A retinal imaging method

Abstract

The present invention provides a retinal imaging method, which first adjusts the pupil position to obtain a better imaging position, thereby improving the imaging quality; firstly, the position of the pupil center is obtained, and the pupil center is adjusted to be initially positioned at a preset target pupil center, and then the position of the pupil center is further adjusted according to the dot matrix distribution of the wavefront image, so that the pupil center is secondarily positioned at the preset target pupil center, and the two methods are used to perform positioning in sequence, so that the result of pupil positioning is more accurate, thereby achieving a better retinal imaging effect.

Description

Technical Field

[0001] The present invention relates to the technical field of retinal imaging, and more particularly to a retinal imaging method. Background Art

[0002] The retina is an important part of the human eye. Currently, more than one billion people in the world suffer from retinal diseases. In order to more effectively treat retinal diseases, the optimization of treatment devices is inevitable. In other words, high-resolution retinal imaging devices are of great significance for the diagnosis and efficacy evaluation of retinal diseases.

[0003] AO-OCT (Adaptive Optics-Optical Coherence Tomography) technology is applied to ophthalmic imaging, that is, the application of cutting-edge adaptive optics technology to combine low-coherence interferometry and confocal scanning microscopy with modern computer image processing technology to display retinal imaging at the cellular level.

[0004] The system can eliminate human eye aberrations and obtain high-resolution images of the retina that are close to the theoretical limit. It can identify early lesions by extracting and comparing image features of retinal cell morphology and density distribution. The system is marker-free and does not require the injection of specific contrast agents. During the observation and detection process, it does not cause any harm to the human body.

[0005] However, the high complexity of the current AO-OCT system makes the operation method difficult, resulting in poor image quality and low acquisition efficiency. In addition, it is difficult to accurately align the position of the center of the human pupil with the optical axis of the imaging device during imaging, and position deviation will affect the imaging effect. Summary of the invention

[0006] In view of this, in order to solve the above problems, the present invention provides a retinal imaging method, and the technical solution is as follows:

[0007] A retinal imaging method, first adjusting the pupil position, includes:

[0008] S1, obtaining the position of the pupil center, and adjusting the pupil center to be initially positioned at a preset target pupil center;

[0009] S2, acquiring a wavefront image, wherein the wavefront image is a dot matrix image formed by incident light rays reflected from the retina and converged in a wavefront detector;

[0010] S3, based on the dot matrix distribution in the wavefront image, further adjusting the position of the pupil center so that the pupil center is repositioned at a preset target pupil center.

[0011] Preferably, in the above retinal imaging method, step S1 comprises:

[0012] At least two light detectors are used to collect images of the eye area, and the collection angles of each light detector are different;

[0013] Analyzing and processing the eye region image to identify the pupil center;

[0014] Comparing the offset between the pupil center and a preset target pupil center;

[0015] Based on the offset, the pupil center is adjusted to move toward a preset target pupil center.

[0016] Preferably, in the above-mentioned retinal imaging method, the specific method of step S3 is: when the dot matrix distribution does not meet the preset conditions, adjust the position of the pupil center, and then repeat steps S2 and S3.

[0017] Preferably, in the above-mentioned retinal imaging method, the preset conditions are: the degree of similarity with the area or contour of the target pupil dot matrix distribution; when adjusting the position of the pupil center, the adjustment direction of the pupil position is determined according to the difference between the dot matrix distribution of the target pupil and the real-time dot matrix distribution.

[0018] Preferably, in the above retinal imaging method, diopter compensation is automatically performed after step S3, the method being:

[0019] Selecting at least two diopters, compensating the optical path, and obtaining a corresponding wavefront image;

[0020] Evaluate the quality of each wavefront image;

[0021] A wavefront image whose quality meets the preset requirements is selected, and the optical path is automatically compensated with its corresponding diopter.

