Skin aging grid scoring system and method

Abstract

The present invention relates to the technical field of skin science assessment, and specifically to a grid-based scoring system and method for skin aging, comprising: a data acquisition module that uses a confocal microscope and a piezoelectric sensor array to collect microstructural images and elastic parameters of the skin surface; a feature extraction module that extracts fractal dimension features, the number of microcracks, and the elastic modulus; a physical information neural network module that maps physical features to a topological feature space, identifies dynamic patterns of skin aging through chaotic system analysis, and generates an aging stress index through quantum probability field calculation; a grid-based scoring module that generates a grid-based skin aging scoring result based on the aging stress index, achieving high-precision assessment at the micron level, establishing a multi-dimensional comprehensive evaluation system, and introducing physical information neural network technology to map physical features to a topological feature space. The grid-based scoring technology visualizes the spatial distribution of skin aging degrees and effectively distinguishes between two different aging types: photoaging and natural aging.

Description

Technical Field

[0001] The present invention relates to the technical field of skin science assessment, and in particular to a skin aging grid scoring system and method thereof, which are used to achieve accurate quantitative assessment of the degree of skin aging. Background Art

[0002] Skin aging is a complex physiological process, characterized by increased epidermal wrinkles, decreased skin elasticity, and the formation of microcracks. Traditional methods for assessing skin aging, including expert visual scoring and simple instrumental measurements, suffer from significant subjectivity, low precision, and poor repeatability.

[0003] With the advancement of medical imaging and sensor technologies, image analysis-based skin assessment methods have gradually emerged. However, existing technologies still have the following shortcomings: First, assessment accuracy is limited, typically only achieving millimeter-level resolution, which cannot capture microscopic skin changes; second, assessment methods are single-minded, often based on a single parameter (such as wrinkle depth or skin elasticity), lacking multi-dimensional comprehensive analysis; and third, assessment results lack spatial resolution, typically only providing an average value for the entire region and failing to reflect local differences.

[0004] Furthermore, existing technologies haven't fully leveraged advanced mathematical models and artificial intelligence algorithms to deeply analyze the characteristics of skin aging, resulting in a significant gap between assessment results and clinical practice. In particular, at the microscopic level, the nonlinear characteristics and complex dynamic changes of skin aging have yet to be effectively characterized and quantified. Summary of the Invention

[0005] The purpose of the present invention is to provide a skin aging grid scoring system and method thereof, aiming to overcome the shortcomings of the existing technology and achieve high-precision, multi-dimensional, grid-based assessment of the degree of skin aging.

[0006] The present invention proposes a skin aging grid scoring system, including:

[0007] A data acquisition module, used to collect microstructure images and elastic parameters of the skin surface;

[0008] a feature extraction module, communicatively connected to the data acquisition module, for extracting fractal dimension features from the microstructure image, and extracting the number of microcracks and elastic modulus from the elastic parameters;

[0009] a physical information neural network module, communicatively connected to the feature extraction module, configured to receive the fractal dimension feature, the number of microcracks, and the elastic modulus, map the fractal dimension feature, the number of microcracks, and the elastic modulus to a topological feature space, identify dynamic patterns of skin aging, and generate an aging stress index;

[0010] A gridded scoring module is communicatively connected to the physical information neural network module and is used to generate a gridded skin aging scoring result based on the aging stress index.

[0011] Preferably, the data acquisition module includes:

[0012] a confocal microscope unit for acquiring high-resolution images of the skin's surface microstructure;

[0013] A piezoelectric sensor array unit for measuring elastic parameters of the skin surface;

[0014] an image preprocessing unit, communicatively connected to the confocal microscope unit, for performing noise reduction, enhancement and standardization processing on the high-resolution image;

[0015] The elastic parameter preprocessing unit is communicatively connected to the piezoelectric sensor array unit and is used for filtering and standardizing the elastic parameters.

[0016] Preferably, the confocal microscope unit adopts a 633nm laser source and a 100x objective lens to collect skin surface microstructure images with a resolution of 0.2μm; the piezoelectric sensor array unit includes N rows and M columns of piezoelectric sensors to collect skin elasticity parameters with a pressure range of 0-10N.

