assign the failed calculated angle to the expected fixation angle and filter it out, and do not use it as the basis for inducing the eyeball by changing the position of the fixation light.

[0074] S2: Take the current actual fixation angle of the eyeball that has been calculated successfully once or continuously for multiple times and obtain the average value of the actual fixation angle; specifically, in this specific implementation, take the average value of the actual fixation angles that have been calculated successfully 15 times continuously, and judge whether the deviation from the expected fixation angle is greater than 1.5°.

[0075] The eyeball fixation angle R x The change of will cause a change in the y value of the fixation point on the objective plane. Similarly, the eyeball fixation angle R y The change of will cause a change in the x value of the fixation point on the objective plane. Considering the relationship between the human eye reaction time and the change frequency of the fixation light, after a period of time since the last change of the fixation light, start to record the actual fixation angles (R x , R y ) that have been calculated successfully. When the number of recorded data reaches n, calculate the average value of the n data Judge its deviation from the expected fixation angle. If the deviation value is not greater than 1.5°, it is considered that the current fixation angle is close to the expected angle, the position of the fixation light remains unchanged, and return to S1 to continue calculating and monitoring the change of the current fixation angle; if the deviation is greater than 1.5°, enter S3 for subsequent steps;

[0076]

[0077] In this specific implementation, the deviation threshold between the average value of the actual fixation angle and the expected fixation angle is selected as 1.5°. Those skilled in the art can select other thresholds in specific implementations, such as 0.5°, 0.8°, 1.0°, 1.6°, 1.7°, 1.8°, etc. The smaller the deviation threshold, the more accurate it is. Subsequently, due to the improvement of the device accuracy, a smaller deviation threshold can be selected. What the present invention protects is the idea of comparing the deviation value between the average value of the actual fixation angle and the expected fixation angle with the deviation threshold. Selecting other deviation thresholds is within the protection scope of the present invention.

[0078] S3: According to if the deviation value between the average value of the actual fixation angle and the expected fixation angle is greater than the deviation threshold, enter S4; otherwise, enter S2 for continuous induction;

[0079] S4: Calculate the spatial position deviation between the actual fixation point of the eyeball and the expected fixation point on the objective lens plane based on the mean value of the actual fixation angles calculated successfully for multiple consecutive times in S2 and the expected fixation angle; the S4 includes:

[0080] S401. Based on the mean value of the actual fixation angles obtained in S2 Calculate the spatial position of the actual fixation point on the objective lens plane by using the triangular relationship formed by the connection line between the pupil center and the objective lens center, the connection line between the pupil center and the actual fixation point, and the connection line between the actual fixation point and the objective lens center. The calculation method is as follows:

[0081]

[0082] S402. Based on the expected fixation angle Calculate the spatial position of the expected fixation point on the objective lens plane by using the triangular relationship formed by the connection line between the pupil center and the objective lens center, the connection line between the pupil center and the expected fixation point, and the connection line between the expected fixation point and the objective lens center:

[0083]

[0084] where D is the working distance during OCT scanning after the robot pupil tracking and alignment is completed, is the spatial position of the actual fixation point of the eyeball on the objective lens plane, (x exp , y exp ) is the spatial position of the expected fixation point on the objective lens plane;

[0085] S403. Calculate the deviation between the spatial positions of the actual fixation point and the expected fixation point of the eyeball. The calculation method is as follows:

[0086]

[0087] S5: Calculate the relative step length that the fixation light needs to move according to the spatial position deviation; the specific calculation method is:

[0088]

[0089] where ΔL x’ and ΔL y’ are the relative step lengths that the fixation light needs to move in the x-axis and y-axis directions respectively, D M is the diameter of the objective lens, x' and y' are the x-axis and y-axis on the induction screen coordinate system, N x‘ and N y’ are the total number of pixels of the induction screen in the x' and y' directions respectively.

[0090] S6: The fixation light moves from the current position according to the calculated relative step length, inducing the actual fixation point of the eyeball to move towards the desired fixation point, thereby achieving precise induction of the eyeball fixation angle and fixation position; the fixation light moves based on the current position according to the relative step length to be moved calculated in S4, inducing the actual fixation point of the eyeball to move towards the desired fixation point.

[0091]

[0092] Among them, and are the horizontal and vertical coordinates of the current coordinate position of the fixation light, and L x‘ and L y’ are the horizontal and vertical coordinates of the coordinate position of the fixation light after movement.

[0093] During the above process, the calculation of the current actual fixation angle of the eyeball is carried out at any time during the eyeball induction process, and the calculation and change of the fixation light position are two independent threads. After the induction starts, the induction needs to continue. In S3 above, if it returns to S2, it means that the current induction accuracy has reached the standard of precise induction, and it can continue to perform S2 for continuous induction. Once it is detected that the deviation value between the mean value of the actual fixation angle and the desired fixation angle is greater than the deviation threshold, it indicates that the fixation position of the human eye has changed, and the induction accuracy is insufficient, and subsequent steps need to be carried out, that is, it will enter S4 for the next step to achieve precise induction. Those skilled in the art should know that the end of the induction process requires manual intervention.