[0022] Preferably, in the above retinal imaging method, the automatic refractive compensation includes automatic spherical lens and / or astigmatism compensation, and the automatic spherical lens or astigmatism compensation is specifically:

[0023] Select the initial compensation value and step size;

[0024] Adding and subtracting the step length from the initial compensation value to obtain a first compensation value and a second compensation value respectively;

[0025] Compensating the optical path with the first compensation value, the initial compensation value and the second compensation value respectively, and obtaining a corresponding wavefront image;

[0026] Evaluate the quality of each wavefront image;

[0027] Select the wavefront image with the best quality, use its corresponding compensation value as the new initial compensation value, redefine the step size, and return to the step of adding and subtracting the step size from the initial compensation value to obtain the first compensation value and the second compensation value respectively, until the step size or the wavefront image quality meets the preset requirements.

[0028] Preferably, in the above retinal imaging method, the automatic refractive compensation includes automatic spherical lens and / or astigmatism compensation, and the automatic spherical lens or astigmatism compensation is specifically:

[0029] Selecting multiple sample compensation values ​​at fixed intervals within the compensation value domain;

[0030] Compensating the optical path with the compensation value of each sample respectively, and obtaining the corresponding wavefront image;

[0031] Evaluate the quality of each wavefront image;

[0032] The wavefront image with the best quality is selected, and its corresponding compensation value plus and minus the fixed interval is used as the new compensation value range endpoint, and the interval is re-determined, and the step of selecting multiple sample compensation values ​​at fixed intervals within the compensation value range is returned to execute until the interval or wavefront image quality meets the preset requirements.

[0033] Preferably, in the above retinal imaging method, after the automatic diopter compensation, adaptive optical aberration compensation is performed, including:

[0034] Acquire a wavefront image;

[0035] Calculating the offset of the dot matrix in the wavefront image relative to the dot matrix of the reference plane wave, and determining the aberration compensation feedback value according to the offset of the dot matrix;

[0036] The optical path is compensated with the aberration compensation feedback value to realize adaptive optical aberration compensation.

[0037] Preferably, in the above-mentioned retinal imaging method, the retinal imaging is retinal multimodal imaging, including synchronous imaging of a first modality image and a second modality image of the retina, and the method is:

[0038] When performing eye movement tracking, a first modality sub-image is obtained, and it is registered with a preset first modality reference image to obtain an offset of the eyeball; the first modality sub-image is a segment of a frame of the first modality image;

[0039] A first control signal is generated based on the offset, and the optical path state is adjusted according to the first control signal to perform eye movement tracking; and a position of the second modality sub-image in the second modality image is determined based on the offset; the second modality sub-image is a portion of the second modality image synchronously acquired when acquiring the first modality sub-image;

[0040] The second modality sub-images are accumulated according to the positions and combined into a second modality image for output.

[0041] Preferably, in the above-mentioned retinal imaging method, the retinal imaging method further comprises:

[0042] After the diopter compensation is completed, the process returns to step S1 and repeats pupil positioning and diopter compensation.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] A retinal imaging method provided by the present invention first adjusts the pupil position to obtain a better imaging position, thereby improving the quality of imaging. Specifically, the position of the pupil center is first obtained, and the pupil center is adjusted to be initially positioned at a preset target pupil center, and then the position of the pupil center is further adjusted according to the dot matrix distribution of the wavefront image, so that the pupil center is secondarily positioned at the preset target pupil center. The two methods are used to perform positioning successively, so that the result of pupil positioning is more accurate, thereby achieving a better retinal imaging effect.

[0045] In addition, the present invention also performs aberration compensation after the pupil center is positioned. After the aberration compensation, the influence of human eye aberration can be eliminated during retinal imaging, thereby further improving the imaging quality. When performing aberration compensation, the diopter compensation is automatically performed first, and then the adaptive optical aberration compensation is performed, which simplifies the operation process, reduces the operation difficulty, avoids the tedious process of manually inputting the diopter parameters and the problem of inaccurate diopter data obtained externally, and is more suitable for large-scale promotion and application. Moreover, the diopter compensation provides a better initial condition for the adaptive optical aberration compensation to obtain a more stable convergence effect.

