A method for optimizing optical coherence tomography data for scar diagnosis

Abstract

The present invention relates to the field of image processing technology, and in particular to a method for optimizing optical coherence tomography data for scar diagnosis. The method first quantifies the clarity of multiple depth intervals based on the tissue distribution of the first OCT microscopic image data obtained by the first scanning parameter, and then determines the overall image clarity; then, the second scanning parameter determined according to the overall image clarity, the ratio of the area of ​​the non-scar region to the entire scar image area, and the wavelength of the first light source is used to obtain the original OCT microscopic image data of the scar again; the original OCT microscopic image data is preprocessed, and the multiple scar images in the obtained second OCT microscopic image data are fused to obtain an optimized image. This method can make the optimized image reflect the overall scar information more clearly and comprehensively, and the resolution of each position of the scar image is very clear, providing a basis for doctors to evaluate the nature, maturity and possible healing of the scar.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an optical coherence tomography data optimization method for scar diagnosis. Background Art

[0002] Scars are natural products of the human body's trauma repair process. When the skin or body tissue is damaged, such as surgical incisions, burns, cuts, abrasions or other types of trauma, the body will automatically start the repair mechanism. In this process, the damaged skin tissue will be replaced by new tissue, and the newly formed tissue often differs from the surrounding normal skin in color, texture and elasticity, forming a scar. Optical coherence tomography (OCT) is used to collect image data of the skin scar area, and the image data is analyzed to obtain the microstructure of the scar tissue, such as the thickness of the epidermis and dermis, blood vessel distribution, collagen fiber arrangement and tissue density. By analyzing the image data, doctors can evaluate the nature, maturity and possible healing of the scar to provide a basis for clinical diagnosis and treatment.

[0003] When scar images are generated by OCT equipment, due to the diversity of scars themselves, the thickness, pigmentation, and tissue distribution of scars at different locations on the same patient are different. The corresponding macroscopic images have obvious contrast with normal tissues. However, the acquired OCT images have poor image quality at the microscopic level because scar tissue has different density, texture, vascular distribution, and optical scattering characteristics, which affects the reliability of subsequent image analysis and diagnostic basis. Therefore, an image optimization method is urgently needed to obtain higher local quality image data. Summary of the invention

[0004] In order to solve the technical problem that the acquired image quality is poor and affects the reliability of subsequent image analysis and diagnosis basis, the purpose of the present invention is to provide an optical coherence tomography data optimization method for scar diagnosis. The technical scheme adopted is as follows:

[0005] Acquire first OCT microscopic image data of a scar through a first scanning parameter of an OCT device, wherein the first OCT microscopic image data includes a plurality of scar images, the first scanning parameter includes a first light source wavelength, a second scanning depth, and a third scanning speed, and the scar image includes a scar area and a non-scar area;

[0006] quantifying the clarity of multiple depth intervals of each scar image in the first OCT microscopic image data based on the tissue distribution of each scar image in the first OCT microscopic image data;

[0007] Determining the overall image clarity of each scar image in the first OCT microscopic image data based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data;

[0008] Determining second scanning parameters of the OCT device according to the overall image clarity of each scar image in the first OCT microscopic image data, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and the wavelength of the first light source, wherein the second scanning parameters include a plurality of second scanning wavelengths;

[0009] Based on the second scanning parameters of the OCT device, original OCT microscopic image data of the scar is acquired again;

[0010] Preprocessing the original OCT microscopic image data to obtain second OCT microscopic image data;

[0011] The multiple scar images in the second OCT microscopic image data are fused to obtain an optimized image.

[0012] In some embodiments, quantifying the clarity of multiple depth regions of each scar image in the first OCT microscopic image data based on the tissue distribution of each scar image in the first OCT microscopic image data includes:

[0013] Performing edge detection on each scar image in the first OCT microscopic image data to obtain a scar region image, wherein the scar region image includes a plurality of pixel points;

[0014] A segmentation window with a width of five is selected, and the segmentation window is translated to divide the scar area in each scar area image into a plurality of sub-scar areas, wherein the central axis of the segmentation window intersects with the pixel points of the plurality of sub-scar areas in a translation order, and the pixel points of the plurality of sub-scar areas intersecting with the central axis of the segmentation window in a translation order are the middle pixel points of the corresponding sub-scar areas, and the middle pixel points in each sub-scar area constitute a segmentation straight area, and the width of the segmentation straight area is one;

[0015] Based on the segmentation degree in each segmented straight region, determining a fault segmentation point of each segmented straight region, wherein the fault segmentation point divides each segmented straight region into an upper half and a lower half, and the fault segmentation point is a pixel point with the largest segmentation degree in each segmented straight region;

[0016] Performing spline interpolation processing on the fault segmentation points to determine a segmentation curve of the scar region image;

[0017] The segmentation curve is translated in a vertical direction to obtain a depth curve of the scar region image, wherein the depth curve divides the scar region into a plurality of depth regions;

[0018] According to the average of the ordered chaotic gradients, disordered chaotic gradients and edge gradients of multiple depth regions, the clarity of multiple depth regions is determined, and the clarity of multiple depth regions of each scar image in the first OCT microscopic image data is obtained, wherein the average of the edge gradients satisfies: edge detection is performed on multiple depth regions to obtain edge information within each depth region, the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is calculated, and the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is the average of the edge gradients of the depth region.

[0019] In some embodiments, determining the second scanning parameter of the OCT device according to the overall image clarity of each scar image in the first OCT microscopic image data, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and the wavelength of the first light source includes:

[0020] Determine the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image;

[0021] determining a weighting coefficient based on the ratio and the overall image clarity of each scar image in the first OCT microscopic image data;

[0022] Calculating the product of the first light source wavelength and the weighting coefficient to obtain an adjusted wavelength corresponding to each scar image in the first OCT microscopic image data;

[0023] Based on the adjusted wavelength corresponding to each scar image in the first OCT microscopic image data, a second scanning parameter of the OCT device is determined.