[0017] Preferably, the feature extraction module includes:

[0018] A fractal dimension calculation unit, used for calculating the fractal dimension characteristics of the microstructure image using a nonlinear iterative box counting method;

[0019] a microcrack analysis unit for calculating the number of microcracks based on the electrical signal distribution of the piezoelectric sensor array;

[0020] The elastic modulus calculation unit is used to calculate the skin elastic modulus using the electrical signal measured by the piezoelectric sensor array.

[0021] Preferably, the fractal dimension calculation unit calculates the fractal dimension by the following steps:

[0022] The skin surface area was divided into 10 μm × 10 μm grid cells;

[0023] A nonlinear iterative box counting method is applied to each grid cell;

[0024] Establish a linear relationship between the box center spacing and the number of grid divisions;

[0025] The fractal dimension was calculated by fitting the equation of a straight line.

[0026] Preferably, the physical information neural network module includes:

[0027] A topological feature mapping unit, configured to map the fractal dimension feature, the number of microcracks, and the elastic modulus to a topological feature space;

[0028] a chaotic system analysis unit, communicatively connected to the topological feature mapping unit, for identifying dynamic patterns of skin aging based on the topological feature space;

[0029] A quantum probability field calculation unit is communicatively connected to the chaotic system analysis unit and is used to generate an aging stress index based on the dynamic mode.

[0030] Preferably, the topological feature mapping unit represents the physical features of each grid cell as a set of points in a topological space, and realizes feature conversion by mapping that maintains the topological structure; the chaotic system analysis unit distinguishes between photoaging and natural aging by identifying the attractor structure in the phase space; the quantum probability field calculation unit represents the grid cell as a quantum state, and describes the non-local interaction between distant grid cells through quantum correlation measurement.

[0031] Preferably, the grid scoring module includes:

[0032] a scoring generating unit, configured to map the aging stress index to a scoring scale of 0-10;

[0033] a visualization unit, communicatively connected to the score generating unit, and configured to generate a heat map of the aging score;

[0034] A regional evaluation unit, connected to the score generation unit for calculating average score values ​​of different functional areas;

[0035] The verification and tuning unit is in communication with the regional evaluation unit and is used to compare the system score with the clinical expert score and optimize the system parameters based on the comparison results.

[0036] Preferably, the verification and tuning unit optimizes system parameters by:

[0037] Establish a standard sample library covering different age groups and different aging types;

[0038] The consistency index between the system score and the clinical expert score was calculated;

[0039] Dynamically adjusting parameters of the physical information neural network based on the consistency index;

[0040] Regular assessments of the same subjects were performed to verify the accuracy of the system's predictions of the aging process.

[0041] The skin aging grid scoring method includes the following steps:

[0042] Collect microstructural images and elastic parameters of the skin surface;

[0043] extracting fractal dimension features from the microstructure image, and extracting the number of microcracks and elastic modulus from the elastic parameters;

[0044] Inputting the fractal dimension feature, the number of microcracks and the elastic modulus into a physical information neural network;

[0045] In the physical information neural network, the fractal dimension feature, the number of microcracks and the elastic modulus are mapped to a topological feature space;

[0046] identifying dynamic patterns of skin aging based on the topological feature space;

[0047] generating an aging stress index based on the dynamic model;

[0048] A gridded skin aging score result is generated based on the aging stress index.

[0049] The beneficial effects of the present invention include:

[0050] 1. It achieves high-precision assessment at the micron level. Through 10μm×10μm micro-grid division, it greatly improves the spatial resolution of skin aging assessment and can capture changes in skin structure at the microscopic level.

[0051] 2. A multi-dimensional comprehensive evaluation system was established. By combining three key parameters, namely fractal dimension characteristics, number of microcracks and elastic modulus, the system comprehensively characterizes the aging state of the skin, avoiding the limitations of single parameter evaluation.

[0052] 3. Innovatively introduced physical information neural network technology to map physical features into topological feature space, achieving accurate identification and quantification of dynamic patterns of skin aging, breaking through the static limitations of traditional evaluation methods.