[0046] In addition, the present invention can perform real-time eye tracking in multimodality, and can eliminate the influence of eye drift, jumping and blinking on imaging, so as to obtain a stable image of the target field of view of the required modality, and divide a frame of image into multiple segments for processing, which can speed up the processing speed and improve the real-time performance of eye tracking of scanning optical imaging systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] 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 or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0048] Figure 1 A schematic diagram of a retinal imaging method according to an embodiment of the present invention;

[0049] Figure 2 A schematic diagram of initial pupil positioning provided by an embodiment of the present invention;

[0050] Figure 3 A schematic diagram of pupil secondary positioning provided by an embodiment of the present invention;

[0051] Figure 4 A schematic diagram of data processing provided by an embodiment of the present invention;

[0052] Figure 5 A schematic diagram of splicing A-line data into a B-Scan based on tracking provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] 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.

[0055] refer to Figure 1 , Figure 1 A schematic flow chart of a retinal imaging method provided in an embodiment of the present invention.

[0056] Specifically, the pupil position is adjusted first, including:

[0057] S1, obtaining the position of the pupil center, and adjusting the pupil center to be initially positioned at a preset target pupil center.

[0058] S2, acquiring a wavefront image, wherein the wavefront image is a dot matrix image formed by incident light rays being reflected by the retina and converged in a wavefront detector.

[0059] S3, based on the dot matrix distribution in the wavefront image, further adjusting the position of the pupil center so that the pupil center is repositioned at a preset target pupil center.

[0060] In this embodiment, the pupil center is accurately located at a preset target pupil center through at least two pupil positionings. The target pupil center includes but is not limited to being determined by a calibration method. After the pupil center is accurately positioned, it is helpful to improve the quality of subsequent retinal imaging and obtain the required high-quality retinal imaging image.

[0061] Optionally, in another embodiment of the present invention, step S1 obtains the position of the pupil center and adjusts the pupil center to be initially positioned at a preset target pupil center, specifically:

[0062] At least two light detectors are used to collect eye area images, and the collection angles of each light detector are different.

[0063] The eye region image is analyzed and processed to identify the pupil center.

[0064] The offset between the pupil center and a preset target pupil center is compared.

[0065] Based on the offset, the pupil center is adjusted to move toward a preset target pupil center.

[0066] refer to Figure 2 , Figure 2 A schematic diagram of the effect of initial pupil positioning provided in an embodiment of the present invention, wherein a computer obtains at least two eye area images taken by a light detector at different shooting angles, identifies the position of the pupil center from each eye area image, and then compares it with the pre-set target pupil center position to analyze and obtain the offset between the two pupil centers.

[0067] After that, the three-dimensional distance between the pupil center and the target pupil center is calculated, and a control signal is sent to the three-dimensional motion structure to move up, down, forward, backward, left, and right to adjust the position of the pupil center. The above operation is repeated until the offset is less than the set threshold condition. At this time, the initial positioning of the pupil is considered successful.

[0068] It should be noted that the three-dimensional motion structure can be automatically adjusted through a control signal, or can be manually adjusted according to the calculated three-dimensional distance.

[0069] Optionally, in another embodiment of the present invention, in order to maximize the imaging effect of the final retinal imaging, after the pupil is initially positioned successfully, it cannot be guaranteed that the imaging position is in the optimal position, and further precise positioning is required, that is, obtaining a wavefront image, and based on the dot distribution in the wavefront image, further adjusting the position of the pupil center so that the pupil center is repositioned at the preset target pupil center.

[0070] refer to Figure 3 According to the principle of adaptive optics, in the optical system, part of the light reflected by the retina is collected into a wavefront detector through a device such as a beam splitter, so that the wavefront image obtained by the wavefront detector can characterize the wavefront, intensity distribution and other information of the reflected light.

