Method and system for evaluating alopecia areata target area repigmentation based on standardized entity reference cards

Abstract

The application relates to an alopecia areata target area hair regrowth evaluation method and system based on a standardized entity reference card. The standardized entity reference card comprises a card body, scale marks, color correction blocks, direction positioning marks and positioning angle marks. The evaluation method comprises collecting an evaluation image containing an alopecia areata target area and a standardized entity reference card; correcting the image based on the scale marks, the color correction blocks, the direction positioning marks and the positioning angle marks; marking the closed boundary of the alopecia areata target area in the image and extracting various parameters; comparing the current parameters with the previous parameters to calculate the parameter change rate; and performing weighted calculation on the various parameters to obtain the target area hair regrowth quantitative parameter. The standardized reference card unifies the image scale, color and direction, eliminates the error caused by shooting, does not require professional equipment, is suitable for outpatient and home scenes, and realizes alopecia areata change digitization through multi-dimensional index weighted calculation of the hair regrowth quantitative parameter, avoids one-sidedness of a single index result, and realizes comprehensive, accurate, convenient and efficient evaluation.

Description

Technical Field

[0001] This invention relates to a method and system for assessing hair regeneration in alopecia areata target areas based on standardized physical reference cards, and is applicable to the field of medical image processing technology. Background Technology

[0002] Alopecia areata is a common non-scarring hair loss condition, clinically characterized by well-defined patches of hair loss on the scalp or other hairy areas of the body. During the treatment and follow-up of alopecia areata, doctors need to understand not only the severity of the patient's overall hair loss but also the dynamic changes of a specific target lesion before and after treatment. These changes include whether the target area has shrunk, whether new hair growth has occurred within the target area, whether the new hair has gradually changed from white vellus hair to pigmented hair or terminal hair, whether the hair color has darkened, and whether the target area boundary has stabilized.

[0003] Among existing clinical evaluation methods, tools such as the SALT score mainly focus on the proportion of total scalp alopecia, which is suitable for assessing the overall severity of the condition. However, they still have certain limitations for scenarios such as continuous follow-up of individual local target lesions, observation of the efficacy of local treatment, rapid outpatient follow-up, and patients' home-based photo follow-up. Photos taken at different time points are easily affected by shooting distance, shooting angle, lighting conditions, hair parting degree, skin color, mobile phone model, and camera parameters, making it difficult to objectively compare changes in area, hair coverage, and color of the same target area between different follow-up time points.

[0004] In existing technologies, one type of approach, as shown in patent publication number CN118379281B, mainly involves acquiring images of the alopecia areata area on the scalp from a dermoscope, feeding the images into a pre-trained instance segmentation neural network model, segmenting instances based on yellow dots, black dots, and hairs with roots, and further distinguishing short vellus hair, exclamation mark hair, broken hair, or normal hair through geometric calculations. Finally, the number of each feature is counted as the analysis result. While this method can effectively identify hair features, it cannot assess the overall hair regeneration status of the alopecia areata target area and relies on dermoscopy equipment, making it unsuitable for scenarios such as rapid outpatient follow-up examinations or patient self-assessment at home. Another type of approach, as shown in patent publication number CN117994213A, mainly uses a three-dimensional face reconstruction scheme. It acquires face images from different angles, constructs a three-dimensional face reconstruction network model, generates a UV unfolded map containing the entire hair texture, segments the UV unfolded map to identify the alopecia areata area, and then calculates the proportion of the alopecia areata area to the entire scalp area to obtain the alopecia areata assessment result. While this type of solution can provide a comprehensive assessment of the target area of ​​alopecia areata, it also relies on specialized equipment, requiring the acquisition of multi-angle images and three-dimensional reconstruction. The process is too complex and cannot be used for scenarios such as rapid outpatient follow-up examinations or patient self-assessment at home. Summary of the Invention

[0005] To address the shortcomings of the existing technology, this invention proposes a method and system for assessing hair regeneration in the target area of ​​alopecia areata based on a standardized physical reference card.

