Orthopedic surgery postoperative scar recovery dynamic evaluation system based on deep learning

Abstract

The invention relates to the technical field of deep learning, in particular to an orthopaedic surgery postoperative scar recovery dynamic evaluation system based on deep learning, and the system comprises the steps: obtaining gray images of a dermoscope at different sampling moments, and dividing the gray images to obtain a scar region and a capillary region around the scar; determining a recovery state index according to the area change of the microvessel texture distribution characteristic scar region in the microvessel region; when the hypertrophic scar appears, determining the area and the gray scale of the hypertrophic scar in the scar region, and determining the hyperplasia degree of the hypertrophic scar according to the area characteristic and the gray scale characteristic; and adjusting the recovery state index according to the hyperplasia degrees in the images with different gray levels to obtain a recovery evaluation coefficient, and determining a recovery evaluation result according to the recovery evaluation coefficients at different sampling moments. According to the invention, timely evaluation can be carried out in combination with the overall recovery state and the hyperplasia condition of the pathological scar area, and the accuracy and reliability of the evaluation result of scar recovery are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of deep learning technology, specifically to a deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery. Background Technology

[0002] Scarring is an inevitable product of the body's wound healing process. Traditional scar assessment methods mainly rely on doctors' clinical experience or dermoscopy. However, these methods cannot dynamically and comprehensively reflect the scar's recovery status. With continuous technological advancements, such as the development of deep learning combined with dermoscopy imaging, new technologies can capture and analyze images of the wound surface, thereby providing a comprehensive and dynamic analysis of scar recovery.

[0003] Currently, surgical scars in plastic surgery are classified into two types: physiological scars and pathological scars. Physiological scars are those that heal normally without excessive proliferation or defects. However, during the wound healing process, factors such as extracellular matrix metabolic imbalance, abnormal fibroblast activation, and inflammatory responses can lead to the formation of pathological scars. Existing methods for assessing scar recovery after plastic surgery rely on image analysis to identify and label scars. However, this approach cannot provide timely assessment of the overall recovery status and the extent of pathological scar proliferation, leading to inaccurate evaluations of postoperative scar recovery. Summary of the Invention

[0004] To address the technical problem in related technologies that fail to provide timely assessments of the overall recovery status and the proliferation of pathological scars, leading to biased evaluations of postoperative scar recovery in plastic surgery patients, this invention provides a deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery. The specific technical solution adopted is as follows: This invention proposes a deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery. The system includes: The acquisition module is used to acquire grayscale images of dermatoscopes at different sampling times during the scar recovery stage, perform edge detection on the grayscale images to obtain edge images, and divide the scar area and the microvascular area around the scar from the edge images; The recovery analysis module is used to determine the vascular distribution characteristic index of the microvascular region based on the microvascular texture density and distribution uniformity in the microvascular region; and to determine the recovery status index at the sampling time based on the changes in the vascular distribution characteristic index in adjacent grayscale images in time sequence and the area change of the scar region. The hyperplasia analysis module is used to determine whether hyperplastic scars have appeared based on the texture changes of scar areas in different grayscale images. When hyperplastic scars appear, it determines the area and grayscale of the hyperplastic scar in the scar area, and determines the degree of hyperplasia of the hyperplastic scar based on the area and grayscale characteristics. The evaluation module is used to adjust the restoration status index according to the degree of proliferation in different grayscale images to obtain restoration evaluation coefficients, and to determine the restoration evaluation result based on the restoration evaluation coefficients at different sampling times.

[0005] Further, edge detection is performed on the grayscale image to obtain an edge image, including: The grayscale image is subjected to edge detection based on the Canny edge detection algorithm to obtain edge pixels. The grayscale image is binarized based on the edge pixels to obtain an edge image, wherein the pixel value corresponding to the edge pixels is 1, and the pixel value corresponding to other pixels is 0.

[0006] Further, the scar region and the microvascular region surrounding the scar are segmented from the edge image, including: Based on a pre-trained image feature recognition model, the scar area and the microvascular area surrounding the scar are identified.

[0007] Furthermore, based on the microvascular texture density and distribution uniformity in the microvascular region, vascular distribution characteristic indicators of the microvascular region are determined, including: The proportion of edge pixels in the microvascular region is calculated as the microvascular texture density. The distribution uniformity index is determined based on the distance between any two edge pixels and the number distribution of edge pixels in different sub-regions of the microvascular region. The product of the microvascular texture density and distribution uniformity index is calculated and normalized to serve as the vascular distribution characteristic index of the microvascular region.