[0071] Based on this, a real-time wavefront image is obtained and the light intensity distribution in the wavefront image is detected; the wavefront image obtained by the wavefront detector is a dot matrix image formed by the incident light being reflected back from the retina and converged in the wavefront detector. The larger the distribution area of ​​these dots, the more light enters the eyeball, which is more conducive to improving the retinal imaging quality.

[0072] Therefore, the principle of positioning using a wavefront detector is to have as many dots as possible in the wavefront image; based on this, a simple trial method can be used to iteratively optimize by moving the three-dimensional motion structure in a smaller step size, or the moving direction of the three-dimensional motion structure can be determined based on the difference between the target's dot distribution and the real-time dot distribution, as well as the spatial geometric relationship of the optical system, to ensure that the dot distribution in the wavefront image meets the preset conditions, which are: the degree of similarity with the area area or contour of the target pupil dot distribution.

[0073] like Figure 3 As shown in the figure, the real-time dot matrix on the left is missing some dots in the lower left corner relative to the target dot matrix on the right. If the direction of the three-dimensional motion structure is consistent with that of the image, the computer sends instructions to move the three-dimensional motion structure to the lower left corner by an appropriate distance. After multi-step iterative optimization, it is necessary to ensure that the dot distribution of the dot matrix image meets the preset distribution conditions, such as the area of ​​the real-time dot matrix distribution reaches 90% of the area of ​​the target dot matrix distribution or is close to a circular distribution.

[0074] If the dot matrix distribution does not meet the preset conditions, the position of the pupil center is adjusted, and then steps S2 and S3 are repeated until the dot matrix distribution meets the preset conditions.

[0075] Optionally, in another embodiment of the present invention, after step S3, the retinal imaging method further includes: automatically performing diopter compensation, specifically, the method is:

[0076] At least two diopters are selected to compensate the optical path and obtain the corresponding wavefront image.

[0077] Evaluate the quality of each wavefront image.

[0078] A wavefront image whose quality meets the preset requirements is selected, and the optical path is automatically compensated with its corresponding diopter.

[0079] In this embodiment, in order to achieve the optimal retinal imaging effect, real-time aberration compensation is required before retinal imaging and after the pupil center is positioned. The real-time aberration compensation specifically includes automatic refractive power compensation and adaptive optical aberration compensation. Among them, the refractive power belongs to the low-order aberration among the aberrations. The purpose of performing the refractive power compensation first is to provide a better initial condition for the subsequent adaptive optical aberration compensation to obtain a more stable convergence effect.

[0080] Automatic refractive compensation specifically achieves compensation for the eye's refractive power by controlling the compensation value of the compensating lens; and the automatic method can avoid the cumbersome manual input process and the problem of inaccurate refractive power data obtained externally.

[0081] Optionally, in another embodiment of the present invention, the automatic refractive compensation includes automatic spherical lens and / or astigmatism compensation, and the automatic spherical lens or astigmatism compensation is specifically:

[0082] Select the initial compensation value and step size.

[0083] The first compensation value and the second compensation value are obtained by adding and subtracting the step length from the initial compensation value.

[0084] The optical path is compensated for the first compensation value, the initial compensation value and the second compensation value respectively, and a corresponding wavefront image is obtained.

[0085] Evaluate the quality of each wavefront image.

[0086] Select the wavefront image with the best quality, use its corresponding compensation value as the new initial compensation value, redefine the step size, and return to the step of adding and subtracting the step size from the initial compensation value to obtain the first compensation value and the second compensation value respectively, until the step size or the wavefront image quality meets the preset requirements.

[0087] In this embodiment, it is assumed that the initial compensation value is 0 degrees and the initial step length is 100 degrees, and a value in each of the positive and negative directions of the initial compensation value (i.e., adding the initial step length and subtracting the initial step length) is selected with the initial step length, i.e., 0 degrees (equivalent to the initial compensation value), +100 degrees (equivalent to the first compensation value), and -100 degrees (equivalent to the second compensation value), with 0 degrees being the middle value.

