Chloasma diagnosis and treatment effect evaluation system based on image feature analysis

Abstract

The present invention relates to the field of image feature analysis technology, and more specifically, to a chloasma diagnosis and treatment effect evaluation system based on image feature analysis. It includes a spot segmentation unit, a spot area unit, and a score adjustment unit. The evaluation and adjustment module of the present invention fills the deducted uncrushed melanin granule area image into the chloasma area image after laser treatment, compares it through LAB color values, and re-scores the ultra-picosecond laser diagnosis and treatment of chloasma. The effect of the re-scored ultra-picosecond laser diagnosis and treatment of chloasma is evaluated and the ultra-picosecond laser parameters are adjusted. By filling and comparing the uncrushed melanin granule area image, the color change of the entire chloasma area can be comprehensively considered, thereby more accurately evaluating the actual effect of ultra-picosecond laser treatment of chloasma, improving the accuracy of evaluating the comprehensive chloasma diagnosis and treatment effect, and adjusting the ultra-picosecond laser parameters according to the score, reducing the adverse reaction efficiency of the patient's skin and the excessive waste of ultra-picosecond laser resources.

Description

Technical Field

[0001] The present invention relates to the technical field of image feature analysis, and in particular to a chloasma diagnosis and treatment effect evaluation system based on image feature analysis. Background Art

[0002] Melasma is a common acquired pigmentation skin disease that primarily occurs on the face. Traditionally, doctors rely on visual observation and subjective experience to judge the effectiveness of melasma treatment. This is due to differences in evaluation criteria and visual sensitivity among different doctors, which can lead to inconsistent evaluation results, lack of accuracy, and lack of comparability. Furthermore, it is difficult to accurately quantify the color, area, and depth of melasma using the naked eye, and subtle changes are difficult to detect, making it impossible to provide accurate data support for adjusting treatment plans. Therefore, computer vision and image processing technologies are used to analyze the image features of the melasma area to provide clinicians with an objective and quantitative diagnosis and treatment effect evaluation tool.

[0003] When using an ultra-picosecond laser to scan the patient's melasma area, since traditional existing technologies evaluate the effectiveness of melasma diagnosis and treatment by comparing the pixel values ​​and area sizes of the melasma area images after scanning with those before scanning, the data is relatively single. Relying solely on pixel values ​​and area sizes cannot fully reflect the subtle changes in the color of melasma. For example, the color of melasma may change from dark brown to light brown, but the change in pixel values ​​is difficult to accurately reflect the quality of this color transition. Information such as hue and saturation is ignored, which may lead to inaccurate judgment of the treatment effect, making the evaluation of the actual effect of ultra-picosecond laser treatment of melasma incomplete or inaccurate, reducing the accuracy of evaluating the overall melasma diagnosis and treatment effect. Therefore, we provide a melasma diagnosis and treatment effect evaluation system based on image feature analysis. Summary of the Invention

[0004] The purpose of the present invention is to provide a chloasma diagnosis and treatment effect evaluation system based on image feature analysis to solve the problems raised in the above background technology:

[0005] Relying solely on pixel values ​​and area size cannot fully reflect the subtle changes in melasma's color, resulting in inaccurate judgments on treatment effectiveness and reducing the accuracy of comprehensive evaluation of melasma treatment results. Therefore, this case uses segmented 3D coordinate points, cutouts, and infill images to improve the accuracy of comprehensive evaluation of melasma treatment results.

[0006] To achieve the above-mentioned object, the present invention provides a chloasma diagnosis and treatment effect evaluation system based on image feature analysis, comprising a spot segmentation unit, a spot area unit and a score adjustment unit;

[0007] The spot segmentation unit obtains the patient's facial image data to analyze the RGB value of the three-dimensional coordinate point, converts it into a color value in the LAB color space, and calculates an integer threshold to determine whether the three-dimensional coordinate point is marked as a chloasma seed point;

[0008] The spot region unit is configured to receive the three-dimensional coordinates of the chloasma seed points in the segmented spot unit, connect the three-dimensional coordinates of the chloasma seed points to form an original chloasma region, scan the epidermis of the original chloasma region with an ultra-picosecond laser, obtain the three-dimensional coordinates of the whitish chloasma seed points and the non-whitish chloasma seed points, connect them respectively, and then extract the original chloasma region and the region where the melanin particles are not broken;

[0009] The scoring adjustment unit cuts out the image of the chloasma and the broken melanin granule area after laser treatment, and depicts the image of the non-whitened area after laser treatment, fills the image of the broken melanin granule area after laser treatment into the cut-out image of the original chloasma area, and then fills the cut-out image of the non-broken melanin granule area into the image of the chloasma area after laser treatment, compares the LAB color values, and scores and evaluates the effects of ultra-picosecond laser diagnosis and treatment of chloasma.