[0008] Furthermore, based on the distance between any two edge pixels and the distribution of edge pixels in different sub-regions within the microvascular region, a distribution uniformity index is determined, including: The microvascular region is divided into sub-regions of a preset size, and the variance of the number of edge pixels in each sub-region is determined as the dispersion coefficient of the number distribution. Calculate the mean distance and variance between any two edge pixels, and normalize the product of the mean distance and variance as the distance distribution dispersion coefficient. The sum of the dispersion coefficients of the quantity distribution and the dispersion coefficients of the distance distribution is used as the total dispersion index. The negative value of the total dispersion index is calculated and normalized to serve as the distribution uniformity index.

[0009] Furthermore, based on the changes in vascular distribution characteristics and scar area changes in adjacent grayscale images over time, recovery status indicators at the sampling time are determined, including: The difference between the blood vessel distribution feature index in the grayscale image at any sampling time and the previous sampling time is normalized and used as the blood vessel recovery coefficient at the corresponding sampling time. The scar recovery coefficient is obtained by calculating the area difference between the previous sampling time and the corresponding sampling time in the time series and normalizing the result. The sum of the vascular recovery coefficient and the scar recovery coefficient is normalized and used as an indicator of recovery status.

[0010] Furthermore, based on the texture changes of the scar area in different grayscale images, it is determined whether hypertrophic scars have occurred, including: Skeleton extraction is performed on scar regions in images of different grayscale values ​​to determine the scar skeleton pixels in each grayscale image. Scar skeleton pixels that belong to the next frame but not to the previous frame in the grayscale image of two adjacent frames are taken as hyperplastic pixels. The ratio of the number of proliferating pixels to the number of scar skeleton pixels in the corresponding subsequent grayscale image is calculated as the proliferation coefficient. When the hyperplasia coefficient is greater than a preset hyperplasia threshold, hyperplastic scars are determined to have occurred; otherwise, hyperplastic scars are determined not to have occurred.

[0011] Furthermore, based on area and grayscale characteristics, the degree of hypertrophic scarring is determined, including: The growth coefficient is used as an area feature; the gray-scale mean of all grown pixels is calculated, and the negative of the gray-scale mean is normalized and used as a gray-scale feature index. The product of the growth coefficient and the grayscale feature index is normalized and used as the degree of growth.

[0012] Furthermore, based on the degree of proliferation in different grayscale images, the restoration status index is adjusted to obtain the restoration evaluation coefficient, including: The product of the degree of proliferation and the preset influence weight is used as the adjustment index; The difference between the recovery status index and the adjustment index is calculated and used as the recovery assessment coefficient.

[0013] Furthermore, based on the recovery evaluation coefficients at different sampling times, the recovery evaluation results are determined, including: The mean of the recovery evaluation coefficients at all sampling times is calculated to obtain the overall recovery index; When the value of the total recovery index is greater than 0.7, the recovery assessment result is determined to be excellent; otherwise, the recovery assessment result is determined to be average.

[0014] The present invention has the following beneficial effects: This invention acquires grayscale images from dermoscopy at different sampling times during the scar recovery stage to identify scar and microvascular regions, enabling classification and analysis. Compared to direct overall scar analysis, this approach is more refined and reliable, resulting in more accurate analysis results. In the recovery analysis module, vascular distribution characteristic indicators are determined based on the microvascular texture density and uniformity in the microvascular region. Furthermore, recovery status indicators are determined based on temporal changes in these indicators and changes in scar area. These indicators combine microvascular texture and scar area features for enhanced accuracy. In the hyperplasia analysis module, the degree of hyperplasia is determined by analyzing the area and grayscale status of hyperplastic scars. In the evaluation module, the recovery status indicators are adjusted based on the degree of hyperplasia to comprehensively determine the recovery evaluation result. In summary, this invention enables timely evaluation by combining the overall recovery status and the hyperplasia of pathological scar areas, effectively improving the accuracy and reliability of scar recovery evaluation results. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a structural diagram of a deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery, provided in one embodiment of the present invention. Figure 2 This is a grayscale image schematic diagram of a microvascular region provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of the edge image of a microvascular region provided in one embodiment of the present invention; Figure 4 This is a schematic diagram of a grayscale image of a scar area provided in one embodiment of the present invention; Figure 5 This is a schematic diagram of the edge image of a scar region provided in one embodiment of the present invention; Figure 6 This is a schematic diagram of the edge image of a microvascular region under normal conditions, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of the deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery provided by this invention.