[0024] In some embodiments, the preprocessing of the original OCT microscopic image data to obtain the second OCT microscopic image data includes:

[0025] Performing corner point detection on multiple scar images in the original OCT microscopic image data to determine the corner points of the multiple scar images in the original OCT microscopic image data;

[0026] Corner points of multiple scar images in the original OCT microscopic image data are matched, and based on the matched corner points, multiple scar images in the original OCT microscopic image data are aligned to obtain second OCT microscopic image data.

[0027] In some embodiments, the fusing of the plurality of scar images in the second OCT microscopic image data to obtain an optimized image comprises:

[0028] determining the clarity of multiple depth regions of each scar image in the second OCT microscopic image data;

[0029] Based on the clarity of multiple depth regions of each scar image in the second OCT microscopic image data, weighted averaging processing is performed on each scar image in the second OCT microscopic image data to obtain an optimized image.

[0030] In some embodiments, the fault segmentation point of each sub-scar region is determined based on the segmentation degree in each sub-scar region according to the following formula:

[0031] ;

[0032] Among them, i represents the number of the cut straight areas. Indicates the segmentation degree corresponding to the a-th pixel in the i-th segmentation area. It represents the texture distribution difference of the upper part obtained by dividing the i-th straight cutting area by the a-th pixel point in the i-th straight cutting area. It represents the texture distribution difference of the lower half of the i-th straight region divided by the a-th pixel in the i-th sub-segmented straight region. represents the absolute value of the difference between the slope between the fault cutting point of the i-th straight region and the fault cutting point of the i-1th straight region and the slope between the fault cutting point of the i-th straight region and the fault cutting point of the i+1th straight region;

[0033] ;

[0034] ;

[0035] In the formula, It represents the texture distribution difference of the upper part of the i-th straight segmentation area divided by the horizontal line where the a-th pixel point in the i-th straight segmentation area is located. It represents the gray level co-occurrence matrix value of the pixel points in the upper half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na1 represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the upper part of the scar region corresponding to the i-th straight segmentation region divided by the horizontal line where the a-th pixel point in the i-th straight segmentation region is located, It represents the slope of the zth edge in the upper part of the scar region corresponding to the i-th straight cutting region divided by the horizontal line where the a-th pixel point in the i-th straight cutting region is located, and They respectively represent the mean length and slope of the edge in the upper part obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located; It represents the texture distribution difference of the lower half of the i-th straight segmentation area obtained by dividing the i-th straight segmentation area by the horizontal line where the a-th pixel point is located. It represents the gray level co-occurrence matrix value of the pixel points in the lower half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na2 represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the lower half of the scar region corresponding to the i-th cut straight region divided by the horizontal line where the a-th pixel point in the i-th cut straight region is located, It represents the slope of the zth edge in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located, and They respectively represent the mean value of the edge length and the mean value of the edge slope in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located.

[0036] In some embodiments, determining the clarity of multiple depth regions according to the average of the ordered chaotic gradient, the disordered chaotic gradient and the edge gradient of the multiple depth regions to obtain the clarity of multiple depth regions of each scar image in the first OCT microscopic image data includes:

[0037] Taking a pixel point in the depth region as the central pixel point, determining the eight neighborhoods of the pixel point, and calculating the gradient of the central pixel point in the eight neighborhood directions, the magnitude of the gradient is the difference in grayscale value between the central pixel point and the pixel points in the eight neighborhood directions of the central pixel point, and the direction of the gradient is the direction of the line connecting the central pixel point and the pixel points in the eight neighborhood directions of the central pixel point;

[0038] Taking the central pixel point as the center, the window determined by the pixel points in the eight neighborhood directions of the central pixel point is rotated by 90°, 180° and 270° in sequence, and the gradient of the eight neighborhood directions of the central pixel point at each angle is calculated respectively;

[0039] The ordered chaotic gradient of the pixel points in the depth region is calculated according to the following formula:

[0040] ;

[0041] In the formula, It represents the ordered chaotic gradient of the pixel points in the depth area, M represents the number of the eight-neighborhood directions of the central pixel point, and v represents the sequence number of the eight-neighborhood directions of the central pixel point. Represents the variance of the gradient change of the central pixel in the vth direction;

[0042] Calculate the ordered chaotic gradients of all pixels in the j-th depth region, and calculate the ordered chaotic gradients of multiple depth regions;

[0043] According to the maximum inter-class variance of the ordered chaotic gradients in multiple depth regions, the ordered chaotic gradients in the depth region are divided into ordered chaotic gradients and disordered chaotic gradients;

[0044] The clarity of multiple depth regions is determined based on the average of the ordered chaotic gradient, the disordered chaotic gradient and the edge gradient by the following formula, and the clarity of multiple depth regions of each scar image in the first OCT microscopic image data is obtained:

[0045] ;

[0046] In the formula, represents the clarity of the j-th depth region, j represents the sequence number of the depth region in each scar image of the first OCT microscopic image data, represents the number of ordered chaotic gradients, represents the number of disordered gradients, represents the mean value of the edge gradient in the jth depth region, Represents the normalization function.

[0047] In some embodiments, the overall image clarity of each scar image in the first OCT microscopic image data is determined based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data according to the following formula:

[0048] ;

[0049] In the formula, represents the overall image clarity of each scar image, S represents the number of depth regions in each scar image of the first OCT microscopic image data, Represents the number of pixels with non-zero grayscale values ​​in the jth depth region.