[0006] The technical solution adopted in this invention is as follows: On the one hand, this invention provides a method for evaluating the re-hair regeneration of the target area of ​​alopecia areata based on a standardized physical reference card. The standardized physical reference card includes a card body, a scale mark, a color correction block, a direction positioning mark, and a positioning corner mark set on the card body. The method performs size normalization, color correction, direction correction, and position registration on images collected at different follow-up time points, thereby reducing evaluation errors caused by differences in shooting distance, shooting angle, lighting conditions, and acquisition equipment.

[0007] The evaluation methods include: S1. Image Acquisition: The standardized physical reference card is placed close to or partially covers the alopecia areata target area on the head, and evaluation images are acquired. The evaluation images include the alopecia areata target area and the standardized physical reference card. Specifically, the evaluation images are acquired by devices such as mobile phones, tablets, cameras, and dedicated image acquisition equipment.

[0008] S2. Image Processing: Establish the correspondence between pixels and actual lengths in the evaluation image based on scale markers; correct the white balance, brightness, chromaticity, and color difference of the evaluation image based on color correction blocks; correct the shooting direction of the evaluation image based on orientation positioning markers; correct the rotation, translation, scaling, and perspective transformation parameters of the evaluation image based on positioning corner marks; normalize the size of the evaluation image through scale markers, color correction blocks, orientation positioning markers, and positioning corner marks, and perform color correction such as white balance, brightness, chromaticity, and color difference. At the same time, correct the shooting direction of the evaluation image, and identify and unify the rotation, translation, scaling, and perspective distortion of the image, so as to standardize and correct the data deviations in the evaluation image caused by different follow-up batches of shooting distance, lighting, and equipment, and ensure the objectivity and accuracy of the evaluation comparison between evaluation images.

[0009] S3. Parameter extraction: Delineate the closed boundary of the alopecia areata target area in the processed evaluation image and calculate the area of ​​the alopecia areata target area. At the same time, extract at least two of the following features from the alopecia areata target area: hair coverage feature, hair density feature, hair thickness feature, and hair color feature. S4. The parameters extracted from the current evaluation image are compared with the parameters extracted from the previous or first-time acquired evaluation image to calculate the change parameters of the alopecia areata target area. The change parameters include at least three of the following: target area change rate, hair coverage change rate, terminal hair ratio change rate, and hair color recovery score. By calculating the area change of the alopecia areata target area and extracting various features of the hair inside, the changes in parameters such as area, hair density, terminal hair ratio, and hair color of the alopecia areata target area can be calculated, thereby comprehensively and effectively evaluating the recovery of the alopecia areata target area.

[0010] S5. Calculate the weighted values ​​of each parameter in the variable parameters, and then sum the weighted calculation results to obtain the target area hair regrowth quantification parameters. By calculating the weighted values ​​of multiple variable parameters to obtain the final target area hair regrowth quantification parameters, the recovery status of alopecia areata in the patient's local area can be digitized and standardized, avoiding the bias caused by a single indicator, ensuring the accuracy of the assessment results, and also providing a direct view of the patient's alopecia areata hair regrowth status, thus improving assessment efficiency.

[0011] Furthermore, in step S3, the method for calculating the target area of ​​alopecia areata includes: a. Delineate the closed boundary of the bald patch target area using any one or a combination of methods such as manual delineation, semi-automatic edge detection, color segmentation, interactive segmentation, and deep learning segmentation, and generate the corresponding binary mask. b. Based on the binary mask of the bald spot target area, count the total number of pixels inside the closed boundary, and calculate the area of ​​the bald spot target area according to the correspondence between pixels and actual length.

[0012] Furthermore, in step S4, the method for calculating the target area change rate is as follows: Target area change rate SAR = (A0 - A t ) / A0; Where A0 represents the area of ​​the alopecia areata target region in the previous or first acquired evaluation image, A t This represents the area of ​​the alopecia areata target region in the currently acquired evaluation image.