[0088] The compensation mirror is controlled to compensate the optical path for the first compensation value, the initial compensation value, and the second compensation value respectively, and three compensations are performed. Each compensation simultaneously obtains the wavefront detection signal of the wavefront detector, and the wavefront images of each compensation are generated according to the wavefront detection signal. A measurement method for evaluating the wavefront quality is selected, such as the sum of squares of all pixels in the wavefront image. The larger the sum of squares, the closer the compensated light is to the plane wave. Then, the direction of the next search is determined from these three values:

[0089] If the wavefront quality of the intermediate value is the best, the step size is reduced and the above steps are repeated to search until it converges to below the preset threshold. For example, if the wavefront quality of 0 degrees is the best, the initial step size is reduced to 50 degrees, that is, 0 degrees, +50 degrees, and -50 degrees are selected, and the three compensation values ​​are compensated to the optical path by controlling the compensation mirror, and the above steps are repeated until it converges to below the preset threshold. Converging to below the preset threshold specifically means that the step size meets the preset accuracy requirements (such as the step size is less than 25°) or the wavefront image quality meets the preset requirements.

[0090] If the wavefront quality of the intermediate value is not optimal, the optimal value is used as the new intermediate value, and the step size remains unchanged. For example, if the wavefront quality of +100 degrees is optimal, then +100 degrees, 0 degrees, and +200 degrees are selected, and the three compensation values ​​are compensated for the optical path by controlling the compensation mirror, and the above steps are repeated until it converges to below the preset threshold.

[0091] It should be noted that, assuming that the initial compensation value is 0 degrees and the initial step size is 100 degrees, multiple values ​​(taking two values ​​as an example) in the positive and negative directions of the initial compensation value (i.e., adding the initial step size and subtracting the initial step size) can also be selected with the initial step size, i.e., 0 degrees, +100 degrees, +200 degrees, -100 degrees, and -200 degrees, with 0 degrees being the middle value.

[0092] It should be noted that the compensation ideas for automatic spherical compensation and automatic astigmatism compensation are the same. You can compensate only one of them. If you want to compensate for both, you can compensate for one after the other is completed, for example, compensate for the spherical lens first and then compensate for the astigmatism.

[0093] Optionally, in another embodiment of the present invention, the automatic refractive compensation includes automatic spherical lens and / or astigmatism compensation, and the automatic spherical lens or astigmatism compensation is specifically:

[0094] A plurality of sample compensation values ​​are selected at fixed intervals within the compensation value domain.

[0095] The optical path is compensated for each sample compensation value and the corresponding wavefront image is obtained.

[0096] Evaluate the quality of each wavefront image.

[0097] The wavefront image with the best quality is selected, and its corresponding compensation value plus and minus the fixed interval is used as the new compensation value range endpoint, and the interval is re-determined, and the step of selecting multiple sample compensation values ​​at fixed intervals within the compensation value range is returned to execute until the interval or wavefront image quality meets the preset requirements.

[0098] In this embodiment, assuming that the compensation value range of the spherical mirror is -400 degrees to +400 degrees with a fixed interval of 200 degrees, the compensation value range of -400 degrees to +400 degrees is segmented based on a fixed interval of 200 degrees to obtain multiple sample compensation values ​​of -400 degrees, -200 degrees, 0 degrees, +200 degrees, and +400 degrees.

[0099] By controlling the compensation mirror to compensate the light path with sample compensation values ​​of -400 degrees, -200 degrees, 0 degrees, +200 degrees, and +400 degrees, multiple compensations are performed. Each compensation simultaneously obtains the wavefront detection signal of the wavefront detector. The wavefront images at the time of respective compensation are generated according to the wavefront detection signal. A measurement method for evaluating the wavefront quality is selected, such as the sum of squares of all pixels in the wavefront image. The larger the sum of squares, the closer the compensated light is to the plane wave. Then, the direction of the next search is determined among these sample compensation values:

[0100] If the wavefront quality at 0 degrees is optimal, -200 degrees and +200 degrees are determined as the new endpoints of the compensation value range, and the new compensation value range is -200 degrees to +200 degrees. Assuming that the re-determined interval is 100 degrees, the compensation threshold from -200 degrees to +200 degrees is segmented based on a fixed interval of 100 degrees to obtain multiple sample compensation values ​​of -200 degrees, -100 degrees, 0 degrees, +100 degrees, and +200 degrees.