[0010] As a further improvement of this technical solution, the score adjustment unit includes a score evaluation module and an evaluation adjustment module;

[0011] The scoring and evaluation module obtains the patient's facial image data after laser treatment again, extracts the post-laser chloasma area image at the same coordinate position as the original chloasma area from the patient's post-laser facial image data, and then extracts the post-laser shattered melanin granule area image at the same coordinate position as the shattered melanin granule area from the post-laser chloasma area image, depicts it in the post-laser chloasma area image according to the coordinate position of the non-whitened area, fills the post-laser shattered melanin granule area image into the extracted original chloasma area image, compares the LAB color values ​​of the post-laser shattered melanin granule area image with the extracted original chloasma area image, scores the ultra-picosecond laser diagnosis and treatment of chloasma and evaluates the effect according to the comparison result, continuously compares the LAB color values ​​and scores during multiple treatments, and can clearly see the dynamic changes of the patient's condition. If the score improvement is not obvious in a certain time, the cause can be analyzed in time, such as insufficient laser energy or improper patient care, and timely intervention can be made.

[0012] As a further improvement of the present technical solution, the evaluation and adjustment module fills the subtracted unbroken melanin granule area image into the post-laser chloasma area image. By comparing the subtracted unbroken melanin granule area image with the LAB color value of the depicted post-laser non-whitened area image, the subtle color difference before and after treatment can be captured, avoiding ignoring some potential incomplete treatment or abnormal color changes due to local observation. The ultra-picosecond laser diagnosis and treatment of chloasma is re-scored based on the comparison results, and the effect of the re-scored ultra-picosecond laser diagnosis and treatment of chloasma is evaluated and the ultra-picosecond laser parameters are adjusted. If the effect evaluation and parameter adjustment are not performed, incomplete treatment and residual chloasma may occur; or over-treatment may cause damage to normal skin tissue. Adjusting parameters through scientific effect evaluation can avoid the occurrence of these two extreme situations.

[0013] A comprehensive evaluation of the treatment effect of melasma was performed using the scored ultra-picosecond laser treatment of melasma and the re-scored ultra-picosecond laser treatment of melasma. The first score can reflect the initial improvement of melasma after the initial ultra-picosecond laser treatment, covering the initial fragmentation of melanin by the laser, and the early changes in color and area. The re-scoring can show the further improvement of melasma after a period of time or subsequent treatment, as well as the development of potential problems after the first treatment.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] In the chloasma diagnosis and treatment effect evaluation system based on image feature analysis, the evaluation and adjustment module fills the deducted unbroken melanin granule area image into the chloasma area image after laser treatment, compares it through LAB color value, and re-scores the ultra-picosecond laser diagnosis and treatment of chloasma. The effect of the re-scored ultra-picosecond laser diagnosis and treatment of chloasma is evaluated and the ultra-picosecond laser parameters are adjusted. By filling and comparing the image of the unbroken melanin granule area, the color changes of the entire chloasma area can be comprehensively considered, including the laser direct action area and the surrounding less affected areas, so as to more comprehensively and accurately evaluate the actual effect of ultra-picosecond laser treatment of chloasma, improve the accuracy of evaluating the overall chloasma diagnosis and treatment effect, and adjust the ultra-picosecond laser parameters according to the comparison score to avoid changes caused by local neglect of potential unbroken melanin granules, reduce the adverse reaction efficiency of the patient's skin, and reduce the excessive waste of ultra-picosecond laser resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a block diagram of the overall system structure of the present invention;

[0017] Figure 2 It is a module block diagram of the present invention.

[0018] The meaning of each number in the figure is:

[0019] 10. Spot segmentation unit; 11. Modeling and segmentation module; 12. Feature fusion module; 13. Chloasma module;

[0020] 20. Spot area unit; 21. Spot connection module; 22. Area cutout module;

[0021] 30. Rating adjustment unit; 31. Rating evaluation module; 32. Evaluation adjustment module. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0023] Example 1

[0024] The present invention provides a chloasma diagnosis and treatment effect evaluation system based on image feature analysis, please refer to Figure 1-Figure 2 , including a spot segmentation unit 10, a spot area unit 20 and a score adjustment unit 30;

[0025] The spot segmentation unit 10 includes a modeling and segmentation module 11 and a feature fusion module 12;