[0050] In some embodiments, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image is determined according to the following formula:

[0051] ;

[0052] In the formula, represents the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, SS represents the area of ​​the non-scar region of each scar image in the first OCT microscopic image data, and S represents the entire scar image area of ​​each scar image in the first OCT microscopic image data.

[0053] In some embodiments, a weighting coefficient is determined based on the ratio and the overall image clarity of each scar image in the first OCT microscopic image data by the following formula, and the product of the first light source wavelength and the weighting coefficient is calculated to obtain an adjusted wavelength corresponding to each scar image in the first OCT microscopic image data:

[0054] ;

[0055] Wherein, B represents the adjustment wavelength corresponding to each scar image in the first OCT microscopic image data, represents the wavelength of the first light source, represents the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and W represents the overall image clarity of each scar image.

[0056] The present invention has the following beneficial effects:

[0057] The present invention provides an optical coherence tomography data optimization method for scar diagnosis. The method first obtains first OCT microscopic image data of a scar through a first scanning parameter of an OCT device as a basis for optimization; then, based on the tissue distribution of each scar image in the first OCT microscopic image data, quantifies the clarity of multiple depth intervals of each scar image in the first OCT microscopic image data to grasp the clarity of each depth portion of the scar image; then, based on the clarity of multiple depth areas of each scar image in the first OCT microscopic image data, determines the overall image clarity of each scar image in the first OCT microscopic image data to grasp the overall clarity of the scar image, which is convenient for guiding parameter setting for the next scan; then, based on the overall image clarity of each scar image in the first OCT microscopic image data, the ... The second scanning parameter of the OCT device is determined based on the ratio of the area of ​​the non-scar region of the image to the area of ​​the entire scar image and the wavelength of the first light source; then based on the second scanning parameter of the OCT device, the original OCT microscopic image data of the scar is acquired again, and the second scanning parameter has multiple scanning wavelengths, which can perform targeted scanning according to the clarity of the overall image of each scar image in the first OCT microscopic image data to ensure the resolution of the scar image obtained in the next scan; then the original OCT microscopic image data is preprocessed to obtain the second OCT microscopic image data, so that the scar information and blank background information contained in each scar image in the second OCT microscopic image data are close to each other, so as to prevent interference with subsequent optimization; finally, the multiple scar images in the second OCT microscopic image data are fused to obtain an optimized image with clearer and more comprehensive scar information. This method can adjust and optimize the scanning parameters of the next OCT device based on the current scanning results to obtain higher local quality image data. The optimized image can reflect the overall scar information more clearly and comprehensively. The resolution of each position of the scar image is relatively clear, providing a basis for doctors to evaluate the nature, maturity and possible healing of the scar. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A method flow chart of an optical coherence tomography data optimization method for scar diagnosis provided by an embodiment of the present invention;

[0060] Figure 2A schematic diagram of the effect of scar images in the optical coherence tomography data optimization method for scar diagnosis provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] 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 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.

[0062] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0063] The specific scheme of the optical coherence tomography data optimization method for scar diagnosis provided by the present invention is described in detail below with reference to the accompanying drawings.

[0064] Figure 1 This is a flow chart of the method for optimizing the optical coherence tomography data for scar diagnosis provided by an embodiment of the present invention. Figure 1 , the method comprises the following steps:

[0065] Step 101, obtaining first OCT microscopic image data of a scar through a first scanning parameter of an OCT device, wherein the first OCT microscopic image data includes multiple scar images, the first scanning parameter includes a first light source wavelength, a second scanning depth and a third scanning speed, and the scar image includes a scar area and a non-scar area.

[0066] The first OCT microscopic image data acquired by the OCT device includes multiple scar images, such as Figure 2 As shown, each scar image includes a scar area and a non-scar area. The non-scar area includes relevant information of the dermis of normal skin. The collagen fibers of normal skin are arranged in an orderly and relatively uniform manner. The scar area includes relevant information of the dermis of the scar tissue, which is different from the imaging of normal skin. The dermis of the scar tissue may have irregular texture and structure, and the thickness of the scar tissue at different locations is different. The collagen fibers of the scar may present an uneven scattering pattern. Therefore, the image clarity of the scar tissue at different locations in the scar area is different, that is, the resolution is different.

[0067] When the OCT device collects scar images, the scanning parameters of the OCT device need to be adjusted. The collected image information is adjusted by adjusting the wavelength of the light source, the scanning depth and the scanning speed. Shorter wavelengths provide higher resolution, but weaker penetration; longer wavelengths have stronger penetration, but lower resolution. When collecting scar image information, due to the different hyperplasia of the scar itself, the scar has different image information at different locations of its thickness, so the fixed light source wavelength cannot meet the imaging resolution requirements of different scar areas. Short-wavelength light sources can obtain clearer images, but internal information cannot be obtained in areas with thicker scars. Long-wavelength light sources have a large penetration depth, but the overall resolution of the collected images is poor, which affects the subsequent doctor's judgment. In order to solve the above problems, the scanning parameters of the next OCT device are adjusted and optimized through the current scanning results to obtain higher local quality image data, so that the scar information is more comprehensive and clear.

[0068] The OCT device scans the same scar tissue multiple times to obtain sufficient scar images, so the first OCT microscopic image data includes multiple scar images.

[0069] In some embodiments, in order to ensure the quality of the first OCT microscopic image data, the scar area needs to be cleaned before scanning to remove external substances that may affect the imaging, such as grease, cosmetics, etc.; after turning on the OCT device, preheating and self-test are required, and the device is calibrated to ensure the alignment of the optical elements and the stability of the light source, so as to ensure the subsequent stable operation of the device.

[0070] In some embodiments, the first scanning parameters including the first light source wavelength, the second scanning depth and the third scanning speed may be set according to the nature and thickness of the scar to be scanned.