[0101] By controlling the compensation mirror, the optical path is compensated with sample compensation values ​​of -400 degrees, -200 degrees, 0 degrees, +200 degrees, and +400 degrees respectively, and the above steps are repeated until the interval meets the preset accuracy requirements (such as the interval is less than 10 degrees) or the wavefront image quality meets the preset requirements.

[0102] It should be noted that the compensation ideas for automatic spherical compensation and automatic astigmatism compensation are the same. You can compensate only one of them. If you want to compensate for both, you can compensate for one after the other is completed, for example, compensate for the spherical lens first and then compensate for the astigmatism.

[0103] It should be noted that after the diopter compensation is completed, the process returns to step S1 to repeat pupil positioning and diopter compensation to optimize the effects of pupil positioning and diopter compensation, and then adaptive optical aberration compensation is performed.

[0104] Optionally, in another embodiment of the present invention, adaptive optical aberration compensation may be performed directly after diopter compensation, and diopter compensation provides a better initial condition for adaptive optical aberration compensation to obtain a more stable convergence effect.

[0105] Perform adaptive optics aberration compensation, including:

[0106] Acquire a wavefront image.

[0107] The offset of the dot matrix in the wavefront image relative to the dot matrix of the reference plane wave is calculated, and the aberration compensation feedback value is determined according to the offset of the dot matrix.

[0108] The optical path is compensated with the aberration compensation feedback value to realize adaptive optical aberration compensation.

[0109] In this embodiment, according to the principle of adaptive optics, the method of adaptive optical aberration compensation is to obtain the wavefront detection signal of the wavefront detector in real time, generate a wavefront image based on the wavefront detection signal, detect the dot position in the image and calculate the offset of the dot matrix relative to the reference plane wave, and then multiply it with the control matrix obtained by correction to obtain feedback that should be sent to the compensation mirror.

[0110] It should be noted that this process is a closed-loop control process, and a relatively stable convergence process can be obtained by adjusting the gain.

[0111] Optionally, in another embodiment of the present invention, after the adaptive optical system stabilizes the imaging, with the assistance of devices such as a head rest and a fixation point, the movement of the eyeball is restricted to a smaller range, which is mainly divided into drifting, jumping and blinking.

[0112] In some application scenarios, in order to obtain a stable image of the target field of view, real-time eye tracking is also required. The real-time retinal image reflected from the retina is used to track eye movements. Specifically:

[0113] The retinal imaging is retinal multimodal imaging, including synchronous imaging of a first modal image and a second modal image of the retina, and the method is:

[0114] When tracking eye movement, a first modality sub-image is obtained, and it is registered with a preset first modality reference image to obtain an offset of the eyeball; the first modality sub-image is a segment of a frame of the first modality image.

[0115] A first control signal is generated based on the offset to adjust the optical path state to perform eye movement tracking; and a position of the second modality sub-image in the second modality image is determined based on the offset; the second modality sub-image is a portion of the second modality image synchronously acquired when acquiring the first modality sub-image.

[0116] The second modality sub-images are accumulated according to the positional relationship and combined into a second modality image output.

[0117] In this embodiment, the first modality image is a two-dimensional reflection image, and the second modality image is an OCT image or a fluorescence image. Figure 4 , Figure 4A schematic diagram of data processing provided for an embodiment of the present invention. In order to improve the real-time performance of eye tracking, a frame of image data can be divided into multiple segments and sent to a computing unit for processing in sequence. Assuming that the system is scanned by lines, each frame contains 512 lines of data, which can be sent to the computing unit immediately after 32 or 16 lines of data are collected. In this way, the tracking feedback frequency is increased to 16 or 32 times compared to the method of waiting until a frame of data is completely collected before sending.