[0026] The modeling and segmentation module 11 obtains the patient's facial image data through the camera device, and then establishes the patient's facial three-dimensional model data based on the obtained patient's facial image data, and divides the patient's facial three-dimensional model data into n three-dimensional coordinate points p i =(x i ,y i ,z i ), record the number of three-dimensional data points n, where p i Refers to the ith three-dimensional coordinate point on the patient's face, x i 、y i 、z i They refer to the x-axis, y-axis, and z-axis of the three-dimensional coordinate point, respectively. The RGB value c of the three-dimensional coordinate point is analyzed by the three-dimensional model data of the patient's face. i =(r i ,g i ,b i ), where r i Represents the red component value of the i-th three-dimensional coordinate point. The range of the red component value is (0-255). i Represents the blue component value of the i-th three-dimensional coordinate point. The range of the blue component value is (0-255). iRepresents the green component value of the i-th three-dimensional coordinate point. The range of the green component value is (0-255);

[0027] The process of establishing the patient's facial 3D model data:

[0028] First, the extracted patient facial image data (including multi-angle image data of the patient's face) is collected, and feature points are extracted from each acquired image data. The extracted feature points include corner points and edge points.

[0029] Secondly, feature matching is performed on different image data to find the corresponding points (matching feature points) of the same 3D point in different images;

[0030] Finally, the camera's internal parameters (such as focal length, principal point position, etc.) and external parameters (such as camera position and posture) are determined. This is the basis for 3D reconstruction. Based on the matched feature points and camera parameters, the 3D coordinates of each matched feature point are calculated using the existing triangulation principle, thereby reconstructing a 3D model of the patient's face.

[0031] The process of obtaining the RGB values ​​of the three-dimensional coordinate points through the three-dimensional model data of the patient's face:

[0032] When building the patient's facial 3D model data, the internal and external parameters of the camera used to capture the facial image (texture image) are also recorded. The mapping relationship between the 3D model surface points and the texture image pixels can be established through the internal and external parameters of the camera.

[0033] Since each 3D coordinate point obtained by segmentation is projected onto the texture image using the above mapping relationship, the corresponding 2D pixel coordinate is obtained. According to the projected 2D pixel coordinate, the color value (RGB value) of the pixel is extracted from the texture image.

[0034] When the RGB value of the three-dimensional coordinate point is known, the color conversion function is used to convert the RGB value of the three-dimensional coordinate point into the color value in the LAB color space to obtain the LAB color value ysz of the three-dimensional coordinate point. LAB , specific algorithm formula: LAB i =T RGB→LAB (c i ), where T RGB→LAB Refers to the color conversion function, LAB i Refers to the color value of the i-th three-dimensional coordinate point in the LAB color space, and the curvature characteristics of the three-dimensional coordinate point are calculated based on the three-dimensional coordinate point through a multivariate function;

[0035] Principle of calculating the curvature characteristics of three-dimensional coordinate points:

[0036] Collect the three-dimensional coordinate points and calculate the curvature characteristics of the three-dimensional coordinate points to obtain the calculated curvature characteristics H(pi ), specific algorithm formula:

[0037] in, and The 1 in is a constant. and The 2 on the left are constants, and θ refers to the partial differential symbol. Since the function depends on multiple independent variables, when studying the rate of change of the function with respect to a certain independent variable, it is necessary to keep the other independent variables unchanged. θ is the symbol used to represent this partial change. θx represents a small change in the independent variable x; θy represents a small change in the independent variable y, corresponding to a very small displacement in the y-axis direction; θz represents the corresponding small change in the dependent variable z due to a small change in the independent variable x or y; is the first-order partial derivative of z with respect to x, the rate of change of z with x while keeping y unchanged; is the first-order partial derivative of z with respect to y, the rate of change of z with y while keeping x unchanged; is the second-order partial derivative of z with respect to x, that is, first find the first-order partial derivative of x Calculate the partial derivative of x again, which reflects the degree of curvature of the patient's face in the x direction; is the second-order partial derivative of z with respect to y. First, find the first-order partial derivative of y. Then calculate the partial derivative of y, which reflects the curvature of the patient's facial surface in the y direction; It is a mixed second-order partial derivative of z. The partial derivative is first calculated with respect to y and then with respect to x. It describes the curvature characteristics of the patient's facial surface in the mixed xy direction. By calculating the curvature characteristics of the three-dimensional coordinate points, it can help identify abnormalities on the skin surface, which may correspond to areas of melasma lesions.

