An Automatic OCT Image Segmentation Method and Related Devices

Abstract

The present invention discloses an automatic segmentation method for OCT images and related devices. The method includes: an extraction step of extracting the stratification line information of the retinal pigment epithelium (RPE) and the choroid-sclera interface (CSI); a pixel acquisition step of extracting the pixel values near the RPE position and the CSI position of each column of the B-scan image; and an interpolation step of performing interpolation and calculation according to the OCT propagation attenuation model by using the preset pixel values near the RPE position and the CSI position to obtain an initial segmentation result of the choroidal blood vessels. The automatic segmentation method for choroidal blood vessels in OCT images proposed by the present invention, based on the light propagation attenuation model of OCT, can not only improve the segmentation accuracy of choroidal blood vessels in OCT images, but also has good processing ability for B-scan images with uneven blood vessel brightness, unclear blood vessel boundaries, and poor contrast. Moreover, the calculation is simple, no denoising preprocessing is required, the speed of automatic segmentation of choroidal blood vessels can be improved, and an OCT-A image with a clear choroidal blood vessel structure can also be generated.

Description

Technical Field

[0001] The present invention relates to image segmentation technology, and in particular to a method and related device for automatically segmenting choroidal blood vessels in OCT images. Background Art

[0002] OCT refers to optical coherence tomography technology. Using OCT for choroidal blood vessel segmentation is an important preliminary step in quantitatively analyzing human choroidal parameters. The En-face image (C-scan image) of OCT can display choroidal blood vessels. However, there are retinal blood vessel projections in the En-face image, and it is difficult to display smaller blood vessels in the choroid, which affects the observation and quantification of choroidal blood vessels. Therefore, using an effective image segmentation technology for choroidal blood vessels to segment choroidal blood vessels and calculate OCT angiography images (OCT-A images) is crucial for observing and quantifying choroidal blood vessels.

[0003] Existing image segmentation technologies for choroidal blood vessels mostly adopt methods such as threshold segmentation, contour extraction, or deep learning. However, the above segmentation technologies have problems such as low accuracy, slow speed, and poor imaging quality. Summary of the Invention

[0004] The purpose of the present invention is to achieve the technical effects of improving the image segmentation accuracy and speed of choroidal blood vessels and improving the quality of OCT angiography images.

[0005] An automatic segmentation method for OCT images includes:

[0006] An extraction step of extracting the stratification line information of the retinal pigment epithelium layer RPE and the choroid-sclera interface CSI;

[0007] A pixel acquisition step of extracting the pixel values near the RPE position and the CSI position of each column of the B-scan image;

[0008] An interpolation step of performing interpolation and calculation according to the OCT propagation attenuation model using the preset pixel values near the RPE position and the CSI position to obtain an initial segmentation result of choroidal blood vessels.

[0009] Optionally, before the extraction step, it further includes:

[0010] A construction step including constructing the light propagation attenuation model of the OCT.

[0011] Optionally, after the interpolation step, it further includes: a binarization step including binarizing the initial segmentation result of choroidal blood vessels and performing morphological operations to obtain a choroidal blood vessel segmentation result with smooth edges and generating an OCT-A image.

[0012] Optionally,

[0013] The interpolation step is specifically implemented as follows: using the pixel values near the RPE position and the CSI position to interpolate the pixels at the positions between the RPE and the CSI for each column;

[0014] According to the OCT imaging model, use the pixel values near the RPE position and the CSI position to fit the propagation curve of light only in other choroid tissues, and interpolate the pixels between the RPE and the CSI for each column according to the pixel values of the fitted curve.

[0015] Optionally, the operations in the interpolation step are specifically implemented as follows:

[0016] Subtract the actual observed B-scan image from the interpolation result to obtain the initial segmentation result of choroidal blood vessels.

[0017] Optionally, the binarization step is specifically implemented as follows:

[0018] Perform Gaussian smoothing on the initial segmentation result of choroidal blood vessels;

[0019] Use the maximum inter-class variance algorithm to binarize the initial segmentation result of choroidal blood vessels to obtain an intermediate binarization result;

[0020] Perform morphological operations on the intermediate binarization result to obtain a segmented result of choroidal blood vessels with smooth edges. And use the segmentation result to generate an OCT-A image.