[0053] 4. Through grid scoring technology, the spatial distribution of skin aging degree is visualized, which can accurately distinguish the aging differences in different areas and provide a scientific basis for precise anti-aging treatment.

[0054] 5. It can effectively distinguish between two different types of aging: photoaging and natural aging, providing an accurate diagnostic basis for targeted treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Schematic diagram of the overall architecture of the skin aging grid scoring system of the present invention;

[0056] Figure 2 Schematic diagram of the structure of the data acquisition module of the present invention;

[0057] Figure 3 Schematic diagram of the structure of the feature extraction module of the present invention;

[0058] Figure 4 This is a schematic diagram of the structure of the physical information neural network module of the present invention;

[0059] Figure 5 This is a schematic diagram of the structure of the grid scoring module of the present invention;

[0060] Figure 6 This is a flow chart of the skin aging grid scoring method of the present invention;

[0061] Figure 7 This is a schematic diagram of the fractal dimension calculation of the present invention;

[0062] Figure 8 This is a schematic diagram of topological feature mapping of the present invention;

[0063] Figure 9 This is a schematic diagram of the chaotic system analysis of the present invention;

[0064] Figure 10 This is an example of a gridded scoring heat map of the present invention. DETAILED DESCRIPTION

[0065] Please refer to Figures 1-10 , the specific implementation of the present invention is further described in detail below with reference to the accompanying drawings.

[0066] like Figure 1 As shown, the skin aging grid scoring system provided by the present invention includes a data acquisition module 1, a feature extraction module 2, a physical information neural network module 3 and a grid scoring module 4.

[0067] Data acquisition module 1 is used to collect microstructural images and elastic parameters of the skin surface. Feature extraction module 2 is communicatively connected to data acquisition module 1 and is used to extract fractal dimension features from the microstructural images and the number of microcracks and elastic modulus from the elastic parameters. Physical information neural network module 3 is communicatively connected to feature extraction module 2 and is used to receive fractal dimension features, number of microcracks, and elastic modulus, map these physical features into a topological feature space, identify dynamic patterns of skin aging, and generate an aging stress index. Gridded scoring module 4 is communicatively connected to physical information neural network module 3 and is used to generate a gridded skin aging score based on the aging stress index.

[0068] like Figure 2 As shown, the data acquisition module 1 includes a confocal microscope unit 11 , a piezoelectric sensor array unit 12 , an image preprocessing unit 13 and an elastic parameter preprocessing unit 14 .

[0069] The confocal microscope unit 11 is used to capture high-resolution images of the skin's surface microstructure. Preferably, the confocal microscope unit 11 utilizes a 633nm laser source and a 100x objective lens, capable of capturing images of the skin's surface microstructure with a resolution of 0.2μm. In one embodiment of the present invention, the confocal microscope unit 11 scans a 1mm×1mm area at a time, with a depth range of 0-200μm and a step size of 5μm. This setup fully captures microstructural changes from the epidermis to the dermis, providing high-quality raw data for subsequent analysis.

[0070] The piezoelectric sensor array unit 12 is used to measure elastic parameters of the skin surface. Preferably, the piezoelectric sensor array unit 12 includes N rows and M columns of piezoelectric sensors, which are used to collect skin elastic parameters within a pressure range of 0-10N. In one specific embodiment of the present invention, N = 10 and M = 10, forming a 10×10 sensor array with a sensor pitch of 100μm, a sensitivity of 0.5mV / N, and a sampling rate of 100Hz. This configuration precisely corresponds to the scanning area of ​​the confocal microscope, ensuring spatial consistency between the elastic parameters and the image data.

[0071] The image preprocessing unit 13 is in communication with the confocal microscope unit 11 and is used to perform noise reduction, enhancement, and normalization on high-resolution images. In one embodiment of the present invention, image preprocessing includes the following steps: first, image noise is removed using a Gaussian filter with a filter kernel size of 5×5 and a standard deviation σ = 1.5; second, image contrast is enhanced through histogram equalization; and finally, image grayscale values ​​are normalized to the range [0, 1] to facilitate subsequent processing. These preprocessing steps effectively improve image quality and enhance the discernibility of microstructures.