[0038] The feature fusion module 12 sets the color threshold μ of the target color spot in the LAB color space. spot and the LAB color value ysz of the three-dimensional coordinate point in the modeling and segmentation module 11 LAB Calculate the color Euclidean distance and obtain the calculated color Euclidean distance || ysz LAB -μ spot ||2, the color Euclidean distance ||ysz will be calculated LAB -μ spot ||2 and the calculated curvature feature H(p i ) is input into the linear regression model, which learns the input data and outputs the corresponding weight coefficients α and β. According to the calculated color Euclidean distance || ysz LAB -μ spot ||2. Calculated curvature feature H(p i) and weight coefficients α and β to perform multimodal feature fusion and obtain the multimodal fusion feature value F i =α·||ysz LAB -μ spot ||2+β·H(p i ), where α and β are used to balance the color Euclidean distance and curvature features to the multimodal feature fusion F i The weight coefficient of the impact degree;

[0039] The spot segmentation unit 10 further includes a chloasma module 13;

[0040] The chloasma module 13 converts the LAB color value ysz of the three-dimensional coordinate point in the modeling and segmentation module 11 into LAB And the calculated curvature characteristics H(p i ) is standardized, and then the LAB color value of the standardized three-dimensional coordinate point and the standardized curvature feature are fused to obtain the fused feature value, which is used as the dimension d of the data. The integer threshold MinPts is calculated according to the dimension of the data. The specific algorithm formula is: MinPts = d + 1. The integer threshold MinPts specifies the minimum number of neighborhood points required for a three-dimensional coordinate point to become a core point, and the number of points in the neighborhood ly is recorded;

[0041] Using the set multimodal fusion feature threshold and the multimodal fusion feature value F in the feature fusion module 12 i , the number of points ly in the neighborhood and the integer threshold MinPts determine whether the three-dimensional coordinate point is marked as a chloasma seed point. When the eigenvalue F of the multimodal fusion i When the value is greater than the set multimodal fusion feature threshold and the number of points ly in the neighborhood is greater than or equal to the integer threshold MinPts, the three-dimensional coordinate point is marked as a chloasma seed point and the three-dimensional coordinate hp of the chloasma seed point is recorded. i =(x i ,y i ,z i ) and the three-dimensional coordinates of the melasma seed point slide pressure diagnosis fading rate sg, the RGB value of the melasma seed point hc i =(r i ,g i ,b i ), chloasma seed point LAB color value hysz LAB ;

[0042] The spot area unit 20 includes a spot connection module 21 and an area cutout module 22;

[0043] The spot connection module 21 connects the three-dimensional coordinates of the melasma seed points in the melasma module 13 to form the original melasma area. The three-dimensional coordinates of the melasma seed points in the original melasma area and the slide pressure diagnosis fading rate sg of the three-dimensional coordinates of the melasma seed points, the RGB value of the melasma seed points, and the LAB color value of the melasma seed points are used to analyze whether the original melasma area is in the stable period. If the analysis shows that it is in the stable period, the patient's facial type is obtained from the medical database. If it is found that the original melasma area is a simple pigment type, it means that there are a large number of melanin particles in the epidermis of the original melasma area. The epidermis of the original melasma area is scanned by ultra-picosecond laser. When the laser energy acts on the epidermal melanin particles, the laser energy will shatter the melanin particles in the epidermis of the original melasma area. At this time, the original melasma area will show a temporary whitening or frosting reaction. The 3D coordinates of the whitened melasma seed points are recorded. The 3D coordinates of the whitened melasma seed points are connected with a red marker line to form an area where the melanin particles are shattered. The 3D coordinates of the non-whitened melasma seed points are then connected with a blue marker line to form a non-whitened area. The LAB color value of the non-whitened area is recorded.

[0044] Table 1 is a table for determining the specific situation of the stable period and the active period:

[0045]

[0046] Among them, a refers to a color channel from green to red; b refers to a color channel from blue to yellow;

[0047] When the red channel in the RGB value is enhanced, the slide pressure diagnosis fading rate in the fading rate is greater than 70%, and the a value and b value in the LAB color value are enhanced, it is determined to be in the active stage;

[0048] When there is no significant erythema in the RGB value, the fading rate is less than 30%, and the LAB color value is less than 30%, it is judged to be in the stable stage;

[0049] The region cutout module 22 cuts out the original chloasma region in the spot connection module 21 from the patient's facial image in the modeling and segmentation module 11 to obtain a cutout original chloasma region image, records the LAB color value of the cutout original chloasma region image, and then cuts out the region image of the unbroken melanin granules at the same coordinate position as the region of the crushed melanin granules in the spot connection module 21 from the cutout original chloasma region image to obtain a cutout unbroken melanin granule region image, and records the LAB color value of the cutout unbroken melanin granule region image. At this point, the first ultra-picosecond laser diagnosis and treatment is completed;

[0050] Steps to extract the original melasma area image:

[0051] Step 1: Collect a color patient facial image and convert it into a grayscale image to reduce computational complexity. Use Gaussian filtering to remove noise from the color patient facial image to obtain a denoised facial image. Then, perform histogram equalization to enhance the contrast of the denoised facial image to obtain an enhanced facial image, making the original melasma area more distinct.