[0118] The specific tracking process is as follows:

[0119] The data acquisition module immediately sends a predetermined short segment of image data to the computer unit after obtaining it. After obtaining the data, the computing unit performs preprocessing such as interpolation and desintoling. Before starting tracking, it selects one or more frames of image data from the video stream and processes them as reference frames. After starting tracking, each preprocessed short segment of image data is aligned with the selected reference frame image data in real time to obtain the offset of the eyeball, and then generates a first control signal based on the offset. According to the first control signal, the state of the optical path is adjusted to realize eyeball movement tracking, so as to achieve the purpose of stabilizing the imaging area. The state of the optical path can be adjusted by adjusting the deflection angle of the galvanometer in the optical path.

[0120] For the point scanning retinal imaging structure, the first modality image is a two-dimensional reflection image, and the second modality image is an OCT image. When the acquisition frequencies of the light detector and the OCT detector are inconsistent, the specific processing method is: determine the field of view to be imaged based on the reference frame image selected by tracking, and the light detector and the OCT detector synchronously collect image data, that is, each line of reflection image data collected by the light detector corresponds to an OCT A-line in the same time period. After obtaining every 16 or 32 lines of reflection image data and the corresponding A-line image data, the offset of the eyeball is calculated according to the reflection image data and finally fed back to the corresponding galvanometer. At the same time, the corresponding A-line image data collected is placed in the corresponding position of the three-dimensional reference system determined based on the tracking reference frame according to the tracking offset. Each time a frame of reflection image is tracked, a corresponding B-scan is collected, and then the scanning of the next position is entered. If the tracking fails, the corresponding data is discarded and the position remains unchanged. The scan is repeated until the tracking is successful, and finally the three-dimensional OCT data of the entire field of view is spliced ​​according to the position relationship. Figure 5 , Figure 5 A schematic diagram of splicing A-line data into a B-Scan based on tracking provided by an embodiment of the present invention.

[0121] In the case where the acquisition frequencies of the optical detector and the OCT detector are the same, the processing method is different in that each time a line of reflected image tracking is completed, a corresponding OCT B-scan acquisition is completed. After obtaining every 16 or 32 lines of reflected data and the corresponding B-scan data, data transmission and processing can be performed.

[0122] For the line scanning retinal imaging structure, the first modality image is a two-dimensional reflection image, and the second modality image is an OCT image. The acquisition frequency of the light detector and the OCT detector is the same. The difference from the point scanning processing method is that each time a reflection line is tracked, an OCT B-scan is acquired accordingly.

[0123] For the retinal imaging structure of fluorescence scanning, the first modal image is a two-dimensional reflection image, and the second modal image is a fluorescence image. Each time a line of reflection image tracking is completed, a line of fluorescence image acquisition is completed. Each line of reflection image data corresponds to a line of fluorescence image data in the same time period. After obtaining each segment (such as 16 or 32 lines) of reflection image data and the corresponding fluorescence image data, the offset of the eyeball is calculated based on the reflection image data and finally fed back to the corresponding galvanometer. At the same time, according to the tracking offset, each segment of fluorescence image data is placed in the corresponding position of the reference system determined based on the tracking reference frame. Multiple segments of fluorescence image data are spliced ​​into one piece of fluorescence image data according to the positional relationship. After that, multiple pieces of fluorescence image data can be accumulated according to the positional relationship, and finally superimposed into a total fluorescence image output to achieve image enhancement effect.

[0124] Imaging systems that use real-time tracking technology can solve problems such as image distortion and misalignment caused by eye movement. For slower OCT imaging, eye movement can lead to inaccurate positions of each B-scan, and complex and slow post-processing is often required to reduce these effects. Therefore, if the real-time tracking method described above is used, valuable image data can be displayed in real time during the acquisition process, without having to wait until post-processing is completed to see the desired results.