[0020] Please see Figure 1 The diagram illustrates a structural diagram of a deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery, according to an embodiment of the present invention. The system includes: an acquisition module 101, a recovery analysis module 102, a proliferation analysis module 103, and an assessment module 104. Further detailed analysis will follow.

[0021] The acquisition module 101 is used to acquire grayscale images of dermatoscope at different sampling times during the scar recovery stage, perform edge detection on the grayscale images to obtain edge images, and divide the scar area and the microvascular area around the scar from the edge images.

[0022] Scarring is an inevitable product of the body's wound healing process. Traditional scar assessment methods mainly rely on doctors' clinical experience or dermoscopy. However, these methods cannot dynamically and comprehensively reflect the scar's recovery status. With the continuous advancement of technology, such as the development of 3D imaging, new technologies can capture and analyze images of the wound surface, thereby providing a comprehensive and dynamic analysis of scar recovery.

[0023] Currently, surgical scars in plastic surgery are classified into two types: physiological scars and pathological scars. Physiological scars are those that heal normally without excessive proliferation or defects. However, during the wound healing process, factors such as extracellular matrix metabolic imbalance, abnormal fibroblast activation, and inflammatory responses can lead to the formation of pathological scars. Existing methods for assessing scar recovery after plastic surgery rely on image analysis to identify and label scars. However, these methods cannot provide timely assessments of the overall recovery status and the extent of pathological scar proliferation, resulting in inaccurate evaluations of postoperative scar recovery.

[0024] To address the aforementioned issues, in this embodiment of the invention, a dermoscope is used to acquire images of the skin surface. Subsequently, image preprocessing and grayscale conversion are performed to obtain grayscale images. Recovery and proliferation analyses are then conducted using the texture features in the grayscale images, thereby enabling accurate assessment of postoperative scar recovery in patients undergoing plastic surgery.

[0025] In this embodiment of the invention, edge detection is performed on the grayscale image to determine the edge image. Further, in some embodiments of the invention, the edge detection method can specifically use the Canny edge detection algorithm, that is, the grayscale image is edge detected based on the Canny edge detection algorithm to obtain edge pixels; the grayscale image is binarized based on the edge pixels to obtain the edge image, wherein the pixel value corresponding to the edge pixels is 1, and the pixel value corresponding to other pixels is 0.

[0026] The Canny edge detection algorithm can detect the edges of an image, thereby identifying edge pixels. These edge pixels are then subjected to binary processing to obtain an edge image. Specifically, in some embodiments of this invention, pixels with a value of 1 are considered edge pixels, with a grayscale value of 255 in the edge image, while non-edge pixels have a pixel value of 0, corresponding to a grayscale value of 0. This enables binary image analysis, facilitating subsequent analysis based on edge features.

[0027] Since the scar area mainly comprises two parts—the scar area awaiting recovery after surgical wound suturing and the affected microvascular area surrounding the scar area—scars exhibit dilated and disordered vascular distribution in dermoscopic images. In contrast, normal skin has uniformly distributed microvessels without abnormal dilation. Therefore, in this embodiment of the invention, the area with dilated and disordered vascular distribution is referred to as the microvascular area, which is mainly distributed around the scar area and presents an irregular shape. To obtain more accurate assessment results, separate analyses are necessary.

[0028] See Figures 2 to 5 , Figure 2 This is a grayscale image schematic diagram of a microvascular region provided in one embodiment of the present invention. Figure 3 This is a schematic diagram of the edge image of a microvascular region provided in one embodiment of the present invention. Figure 4 This is a schematic diagram of a grayscale image of a scar region provided in one embodiment of the present invention. Figure 5 This is a schematic diagram of the edge image of a scar region provided in one embodiment of the present invention. Figures 2 to 5 It can be seen that the scar area presents an area with obvious knife-like texture, while the microvascular area is an abnormal area caused by factors such as inflammatory response, and is usually located around the scar area.

[0029] The scar region and the microvascular region surrounding the scar are segmented from the edge image, including: the scar region and the microvascular region surrounding the scar are identified based on a pre-trained image feature recognition model.