[0021] Optionally, constructing the light propagation attenuation model of the OCT is specifically implemented as follows:

[0022] Use the Beer-Lambert law to describe the law of attenuation during propagation in the choroid tissue;

[0023] Solve the backscattering parameters of light by taking the scattering part as a component of the incident light, and regard the choroidal blood vessels and the choroid tissue as homogeneous media to calculate the first absorption coefficient and the second absorption coefficient, as well as the first scattering coefficient and the second scattering coefficient to fit the attenuation curve;

[0024] Use the first absorption coefficient and the first scattering coefficient to indicate the attenuation law of the visual choroidal blood vessels;

[0025] Use the second absorption coefficient and the second scattering system to indicate the attenuation law of the choroid tissue.

[0026] Optionally, extracting the stratification line information of the retinal pigment epithelium layer RPE and the choroid scleral interface CSI is specifically implemented as follows:

[0027] Use the shortest path search algorithm to extract the stratification line information of the RPE and the CSI respectively.

[0028] A computing device, comprising: at least one processor; and

[0029] A memory communicatively connected to the at least one processor; wherein,

[0030] the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method described in the embodiments of the present invention.

[0031] A readable storage medium stores computer-executable instructions for executing the method described in the embodiments of the present invention.

[0032] The automatic choroidal vessel segmentation method for OCT images proposed by the present invention is based on the light propagation attenuation model of OCT. It can not only improve the segmentation accuracy of choroidal vessels in OCT images, but also has good processing ability for B-scan images with uneven vessel brightness, unclear vessel boundaries, and poor contrast. Moreover, the calculation is simple, no denoising preprocessing is required, which can improve the speed of automatic choroidal vessel segmentation. It can also generate OCT-A images with clear choroidal vessel structures and effectively reduce the influence of retinal vessel shadows on the choroidal vessel OCT-A images, thereby achieving the technical effects of improving the image segmentation accuracy and speed of choroidal vessels and enhancing the quality of OCT angiography images. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0034] Figure 1 is a flowchart of the steps of the OCT image automatic segmentation method according to an embodiment of the present invention;

[0035] Figure 2 is a flowchart of the steps of the OCT image automatic segmentation method according to an embodiment of the present invention;

[0036] Figure 3 is a flowchart of the light propagation process in OCT imaging according to an embodiment of the present invention;

[0037] Figure 4a is a schematic diagram of the A-scan light intensity curve when light only propagates in choroidal vessels or other choroidal tissues according to an embodiment of the present invention;

[0038] Figure 4b is a schematic diagram of the A-scan light intensity curve when light propagates in choroidal vessels and other choroidal tissues according to an embodiment of the present invention;

[0039] Figure 5 is the actually observed B-scan image according to an embodiment of the present invention;

[0040] Figure 6a Flowchart of the interpolation step of the OCT image automatic segmentation method according to an embodiment of the present invention;

[0041] Figure 6b Interpolation result image between the RPE and CSI positions according to an embodiment of the present invention;

[0042] Figure 7 Flowchart of the binarization step of the OCT image automatic segmentation method according to an embodiment of the present invention;

[0043] Figure 8 Schematic structural diagram of a computing device according to an embodiment of the present application;

[0044] Figure 9 Schematic structural diagram of a readable medium according to an embodiment of the present application;

[0045] Figure 10 Initial segmentation result image of choroidal blood vessels according to an embodiment of the present invention;

[0046] Figure 11 Segmentation result image of choroidal blood vessels after image binarization according to an embodiment of the present invention;

[0047] Figure 12 Comparison diagram of the OCT-A image (left) and En-face image (right) of choroidal blood vessels according to an embodiment of the present invention. Detailed implementation manners

[0048] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0049] Choroidal blood vessel segmentation is an important preliminary step when quantitatively analyzing human choroidal parameters using optical coherence tomography (OCT). Although the En-face image of OCT can show choroidal blood vessels, there are retinal blood vessel projections in the En-face image, and it is difficult to show smaller blood vessels in the choroid, which affects the observation and quantification of choroidal blood vessels. Therefore, using an effective choroidal blood vessel segmentation technology to segment choroidal blood vessels and calculate the OCT angiography image is very important for observing and quantifying choroidal blood vessels.