[0013] Furthermore, in step S4, the method for calculating the rate of change in hair coverage is as follows: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. b. Perform morphological opening and closing operations and small-area connected component clearing on the binary mask of the hair region, and then calculate the area of ​​the hair region; c. Based on the area of ​​the hair region and the area of ​​the alopecia areata target region, the hair coverage rate C of the alopecia areata target region is calculated as: area of ​​hair region / area of ​​alopecia areata target region; d. Based on the hair coverage C0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair coverage C of the alopecia areata target area in the current evaluation image. t The calculated hair coverage change rate HCR=C t - C0.

[0014] Furthermore, in step S4, the method for calculating the rate of change in the final wool ratio is as follows: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. Then, morphological opening and closing operations and small-area connected component clearing are performed on the binary mask of the hair region, and the area of ​​the hair region is calculated. b. Thin the image of the binary mask of the hair region, extract the single-pixel hair skeleton, and measure the hair shaft width segment by segment along the normal direction of the hair skeleton. Then, based on the preset hair shaft width threshold, divide the hair pixels into terminal hair pixels and vellus hair pixels. c. Calculate the area of ​​the hair-terminal region formed by the hair-terminal pixels, and use this area to calculate the hair-terminal ratio P of the alopecia areata target area = area of ​​the hair-terminal region / area of ​​the hair region. Then, based on the hair-terminal ratio P0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair-terminal ratio P of the alopecia areata target area in the current evaluation image... t Calculate the final gross percentage change rate (TPR) = P t - P0.

[0015] Furthermore, in step S4, the calculation method for the hair color restoration score includes: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels, and a binary mask of the hair region formed by the hair pixels is generated. At the same time, based on the closed boundary of the alopecia areata target area, a stable region is selected in the adjacent normal hair region outside the alopecia areata target area and a corresponding normal hair reference mask is generated. b. Based on the binary mask of the hair region and the reference mask of normal hair, extract the target area hair pixels and normal hair pixels respectively, and convert the image to the CIE Lab color model to obtain the target area hair color vector (L). t a t b t ) and normal hair color vector (L n a n b n Then, based on the color vectors of the target hair and normal hair, the color difference value between the hair and normal hair in the alopecia areata target area is calculated. ; c. Preset color difference value scoring boundary (ΔE) min ΔE max ), calculate hair color restoration score And when ΔE≤ΔE min When CRS=1, and when ΔE≥ΔE max At that time, CRS=0.

[0016] Furthermore, in step S5, among the various change parameters involved in calculating the quantitative parameters of target area hair recovery, the weight of the target area change rate is 0.45–0.65, the weight of the hair coverage change rate is 0.15–0.3, the weight of the final hair ratio change rate is 0.04–0.12, and the weight of the hair color recovery score is 0.01–0.06.

[0017] Furthermore, the variation parameters also include at least one of the following: hair density change rate, hair shaft thickness change parameter, and target area boundary stability parameter. Specifically, the weight of the hair density change rate is 0.03 to 0.1, the weight of the hair shaft thickness change parameter is 0.02 to 0.08, and the weight of the target area boundary stability parameter is 0.05 to 0.15.

[0018] Furthermore, the scale markers include at least one of the following: scale lines, grids, circles, and rectangles with known dimensions; the color correction blocks include at least two of the following: black blocks, white blocks, gray blocks, skin color reference blocks, and hair color reference blocks; the directional positioning markers include at least one of the following: arrows for indicating direction, markings for indicating the top of the patient's head, markings for indicating the frontal side of the patient, markings for indicating the occipital side of the patient, and asymmetrical patterns; and the positioning corner markers include at least three non-collinear positioning markers.