[0030] Since the scar area and microvascular area have obvious texture features, a big data model can be directly trained based on the labeled image to obtain a pre-trained image feature recognition model. Then, based on this model, efficient recognition of the scar area and microvascular area can be achieved. This is well known to those skilled in the art and will not be elaborated here.

[0031] The recovery analysis module 102 is used to determine the vascular distribution characteristic index of the microvascular region based on the microvascular texture density and distribution uniformity in the microvascular region; and to determine the recovery status index at the sampling time based on the changes in the vascular distribution characteristic index in adjacent grayscale images in time series and the area change of the scar region.

[0032] In this embodiment of the invention, the recovery status can be analyzed by analyzing the changes in the microvascular texture distribution characteristics in the microvascular region obtained by dermatoscopy, as well as the changes in the scar itself. In this embodiment of the invention, the vascular distribution feature index characterizes the texture distribution of the microvascular region, and the grayscale state index characterizes the recovery status at the corresponding sampling time.

[0033] Furthermore, in some embodiments of the present invention, the vascular distribution characteristic index of the microvascular region is determined based on the microvascular texture density and distribution uniformity in the microvascular region, including: calculating the proportion of edge pixels in the microvascular region as the microvascular texture density; determining the distribution uniformity index based on the distance between any two edge pixels and the distribution of edge pixels in different sub-regions of the microvascular region; and calculating the product of the microvascular texture density and the distribution uniformity index, and normalizing it as the vascular distribution characteristic index of the microvascular region.

[0034] See Figure 6 , Figure 6 This is a schematic diagram of the edge image of a microvascular region under normal conditions, provided in an embodiment of the present invention; Figure 6 and Figure 3 By comparison, it is clear that the distribution of microvascular texture is more even under normal circumstances, while the uneven distribution of blood vessels due to scar wounds results in a lower blood vessel density under abnormal circumstances.

[0035] In this embodiment of the invention, the proportion of edge pixels in the microvascular region is calculated as the microvascular texture density. The higher the microvascular texture density, the more it conforms to the normal distribution characteristics of blood vessels.

[0036] In this embodiment of the invention, the analysis of distribution uniformity requires judgment based on the distance and number distribution of edge pixels in different sub-regions. That is, the microvascular region is divided into different sub-regions, and then the uniformity analysis is performed based on the number distribution of edge pixels in each sub-region.

[0037] Based on the distance between any two edge pixels and the distribution of edge pixels in different sub-regions of the microvascular region, a distribution uniformity index is determined, including: dividing the microvascular region into sub-regions of a preset size on an average basis, determining the variance of the number of edge pixels in each sub-region as the number distribution dispersion coefficient; calculating the mean and variance of the distance between any two edge pixels, normalizing the product of the mean and variance as the distance distribution dispersion coefficient; and using the sum of the number distribution dispersion coefficient and the distance distribution dispersion coefficient as the total dispersion index, calculating the negative of the total dispersion index, and normalizing it as the distribution uniformity index.

[0038] The preset size refers to the size of the sub-region. In this embodiment of the invention, the preset size can be determined to be 5×5, that is, the microvascular region is divided into different 5×5 sub-regions. Of course, it can also be adjusted to different forms such as 7×7 according to the actual situation, and there is no limitation on this.

[0039] Discreteness analysis is performed on the number of edge pixels in each sub-region. Specifically, the variance of the number of edge pixels in all sub-regions is calculated as the dispersion coefficient of the number distribution. The larger the value of the dispersion coefficient of the number distribution, the more uneven the distribution of the number of edge pixels in different sub-regions, that is, the lower the distribution uniformity.

[0040] The distances between different edge pixels are analyzed and divided into two categories: distance mean and distance variance. The distance mean represents the density feature, and the larger the distance mean, the smaller the density distribution. The distance variance represents the distance uniformity feature, and the larger the distance variance, the more uneven the distance distribution. Therefore, the product of the distance mean and the distance variance is normalized and used as the distance distribution dispersion coefficient.

[0041] Since the larger the values ​​of the coefficient of variation of the quantity distribution and the coefficient of variation of the distance distribution, the lower the overall distribution uniformity is, in this embodiment of the invention, the sum of the coefficient of variation of the quantity distribution and the coefficient of variation of the distance distribution is directly calculated as the total dispersion index. The negative value of the total dispersion index is calculated and normalized as the distribution uniformity index, so that the larger the value of the distribution uniformity index, the more it conforms to the characteristics of a uniform distribution.