[0072] The elastic parameter preprocessing unit 14 is in communication with the piezoelectric sensor array unit 12 and is used to filter and normalize the elastic parameters. In one embodiment of the present invention, the elastic parameter preprocessing includes median filtering to remove outliers with a time window of 0.5 seconds, followed by parameter normalization using the Z-score method. This preprocessing method can effectively reduce the impact of measurement noise and improve the reliability of the elastic parameters.

[0073] like Figure 3 As shown, the feature extraction module 2 includes a fractal dimension calculation unit 21 , a microcrack analysis unit 22 and an elastic modulus calculation unit 23 .

[0074] The fractal dimension calculation unit 21 is used to calculate the fractal dimension characteristics of the microstructure image using a nonlinear iterative box counting method. In one embodiment of the present invention, the specific steps of the fractal dimension calculation are as follows:

[0075] First, the skin surface area was divided into 10 μm × 10 μm grid cells;

[0076] Then, a nonlinear iterative box counting method is applied to each grid cell;

[0077] Next, a linear relationship between the box center spacing and the number of grid divisions is established;

[0078] Finally, the fractal dimension is calculated by fitting the linear equation.

[0079] The calculation process can be expressed by the following mathematical formula:

[0080] For a given grid cell image, we use boxes of different sizes to cover the micro-wrinkle structure in the image and record the number of boxes required. Let the number of boxes with a side length of ε be N(ε), then the fractal dimension Df can be calculated by the following formula:

[0081] ,

[0082] in: is the fractal dimension, which characterizes the complexity of the skin microwrinkle structure; ε is the side length of the box, in micrometers (μm); N(ε) is the number of boxes required to cover the microwrinkle structure; log represents the natural logarithm.

[0083] In actual skin aging assessment applications, as age increases, the micro-wrinkle structure on the skin surface becomes more complex, and the fractal dimension Normally, healthy young skin The value is about 1.2-1.4, while the The value can reach 1.6-1.8. We use a nonlinear iterative method to set the initial grid edge length to 1.0μm and halve it each iteration until the number of boxes is 1. This can obtain a more accurate fractal dimension value, thereby accurately assessing the degree of aging of the skin microstructure.

[0084] The microcrack analysis unit 22 is used to calculate the number of microcracks based on the distribution of electrical signals from the piezoelectric sensor array. In one embodiment of the present invention, microcrack analysis is based on the fluctuation characteristics of the electrical signals generated by the piezoelectric sensor array. Specifically, when pressure is applied to the skin surface containing microcracks, the electrical signal at the microcracks will exhibit significant fluctuations. By setting a threshold value τ (preferably, τ = 2.5 times the standard deviation), abnormal fluctuation points in the signal can be identified, thereby determining the location and number of microcracks.

[0085] The calculation formula for the number of microcracks is as follows:

[0086] ,

[0087] in: is the number of microcracks in the kth grid element; is the electrical signal value of the sensor in row i and column j, in millivolts (mV); is the average value of all sensor electrical signals, in millivolts (mV); is the standard deviation in millivolts (mV); τ is the threshold coefficient, dimensionless, with an optimal value of 2.5; δ is the indicative function, which takes the value 1 when the condition is met and 0 otherwise; N is the number of rows in the sensor array; and M is the number of columns in the sensor array.

[0088] In the assessment of skin aging, microcracks are an important indicator of aging. Young and healthy skin usually has fewer microcracks ( <5, while the number of microcracks in aging skin, especially photoaging skin, increases significantly ( >15). This parameter is particularly important for assessing the degree of skin aging after long-term exposure to ultraviolet rays.

[0089] The elastic modulus calculation unit 23 is used to calculate the skin elastic modulus using the electrical signal measured by the piezoelectric sensor array. In one embodiment of the present invention, based on the relationship between pressure and deformation, the elastic modulus E can be calculated using the following formula:

[0090] ,

[0091] Where: E is the elastic modulus in Pascals (Pa); F is the applied pressure in Newtons (N); A is the contact area in square meters (m²); L is the skin thickness in meters (m); and ΔL is the deformation in meters (m).