[0071] In some embodiments, during scanning, the probe of the OCT device needs to be aimed at the scar area to ensure that the probe is perpendicular to the skin surface.

[0072] Step 102: quantify the clarity of each scar image in the first OCT microscopic image data in multiple depth intervals based on the tissue distribution of each scar image in the first OCT microscopic image data.

[0073] By quantifying the clarity of multiple depth intervals of each scar image in the first OCT microscopic image data, the scar image can be divided into multiple depth intervals, and the clarity of different positions of the scar image can be obtained according to the stratification of normal skin and scar tissue in the multiple depth intervals.

[0074] In some embodiments, step 102 may specifically include:

[0075] Step 1021 , performing edge detection on each scar image in the first OCT microscopic image data to obtain a scar region image, wherein the scar region image includes a plurality of pixel points.

[0076] The scar area image presents relevant information of the scar tissue, preparing for the subsequent grasp of the clarity of the scar image.

[0077] In some embodiments, the edge detection may be Canny edge detection.

[0078] Step 1022, select a segmentation window with a width of five, translate the segmentation window, and divide the scar area in each scar area image into multiple sub-scar areas, the central axis of the segmentation window intersects with the pixel points of the multiple sub-scar areas in a translation order, the pixel points of the multiple sub-scar areas intersecting with the central axis of the segmentation window in a translation order are the middle pixel points of the corresponding sub-scar areas, the middle pixel points in each sub-scar area constitute a segmentation straight area, and the width of the segmentation straight area is one.

[0079] In step 1022, the scar region is divided into a plurality of sub-scar regions in the horizontal direction, and the divided straight regions represent the corresponding sub-scar regions for subsequent layered judgment.

[0080] Step 1023, based on the segmentation degree in each segmented straight region, determine the fault segmentation point of each segmented straight region, the fault segmentation point divides each segmented straight region into an upper half and a lower half, and the fault segmentation point is the pixel point with the largest segmentation degree in each segmented straight region.

[0081] The scar area in each straight cutting area may be different, and stratification may occur. Based on the degree of segmentation in each straight cutting area, that is, the stratification, the fault cutting point of each straight cutting area is determined to be where the difference occurs, which best reflects the boundary of the stratification.

[0082] In some embodiments, the fault segmentation point of each sub-scar region is determined based on the segmentation degree in each sub-scar region according to the following formula:

[0083] ;

[0084] Among them, i represents the number of the cut straight areas. Indicates the segmentation degree corresponding to the a-th pixel in the i-th segmentation area. It represents the texture distribution difference of the upper part obtained by dividing the i-th straight cutting area by the a-th pixel point in the i-th straight cutting area. It represents the texture distribution difference of the lower half of the i-th straight region divided by the a-th pixel in the i-th sub-segmented straight region. It represents the absolute value of the difference between the slope between the fault cutting point of the i-th straight cutting area and the fault cutting point of the i-1th straight cutting area and the slope between the fault cutting point of the i-th straight cutting area and the fault cutting point of the i+1th straight cutting area. When the texture distribution difference of three adjacent straight cutting areas is similar, the difference between the slope between the fault cutting point of the i-th straight cutting area and the fault cutting point of the i-1th straight cutting area and the slope between the fault cutting point of the i-th straight cutting area and the fault cutting point of the i+1th straight cutting area is 0. There may be a situation where the denominator is 0. In order to avoid the denominator being 0, use Perform anti-zero adjustment;

[0085] ;

[0086] ;

[0087] In the formula, It represents the texture distribution difference of the upper part of the i-th straight segmentation area divided by the horizontal line where the a-th pixel point in the i-th straight segmentation area is located. It represents the gray level co-occurrence matrix value of the pixel points in the upper half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na1 represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the upper part of the scar region corresponding to the i-th straight segmentation region divided by the horizontal line where the a-th pixel point in the i-th straight segmentation region is located, It represents the slope of the zth edge in the upper part of the scar region corresponding to the i-th straight cutting region divided by the horizontal line where the a-th pixel point in the i-th straight cutting region is located, and They respectively represent the mean length and slope of the edge in the upper part obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located; It represents the texture distribution difference of the lower half of the i-th straight segmentation area obtained by dividing the i-th straight segmentation area by the horizontal line where the a-th pixel point is located. It represents the gray level co-occurrence matrix value of the pixel points in the lower half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na2 represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the lower half of the scar region corresponding to the i-th cut straight region divided by the horizontal line where the a-th pixel point in the i-th cut straight region is located, It represents the slope of the zth edge in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located, and They respectively represent the mean value of the edge length and the mean value of the edge slope in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located.

[0088] The above determination of the fault segmentation point of each sub-scar region is an iterative method to find the most suitable fault segmentation point for each sub-scar region from the pixel points of each sub-scar region, that is, the segmentation degree corresponding to the a-th pixel point in the i-th segmentation region The point with the largest value. The segmentation degree corresponding to the a-th pixel point in the i-th segmentation area It represents the segmentation ability of the ath pixel in the ith segmentation region to segment the ith segmentation region into two parts with obvious texture distribution differences, The larger the value, the stronger the segmentation ability; The absolute value of the difference between the texture distribution difference of the upper part and the texture distribution difference of the lower part obtained by dividing the i-th segmentation straight area by the a-th pixel point in the i-th segmentation straight area indicates the obvious degree of stratification between the upper and lower parts. The larger the value, the more obvious the stratification; Reflects the change in the slope of the fault dividing point in the two dividing straight areas before and after the i-th dividing straight area. The larger the value, the greater the slope change. Adjustment Thus, the segmentation degree corresponding to the ath pixel point in the ith straight cutting area is obtained by changing the slope of the fault cutting point in the two cutting straight cutting areas before and after the ith straight cutting area to the obvious degree of stratification of the upper and lower parts of the ith straight cutting area. .