[0094] The present invention obtains the eyeball fixation angle based on the camera recognition result, calculates the position coordinates of the actual fixation point on the incident plane through the angle, compares it with the position coordinates at the incident plane corresponding to the desired fixation angle, calculates the relative step length that the fixation light needs to move through the deviation of the two fixation points in the spatial position, and the fixation light moves based on the current position, inducing the actual fixation point of the eyeball to move towards the desired fixation point, making the actual fixation point approach the desired fixation point, achieving precise induction of the eyeball angle and fixation position during the scanning process of the free-form ophthalmic robot OCT. By effectively controlling the movement of the eyeball, problems such as blurred scanning images caused by large-scale movement of the eyeball are overcome, thereby improving the clarity and accuracy of the scanning images. In the specific implementation of the present invention, the above technical solution was tested on the human eye.

[0095] In this specific implementation, after the pupil tracking and alignment of the used free-form ophthalmic robot OCT are completed, the working distance D from the objective lens to the human eye is 50 mm, the diameter D M of the objective lens is 53.3 mm, the annular near-infrared LED array used around the objective lens contains a total of 10 LED lamp beads, the size of the induction screen is 480×480 (pix 2 )), and the size of the rotating cross pattern used as the fixation light on the induction screen is approximately 5×5 (pix2 ) Calculate the fixation light moving step and induce the eyeball using the above parameters.

[0096] Figure 3 It shows the schematic diagram of the angle calculation principle of the present invention and describes how to calculate the fixation angle using the pupil center and corneal reflection (PCCR). Figure 4 (a) represents the pupil-eye image of the detection frame, Figure 4 (b) represents the capture of the pupil center in (a), Figure 4 (c) represents the geometric center of the reflection capture of the annular LED array in (a), Figure 4 (d) represents the fixation monitoring of (a), which is the result obtained from (b) and (c), Figure 4 (e) represents the fixation monitoring when the directions are inconsistent, Figure 4 (f) represents the fixation monitoring when the directions are aligned. The annular near-infrared LED array around the objective lens illuminates the eye, generating a bright spot reflection on the cornea. By segmenting each reflection spot and performing elliptical fitting, the geometric center G of the LED reflection can be calculated. Denote the projections of the geometric center G of the LED reflection and the pupil center P on the imaging plane as G' and P'. According to the working distance and the corneal curvature radius, reconstruct the visual axis of the eye as the line defined by G and P. When the directions are aligned, that is, when the fixation angle is 0, the two projection points G' and P' coincide. Otherwise, the line connecting the two points indicates the fixation direction, and the fixation angle R can be calculated based on the connection line. x and R y .

[0097] Figures 5 - 7 It shows the schematic diagram of the calculation of the spatial position of the fixation point of the present invention; Figure 5 It shows the fixation point O of the eyeball fixation direction on the objective lens plane; Figure 6 It shows the fixation point O on the projection in the y direction of the objective lens plane y ; Figure 7 It shows the fixation point O on the projection in the x direction of the imaging plane x .

[0098] Figure 8 and Figure 9 Taking the desired fixation angle as an example, it shows the curve comparison diagram of the actual average fixation angle and the corresponding angles Rlx and Rly of the fixation light position during the eyeball induction process of the present invention. From the curve comparison diagram, it can be observed that regardless of the value of the desired fixation angle, the actual average fixation angle always shows the same change trend as the corresponding angles Rlx and Rly of the fixation light position with a certain time delay and tends to the desired fixation angle after oscillating for a certain time. From the start to the end of the induction process, the average value of the actual fixation angle of the eyeball Basically conforms to the expected viewing angle.

[0099] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. An ophthalmic robot optical coherence tomography eye gaze induction method, characterized in that: include: S1: After the eye induction begins, turn on the fixation light and calculate the actual eye gaze angle based on the camera recognition results; S2: Take the current actual gaze angle of the eyeball that is successfully calculated once or multiple times and obtain the average of the actual gaze angles; S3: If the deviation between the actual gaze angle or the mean of the actual gaze angle and the expected gaze angle is greater than the deviation threshold, enter S4; otherwise, enter S2 for continuous induction; S4: Calculate the spatial position deviation between the actual gaze point of the eyeball and the expected gaze point on the objective lens plane according to the average of the actual gaze angles calculated in S2 or the multiple consecutive angle calculations that succeeded, and the expected gaze angle; S5: Calculate the relative step length of the fixation light according to the spatial position deviation; S6: The fixation light moves from the current position according to the calculated relative step length, inducing the actual eye gaze point to move toward the desired gaze point, thereby achieving accurate induction of the eye gaze angle and gaze position.

2. The method for inducing eye gaze using optical coherence tomography of an ophthalmic robot according to claim 1, characterized in that: The S1 is specifically: S101. After the tracking and alignment of the pupil of the eye is completed, the coordinates of the geometric center of the LED array reflection relative to the pupil center projected on the imaging plane are calculated using the pupil center and the reflection of the infrared LED array on the cornea. S102. According to coordinates Calculate the actual gaze angle of the eyeball at this time (R x ,R y ), and filter out the failure results of the actual gaze angle calculation, where R x It represents the rotation angle of the eyeball with the x-axis on the imaging plane as the axis, R y It represents the rotation angle of the eyeball around the y-axis on the imaging plane.