[0092] In actual skin evaluation applications, F and ΔL can be obtained by converting the electrical signals of the piezoelectric sensor array, while A and L are known parameters. For standard test conditions, we set (corresponding to a 10μm×10μm grid unit), L is usually set according to the test site, and a typical value for facial skin is 1-2mm.

[0093] Skin elastic modulus is a key indicator for assessing the degree of aging. While the elastic modulus of healthy young skin typically ranges from 0.1 to 0.3 MPa, it can increase to 0.5 to 0.8 MPa in aged skin. This increase in elastic modulus indicates a loss of elasticity and a harderening of the skin, a typical sign of aging.

[0094] like Figure 4 As shown, the physical information neural network module 3 includes a topological feature mapping unit 31, a chaotic system analysis unit 32 and a quantum probability field calculation unit 33.

[0095] The topological feature mapping unit 31 is used to map the fractal dimension features, the number of microcracks, and the elastic modulus to the topological feature space. In one embodiment of the present invention, the topological feature mapping adopts the principle of homology equivalence, regards the physical features as manifolds in the topological space, and realizes feature conversion by mapping that maintains the topological structure. Specifically, for each grid cell, we construct a three-dimensional physical feature vector v=[Df, Ck, E], where Df is the fractal dimension, Ck is the number of microcracks, and E is the elastic modulus. Then, it is converted into a topological feature vector t through the nonlinear mapping function Φ:

[0096] ,

[0097] Where: t is the topological eigenvector with dimension n; is a nonlinear mapping function; is the i-th nonlinear basis function; v is the physical eigenvector, v=[Df,Ck,E]; n is the dimension of the topological feature space.

[0098] We usually choose radial basis function (RBF) as the nonlinear basis function, which is in the form of:

[0099] ,

[0100] in: is the output value of the i-th basis function; γ is a scaling parameter, dimensionless, that controls the width of the function; is a predefined center point with the same dimension as v; Indicates v and The Euclidean distance between them.

[0101] In skin aging assessment applications, the γ parameter is preferably set to 0.1. This value, determined through extensive experimentation, ensures smooth mapping while maintaining feature differentiation. The dimension n of the topological feature vector t is preferably set to 7. This setting enhances the expressive power of the features and more comprehensively describes the microscopic characteristics of skin aging.

[0102] Chaotic system analysis unit 32 is in communication with topological feature mapping unit 31 and is configured to identify dynamic patterns of skin aging based on the topological feature space. In one embodiment of the present invention, chaotic system analysis treats the skin aging process as a dynamic chaotic system and distinguishes different types of aging patterns through phase space reconstruction and attractor structure identification.

[0103] Specifically, the phase space is first constructed based on the topological eigenvector t:

[0104] ,

[0105] in: is a point in the phase space, representing the state of the system at time i; is the topological eigenvector at time i; τ is the time delay parameter, which represents the sampling interval; m is the embedding dimension, which represents the dimension of the phase space.

[0106] In skin aging assessment applications, the optimal τ value is determined using the mutual information function, typically 1 to 3 sampling intervals. The optimal m value is determined using the false nearest neighbor method, typically 3 to 5. These parameter settings effectively reconstruct the phase space of the skin aging system and reveal its inherent dynamics.

[0107] Then, the Lyapunov exponent λ is calculated to quantify the degree of chaos in the system:

[0108] ,

[0109] Where: λ is the dimensionless Lyapunov exponent, which characterizes the sensitivity of the system to initial conditions; X(t) and X'(t) are the states of two trajectories with a small initial distance in phase space at time t; ||X(t)-X'(t)|| represents the distance between the two trajectories at time t; and t is time.

[0110] In skin aging assessment, different aging types exhibit different λ values. Photoaging (primarily caused by UV rays) typically has a larger λ value (λ>0.5), indicating a more unstable system and more dramatic changes in the skin's microstructure. In contrast, natural aging (primarily caused by aging) has a smaller λ value (λ<0.5), indicating a relatively stable system and slower changes in the skin's microstructure. This characteristic enables us to distinguish different types of skin aging, providing a scientific basis for targeted treatment.