[0050] The following are the existing technologies for choroidal blood vessel segmentation and the corresponding problems:

[0051] (1) Since the method of manual segmentation is cumbersome and time-consuming, an automatic choroidal vessel segmentation technique for OCT images has been proposed. For the existing automatic choroidal vessel segmentation techniques for OCT images, the quality of the OCT-A images obtained from the choroidal vessel segmentation results of the existing schemes is poor, indicating that the accuracy of the existing choroidal vessel segmentation schemes needs to be improved. Moreover, since retinal vessels will produce shadows in B-scan images, and the gray values of the shadow parts are relatively low, they are easily misclassified as vessels, so the choroidal vessel OCT-A images often contain the projections of retinal vessels.

[0052] (2) In the case where the blood vessel brightness in the B-scan image is uneven, the blood vessel boundaries are unclear, and the contrast of the B-scan image is poor, the methods based on threshold segmentation and the methods based on contour extraction sometimes cannot obtain good choroidal vessel segmentation results. And both of these two types of methods require good preprocessing or postprocessing to denoise the input image and refine the segmentation results respectively.

[0053] (3) The methods based on deep learning require a large amount of labeled data, and have a high computational cost and a slow running speed.

[0054] Based on this, the present invention proposes an OCT image segmentation technique for choroidal vessels to improve the segmentation accuracy and speed of choroidal vessels in OCT images, and can also generate choroidal vessel OCT-A images with clear structures and without projections of retinal vessels.

[0055] The embodiments of the present invention provide an automatic OCT image segmentation method and related devices, achieving the technical effects of improving the image segmentation accuracy and speed of choroidal vessels and improving the quality of OCT angiography images.

[0056] An automatic OCT image segmentation method of the present invention includes the following steps:

[0057] S11: An extraction step of extracting the stratification line information of the retinal pigment epithelium (RPE) and the choroid-sclera interface (CSI);

[0058] The choroid is a connective tissue located between the retina and the sclera, containing various tissues and structures such as blood vessels, melanocytes, fibroblasts, immune cells, neurons, and collagen fibers. The choroid has rich blood flow and is an important part of the blood supply to the outer layer of the retina and the macular area, playing a crucial role in the metabolic activities of the eye. Therefore, the structural and functional abnormalities of the choroid are related to many eye diseases, such as age-related macular degeneration (AMD), polypoidal choroidal vasculopathy (PCV), choroidal neovascularization (CNV), etc.

[0059] Optical Coherence Tomography (OCT) is a non-invasive imaging technique that can detect reflection or scattering signals at different depths in biological tissues, and has the advantages of non-contact, high resolution, and high speed. In OCT imaging, the depth tomographic image of a point is called an A-scan image; the two-dimensional cross-sectional image is generated by laterally scanning the beam across the sample to be measured and collecting continuous A-scans, and is called a B-scan image; the image of the sample to be measured at a certain depth position parallel to the surface of the sample to be measured obtained by scanning the sample to be measured in both the lateral and longitudinal directions is called an En-face image or a C-scan image.

[0060] Quantitative analysis of choroidal parameters, including measuring the total volume of choroidal blood vessels, the distribution of blood vessels in the choroid, etc. Among them, segmentation of choroidal blood vessels is an important prior step, and choroidal blood vessel segmentation is necessary for generating OCT Angiography (OCT-A) images of choroidal blood vessels.

[0061] Optionally, before the extraction step, it further includes:

[0062] The construction step includes: constructing the light propagation attenuation model of the OCT. Specifically, referring to Figure 2 :

[0063] S21: Using the Beer-Lambert law to describe the attenuation law during propagation in choroidal tissue;

[0064] The imaging principle of OCT is based on the optical coherence principle, with a Michelson interferometer as the core optical structure. The basic process of OCT imaging is that light is incident on the sample to be measured, the backscattered light scattered by the sample to be measured is coupled into the sample arm, and then interferes with the reference light that has traveled a fixed optical path length along the reference arm. The interference pattern data is obtained through photoelectric conversion by the receiving unit of the interferometer, and the tomographic image in the depth direction of the sample to be measured is reconstructed through data processing. According to the above basic process of OCT imaging, a beam of light from incidence on the sample to emergence and interference with the reference light will experience processes such as Figure 3 shown in (1) absorption attenuation, (2) backscattering, and (3) absorption attenuation of backscattered light.