[0052] Step ②: Binarize the enhanced facial image to preliminarily separate the original melasma area from the facial image background, obtaining a binary image of the original melasma area. Apply an erosion operation to remove small noise points in the binary image of the original melasma area and fill small holes in the original melasma area to make the original melasma area more complete.

[0053] Step 3: Use the findContours function to find the contours in the binary original melasma area image, and select the contours that meet the characteristics of the original melasma area based on the area and perimeter characteristics of the contours;

[0054] Step 4: Create a mask of the same size as the color patient facial image based on the filtered contours (the mask is essentially a binary image, which has the same size as the color patient facial image. In this binary image, the pixel value has only two states, usually represented by 0 and 255 (in different programming languages ​​and libraries, it may also be represented by Boolean values ​​False and True). In the operation of extracting the original melasma area from the patient's facial image, the pixel value corresponding to the original melasma area in the mask will be set to 255 (white), while the pixel value corresponding to other non-original melasma areas will be 0 (black)). Mark the original melasma area as white and other areas as black. Apply the mask to the color patient facial image to extract the original melasma area image;

[0055] Steps to remove the area where the melanin particles are not broken:

[0056] Step 1: Collect the extracted original melasma region image and convert it into a grayscale image to reduce computational complexity; use Gaussian filtering to remove noise from the extracted original melasma region image to obtain a denoised original melasma region image; then use histogram equalization to enhance the contrast of the denoised original melasma region image to obtain an enhanced original melasma region image, making the unbroken melanin granule area more prominent;

[0057] Step ②: Binarize the enhanced original melasma region image to preliminarily separate the shattered melanin granule region from the cut-out original melasma region image background, obtaining a binary image of the shattered melanin granule region. Erosion is then performed to remove small noise points in the binary image of the shattered melanin granule region, and small holes in the shattered melanin granule region are filled to make the unshattered melanin granule region more complete.

[0058] Step 3: Use the findContours function to find the contours in the binary image of the unbroken melanin granules, and select the contours that meet the characteristics of the unbroken melanin granules based on the area and perimeter of the contours;

[0059] Step ④: Create a mask of the same size as the original melasma area image based on the filtered contour (the mask is essentially a binary image, which has the same size as the original melasma area image. In this binary image, the pixel value has only two states, usually represented by 0 and 255 (in different programming languages ​​and libraries, it may also be represented by Boolean values ​​False and True). In the operation of extracting the unbroken melanin granule area from the extracted original melasma area image, the pixel value corresponding to the crushed melanin granule area in the mask will be set to 255 (white), while the pixel value corresponding to other non-unbroken melanin granule areas will be 0 (black)). Mark the unbroken melanin granule area as white and other areas as black. Apply the mask to the extracted original melasma area image to extract the unbroken melanin granule area image.

[0060] The rating adjustment unit 30 includes a rating evaluation module 31 and an evaluation adjustment module 32;

[0061] When the patient needs to undergo ultra-picosecond laser treatment again after a period of recovery, the modeling and segmentation module 11 is used to obtain the patient's facial image data after laser treatment through the camera device again. The scoring and evaluation module 31 receives the patient's facial image data after laser treatment in the modeling and segmentation module 11, and extracts the image of the chloasma area after laser treatment at the same coordinate position as the original chloasma area in the spot connection module 21 from the patient's facial image data after laser treatment, and then extracts the image of the shattered melanin granule area after laser treatment at the same coordinate position as the shattered melanin granule area in the spot connection module 21 from the chloasma area image after laser treatment, and records the shattered melanin granule area after laser treatment. The LAB color value of the domain image is depicted in the chloasma region image after laser treatment based on the coordinate position of the non-whitened area in the chloasma module 13 to obtain the depicted non-whitened area image after laser treatment. The image of the crushed melanin granules region after laser treatment is filled into the original chloasma region image cut out in the region cutout module 22. The LAB color value of the image of the crushed melanin granules region after laser treatment is compared with the LAB color value of the cutout original chloasma region image. Based on the comparison result, the ultra-picosecond laser treatment of chloasma is scored to obtain the scored ultra-picosecond laser treatment of chloasma, and then the effect of the scored ultra-picosecond laser treatment of chloasma is evaluated.