[0089] The formula can be used to find and , and Respectively reflect the gray level co-occurrence matrix values ​​of the pixels in the upper and lower parts of the scar area corresponding to the i-th cut straight area divided by the horizontal line where the a-th pixel point in the i-th cut straight area is located, The larger the value, the more obvious the texture features of the upper part are. The larger the value, the more obvious the texture features of the lower part are. and They respectively reflect the size of the difference between the upper and lower parts of the sub-scar area corresponding to the i-th cut straight area divided by the horizontal line where the a-th pixel point in the i-th cut straight area is located, that is, the size of the texture fluctuation. and The larger the value, the greater the difference. and Respectively reflect the cosine value in the horizontal direction of the difference between the upper and lower parts of the sub-scar area corresponding to the i-th cut straight area divided by the horizontal line where the a-th pixel point is located, and The larger the value, the more obvious the difference is. and Used to evaluate the difference between the upper and lower parts. and The larger the value, the greater the difference. and Reflects the average difference between the upper and lower parts. and Indicates the average level of the difference between the upper and lower parts. and Reflects the average of the difference characteristics between the upper and lower parts, and Respectively represent the sum of the differences between the differences in the upper and lower parts and the average differences in each part, which can reflect the difference between the differences in each part and the average difference level, respectively. and Multiplying together, the overall texture distribution difference between the upper and lower parts is reflected.

[0090] In some embodiments, by finding the segmentation degree corresponding to the a-th pixel point in the i-th segmentation straight area The point with the largest value is used to determine the fault segmentation point of each sub-scar area.

[0091] Step 1024: Perform spline interpolation processing on the fault segmentation points to determine the segmentation curve of the scar region image.

[0092] Spline interpolation processing can reasonably supplement the number of fault segmentation points, which is convenient for determining the segmentation curve of the scar area image, thereby connecting the differences in the appearance of various parts of the scar area and realizing the preliminary stratification of the scar area.

[0093] Step 1025 , performing translation processing on the segmentation curve in the vertical direction to obtain a depth curve of the scar region image, wherein the depth curve segments the scar region into a plurality of depth regions.

[0094] The depth curve divides the scar area into multiple depth areas, thereby completing all stratification of the scar area.

[0095] Step 1026, determining the clarity of multiple depth regions based on the average of the ordered chaotic gradients, the disordered chaotic gradients and the edge gradients of the multiple depth regions, and obtaining the clarity of multiple depth regions of each scar image in the first OCT microscopic image data, wherein the average of the edge gradients satisfies: edge detection is performed on multiple depth regions to obtain edge information within each depth region, and the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is calculated; the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is the average of the edge gradients of the depth region.

[0096] The ordered chaotic gradient can reflect the effective information of scar tissue, the disordered chaotic gradient can reflect the noise in the scar image, and the mean of the edge gradient can reflect the average level of differences in the scar image. According to the mean of the ordered chaotic gradient, the disordered chaotic gradient and the edge gradient in multiple depth areas, the clarity of multiple depth areas of each scar image in the first OCT microscopic image data is obtained, thereby quantifying the clarity of different depth areas of the scar image.

[0097] In some embodiments, step 1026 may specifically include:

[0098] Step 10261, taking the pixel point in the depth area as the center pixel point, determine the eight-neighborhood of the pixel point, and calculate the gradient of the center pixel point in the direction of the eight-neighborhood. The magnitude of the gradient is the difference in grayscale value between the center pixel point and the pixel points in the direction of the eight-neighborhood of the center pixel point, and the direction of the gradient is the direction of the line connecting the center pixel point and the pixel points in the direction of the eight-neighborhood of the center pixel point.

[0099] The gradient of the eight-neighborhood direction of a pixel can reflect the texture change near the pixel.

[0100] Step 10262, with the central pixel as the center, rotate the window determined by the pixels in the eight-neighborhood direction of the central pixel by 90°, 180°, and 270° in sequence, and calculate the gradient of the eight-neighborhood direction of the central pixel at each angle respectively.

[0101] The gradients in the eight-neighborhood directions of the pixel points at multiple angles are obtained to more fully and comprehensively grasp the texture changes near the pixel points, which is convenient for the subsequent division of effective information and noise.

[0102] Step 10263, calculate the ordered chaotic gradient of the pixel points in the depth region according to the following formula:

[0103] ;

[0104] In the formula, It represents the ordered chaotic gradient of the pixel points in the depth area, M represents the number of the eight-neighborhood directions of the central pixel point, and v represents the sequence number of the eight-neighborhood directions of the central pixel point. Represents the variance of the gradient change of the central pixel in the vth direction.

[0105] is the gradient change in the eight neighborhood directions of the pixel point, Reflects the fluctuation of the gradient changes in v directions, Reflects the fluctuation of the overall gradient change, that is, the ordered chaotic gradient of the pixels in the depth area , so as to obtain effective information about scar tissue.

[0106] Step 10264, calculate the ordered chaotic gradients of all pixels in the j-th depth region, and calculate the ordered chaotic gradients of multiple depth regions.

[0107] The ordered chaotic gradients of multiple depth regions are calculated to obtain the effective information of scar tissue in the entire scar image.

[0108] Step 10265, according to the maximum inter-class variance of the ordered chaotic gradients in multiple depth regions, the ordered chaotic gradients of the depth region are divided into ordered chaotic gradients and disordered chaotic gradients.

[0109] Dividing the ordered chaotic gradient and the disordered chaotic gradient by the maximum inter-class variance can ensure the effective information of scar tissue and the accurate division of noise.