[0019] On the other hand, the present invention also provides a system for assessing hair repositioning in the target area of ​​alopecia areata, for performing the above-described hair repositioning assessment method, the system comprising: A standardized entity reference card is used to simultaneously enter the acquisition range of the evaluation image with the alopecia areata target area and provide scale, color, orientation, and position references; The image acquisition module is used to acquire evaluation images containing the alopecia areata target area and a standardized physical reference card; The image processing module is used to identify scale markers, color correction blocks, orientation markers, and positioning corner marks in the standardized entity reference card, and to correct and evaluate the scale, color, orientation, and position of the image. The parameter extraction module is used to extract at least one of the following features in the evaluation image: hair coverage features, hair density features, hair thickness features, and hair color features. The parameter calculation module is used to calculate the quantification parameters of the hair regeneration in the alopecia areata target area based on the parameter features acquired from the previous or first-time evaluation image and the current evaluation image.

[0020] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art: 1. By using standardized physical reference cards in image acquisition and processing, images with different scales, colors, orientations, and positions can be standardized to eliminate errors caused by differences in shooting distance, angle, lighting, and equipment. This eliminates the need for specialized equipment or facilities for image acquisition, making it convenient, cost-effective, and applicable to scenarios such as rapid outpatient follow-up and patient self-assessment at home, thereby improving assessment efficiency and practicality. Second, by collecting and calculating changes in multiple parameters such as the area of ​​the alopecia areata target zone, hair density, terminal hair ratio, and hair color, and then weighting these parameters according to their respective weight values, the final quantitative parameters for target zone hair regeneration are obtained. This standardizes and digitizes the changes in the patient's alopecia areata target zone, which not only intuitively reflects the patient's recovery status and improves the convenience and efficiency of the assessment process, but also avoids the bias and unreliability of assessment results caused by a single indicator, ensuring the accuracy and comprehensiveness of the assessment results. Attached Figure Description

[0021] The following sections will describe some specific embodiments of the invention in a detailed manner, by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar components or parts. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a schematic diagram of the structure of a standardized entity reference card in one embodiment of the present invention; Figure 2 yes Figure 1 Flowchart of the embodiment shown; The annotations in the attached figures are explained as follows: 1. Card body; 11. Scale mark; 12. Color correction block; 13. Orientation positioning mark; 14. Positioning corner mark. Detailed Implementation

[0022] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. The technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0023] Reference Appendix Figures 1-2On the one hand, this embodiment provides a method for assessing hair regeneration in the target area of ​​alopecia areata based on a standardized physical reference card. The standardized physical reference card includes a card body 1, scale markers 11, color correction blocks 12, directional positioning markers 13, and positioning angle markers 14 set on the card body 1. Specifically, the scale markers 11 include at least one of scale lines, grids, circles, and rectangles with known dimensions; the color correction blocks 12 include at least two of black blocks, white blocks, gray blocks, skin color reference blocks, and hair color reference blocks; the directional positioning markers 13 include arrows for indicating direction and arrows for indicating the top of the patient's head. The system includes at least one of the following: a marker for the forehead, a marker for the occipital region, and an asymmetrical pattern; a positioning marker 14 includes at least three non-collinear positioning marks; in this embodiment, the scale marker 11 is set as a scale line; the color correction block 12 includes three skin tone reference blocks (black, white, gray, and yellowish-brown) and one gray hair color reference block; the direction positioning marker 13 includes an arrow indicating direction, a marker for the top of the patient's head, a marker for the forehead, and a marker for the occipital region; and the positioning marker 14 includes four non-collinear positioning marks. The various markers on the card body 1 are used to normalize the size, correct the color, correct the direction, and register the position of images acquired at different follow-up time points, reducing evaluation errors caused by differences in shooting distance, shooting angle, lighting conditions, and acquisition equipment.

[0024] The evaluation methods include: S1. Image Acquisition: The standardized physical reference card is placed close to or partially covers the alopecia areata target area on the head, and evaluation images are acquired. The evaluation images include the alopecia areata target area and the standardized physical reference card. Specifically, the evaluation images are acquired by devices such as mobile phones, tablets, cameras, and dedicated image acquisition equipment, which makes it convenient for users to complete the acquisition using ordinary devices, taking into account multiple scenarios such as home and outpatient clinics, and reducing reliance on professional equipment.