[0065] For the process of light absorption attenuation, the present invention considers that the Beer-Lambert law is the law of light absorption when passing through a substance, so the attenuation law of light during propagation in choroidal tissue can be described by the Beer-Lambert law, as shown in formula (1):

[0066] T = I / I0 = e -α′cd (1)

[0067] Among them, T is the transmittance, I0 is the intensity of the incident light, I is the intensity of the transmitted light, α' is the light absorption coefficient of the substance, c is the concentration, and d is the propagation distance of light in the choroid.

[0068] S22: Solve the backscattering parameters of light by taking the scattered part as a component of the incident light, and regard the choroidal blood vessels and choroidal tissue as a homogeneous medium;

[0069] For the backward scattering process of light, it is approximately considered that the scattered part is a component of the incident light. Then the backward scattering process is shown in formula (2):

[0070] I′=k*I0 (2)

[0071] Among them, I0 is the intensity of the incident light, I’ is the intensity of the backward scattered light, and k is the backward scattering coefficient.

[0072] S23: Calculate the first absorption coefficient and the second absorption coefficient, as well as the first scattering coefficient and the second scattering coefficient to fit the attenuation curve;

[0073] Use the first absorption coefficient and the first scattering coefficient to indicate the attenuation law of the choroidal blood vessels;

[0074] Use the second absorption coefficient and the second scattering coefficient to indicate the attenuation law of the choroidal tissue.

[0075] In this embodiment, according to the OCT imaging principle and the Beer-Lambert law, it is considered that the choroidal blood vessels and other choroidal tissues are approximately homogeneous media, which respectively have different absorption coefficients α1 or α2 and different scattering coefficients k1 or k2, and an optical propagation attenuation model of the OCT imaging process is established.

[0076] Denote R as the ratio of the intensity of the outgoing light to the intensity of the incident light. If the same homogeneous medium is observed using OCT, the optical propagation process involved in the process shown in Figure 4 is shown in formula (3):

[0077]

[0078] Since α′ and c in the above formula are physical units and constants, the absorption coefficient α = 2α′c is denoted, then as shown in formula (4):

[0079] R=k*e -αd (4)

[0080] In this embodiment, considering that the choroidal blood vessels and other choroidal tissues are both approximately homogeneous media, when light propagates only in the choroidal blood vessels or other choroidal tissues, there are two different first absorption coefficients α1 or second absorption coefficients α2 and different first scattering coefficients k1 or second scattering coefficients k2 for the attenuation of light. Since in the B-scan image, the brightness of the choroidal blood vessel part is lower than that of other choroidal tissues, the present invention believes that the choroidal blood vessels have a larger absorption coefficient α1 or a smaller scattering coefficient k1. If light propagates only in the choroidal blood vessels or other choroidal tissues, the light intensity of an A-scan of OCT is approximately as Figure 4a shown by the dot or asterisk curve in

[0081] If light passes through both choroidal blood vessels and other choroidal tissues during propagation, the attenuation of light combines two different first absorption coefficients α1 and second absorption coefficients α2, as well as the first scattering coefficient k1 and the second scattering coefficient k2. The light intensity of an A-scan of OCT passing through a blood vessel is approximately as Figure 4b shown by the triangle curve in

[0082]

[0083] where d1 and d2 are the upper and lower boundaries of the choroidal blood vessels and other choroidal tissues respectively.

[0084] In Figure 4b it can be seen that when d1 ≤ d ≤ d2, when d is small, the value of the curve passing through the two tissues is closer to the curve passing through only other choroidal tissues, and when d is large, the value of the curve passing through the two tissues is significantly different from the curve passing through only other choroidal tissues. This attenuation law explains the cause of the phenomenon that in the B-scan image, as Figure 5 shown, the upper part of the blood vessel is brighter and the upper boundary is more blurred, while the lower part of the blood vessel is darker and the lower boundary is clearer.