[0110] Step 10266, determining the clarity of multiple depth regions based on the average of the ordered chaotic gradient, the disordered chaotic gradient and the edge gradient by the following formula, and obtaining the clarity of multiple depth regions of each scar image in the first OCT microscopic image data:

[0111] ;

[0112] In the formula, represents the clarity of the j-th depth region, j represents the sequence number of the depth region in each scar image of the first OCT microscopic image data, represents the number of ordered chaotic gradients, represents the number of disordered gradients, represents the mean value of the edge gradient in the jth depth region, Represents the normalization function.

[0113] Reflects the clarity of different depth areas of scar images. The larger the scar, the clearer the image will be. Reflects the average level of differences in scar images, Represents the number of ordered chaotic gradients, which is obtained by dividing the ordered chaotic gradients and the disordered chaotic gradients. Reflects the proportional relationship between the number of ordered chaotic gradients and the number of disordered chaotic gradients. The number of ordered chaotic gradients is used to regulate , and finally the clarity of multiple depth regions is obtained through normalization.

[0114] Step 103: determining the overall image clarity of each scar image in the first OCT microscopic image data based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data.

[0115] The clarity of multiple depth areas obtained needs to be fused to reflect the overall image clarity.

[0116] In some embodiments, the overall image clarity of each scar image in the first OCT microscopic image data is determined based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data according to the following formula:

[0117] ;

[0118] In the formula, represents the overall image clarity of each scar image, S represents the number of depth regions in each scar image of the first OCT microscopic image data, Represents the number of pixels with non-zero grayscale values ​​in the jth depth region.

[0119] The quantitative result reflecting the clarity of the scar image, S is the number of depth regions in each scar image of the first OCT microscopic image data, is the number of pixels with non-zero grayscale values ​​in the j-th depth region. A non-zero grayscale value indicates that the pixel has information rather than a blank background. It is the clarity of the j-th depth area after adjusting the number of pixels with non-zero grayscale values, which can more accurately reflect the clarity of the j-th depth area. is the result of the clarity fusion of S depth regions of the scar image, that is, the quantitative result reflecting the clarity of the scar image .

[0120] Step 104, determining second scanning parameters of the OCT device according to the overall image clarity of each scar image in the first OCT microscopic image data, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and the wavelength of the first light source, wherein the second scanning parameters include multiple second scanning wavelengths.

[0121] The clarity of the scar image obtained by the current scan is used to adjust the wavelength of the first light source to obtain the second scanning wavelength. The next targeted scan can be performed according to the clarity of the overall image of each scar image in the first OCT microscopic image data to ensure the resolution of the scar image obtained by the next scan.

[0122] In some embodiments, step 104 may specifically include:

[0123] Step 1041 , determining the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image.

[0124] The ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image is used to adjust the wavelength of the first light source.

[0125] In some embodiments, step 1041 is implemented according to the following formula:

[0126] ;

[0127] In the formula, It represents the ratio of the non-scar area of ​​each scar image in the first OCT microscopic image data to the entire scar image area, reflecting the overall proportion of the non-scar area. SS represents the non-scar area of ​​each scar image in the first OCT microscopic image data, S represents the entire scar image area of ​​each scar image in the first OCT microscopic image data, and 0.6 is used to adjust the ratio. Finally, the overall proportion of the non-scar area is obtained. .

[0128] Step 1042: Determine a weighting coefficient based on the ratio and the overall image clarity of each scar image in the first OCT microscopic image data. The weighting coefficient is a key coefficient for adjusting the wavelength of the first light source.

[0129] Step 1043: Calculate the product of the wavelength of the first light source and the weighting coefficient to obtain an adjusted wavelength corresponding to each scar image in the first OCT microscopic image data.

[0130] In some embodiments, step 1042 and step 1043 are implemented by the following formula:

[0131] ;

[0132] Wherein, B represents the adjusted wavelength corresponding to each scar image in the first OCT microscopic image data, that is, the wavelength of the light source that can be used for the next scan, represents the wavelength of the first light source, is the weighting coefficient for adjusting the wavelength of the first light source, represents the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image. It may be positive or negative, so use Change The symbol before , W represents the overall image clarity of each scar image.

[0133] Step 1044, based on the adjusted wavelength corresponding to each scar image in the first OCT microscopic image data, determine the second scanning parameter of the OCT device to facilitate the next scan.

[0134] Step 105: acquiring original OCT microscopic image data of the scar again based on the second scanning parameters of the OCT device.

[0135] Each scar image in the obtained original OCT microscopic image data can more clearly reflect the scar information at a certain position.

[0136] Step 106: preprocess the original OCT microscopic image data to obtain second OCT microscopic image data.

[0137] Preprocessing can make the scar information and blank background information contained in each scar image in the second OCT microscopic image data close to the same, thereby preventing interference with subsequent optimization.

[0138] In some embodiments, the specific implementation of step 106 includes:

[0139] Step 1061 , performing corner point detection on a plurality of scar images in the original OCT microscopic image data, and determining the corner points of the plurality of scar images in the original OCT microscopic image data.

[0140] Corner point detection can determine the location of the scar area, which is convenient for subsequent alignment processing.

[0141] Step 1062, matching the corner points of the multiple scar images in the original OCT microscopic image data, and aligning the multiple scar images in the original OCT microscopic image data based on the matched corner points to obtain second OCT microscopic image data.

[0142] Based on the matched corner points, multiple scar images in the original OCT microscopic image data are aligned, so that the scar information and blank background information contained in each scar image in the second OCT microscopic image data are close to the same, preventing interference with subsequent optimization and facilitating subsequent fusion processing.

[0143] Step 107: fusing the multiple scar images in the second OCT microscopic image data to obtain an optimized image.