3. The ophthalmic robot optical coherence tomography eye gaze according to claim 2 The induction method is characterized in that The S101 is specifically as follows: illuminating the eye through the annular near-infrared LED array around the objective lens to form a circular reflective bright spot on the cornea, segmenting the circular bright spot through the image threshold method, and performing ellipse fitting on the segmented bright spot to obtain the ellipse center (x g ,y g ) is the geometric center of the LED array reflection; the pupil center (x p ,y p ) as the coordinate origin, according to the resolution factor r and the corneal curvature radius R eye , convert the coordinates on the image to physical space coordinates, and calculate the coordinates of the geometric center of the LED array reflection relative to the pupil center projected on the imaging plane The calculation is as follows: If θ=0, then: If θ≠0, then:

4. The ophthalmic robot optical coherence tomography eye gaze induction method according to claim 3, characterized in that: In S102, the coordinates of the geometric center of the LED array reflection calculated in S101 projected on the imaging plane relative to the pupil center are used. Calculate the actual gaze angle (R x ,R y ), the calculation formula is:

5. The ophthalmic robot optical coherence tomography eye gaze induction method according to claim 1, characterized in that: The S4 includes: S401. According to the actual gaze angle average value obtained in S2, the spatial position of the actual gaze point on the objective lens plane is calculated using the triangular relationship formed by the line between the pupil center and the objective lens center, the line between the pupil center and the actual gaze point, and the line between the actual gaze point and the objective lens center. The calculation method is: in, is the spatial position of the actual eye gaze point on the objective lens plane, Represent the actual gaze angle (R x ,R y )’s mean; S402. According to the desired gaze angle, the spatial position of the desired gaze point on the objective lens plane is calculated using the triangular relationship formed by the line between the pupil center and the objective lens center, the line between the pupil center and the desired gaze point, and the line between the desired gaze point and the objective lens center: Where D is the working distance when the robot performs OCT scanning after pupil tracking and alignment is completed. is the expected gaze angle, (x exp ,y exp ) is the spatial position of the desired fixation point on the objective lens plane; S403. Calculate the spatial position deviation between the actual eye gaze point and the expected eye gaze point on the objective lens plane, and the calculation method is:

6. The ophthalmic robot optical coherence tomography eye gaze according to claim 5 The induction method is characterized in that The calculation method of S5 is specifically as follows: Where, ΔL x’ and ΔL y’ are the relative steps of the fixation light in the x-axis and y-axis directions, D M is the diameter of the objective lens, x' and y' are the x-axis and y-axis of the induced screen coordinate system, N x‘ and N y’ are the total number of pixels of the induction screen in the x' and y' directions respectively.

7. The ophthalmic robot optical coherence tomography eye gaze induction method according to claim 6, characterized in that: In S6, the fixation lamp moves according to the relative step length calculated in S5, taking the current position as a reference, to induce the actual fixation point of the eyeball to move toward the desired fixation point. in, and are the horizontal and vertical coordinates of the current fixation light coordinate position, L x‘ and L y’ are the horizontal and vertical coordinates of the coordinate position of the fixation light after movement.

8. The method for inducing eye gaze using optical coherence tomography of an ophthalmic robot according to claim 1, characterized in that: In S3, the deviation threshold is 1.5°.

9. An ophthalmic robot optical coherence tomography eye gaze induction device, characterized in that: The device is used to induce the eye gaze direction in real time during ophthalmic optical coherence tomography, and the specific induction method adopts the ophthalmic robot optical coherence tomography eye gaze induction method as described in any one of claims 1 to 8; The device includes: an infrared LED array 1, a side camera 2, an objective lens 3, a short-wave pass dichroic mirror 4, a long-wave pass dichroic mirror 6, a center camera 8, an induction screen 9 and an optical path of an OCT scanning arm; The objective lens 3, the short-wave pass dichroic mirror 4, the long-wave pass dichroic mirror 6, the center camera 8, and the induction screen 9 are arranged in sequence; The infrared LED array 1 is located around the objective lens 3; there are two side cameras 2, which are respectively arranged on both sides of the objective lens 3; The optical path of the OCT scanning arm is arranged below the short-wave pass dichroic mirror 4 , and includes a galvanometer mirror 11 and a fiber collimator 12 which are arranged in sequence.

10. The ophthalmic robot optical coherence tomography eye gaze induction device according to claim 9, characterized in that: The device also includes an intermediate optical path correction module 15, an intermediate optical path correction module 27 and an OCT optical path correction module 10. The intermediate optical path correction module 15 is located between the short-wave pass dichroic mirror 4 and the long-wave pass dichroic mirror 6, the intermediate optical path correction module 27 is located between the long-wave pass dichroic mirror 6 and the central camera 8, and the OCT optical path correction module 10 is located between the short-wave pass dichroic mirror 4 and the galvanometer 11.

Images