[0085] Furthermore, the extraction of the stratification line information of the retinal pigment epithelium layer RPE and the choroid-sclera interface CSI is specifically implemented as follows:

[0086] The stratification line information of the RPE and CSI is extracted using the shortest path search algorithm respectively, and of course, it is not limited to this algorithm.

[0087] S12: Pixel acquisition step, extracting the pixel values near the positions of the RPE and CSI in each column of the B-scan image;

[0088] Since the RPE and CSI stratification lines do not pass through the choroidal blood vessels, after obtaining the pixel values near these two positions, the propagation law of light in other choroidal tissues can be approximately fitted according to the pixel values near these two positions, that is, the curve shown by the dots is fitted through the pixel values near the RPE and CSI stratification lines. Therefore, optionally, after fitting the light intensity curve of light propagating only in other choroidal tissues, subtracting the light intensity curve of the A-scan passing through the blood vessels can suppress the brightness of other choroidal tissues and retain and highlight the brightness of the choroidal blood vessel tissue. Figure 5 After fitting the light intensity curve of light propagating only in other choroidal tissues, subtracting the light intensity curve of the A-scan passing through the blood vessels can suppress the brightness of other choroidal tissues and retain and highlight the brightness of the choroidal blood vessel tissue.

[0089] S13: Interpolation step. According to the OCT propagation attenuation model, using the preset pixel values at the RPE position and the CSI position, interpolation and calculation are performed to obtain the initial segmentation result of the choroidal blood vessels.

[0090] It should be noted that, optionally, the initial segmentation result of the choroidal blood vessels is obtained by subtracting the actually observed B-scan image from the interpolation result, but it is not limited to this.

[0091] Reference Figure 6a , the specific implementation of the S13 interpolation step is as follows:

[0092] S61: Use the pixel values near the RPE position and the CSI position to interpolate the pixels at the positions between the RPE and the CSI in each column.

[0093] S62: According to the OCT imaging model, use the pixel values near the RPE position and the CSI position to fit the propagation curve of light only in other choroidal tissues, and interpolate the pixels between the RPE and the CSI in each column according to the pixel values of the fitted curve.

[0094] According to the OCT imaging model in S21 - S23, use the pixel values near the RPE position and the CSI position to fit the propagation curve of light only in other choroidal tissues, and interpolate the pixels between the RPE and the CSI in each column according to the pixel values of the fitted curve, as shown in 6b.

[0095] The specific implementation of the operation step in S13 is to use the interpolated B-scan image shown as Figure 6b subtract the actually observed B-scan image shown as Figure 5 to suppress the brightness of other choroidal tissues in the B-scan image and retain only the choroidal blood vessel tissue, achieving a better initial segmentation result of the choroidal blood vessels.

[0096] Regarding the binarization step in S13, it includes: binarizing the initial segmentation result of the choroidal blood vessels and performing morphological operations to obtain a choroidal blood vessel segmentation result with smooth edges and generating an OCT-A image. More specifically, and optionally, the specific implementation of this binarization step is

[0097] Reference Figure 7 ,

[0098] S71: Perform Gaussian smoothing on the initial segmentation result of choroidal blood vessels;

[0099] The edges of the initial segmentation result of choroidal blood vessels are relatively rough and blurred. As shown in 10, if better segmentation results are to be obtained, binarization and edge smoothing of the initial segmentation result of choroidal blood vessels are required. Gaussian smoothing is performed on the initial segmentation result of choroidal blood vessels to reduce the noise of the initial segmentation result.

[0100] S72: Use the Otsu algorithm to binarize the initial segmentation result of choroidal blood vessels to obtain an intermediate binarization result, and of course it is not limited to this algorithm.

[0101] S73: Perform morphological operations on the intermediate binarization result to achieve a segmentation result of choroidal blood vessels with smooth edges.