[0144] The optimized images can reflect the overall scar information more clearly and comprehensively, and the resolution of each location of the scar image is clear. The optimized images can provide a basis for doctors to evaluate the nature, maturity and possible healing of scars.

[0145] In some embodiments, step 107 may specifically include:

[0146] Step 1071, determine the clarity of multiple depth areas of each scar image in the second OCT microscopic image data to provide weights for subsequent weighted averaging processing. The weighting coefficients of different depth areas of each scar image are different, so that the scar information at each position of the scar image can be improved. The higher the clarity of a certain depth area in different scar images, the better the performance and the greater the weight.

[0147] Step 1072, based on the clarity of multiple depth areas of each scar image in the second OCT microscopic image data, weighted averaging processing is performed on each scar image in the second OCT microscopic image data to obtain an optimized image, thereby more clearly and comprehensively reflecting the scar information at different locations and improving the local image quality.

[0148] In summary, the optical coherence tomography data optimization method for scar diagnosis provided in an embodiment of the present invention can adjust and optimize the scanning parameters of the next OCT device according to the current scanning results to obtain image data with higher local quality. The optimized image can reflect the overall scar information more clearly and comprehensively, and the resolution of each position of the scar image is relatively clear, providing a basis for doctors to evaluate the nature, maturity and possible healing of the scar.

[0149] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0150] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for optimizing optical coherence tomography data for scar diagnosis, characterized in that: The method comprises: Acquire first OCT microscopic image data of a scar through a first scanning parameter of an OCT device, wherein the first OCT microscopic image data includes a plurality of scar images, the first scanning parameter includes a first light source wavelength, a second scanning depth, and a third scanning speed, and the scar image includes a scar area and a non-scar area; quantifying the clarity of multiple depth intervals of each scar image in the first OCT microscopic image data based on the tissue distribution of each scar image in the first OCT microscopic image data; Determining the overall image clarity of each scar image in the first OCT microscopic image data based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data; Determining second scanning parameters of the OCT device according to the overall image clarity of each scar image in the first OCT microscopic image data, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and the wavelength of the first light source, wherein the second scanning parameters include a plurality of second scanning wavelengths; Based on the second scanning parameters of the OCT device, original OCT microscopic image data of the scar is acquired again; Preprocessing the original OCT microscopic image data to obtain second OCT microscopic image data; The multiple scar images in the second OCT microscopic image data are fused to obtain an optimized image.

2. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 1, characterized in that: The step of quantifying the clarity of multiple depth regions of each scar image in the first OCT microscopic image data based on the tissue distribution of each scar image in the first OCT microscopic image data comprises: Performing edge detection on each scar image in the first OCT microscopic image data to obtain a scar region image, wherein the scar region image includes a plurality of pixel points; A segmentation window with a width of five is selected, and the segmentation window is translated to divide the scar area in each scar area image into a plurality of sub-scar areas, wherein the central axis of the segmentation window intersects with the pixel points of the plurality of sub-scar areas in a translation order, and the pixel points of the plurality of sub-scar areas intersecting with the central axis of the segmentation window in a translation order are the middle pixel points of the corresponding sub-scar areas, and the middle pixel points in each sub-scar area constitute a segmentation straight area, and the width of the segmentation straight area is one; Based on the segmentation degree in each segmented straight region, determining a fault segmentation point of each segmented straight region, wherein the fault segmentation point divides each segmented straight region into an upper half and a lower half, and the fault segmentation point is a pixel point with the largest segmentation degree in each segmented straight region; Performing spline interpolation processing on the fault segmentation points to determine a segmentation curve of the scar region image; The segmentation curve is translated in a vertical direction to obtain a depth curve of the scar region image, wherein the depth curve divides the scar region into a plurality of depth regions; According to the average of the ordered chaotic gradients, disordered chaotic gradients and edge gradients of multiple depth regions, the clarity of multiple depth regions is determined, and the clarity of multiple depth regions of each scar image in the first OCT microscopic image data is obtained, wherein the average of the edge gradients satisfies: edge detection is performed on multiple depth regions to obtain edge information within each depth region, the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is calculated, and the average of the gradient amplitudes between the edge pixels and the non-edge pixels within each depth region is the average of the edge gradients of the depth region.

3. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 1, characterized in that: Determining the second scanning parameter of the OCT device according to the overall image clarity of each scar image in the first OCT microscopic image data, the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and the wavelength of the first light source includes: Determine the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image; determining a weighting coefficient based on the ratio and the overall image clarity of each scar image in the first OCT microscopic image data; Calculating the product of the first light source wavelength and the weighting coefficient to obtain an adjusted wavelength corresponding to each scar image in the first OCT microscopic image data; Based on the adjusted wavelength corresponding to each scar image in the first OCT microscopic image data, a second scanning parameter of the OCT device is determined.

4. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 2, characterized in that: The preprocessing of the original OCT microscopic image data to obtain the second OCT microscopic image data includes: Performing corner point detection on multiple scar images in the original OCT microscopic image data to determine the corner points of the multiple scar images in the original OCT microscopic image data; Corner points of multiple scar images in the original OCT microscopic image data are matched, and based on the matched corner points, multiple scar images in the original OCT microscopic image data are aligned to obtain second OCT microscopic image data.

5. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 2 or 4, characterized in that: The fusing of the plurality of scar images in the second OCT microscopic image data to obtain an optimized image comprises: determining the clarity of multiple depth regions of each scar image in the second OCT microscopic image data; Based on the clarity of multiple depth regions of each scar image in the second OCT microscopic image data, weighted averaging processing is performed on each scar image in the second OCT microscopic image data to obtain an optimized image.

6. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 2, characterized in that: According to the following formula, the fault segmentation point of each sub-scar area is determined based on the segmentation degree in each sub-scar area: ； Among them, i represents the number of the cut straight areas. Indicates the segmentation degree corresponding to the a-th pixel in the i-th segmentation area. It represents the texture distribution difference of the upper part obtained by dividing the i-th straight cutting area by the a-th pixel point in the i-th straight cutting area. It represents the texture distribution difference of the lower half of the i-th straight region divided by the a-th pixel in the i-th sub-segmented straight region. represents the absolute value of the difference between the slope between the fault cutting point of the i-th straight region and the fault cutting point of the i-1th straight region and the slope between the fault cutting point of the i-th straight region and the fault cutting point of the i+1th straight region; ； ； In the formula, It represents the texture distribution difference of the upper part of the i-th straight segmentation area divided by the horizontal line where the a-th pixel point in the i-th straight segmentation area is located. It represents the gray level co-occurrence matrix value of the pixel points in the upper half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na1 represents the number of edges in the upper half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the upper part of the scar region corresponding to the i-th straight segmentation region divided by the horizontal line where the a-th pixel point in the i-th straight segmentation region is located, It represents the slope of the zth edge in the upper part of the scar region corresponding to the i-th straight cutting region divided by the horizontal line where the a-th pixel point in the i-th straight cutting region is located, and They respectively represent the mean length and slope of the edge in the upper part obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located; It represents the texture distribution difference of the lower half of the i-th straight segmentation area obtained by dividing the i-th straight segmentation area by the horizontal line where the a-th pixel point is located. It represents the gray level co-occurrence matrix value of the pixel points in the lower half obtained by dividing the sub-scar area corresponding to the i-th cut straight area by the horizontal line where the a-th pixel point in the i-th cut straight area is located, represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, Na2 represents the number of edges in the lower half obtained by dividing the sub-scar region corresponding to the ith cut straight region by the horizontal line where the ath pixel point in the ith cut straight region is located, represents the length of the zth edge in the lower half of the scar region corresponding to the i-th cut straight region divided by the horizontal line where the a-th pixel point in the i-th cut straight region is located, It represents the slope of the zth edge in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located, and They respectively represent the mean value of the edge length and the mean value of the edge slope in the lower half of the sub-scar area corresponding to the ith cut straight area divided by the horizontal line where the ath pixel point in the ith cut straight area is located.

7. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 6, characterized in that: The method of determining the clarity of multiple depth regions according to the average values ​​of the ordered chaotic gradients, the disordered chaotic gradients and the edge gradients of the multiple depth regions to obtain the clarity of multiple depth regions of each scar image in the first OCT microscopic image data includes: Taking a pixel point in the depth region as the central pixel point, determining the eight neighborhoods of the pixel point, and calculating the gradient of the central pixel point in the eight neighborhood directions, the magnitude of the gradient is the difference in grayscale value between the central pixel point and the pixel points in the eight neighborhood directions of the central pixel point, and the direction of the gradient is the direction of the line connecting the central pixel point and the pixel points in the eight neighborhood directions of the central pixel point; Taking the central pixel point as the center, the window determined by the pixel points in the eight neighborhood directions of the central pixel point is rotated by 90°, 180° and 270° in sequence, and the gradient of the eight neighborhood directions of the central pixel point at each angle is calculated respectively; The ordered chaotic gradient of the pixel points in the depth region is calculated according to the following formula: ； In the formula, It represents the ordered chaotic gradient of the pixel points in the depth area, M represents the number of the eight-neighborhood directions of the central pixel point, and v represents the sequence number of the eight-neighborhood directions of the central pixel point. Represents the variance of the gradient change of the central pixel in the vth direction; Calculate the ordered chaotic gradients of all pixels in the j-th depth region, and calculate the ordered chaotic gradients of multiple depth regions; According to the maximum inter-class variance of the ordered chaotic gradients in multiple depth regions, the ordered chaotic gradients in the depth region are divided into ordered chaotic gradients and disordered chaotic gradients; The clarity of multiple depth regions is determined based on the average of the ordered chaotic gradient, the disordered chaotic gradient and the edge gradient by the following formula, and the clarity of multiple depth regions of each scar image in the first OCT microscopic image data is obtained: ； In the formula, represents the clarity of the j-th depth region, j represents the sequence number of the depth region in each scar image of the first OCT microscopic image data, represents the number of ordered chaotic gradients, represents the number of disordered gradients, represents the mean value of the edge gradient in the jth depth region, Represents the normalization function.

8. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 7, characterized in that: The overall image clarity of each scar image in the first OCT microscopic image data is determined based on the clarity of multiple depth regions of each scar image in the first OCT microscopic image data according to the following formula: ； In the formula, represents the overall image clarity of each scar image, S represents the number of depth regions in each scar image of the first OCT microscopic image data, Represents the number of pixels with non-zero grayscale values ​​in the jth depth region.

9. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 3 or 8, characterized in that: The ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image is determined according to the following formula: ； In the formula, represents the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, SS represents the area of ​​the non-scar region of each scar image in the first OCT microscopic image data, and S represents the entire scar image area of ​​each scar image in the first OCT microscopic image data.

10. The method for optimizing optical coherence tomography data for scar diagnosis according to claim 3 or 8, characterized in that: The weighting coefficient is determined based on the ratio and the overall image clarity of each scar image in the first OCT microscopic image data by the following formula, and the product of the first light source wavelength and the weighting coefficient is calculated to obtain the adjusted wavelength corresponding to each scar image in the first OCT microscopic image data: ； Wherein, B represents the adjustment wavelength corresponding to each scar image in the first OCT microscopic image data, represents the wavelength of the first light source, represents the ratio of the area of ​​the non-scar region of each scar image in the first OCT microscopic image data to the area of ​​the entire scar image, and W represents the overall image clarity of each scar image.

Images