[0111] The quantum probability field calculation unit 33 is in communication with the chaotic system analysis unit 32 and is configured to generate an aging stress index based on the dynamic pattern. In one embodiment of the present invention, the quantum probability field calculation treats each grid cell as a quantum state and describes the nonlocal interactions between distant grid cells using a quantum correlation metric.

[0112] Specifically, the quantum state of grid cell k can be expressed as:

[0113] ,

[0114] in: is the quantum state of grid cell k; is the standard orthogonal basis of the Hilbert space; d is the dimension of the Hilbert space; is the complex amplitude, indicating the state The probability amplitude of .

[0115] The quantum state satisfies the normalization condition:

[0116] ,

[0117] In skin aging assessment applications, d is preferably set to 32 to provide sufficient state space to express the complexity of the skin microstructure. The value of αi is determined by the results of chaotic system analysis and reflects the probabilistic distribution characteristics of skin aging status.

[0118] The quantum correlation between grid cells can be calculated using the following formula:

[0119] ,

[0120] Where: C(k,l) is the quantum correlation between grid cells k and l, dimensionless; is the joint density matrix of grid cells k and l; and are their respective reduced density matrices; represents the trace operation of the matrix; log represents the logarithm operation of the matrix.

[0121] In skin aging assessment, the quantum correlation metric provides a method for describing the mutual influence of distant skin regions. When C(k,l)>0.3, two grid cells are considered to be significantly correlated, indicating that they may be in a similar aging state or influencing each other. This non-local correlation is more prevalent in aged skin, especially in areas exposed to long-term light.

[0122] Finally, based on the quantum correlation network, the aging stress index S is calculated:

[0123] ,

[0124] in: is the original aging stress index, dimensionless; K is the total number of grid elements; is the weight coefficient, ranging from 0 to 1, reflecting the importance of grid cell k; is the quantum state evaluation function; is the quantum state of grid cell k; is the association weight, ranging from 0 to 1; is the quantum correlation between grid cells k and l, ranging from 0 to 1.

[0125] To ensure the comparability of the aging stress index under different grid unit numbers, the system normalizes the original aging stress index to obtain the standardized aging stress index :

[0126] ,

[0127] in: is the normalized aging stress index of the kth grid cell, ranging from 0 to 1; is the original aging stress index; K is the total number of grid elements; is the number of significant quantum correlations, that is, The number of (k,l) pairs.

[0128] In skin aging assessment applications, Dynamically adjust according to the position of the grid unit in the skin tissue. The value is larger (about 0.6-0.8) because the aging characteristics of the surface skin are more obvious; the deep mesh The value is small (about 0.2-0.4) because the deep layer changes relatively slowly. This weight setting reflects the gradual aging of the skin from the surface to the deep layer.

[0129] like Figure 5 As shown, the grid-based scoring module 4 includes a scoring generation unit 41 , a visualization unit 42 , a region-by-region evaluation unit 43 and a verification and tuning unit 44 .

[0130] The scoring generating unit 41 is used to map the aging stress index to a scoring scale of 0-10. In one embodiment of the present invention, the scoring mapping adopts a piecewise linear function, and the specific formula is as follows:

[0131] ,

[0132] in: is the score of the kth grid cell, ranging from 0 to 10; is the aging stress index of the kth grid element; is a predefined lower threshold; is a predefined upper threshold.

[0133] In the application of skin aging assessment, based on a large amount of experimental data and clinical verification, the preferred setting =0.2, = 0.8. This setting enables the scoring scale to cover the whole range from young healthy skin ( ≈0-2) to moderately aged skin ( ≈3-7) to severely aging skin ( ≈8-10) to provide an intuitive indicator of aging degree.

[0134] The present invention adopts a multi-scale grid division strategy to achieve a balance between high-precision data acquisition and efficient data processing. Specifically, it includes:

[0135] (1) Microscopic scale: The confocal microscope uses a 10μm×10μm microscopic grid for image acquisition and feature analysis, forming a total of 100×100=10,000 microscopic grid cells within a 1mm×1mm scanning area. This microscopic scale ensures high-precision capture of the microstructural features of the skin.