[0102] Use a series of morphological operations such as dilation and erosion to obtain a segmentation result of choroidal blood vessels that is binarized and has relatively smooth edges, as Figure 11 shown. And generate an OCT-A image of choroidal blood vessels according to the segmentation result of choroidal blood vessels. Apply the Figure 1 corresponding steps to all B-scan images of the OCT image, and use the binarization result of choroidal blood vessels of the B-scan obtained in Figure 7 to calculate the average value of the pixels between the RPE layer and the CSI layer to generate an OCT-A image of choroidal blood vessels. Compared with the En-face image, the OCT-A image of choroidal blood vessels calculated by the present invention, as shown in 12, not only has a clearer and richer vascular structure, but also basically removes the projection of retinal blood vessels.

[0103] In summary, the technical effects of the present invention are as follows:

[0104] Based on the proposed light propagation attenuation model of OCT, the present invention improves the segmentation accuracy of choroidal blood vessels in OCT images, can effectively segment choroidal blood vessels in the shadows generated by retinal blood vessels, and therefore can generate an OCT-A image of choroidal blood vessels with a clear structure and remove the projection of retinal blood vessels in the OCT-A image of choroidal blood vessels.

[0105] Compared with the method based on threshold segmentation and the method based on contour extraction, the solution proposed by the present invention also has good processing ability for B-scan images with uneven blood vessel brightness, unclear blood vessel boundaries, and poor contrast, and can segment larger blood vessels more completely.

[0106] Compared with the deep learning-based method, the solution proposed by the present invention can achieve better segmentation results with smaller computational overhead and faster computing speed.

[0107] Figure 8 Shown is the matching Figures 1 - 5 The computing device 80 of the method shown includes:

[0108] It should be noted that Figure 8 The displayed computing device 80 is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0109] Such as Figure 8 As shown, the server is presented in the form of a general-purpose computing device 80. The components of the computing device 80 may include, but are not limited to: the above-mentioned at least one processor 81, the above-mentioned at least one memory 82, and a bus 83 connecting different system components (including the memory 82 and the processor 81).

[0110] The bus 83 represents one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a processor, or a local bus using any bus structure in a variety of bus structures.

[0111] The memory 82 may include a readable medium in the form of volatile memory, such as a random access memory (RAM) 821 and / or a cache memory 822, and may further include a read-only memory (ROM) 823.

[0112] The memory 82 may also include a program / utility 825 having a set (at least one) of program modules 824. Such program modules 824 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. The implementation of a network environment may be included in each or some combination of these examples.

[0113] The computing device 80 can also communicate with one or more external devices 84 (such as a keyboard, a pointing device, etc.), and can also communicate with one or more devices that enable a user to interact with the computing device 80, and / or communicate with any device (such as a router, a modem, etc.) that enables the computing device 80 to communicate with one or more other computing devices. Such communication can be carried out through the input / output (I / O) interface 85. Moreover, the computing device 80 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 88. As shown in the figure, the network adapter 88 communicates with other modules for the computing device 80 through the bus 83. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in combination with the computing device 80, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0114] In some possible implementation manners, the computing device according to the present application may include at least one processor and at least one memory (such as a first server). Among them, the memory stores program code, and when the program code is executed by the processor, the processor is caused to execute the steps in the system privilege opening method according to various exemplary implementation manners of the present application described above in this specification.

[0115] Reference Figure 9 , Figures 1 - 5 The automatic choroidal vessel segmentation method for OCT images in the illustrated and corresponding embodiments can also be implemented through a computer-readable medium 91. Reference Figure 9 , storing computer-executable instructions, that is, the program instructions required to execute the method of the present invention. The computer or high-speed chip executable instructions are used to execute the automatic choroidal vessel segmentation method for OCT images described in the above embodiments.