[0125] The retinal imaging method provided by the present invention is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. 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 method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A retinal imaging method, characterized in that: First adjust the pupil position, including: S1, obtaining the position of the pupil center, and adjusting the pupil center to be initially positioned at a preset target pupil center; S2, acquiring a wavefront image, wherein the wavefront image is a dot matrix image formed by incident light rays reflected from the retina and converged in a wavefront detector; S3, based on the dot matrix distribution in the wavefront image, further adjusting the position of the pupil center so that the pupil center is repositioned at a preset target pupil center; The specific method of step S3 is: when the dot matrix distribution does not meet the preset conditions, adjust the position of the pupil center, and then repeat steps S2 and S3; The preset conditions are: the similarity of the area or contour with the target pupil dot matrix distribution; when adjusting the position of the pupil center, determining the adjustment direction of the pupil position according to the difference between the target pupil dot matrix distribution and the real-time dot matrix distribution; After step S3, spherical lens and / or astigmatism compensation is automatically performed, and the specific method is as follows: (1) selecting an initial compensation value and a step length; obtaining a first compensation value and a second compensation value by adding and subtracting the step length from the initial compensation value; compensating the optical path with the first compensation value, the initial compensation value, and the second compensation value, respectively, and obtaining a corresponding wavefront image; evaluating the quality of each wavefront image; selecting a wavefront image with the best quality, using its corresponding compensation value as a new initial compensation value, re-determining the step length, and returning to execute the step of obtaining the first compensation value and the second compensation value by adding and subtracting the step length from the initial compensation value, respectively, until the step length or the wavefront image quality meets the preset requirements; Or (2) selecting a plurality of sample compensation values ​​at fixed intervals within the compensation value domain; compensating the optical path for each sample compensation value and obtaining a corresponding wavefront image; evaluating the quality of each wavefront image; selecting the wavefront image with the best quality, using its corresponding compensation value plus and minus the fixed interval as the new endpoints of the compensation value domain, and re-determining the interval, and returning to execute the step of selecting a plurality of sample compensation values ​​at fixed intervals within the compensation value domain until the interval or the wavefront image quality meets the preset requirements.

2. A retinal imaging method according to claim 1, characterized in that: The step S1 comprises: At least two light detectors are used to collect images of the eye area, and the collection angles of each light detector are different; Analyzing and processing the eye region image to identify the pupil center; Comparing the offset between the pupil center and a preset target pupil center; Based on the offset, the pupil center is adjusted to move toward a preset target pupil center.

3. A retinal imaging method according to claim 1, characterized in that: After the automatic spherical lens and / or astigmatism compensation is performed, adaptive optical aberration compensation is performed, including: Acquire a wavefront image; Calculating the offset of the dot matrix in the wavefront image relative to the dot matrix of the reference plane wave, and determining the aberration compensation feedback value according to the offset of the dot matrix; The optical path is compensated with the aberration compensation feedback value to realize adaptive optical aberration compensation.

4. A retinal imaging method according to claim 1, characterized in that: The retinal imaging is retinal multimodal imaging, including synchronous imaging of a first modal image and a second modal image of the retina, and the method is: When performing eye movement tracking, a first modality sub-image is obtained, and it is registered with a preset first modality reference image to obtain an offset of the eyeball; the first modality sub-image is a segment of a frame of the first modality image; A first control signal is generated based on the offset, and the optical path state is adjusted according to the first control signal to perform eye movement tracking; and a position of the second modality sub-image in the second modality image is determined based on the offset; the second modality sub-image is a portion of the second modality image synchronously acquired when acquiring the first modality sub-image; The second modality sub-images are accumulated according to their positions in the second modality image, and are combined into a second modality image for output.

5. A retinal imaging method according to any one of claims 1 to 4, characterized in that: The retinal imaging method further comprises: After the spherical lens and / or astigmatism compensation is completed, the process returns to step S1 to repeat pupil positioning and spherical lens and / or astigmatism compensation.

Images