[0136] (2) Mesoscale: To improve computational efficiency and enable effective visualization, the system aggregates 10×10 microgrid cells (100μm×100μm) into a mesoscale grid cell. This results in a total of 10×10=100 mesoscale grid cells within a 1mm×1mm scanning area. The piezoelectric sensor array uses the same spatial resolution as the mesoscale grid cells, i.e., a 10×10 sensor array with a sensor pitch of 100μm.

[0137] (3) Information aggregation: The system aggregates the feature information of 100 micro grid cells within each meso grid cell through the weighted average method, and calculates the comprehensive feature parameters of the meso grid cell for subsequent processing and visualization.

[0138] Visualization unit 42 is in communication with score generation unit 41 and is configured to generate a heat map of the aging score. In one embodiment of the present invention, the heat map is represented by a 10×10 meso-grid matrix, corresponding to a 10×10 piezoelectric sensor array distribution, with colors gradually varying from blue to red, corresponding to scores from low to high.

[0139] To enhance visualization accuracy, the system can also use an interpolation algorithm to expand the 10×10 meso-scale score to a 100×100 micro-scale score, generating a high-precision heat map. Furthermore, the system can generate a 3D surface map to visually display the spatial distribution of aging. This multi-scale visualization approach ensures high assessment accuracy while enabling intuitive and efficient data presentation.

[0140] The piezoelectric sensor array unit 12 consists of 10 rows and 10 columns of piezoelectric sensors, forming a 10×10 sensor array. The sensor spacing is 100 μm, corresponding to the size of the mesoscale grid cells. Each sensor is responsible for collecting the elastic parameters of its corresponding mesoscale grid cell (100 μm × 100 μm). This area contains 100 microscale grid cells (10 × 10 = 100). The elastic parameters collected by the sensors are preprocessed and used as the comprehensive elastic characteristics of the corresponding mesoscale grid cell.

[0141] This multi-scale gridding strategy ensures high-precision data acquisition while effectively reducing computational complexity and hardware costs, achieving an optimal balance between accuracy and efficiency. A microgrid (10μm×10μm) is used for high-precision image feature extraction, while a mesogrid (100μm×100μm) is used for sensor layout and visualization. A clear information aggregation mechanism maintains consistency between the two, ensuring the accuracy and reliability of the overall system assessment.

[0142] The regional assessment unit 43 is in communication with the score generation unit 41 and is used to calculate the average score values ​​for different functional areas. In one embodiment of the present invention, the skin area is divided into four functional areas: forehead, eye area, cheek, and chin. The average score and standard deviation of each area are calculated to generate a regional assessment report.

[0143] The verification and tuning unit 44 is in communication with the regional evaluation unit 43 and is used to compare the system score with the clinical expert score and optimize the system parameters based on the comparison results. In one embodiment of the present invention, the verification and tuning adopts the following method:

[0144] First, a standard sample library is established, encompassing different age groups and different aging types. Preferably, the sample library includes volunteers aged 20-70, with no fewer than 20 participants in each age group, covering both photoaging and natural aging.

[0145] Then, the consistency index κ between the system score and the clinical expert score was calculated:

[0146] ,

[0147] Where: κ is the Kappa consistency index, dimensionless; To observe the consistency, it indicates the proportion of the system score that is consistent with the expert score; is the expected consistency, which represents the expected proportion of random consistency.

[0148] In skin aging assessment applications, a value of κ > 0.75 indicates high agreement, 0.4 < κ < 0.75 indicates moderate agreement, and κ < 0.4 indicates low agreement. Through parameter optimization, this system was able to achieve a κ value of 0.85 or higher, indicating high consistency between the system and expert scores and suitable for use as a clinical diagnostic tool.

[0149] Then, the parameters of the physical information neural network are dynamically adjusted based on the consistency index. Preferably, a grid search method is used to find the optimal parameter combination in a predefined parameter space to maximize the κ value.

[0150] Finally, regular assessments are conducted on the same subjects to verify the accuracy of the system's predictions of the aging process. In one embodiment of the present invention, follow-up assessments are conducted every three months for at least one year. The correlation coefficient, r, between the predicted and measured values ​​is calculated, with a requirement of r > 0.8. This long-term follow-up assessment verifies the system's accuracy in predicting the progression of skin aging and provides an objective basis for evaluating the effectiveness of anti-aging treatments.