[0042] The product of microvascular texture density and distribution uniformity index is calculated and normalized to serve as a vascular distribution characteristic index for the microvascular region. The larger the value of the vascular distribution characteristic index, the more normal the overall microvascular distribution is.

[0043] Furthermore, in some embodiments of the present invention, the recovery status index at a sampling time is determined based on the changes in vascular distribution feature indicators in adjacent grayscale images over time and the changes in the area of ​​the scar region. This includes: normalizing the difference between the vascular distribution feature indicators in the grayscale image at any sampling time and the previous sampling time to obtain the vascular recovery coefficient at the corresponding sampling time; calculating the difference in the area of ​​the scar region between the previous sampling time and the corresponding sampling time over time and normalizing it to obtain the scar recovery coefficient; and normalizing the sum of the vascular recovery coefficient and the scar recovery coefficient to obtain the recovery status index.

[0044] Among them, the vascular recovery coefficient represents the degree to which microvessels tend to return to normal between two adjacent sampling times. The larger the value of the vascular recovery coefficient, the larger the value of the vascular distribution characteristic index, that is, the more normal the overall microvascular distribution is and the more obvious the recovery effect.

[0045] In this embodiment of the invention, it is necessary to analyze the temporal change in the area of ​​the scar region. As healing progresses, the scar gradually shrinks until it scabs over. Therefore, the area of ​​the scar region gradually decreases. The difference in area between two adjacent sampling times is used as the scar recovery coefficient. The larger the scar recovery coefficient, the faster the scar recovers and the more obvious the recovery effect between two adjacent sampling times.

[0046] In summary, the sum of the vascular recovery coefficient and the scar recovery coefficient was calculated and normalized as an indicator of recovery status.

[0047] The hyperplasia analysis module 103 is used to determine whether hyperplastic scars have appeared based on the texture changes of scar areas in different grayscale images. When hyperplastic scars appear, it determines the area and grayscale of the hyperplastic scar in the scar area, and determines the degree of hyperplasia of the hyperplastic scar based on the area and grayscale features.

[0048] When analyzing scars, it is also necessary to conduct a specific analysis of scar hyperplasia. Scar hyperplasia is mainly caused by the scar itself due to factors such as extracellular matrix metabolic imbalance, abnormal activation of fibroblasts, and inflammatory response, resulting in pathological scars. It is necessary to constantly analyze whether scar hyperplasia occurs.

[0049] Furthermore, in some embodiments of the present invention, determining whether hypertrophic scars have occurred based on the texture changes of scar regions in different grayscale images includes: extracting the skeleton of the scar region in different grayscale images to determine the scar skeleton pixels in each grayscale image; taking the scar skeleton pixels belonging to the next frame but not to the previous frame in two adjacent frames as hypertrophic pixels; calculating the ratio of the number of hypertrophic pixels to the number of scar skeleton pixels in the corresponding next frame grayscale image as a hypertrophy coefficient; determining that hypertrophic scars have occurred when the hypertrophy coefficient is greater than a preset hypertrophy threshold, otherwise determining that no hypertrophic scars have occurred.

[0050] In this process, skeleton extraction can identify scar skeleton pixels. Skeleton extraction can specifically use morphological erosion to extract skeleton information, or it can directly use a skeleton extraction algorithm to achieve skeleton analysis. In this embodiment of the invention, the analysis is mainly focused on the scar hyperplasia generated in adjacent images. Therefore, scar skeleton pixels that are present in the later frame but not in the previous frame are taken as hyperplastic pixels.

[0051] The more hyperplastic pixels there are, the more severe the hyperplastic scar is. In this embodiment of the invention, the ratio of the number of hyperplastic pixels to the number of scar skeleton pixels in the corresponding subsequent grayscale image is calculated as the hyperplasia coefficient.

[0052] The preset hyperplasia threshold is a threshold value for the hyperplasia coefficient. In this embodiment of the invention, the preset hyperplasia threshold can be, for example, 0.15. That is, when the hyperplasia coefficient is greater than 0.15, it is determined that hyperplastic scars have occurred; otherwise, it is determined that no hyperplastic scars have occurred, and the skeletal changes may be caused by errors during image analysis.