[0025] S2. Image Processing: Based on scale marker 11, establish the correspondence between pixels and actual length in the evaluation image; based on color correction block 12, correct the white balance, brightness, chromaticity, and color difference of the evaluation image; based on orientation positioning marker 13, correct the shooting direction of the evaluation image; based on positioning corner marker 14, correct the rotation, translation, scaling, and perspective transformation parameters of the evaluation image; through scale marker 11, color correction block 12, orientation positioning marker 13, and positioning corner marker 14, normalize the size of the evaluation image, and perform color correction such as white balance, brightness, chromaticity, and color difference, while correcting the shooting direction of the evaluation image, and identifying and unifying the rotation, translation, scaling, and perspective distortion of the image, so as to standardize and correct the data deviations in the evaluation image caused by shooting distance, lighting, and equipment of different follow-up batches, and ensure the objectivity and accuracy of the evaluation comparison between evaluation images.

[0026] S3. Parameter Extraction: Delineate the closed boundary of the alopecia areata target region in the processed evaluation image and calculate the area of ​​the alopecia areata target region. Simultaneously, extract at least two of the following features from the alopecia areata target region: hair coverage, hair density, hair thickness, and hair color. Specifically, the method for calculating the area of ​​the alopecia areata target region includes: a. Delineate the closed boundary of the bald patch target area using any one or a combination of methods such as manual delineation, semi-automatic edge detection, color segmentation, interactive segmentation, and deep learning segmentation, and generate the corresponding binary mask. b. Based on the binary mask of the bald spot target area, count the total number of pixels inside the closed boundary, and calculate the area of ​​the bald spot target area according to the correspondence between pixels and actual length.

[0027] Furthermore, in step S4, the method for calculating the target area change rate is as follows: Target area change rate SAR = (A0 - A t ) / A0; Where A0 represents the area of ​​the alopecia areata target region in the previous or first acquired evaluation image, A t This represents the area of ​​the alopecia areata target region in the currently acquired evaluation image.

[0028] S4. The parameters extracted from the current evaluation image are compared with the parameters extracted from the previous or first acquired evaluation image to calculate the change parameters of the alopecia areata target area. These change parameters include at least three of the following: target area change rate, hair coverage change rate, terminal hair proportion change rate, and hair color recovery score. By calculating the area change of the alopecia areata target area and extracting various features of the hair within it, changes in parameters such as area, hair density, terminal hair proportion, and hair color of the alopecia areata target area can be calculated, thereby comprehensively and effectively evaluating the recovery status of the alopecia areata target area. Specifically, the calculation method for the hair coverage change rate is as follows: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. b. Perform morphological opening and closing operations and small-area connected component clearing on the binary mask of the hair region, and then calculate the area of ​​the hair region; c. Based on the area of ​​the hair region and the area of ​​the alopecia areata target region, the hair coverage rate C of the alopecia areata target region is calculated as: area of ​​hair region / area of ​​alopecia areata target region; d. Based on the hair coverage C0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair coverage C of the alopecia areata target area in the current evaluation image. t The calculated hair coverage change rate HCR=C t - C0; The method for calculating the rate of change in the final wool percentage is as follows: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. Then, morphological opening and closing operations and small-area connected component clearing are performed on the binary mask of the hair region, and the area of ​​the hair region is calculated. b. Thin the image of the binary mask of the hair region, extract the single-pixel hair skeleton, and measure the hair shaft width segment by segment along the normal direction of the hair skeleton. Then, based on the preset hair shaft width threshold, divide the hair pixels into terminal hair pixels and vellus hair pixels. c. Calculate the area of ​​the hair-terminal region formed by the hair-terminal pixels, and use this area to calculate the hair-terminal ratio P of the alopecia areata target area = area of ​​the hair-terminal region / area of ​​the hair region. Then, based on the hair-terminal ratio P0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair-terminal ratio P of the alopecia areata target area in the current evaluation image... t Calculate the final gross percentage change rate (TPR) = P t - P0; The calculation methods for hair color restoration scores include: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels, and a binary mask of the hair region formed by the hair pixels is generated. At the same time, based on the closed boundary of the alopecia areata target area, a stable region is selected in the adjacent normal hair region outside the alopecia areata target area and a corresponding normal hair reference mask is generated. b. Based on the binary mask of the hair region and the reference mask of normal hair, extract the target area hair pixels and normal hair pixels respectively, and convert the image to the CIE Lab color model to obtain the target area hair color vector (L). t a t b t ) and normal hair color vector (L n a n b n Then, based on the color vectors of the target hair and normal hair, the color difference value between the hair and normal hair in the alopecia areata target area is calculated. ; c. Preset color difference value scoring boundary (ΔE) min ΔE max ), calculate hair color restoration score And when ΔE≤ΔE min When CRS=1, and when ΔE≥ΔE max At that time, CRS=0.