[0062] The ultra-picosecond laser treatment of melasma is scored based on the comparison results, and then the effect of the scored ultra-picosecond laser treatment of melasma is evaluated. Implementation process:

[0063] Collect the LAB color values ​​of the image of the crushed melanin granules area after the laser and the LAB color values ​​of the original melasma area image cut out, extract the L1 (brightness) value, a1 value (red-green axis) and b1 (yellow-blue axis) value from the LAB color values ​​of the image of the crushed melanin granules area after the laser, and then extract the L2 value, a2 value and b2 value from the LAB color value of the original melasma area image cut out, and use the L1 value, a1 value, b1 value of the LAB color value of the image of the crushed melanin granules area after the laser and the L2 value, a2 value, b2 value of the LAB color value of the original melasma area image cut out to calculate ΔL=L1-L2, Δa=a1-a2, Δb=b1-b2, and then calculate the average color difference based on ΔL, Δa, and Δb.

[0064] The set L difference, the set a difference, the set b difference, and the set average color threshold are used to compare with ΔL, Δa, and Δb respectively. When ΔL is greater than the set L difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when Δa is less than the set a difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when Δb is less than the set b difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when ΔE is less than the average color threshold, the score is 4 points;

[0065] The scores are weighted to obtain the scores of ultra-picosecond laser treatment for melasma. Since the full score of ultra-picosecond laser treatment for melasma is set to 10 points, when the score of ultra-picosecond laser treatment for melasma is less than 5 points, it is judged that the effect of ultra-picosecond laser treatment for melasma is poor. When the score of ultra-picosecond laser treatment for melasma is 5-6 points, it is judged that the effect of ultra-picosecond laser treatment for melasma is average. When the score of ultra-picosecond laser treatment for melasma is 7-8 points, it is judged that the effect of ultra-picosecond laser treatment for melasma is good. When the score of ultra-picosecond laser treatment for melasma is 9-10 points, it is judged that the effect of ultra-picosecond laser treatment for melasma is excellent.

[0066] The evaluation and adjustment module 32 fills the image of the uncrushed melanin granule region deducted by the region cutout module 22 into the image of the chloasma region after laser treatment in the scoring and evaluation module 31, compares the LAB color value of the deducted image of the uncrushed melanin granule region with the LAB color value of the image of the non-whitened region after laser treatment, and re-scores the ultra-picosecond laser treatment of chloasma based on the comparison result, obtains the re-scored ultra-picosecond laser treatment of chloasma, evaluates the effect of the re-scored ultra-picosecond laser treatment of chloasma, and adjusts the ultra-picosecond laser parameters based on the re-scored ultra-picosecond laser treatment of chloasma;

[0067] The comprehensive evaluation of the melasma treatment effect is performed using the ultra-picosecond laser treatment of melasma scored in the scoring evaluation module 31 and the ultra-picosecond laser treatment of melasma scored again.

[0068] The process of re-scoring the ultra-picosecond laser treatment for melasma based on the comparison results, evaluating the effect of the re-scored ultra-picosecond laser treatment for melasma, and adjusting the ultra-picosecond laser parameters based on the re-scored ultra-picosecond laser treatment for melasma is as follows:

[0069] Collect the LAB color values ​​of the unbroken melanin granule area image and the LAB color values ​​of the non-whitened area image after the laser, extract the L3 (brightness) value, a3 value (red-green axis) and b3 (yellow-blue axis) value from the LAB color values ​​of the crushed melanin granule area image after the laser, and then extract the L4 value, a4 value and b4 value from the LAB color values ​​of the non-whitened area image after the laser, and use the L3 value, a3 value, b3 value of the unbroken melanin granule area image LAB color value and the L4 value, a4 value, b4 value of the LAB color value of the non-whitened area image after the laser to calculate ΔL2=L3-L4, Δa2=a3-a4, Δb2=b3-b4, and then calculate the average color difference based on ΔL2, Δa2, Δb2

[0070] The set L2 difference, the set a2 difference, the set b2 difference, and the set average color threshold are used to compare with ΔL2, Δa2, and Δb2 respectively. When ΔL2 is greater than the set L2 difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when Δa2 is less than the set a2 difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when Δb2 is less than the set b2 difference, it means that the area of ​​shattered melanin particles after laser treatment is slowly approaching normal skin, and the score is 2 points; when ΔE2 is less than the average color threshold, the score is 4 points;

[0071] The scores were weighted to obtain the re-scored super-picosecond laser diagnosis and treatment of melasma. The full score of the super-picosecond laser diagnosis and treatment of melasma was set to 10 points.