[0116] The readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which the readable program code is carried. Such a propagated data signal can take various forms, including - but not limited to - electromagnetic signals, optical signals, or any suitable combination of the above. The readable signal medium can also be any readable medium other than the readable storage medium, and this readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0117] The program code included on the readable medium can be transmitted by any appropriate medium, including - but not limited to - wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0118] The program code for performing the operations of this application can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0119] The program product can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0120] The program product for system privilege activation in the embodiments of this application can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can run on a computing device. However, the program product of this application is not limited to this. In this document, the readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0121] In summary, the automatic choroidal vessel segmentation method for OCT images proposed by the present invention, based on the light propagation attenuation model of OCT, can not only improve the segmentation accuracy of choroidal vessels in OCT images, but also has good processing ability for B-scan images with uneven vessel brightness, unclear vessel boundaries, and poor contrast. Moreover, the calculation is simple, no denoising preprocessing is required, which can improve the speed of automatic choroidal vessel segmentation, and can also generate OCT-A images with clear choroidal vessel structures, and effectively reduce the influence of retinal vessel shadows on choroidal vessel OCT-A images, so as to achieve the technical effects of improving the image segmentation accuracy and speed of choroidal vessels and improving the quality of OCT angiography images.

[0122] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or multiple blocks.

[0123] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or multiple blocks.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or multiple blocks.

[0125] Although alternative embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include alternative embodiments and all changes and modifications that fall within the scope of the present application.

[0126] Obviously, those skilled in the art can make various changes and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. An automatic segmentation method for OCT images, characterized in that, Comprising: An extraction step of extracting the stratification line information of the retinal pigment epithelium (RPE) and the choroid-sclera interface (CSI); A pixel acquisition step of extracting the pixel values near the RPE position and the CSI position for each column of the B-scan image; An interpolation step of performing interpolation and calculation according to the OCT propagation attenuation model by using the preset pixel values near the RPE position and the CSI position, and obtaining the initial choroidal vessel segmentation result by subtracting the actually observed B-scan image from the interpolation result.

2. The automatic segmentation method of OCT images according to claim 1, wherein Before the extraction step, there is also a construction step, including: constructing the light propagation attenuation model of the OCT.

3. The automatic OCT image segmentation method according to claim 1, characterized in that After the interpolation step, there is also a binarization step, including: binarizing the initial choroidal vessel segmentation result and performing morphological operations to obtain a choroidal vessel segmentation result with smooth edges, and generating an OCT-A image.

4. The automatic OCT image segmentation method according to claim 1, wherein The interpolation step is specifically implemented as: Interpolating the pixels at the positions between the RPE and the CSI for each column by using the pixel values near the RPE position and the CSI position; According to the OCT imaging model, fitting the propagation curve of light only in other choroidal tissues by using the pixel values near the RPE position and the CSI position, and interpolating the pixels between the RPE and the CSI for each column according to the pixel values of the fitted curve.

5. The OCT image automatic segmentation method according to claim 3, wherein The binarization step is specifically implemented as: Performing Gaussian smoothing processing on the initial choroidal vessel segmentation result; Binarizing the initial choroidal vessel segmentation result by using the Otsu algorithm to obtain an intermediate binarization result; Performing morphological operations on the intermediate binarization result to obtain a choroidal vessel segmentation result with smooth edges, and generating an OCT-A image by using the choroidal vessel segmentation result.

6. The automatic OCT image segmentation method according to claim 2, wherein Constructing the light propagation attenuation model of the OCT includes: Using the Beer-Lambert law to describe the law of attenuation during propagation in the choroidal tissue; Solving the backscattering parameters of light by taking the scattering part as a component of the incident light, regarding the choroidal vessels and the choroidal tissue as homogeneous media, and calculating the first absorption coefficient and the second absorption coefficient, as well as the first scattering coefficient and the second scattering coefficient to fit the attenuation curve; Using the first absorption coefficient and the first scattering coefficient to indicate the attenuation law of the regarded choroidal vessels; Using the second absorption coefficient and the second scattering coefficient to indicate the attenuation law of the choroidal tissue.

7. The automatic OCT image segmentation method according to claim 1, wherein Extracting the stratification line information of the retinal pigment epithelium (RPE) and the choroid-sclera interface (CSI) includes: Using the shortest path search algorithm to respectively extract the stratification line information of the RPE and the CSI.

8. A computing device, characterized in that, Comprising: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-7.

9. A readable storage medium, characterized in that, Stored with computer-executable instructions for executing the method according to any one of claims 1-7.

Images