[0053] When hypertrophic scars are present, the degree of hyperplasia is determined based on area and grayscale features, including: using the hyperplasia coefficient as an area feature; calculating the mean grayscale value of all hyperplastic pixels, normalizing the negative of the mean grayscale value as a grayscale feature index; and normalizing the product of the hyperplasia coefficient and the grayscale feature index as the degree of hyperplasia.

[0054] The larger the area of ​​the hypertrophic scar and the lower the gray value, the more severe the hypertrophic scar. The lower the gray value, the darker the overall color of the hypertrophic scar. Newly formed hypertrophic scars are usually red. A dark color indicates that the scar has been present for a long time and the scar characteristics are more severe.

[0055] Based on this calculation, the degree of proliferation is obtained. The higher the value of the degree of proliferation, the more severe the proliferative effect.

[0056] The evaluation module 104 is used to adjust the restoration status index according to the degree of proliferation in different grayscale images to obtain the restoration evaluation coefficient, and to determine the restoration evaluation result based on the restoration evaluation coefficient at different sampling times.

[0057] In this embodiment of the invention, the recovery status index can be adjusted by combining the degree of proliferation, that is, numerical suppression. When the value of the degree of proliferation is larger, it indicates that the recovery status is still poor. At this time, it is necessary to reduce the recovery status index.

[0058] Furthermore, in some embodiments of the present invention, the product of the degree of proliferation and a preset influence weight is used as an adjustment index; the difference between the recovery state index and the adjustment index is calculated as a recovery evaluation coefficient.

[0059] The preset influence weight is the weight of the degree of proliferation on the recovery status index. Specifically, the preset influence weight can be, for example, 0.2. That is, the product of the degree of proliferation and 0.2 is first calculated as the adjustment index, and then the difference between the recovery status index and the adjustment index is used as the recovery evaluation coefficient.

[0060] The recovery assessment coefficient is an evaluation value obtained by combining scar hyperplasia and microvascular regional texture features. The recovery assessment coefficient can characterize the overall state of the scar at the corresponding sampling time, and the recovery assessment of the scar can be realized based on the recovery assessment coefficient.

[0061] The recovery assessment result is determined based on the recovery assessment coefficients at different sampling times, including: calculating the mean of the recovery assessment coefficients at all sampling times to obtain the total recovery index; when the value of the total recovery index is greater than 0.7, the recovery assessment result is determined to be excellent; otherwise, the recovery assessment result is determined to be average.

[0062] Of course, in other embodiments of the present invention, various other numerical classification standards may be set, or the total recovery index may be directly used as the recovery assessment result, without any limitation.

[0063] This invention acquires grayscale images from dermoscopy at different sampling times during the scar recovery stage to identify scar and microvascular regions, enabling classification and analysis. Compared to direct overall scar analysis, this approach is more refined and reliable, resulting in more accurate analysis results. In the recovery analysis module, vascular distribution characteristic indicators are determined based on the microvascular texture density and uniformity in the microvascular region. Furthermore, recovery status indicators are determined based on temporal changes in these indicators and changes in scar area. These indicators combine microvascular texture and scar area features for enhanced accuracy. In the hyperplasia analysis module, the degree of hyperplasia is determined by analyzing the area and grayscale status of hyperplastic scars. In the evaluation module, the recovery status indicators are adjusted based on the degree of hyperplasia to comprehensively determine the recovery evaluation result. In summary, this invention enables timely evaluation by combining the overall recovery status and the hyperplasia of pathological scar areas, effectively improving the accuracy and reliability of scar recovery evaluation results.

[0064] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0065] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery, characterized in that, The system includes: The acquisition module is used to acquire grayscale images of dermatoscopes at different sampling times during the scar recovery stage, perform edge detection on the grayscale images to obtain edge images, and divide the scar area and the microvascular area around the scar from the edge images; The recovery analysis module is used to determine the vascular distribution characteristic index of the microvascular region based on the microvascular texture density and distribution uniformity in the microvascular region; and to determine the recovery status index at the sampling time based on the changes in the vascular distribution characteristic index in adjacent grayscale images in time sequence and the area change of the scar region. The hyperplasia analysis module is used to determine whether hyperplastic scars have appeared based on the texture changes of scar areas in different grayscale images. When hyperplastic scars appear, it determines the area and grayscale of the hyperplastic scar in the scar area, and determines the degree of hyperplasia of the hyperplastic scar based on the area and grayscale characteristics. The evaluation module is used to adjust the restoration status index according to the degree of proliferation in different grayscale images to obtain restoration evaluation coefficients, and to determine the restoration evaluation result based on the restoration evaluation coefficients at different sampling times.