[0072] When the score of ultra-picosecond laser treatment for melasma is less than 5 points, it is judged that the effect of ultra-picosecond laser treatment for melasma is poor, the skin type is re-evaluated, and segmented laser is used instead to promote collagen remodeling;

[0073] If the score of the ultra-picosecond laser treatment for melasma is 5-6 points, the effect of the ultra-picosecond laser treatment for melasma is judged to be average, and the wavelength is changed (e.g., from 532nm to 755nm or 1064nm, which penetrate deeper) and the energy is increased by 20% (skin reaction testing is required);

[0074] When the score of ultra-picosecond laser treatment for melasma is 7-8 points, it is judged that the ultra-picosecond laser treatment for melasma is effective, but the melanin particles are not completely removed. Increase the energy by 10%-15% (if the skin tolerates it well) and shorten the pulse width (such as adjusting from 3ms to 1.5ms to enhance the selective destruction of melanin);

[0075] When the score of ultra-picosecond laser diagnosis and treatment of melasma is 9 to 10 points, it is judged that the ultra-picosecond laser diagnosis and treatment of melasma is extremely effective. The current parameters can be maintained, the energy can be appropriately reduced (to avoid overtreatment), the treatment interval can be extended (such as 6 to 8 weeks), and the pigment metabolism can be observed.

[0076] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A chloasma diagnosis and treatment effect evaluation system based on image feature analysis, characterized by: It includes a spot segmentation unit (10), a spot area unit (20) and a score adjustment unit (30); The spot segmentation unit (10) acquires the patient's facial image data to analyze the RGB value of the three-dimensional coordinate point, converts it into a color value in the LAB color space, and calculates an integer threshold to determine whether the three-dimensional coordinate point is marked as a chloasma seed point; The spot area unit (20) connects the three-dimensional coordinates of the chloasma seed points in the spot segmentation unit (10) to form an original chloasma area, scans the epidermis of the original chloasma area with an ultra-picosecond laser, obtains the three-dimensional coordinates of the whitish chloasma seed points and the non-whitish chloasma seed points, connects them respectively, and then extracts the original chloasma area and the area image of the unbroken melanin particles; The scoring adjustment unit (30) cuts out the images of the chloasma and the crushed melanin granule area after laser treatment, and depicts the image of the non-whitened area after laser treatment, fills the image of the crushed melanin granule area after laser treatment into the cut-out original chloasma area image, and then fills the cut-out non-crushed melanin granule area image into the image of the chloasma area after laser treatment, compares the LAB color values, and scores and evaluates the effect of ultra-picosecond laser diagnosis and treatment of chloasma.

2. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 1, characterized in that: The spot segmentation unit (10) includes a modeling and segmentation module (11) and a feature fusion module (12); The modeling and segmentation module (11) obtains patient facial image data through a camera device and establishes patient facial three-dimensional model data, segments the patient facial three-dimensional model data into n three-dimensional coordinate points, analyzes the RGB values ​​of the three-dimensional coordinate points through the patient facial three-dimensional model data, converts the RGB values ​​of the three-dimensional coordinate points into color values ​​in the LAB color space using a color conversion function, obtains the LAB color values ​​of the three-dimensional coordinate points, and calculates the curvature features of the three-dimensional coordinate points based on the three-dimensional coordinate points using a multivariate function.

3. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 2, characterized in that: The feature fusion module (12) calculates the color Euclidean distance between the color threshold of the set target color spot in the LAB color space and the LAB color value of the three-dimensional coordinate point in the modeling and segmentation module (11), and performs multimodal feature fusion based on the calculated color Euclidean distance and the calculated curvature feature.

4. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 2, characterized in that: The spot segmentation unit (10) further includes a chloasma module (13); The chloasma module (13) performs feature fusion on the LAB color values ​​of the three-dimensional coordinate points and the calculated curvature features in the modeling and segmentation module (11) and uses the fusion as the dimension of the data, and calculates the integer threshold value according to the dimension of the data; The set multimodal fusion feature threshold, multimodal fusion eigenvalue, number of points in the neighborhood and integer threshold are used to determine whether a three-dimensional coordinate point is marked as a chloasma seed point. When the multimodal fusion eigenvalue is greater than the set multimodal fusion feature threshold and the number of points in the neighborhood is greater than or equal to the integer threshold, the three-dimensional coordinate point is marked as a chloasma seed point, and the three-dimensional coordinates of the chloasma seed point and the slide compression diagnosis fading rate of the three-dimensional coordinates of the chloasma seed point are recorded.

5. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 4, characterized in that: The spot area unit (20) includes a spot connection module (21) and an area cutout module (22); The spot connection module (21) connects the three-dimensional coordinates of the chloasma seed points in the chloasma module (13) to form an original chloasma area, and uses the three-dimensional coordinates of the chloasma seed points in the original chloasma area and the slide pressure diagnosis fading rate of the three-dimensional coordinates to analyze whether the original chloasma area is in a stable period. When the analysis shows that it is in a stable period, the epidermis of the original chloasma area is scanned by an ultra-picosecond laser to break the melanin particles in the epidermis of the original chloasma area. At this time, the original chloasma area will temporarily turn white, and the three-dimensional coordinates of the whitened chloasma seed points are recorded. The three-dimensional coordinates of the whitened chloasma seed points are connected with a red marking line to form an area where the melanin particles are broken, and then the three-dimensional coordinates of the non-whitened chloasma seed points are connected with a blue marking line to form an area that is not whitened.

6. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 5, characterized in that: The region cutout module (22) cuts out the original chloasma region in the spot connection module (21) from the patient's facial image in the modeling and segmentation module (11), and then cuts out the region image of the unbroken melanin particles having the same coordinate position as the region of the broken melanin particles in the spot connection module (21) from the cutout original chloasma region image.

7. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 2, characterized in that: The score adjustment unit (30) includes a score evaluation module (31) and an evaluation adjustment module (32); The scoring and evaluation module (31) obtains the patient's facial image data after laser treatment again through the modeling and segmentation module (11), extracts the chloasma area image after laser treatment at the same coordinate position as the original chloasma area in the spot connection module (21) from the patient's facial image data after laser treatment, and then extracts the shattered melanin granule area image after laser treatment at the same coordinate position as the shattered melanin granule area in the spot connection module (21) from the chloasma area image after laser treatment, depicts the chloasma area image after laser treatment according to the coordinate position of the non-whitened area in the chloasma module (13), fills the shattered melanin granule area image after laser treatment into the original chloasma area image extracted in the area extraction module (22), compares the LAB color values ​​of the shattered melanin granule area image after laser treatment with the extracted original chloasma area image, and scores and evaluates the effect of ultra-picosecond laser diagnosis and treatment of chloasma based on the comparison results.

8. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 7, characterized in that: The evaluation and adjustment module (32) fills the image of the unbroken melanin granule region deducted from the region cutout module (22) into the image of the chloasma region after laser treatment in the scoring and evaluation module (31), compares the LAB color value of the deducted image of the unbroken melanin granule region with the image of the non-whitened region after laser treatment, and re-scores the ultra-picosecond laser diagnosis and treatment of chloasma based on the comparison result, evaluates the effect of the re-scored ultra-picosecond laser diagnosis and treatment of chloasma, and adjusts the ultra-picosecond laser parameters; A comprehensive evaluation of the treatment effect of melasma was performed using the scored ultra-picosecond laser treatment of melasma and the re-scored ultra-picosecond laser treatment of melasma.

9. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 3, characterized in that: The implementation principle of calculating the curvature features of the three-dimensional coordinate points in the feature fusion module (12) is as follows: Collect the three-dimensional coordinate points and calculate the curvature characteristics of the three-dimensional coordinate points to obtain the calculated curvature characteristics H(p i ), specific algorithm formula:

10. The chloasma diagnosis and treatment effect evaluation system based on image feature analysis according to claim 6, characterized in that: Steps to extract the original melasma area image: Step 1: Collect the color patient facial image and convert it into a grayscale image, remove the noise in the color patient facial image using Gaussian filtering, and then enhance the contrast of the denoised facial image through histogram equalization; Step 2: Binarize the enhanced facial image to preliminarily separate the original melasma area from the facial image background, and use the erosion operation to remove small noise points in the binary original melasma area image; Step 3: Use the findContours function to find the contours in the binary original melasma area image, and select the contours that meet the characteristics of the original melasma area based on the area and perimeter characteristics of the contours; Step 4: Create a mask of the same size as the color patient facial image based on the filtered contour, mark the original melasma area as white and other areas as black, and then apply the mask to the color patient facial image to extract the original melasma area image.

Images