[0029] S5. Calculate the weighted values ​​of each parameter in the variation parameters, then sum the weighted results to obtain the target area hair restoration quantification parameters. Among the variation parameters involved in calculating the target area hair restoration quantification parameters, the weight of the target area change rate is 0.45–0.65, the weight of the hair coverage change rate is 0.15–0.3, the weight of the final hair proportion change rate is 0.04–0.12, and the weight of the hair color recovery score is 0.01–0.06. Specifically, after obtaining the target area hair restoration quantification parameters, existing SALT assessment standards can be referenced. Standards such as the Precision Scalp and Hair PRO Grading Scale are used to output status levels such as "no significant change", "mild improvement", "moderate improvement", "significant improvement" and "near recovery" based on the quantitative parameters of hair regeneration in the target area, from small to large. By weighting and calculating multiple change parameters, the final quantitative parameters of hair regeneration in the target area can be obtained. This can digitize and standardize the recovery status of alopecia areata in local areas of patients, avoid the one-sidedness of assessment results caused by a single indicator, ensure the accuracy of assessment results, and also intuitively show the hair regeneration status of patients with alopecia areata, thus improving assessment efficiency.

[0030] In a more preferred embodiment, the variation parameters further include at least one of hair density change rate, hair shaft thickness change parameter, and target area boundary stability parameter. Specifically, the weight of hair density change rate is 0.03 to 0.1, the weight of hair shaft thickness change parameter is 0.02 to 0.08, and the weight of target area boundary stability parameter is 0.05 to 0.15.

[0031] On the other hand, the present invention also provides a system for assessing hair repositioning in the target area of ​​alopecia areata, for performing the above-described hair repositioning assessment method, the system comprising: A standardized entity reference card is used to simultaneously enter the acquisition range of the evaluation image with the alopecia areata target area and provide scale, color, orientation, and position references; The image acquisition module is used to acquire evaluation images containing the alopecia areata target area and a standardized physical reference card; The image processing module is used to identify the scale mark 11, color correction block 12, orientation positioning mark 13 and positioning corner mark 14 in the standardized entity reference card, and to correct and evaluate the scale, color, orientation and position of the image. The parameter extraction module is used to extract at least one of the following features in the evaluation image: hair coverage features, hair density features, hair thickness features, and hair color features. The parameter calculation module is used to calculate the quantification parameters of the hair regeneration in the alopecia areata target area based on the parameter features acquired from the previous or first-time evaluation image and the current evaluation image.