2. The deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, Edge detection is performed on the grayscale image to obtain an edge image, including: The grayscale image is subjected to edge detection based on the Canny edge detection algorithm to obtain edge pixels. The grayscale image is binarized based on the edge pixels to obtain an edge image, wherein the pixel value corresponding to the edge pixels is 1, and the pixel value corresponding to other pixels is 0.

3. The deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, The scar region and the microvascular region surrounding the scar are delineated from the edge image, including: Based on a pre-trained image feature recognition model, the scar area and the microvascular area surrounding the scar are identified.

4. The deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery as described in claim 2, characterized in that, Based on the microvascular texture density and distribution uniformity in the microvascular region, the vascular distribution characteristic indicators of the microvascular region are determined, including: The proportion of edge pixels in the microvascular region is calculated as the microvascular texture density. The distribution uniformity index is determined based on the distance between any two edge pixels and the number distribution of edge pixels in different sub-regions of the microvascular region. The product of the microvascular texture density and distribution uniformity index is calculated and normalized to serve as the vascular distribution characteristic index of the microvascular region.

5. The deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery as described in claim 4, characterized in that, Based on the distance between any two edge pixels and the distribution of edge pixels in different sub-regions within the microvascular region, a distribution uniformity index is determined, including: The microvascular region is divided into sub-regions of a preset size, and the variance of the number of edge pixels in each sub-region is determined as the dispersion coefficient of the number distribution. Calculate the mean distance and variance between any two edge pixels, and normalize the product of the mean distance and variance as the distance distribution dispersion coefficient. The sum of the dispersion coefficients of the quantity distribution and the dispersion coefficients of the distance distribution is used as the total dispersion index. The negative value of the total dispersion index is calculated and normalized to serve as the distribution uniformity index.

6. The deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, Based on the changes in vascular distribution characteristics and scar area changes in adjacent grayscale images over time, recovery status indicators at the sampling time are determined, including: The difference between the blood vessel distribution feature index in the grayscale image at any sampling time and the previous sampling time is normalized and used as the blood vessel recovery coefficient at the corresponding sampling time. The scar recovery coefficient is obtained by calculating the area difference between the previous sampling time and the corresponding sampling time in the time series and normalizing the result. The sum of the vascular recovery coefficient and the scar recovery coefficient is normalized and used as an indicator of recovery status.

7. The deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, Based on the texture changes of scar areas in different grayscale images, determine whether hypertrophic scars have occurred, including: Skeleton extraction is performed on scar regions in images of different grayscale values ​​to determine the scar skeleton pixels in each grayscale image. Scar skeleton pixels that belong to the next frame but not to the previous frame in the grayscale image of two adjacent frames are taken as hyperplastic pixels. The ratio of the number of proliferating pixels to the number of scar skeleton pixels in the corresponding subsequent grayscale image is calculated as the proliferation coefficient. When the hyperplasia coefficient is greater than a preset hyperplasia threshold, hyperplastic scars are determined to have occurred; otherwise, hyperplastic scars are determined not to have occurred.

8. The deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery as described in claim 7, characterized in that, The degree of hypertrophic scarring is determined based on area and grayscale characteristics, including: The growth coefficient is used as an area feature; the gray-scale mean of all grown pixels is calculated, and the negative of the gray-scale mean is normalized and used as a gray-scale feature index. The product of the growth coefficient and the grayscale feature index is normalized and used as the degree of growth.

9. The deep learning-based dynamic evaluation system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, Based on the degree of proliferation in different grayscale images, the recovery status index is adjusted to obtain the recovery evaluation coefficient, including: The product of the degree of proliferation and the preset influence weight is used as the adjustment index; The difference between the recovery status index and the adjustment index is calculated and used as the recovery assessment coefficient.

10. The deep learning-based dynamic assessment system for postoperative scar recovery in plastic surgery as described in claim 1, characterized in that, Based on the recovery evaluation coefficients at different sampling times, the recovery evaluation results are determined, including: The mean of the recovery evaluation coefficients at all sampling times is calculated to obtain the overall recovery index; When the value of the total recovery index is greater than 0.7, the recovery assessment result is determined to be excellent; otherwise, the recovery assessment result is determined to be average.

Images