[0032] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art: 1. By using standardized physical reference cards in image acquisition and processing, images with different scales, colors, orientations, and positions can be standardized to eliminate errors caused by differences in shooting distance, angle, lighting, and equipment. This eliminates the need for specialized equipment or facilities for image acquisition, making it convenient, cost-effective, and applicable to scenarios such as rapid outpatient follow-up and patient self-assessment at home, thereby improving assessment efficiency and practicality. Second, by collecting and calculating changes in multiple parameters such as the area of ​​the alopecia areata target zone, hair density, terminal hair ratio, and hair color, and then weighting these parameters according to their respective weight values, the final quantitative parameters for target zone hair regeneration are obtained. This standardizes and digitizes the changes in the patient's alopecia areata target zone, which not only intuitively reflects the patient's recovery status and improves the convenience and efficiency of the assessment process, but also avoids the bias and unreliability of assessment results caused by a single indicator, ensuring the accuracy and comprehensiveness of the assessment results.

[0033] The above embodiments are only for illustrating the technical concept and features of the present invention. Their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be used to limit the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for assessing hair regeneration in target areas of alopecia areata based on standardized entity reference cards, characterized in that: The standardized entity reference card includes a card body (1), a scale mark (11) set on the card body (1), a color correction block (12), a direction positioning mark (13), and a positioning corner mark (14). The evaluation method includes: S1. Image acquisition: The standardized entity reference card is brought close to or partially covers the target area of ​​alopecia areata on the head, and an evaluation image is acquired. The evaluation image includes the target area of ​​alopecia areata and the standardized entity reference card. S2, Image Processing: Based on the scale marker (11), establish the correspondence between pixels and actual length in the evaluation image; based on the color correction block (12), correct the white balance, brightness, chromaticity and color difference of the evaluation image; based on the orientation positioning marker (13), correct the shooting direction of the evaluation image; based on the positioning corner marker (14), correct the rotation, translation, scaling and perspective transformation parameters of the evaluation image; S3. Parameter extraction: Delineate the closed boundary of the alopecia areata target area in the processed evaluation image, calculate the area of ​​the alopecia areata target area, and extract at least two of the following features from the alopecia areata target area: hair coverage feature, hair density feature, hair thickness feature, and hair color feature. S4. The parameters extracted from the current evaluation image are compared with the parameters extracted from the previous or first-time acquired evaluation image to calculate the change parameters of the alopecia areata target area. The change parameters include at least three of the following: target area change rate, hair coverage change rate, terminal hair ratio change rate, and hair color recovery score. S5. Perform weighted calculations on each parameter in the changed parameters, and then add the weighted calculation results to obtain the target area hair quantification parameters.

2. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 1, characterized in that: In step S3, the method for calculating the area of ​​the alopecia areata target region includes: a. Delineate the closed boundary of the bald spot target area using any one or a combination of methods such as manual delineation, semi-automatic edge detection, color segmentation, interactive segmentation, and deep learning segmentation, and generate the corresponding binary mask; b. Based on the binary mask of the bald spot target area, count the total number of pixels inside the closed boundary, and calculate the area of ​​the bald spot target area according to the correspondence between pixels and actual length.

3. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 2, characterized in that: In step S4, the method for calculating the target area change rate is as follows: Target area change rate SAR = (A0 - A t ) / A0; Where A0 is the area of ​​the alopecia areata target region in the previous or first acquired evaluation image, A t This refers to the area of ​​the alopecia areata target region in the currently acquired evaluation image.

4. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 2, characterized in that: In step S4, the method for calculating the hair coverage change rate is as follows: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. b. Perform morphological opening and closing operations and small-area connected component clearing on the binary mask of the hair region, and then calculate the area of ​​the hair region; c. Based on the area of ​​the hair region and the area of ​​the alopecia areata target region, the hair coverage rate C of the alopecia areata target region is calculated as: area of ​​hair region / area of ​​alopecia areata target region; d. Based on the hair coverage C0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair coverage C of the alopecia areata target area in the current evaluation image. t The hair coverage change rate HCR was calculated to be C. t - C0.

5. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 2, characterized in that: In step S4, the method for calculating the final wool ratio change rate is as follows: a. Using grayscale thresholding, skin color chromaticity segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels. The hair region formed by the hair pixels is extracted, and a binary mask of the hair region is generated. Then, morphological opening and closing operations and small-area connected component clearing are performed on the binary mask of the hair region, and the area of ​​the hair region is calculated. b. Thin the image of the binary mask of the hair region, extract the single-pixel hair skeleton, and measure the hair shaft width segment by segment along the normal direction of the hair skeleton. Then, based on the preset hair shaft width threshold, divide the hair pixels into terminal hair pixels and vellus hair pixels. c. Calculate the area of ​​the hair-terminal region formed by the hair-terminal pixels, and use this area to calculate the hair-terminal ratio P of the alopecia areata target area = area of ​​the hair-terminal region / area of ​​the hair region. Then, based on the hair-terminal ratio P0 of the alopecia areata target area in the previous or first acquired evaluation image and the hair-terminal ratio P of the alopecia areata target area in the current evaluation image... t Calculate the final gross percentage change rate (TPR) = P t -P0.

6. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 2, characterized in that: In step S4, the calculation method for the hair color restoration score includes: a. Using grayscale thresholding, skin color segmentation, texture feature classification, or deep learning segmentation algorithms, the pixels in the alopecia areata target area are divided into hair pixels and scalp pixels, and a binary mask of the hair region formed by the hair pixels is generated. At the same time, based on the closed boundary of the alopecia areata target area, a stable region is selected in the adjacent normal hair region outside the alopecia areata target area and a corresponding normal hair reference mask is generated. b. Based on the binary mask of the hair region and the reference mask of normal hair, extract the target area hair pixels and normal hair pixels respectively, and convert the image to the CIE Lab color model to obtain the target area hair color vector (L). t a t b t ) and normal hair color vector (L n a n b n Then, based on the color vectors of the target hair and normal hair, the color difference value between the hair and normal hair in the alopecia areata target area is calculated. ; c. Preset color difference value scoring boundary (ΔE) min ΔE max ), calculate the hair color restoration score And when ΔE≤ΔE min When CRS=1, and when ΔE≥ΔE max At that time, CRS=0.

7. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 1, characterized in that: In step S5, among the various change parameters involved in calculating the target area hair quantification parameters, the weight of the target area change rate is 0.45 to 0.65, the weight of the hair coverage change rate is 0.15 to 0.3, the weight of the terminal hair ratio change rate is 0.04 to 0.12, and the weight of the hair color recovery score is 0.01 to 0.

06.

8. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 1, characterized in that: The variation parameters also include at least one of the following: hair density change rate, hair shaft thickness change parameter, and target area boundary stability parameter.

9. The method for assessing hair regeneration in alopecia areata target areas based on standardized entity reference cards according to claim 1, characterized in that: The scale marker (11) includes at least one of the following: scale lines, grids, circles, and rectangles with known dimensions; the color correction block (12) includes at least two of the following: black block, white block, gray block, skin color reference block, and hair color reference block; the direction positioning marker (13) includes at least one of the following: an arrow for indicating direction, a mark for indicating the top of the patient's head, a mark for indicating the frontal side of the patient, a mark for indicating the occipital side of the patient, and an asymmetrical pattern; the positioning corner mark (14) includes at least three non-collinear positioning marks.

10. A system for assessing hair regeneration in target areas of alopecia areata, characterized in that: The system for performing the hair assessment method according to any one of claims 1 to 9, the system comprising: The standardized entity reference card is used to simultaneously enter the acquisition range of the evaluation image with the alopecia areata target area and provide scale, color, direction, and position references; An image acquisition module is used to acquire evaluation images containing the alopecia areata target area and a standardized entity reference card; The image processing module is used to identify the scale marker (11), color correction block (12), orientation positioning marker (13) and positioning corner mark (14) in the standardized entity reference card, and to correct the scale, color, orientation and position of the evaluation image; The parameter extraction module is used to extract at least one of the following features from the evaluation image: hair coverage features, hair density features, hair thickness features, and hair color features. The parameter calculation module is used to calculate the hair quantification parameters of the alopecia areata target area based on the parameter features acquired from the previous or first-time acquired evaluation image and the current evaluation image.

Images