[0151] The following combination Figure 6 The skin aging grid scoring method of the present invention is described in detail. The method comprises the following steps:

[0152] Step S1: Acquire microstructure images and elastic parameters of the skin surface.

[0153] In one embodiment of the present invention, a confocal microscope is used to capture microstructural images of the skin surface, and a piezoelectric sensor array is used to capture elastic parameters of the skin surface. Preferably, the confocal microscope uses a 633 nm laser source and a 100x objective lens, and the piezoelectric sensor array uses a 10×10 array configuration.

[0154] Step S2: extracting fractal dimension features from the microstructure image, and extracting the number of microcracks and elastic modulus from the elastic parameters.

[0155] In one embodiment of the present invention, a nonlinear iterative box counting method is used to calculate fractal dimension features, a signal fluctuation analysis method is used to identify the number of microcracks, and a pressure-deformation relationship is used to calculate the elastic modulus. Together, these physical features provide a complete representation of the skin's aging state.

[0156] Step S3: Input the fractal dimension characteristics, the number of microcracks and the elastic modulus into the physical information neural network.

[0157] In one embodiment of the present invention, the three physical features are combined into a feature vector [Df, Ck, E], which is then normalized and fed into a physical information neural network. Preferably, a Z-score normalization method is employed to ensure that features of different dimensions can be effectively integrated.

[0158] Step S4: In the physical information neural network, the fractal dimension characteristics, the number of microcracks and the elastic modulus are mapped to the topological feature space.

[0159] In one embodiment of the present invention, radial basis functions are used to implement nonlinear mapping, mapping three-dimensional physical feature vectors to a seven-dimensional topological feature space to enhance the expressive power of features.

[0160] Step S5: Identifying dynamic patterns of skin aging based on the topological feature space.

[0161] In one embodiment of the present invention, two different aging dynamic patterns, photoaging and natural aging, are identified through phase space reconstruction and attractor structure analysis. Preferably, different aging types are distinguished based on the Lyapunov exponent, with a threshold of 0.5.

[0162] Step S6: generating an aging stress index based on the dynamic model.

[0163] In one embodiment of the present invention, quantum probability field theory is used to represent grid cells as quantum states, and quantum correlation metrics are used to describe the non-local interactions between grid cells to comprehensively generate an aging stress index.

[0164] Step S7: generating a gridded skin aging score result based on the aging stress index.

[0165] In one embodiment of the present invention, a piecewise linear mapping is used to convert the aging stress index into a scoring scale of 0-10 to generate a gridded scoring heat map and a regional assessment report.

[0166] Through the above steps, the present invention realizes accurate quantitative evaluation of the degree of skin aging, providing a scientific basis for skin health management and anti-aging treatment.

[0167] It should be noted that the embodiments described above are only some of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

Claims

1. A skin aging grid scoring system, characterized by: include: A data acquisition module, used to collect microstructure images and elastic parameters of the skin surface; a feature extraction module, communicatively connected to the data acquisition module, for extracting fractal dimension features from the microstructure image, and extracting the number of microcracks and elastic modulus from the elastic parameters; a physical information neural network module, communicatively connected to the feature extraction module, configured to receive the fractal dimension feature, the number of microcracks, and the elastic modulus, map the fractal dimension feature, the number of microcracks, and the elastic modulus to a topological feature space, identify dynamic patterns of skin aging, and generate an aging stress index; a gridded scoring module, communicatively connected to the physical information neural network module, for generating a gridded skin aging scoring result based on the aging stress index; The physical information neural network module includes: A topological feature mapping unit, configured to map the fractal dimension feature, the number of microcracks, and the elastic modulus to a topological feature space; a chaotic system analysis unit, communicatively connected to the topological feature mapping unit, for identifying dynamic patterns of skin aging based on the topological feature space; a quantum probability field calculation unit, communicatively connected to the chaotic system analysis unit, and configured to generate an aging stress index based on the dynamic pattern; The topological feature mapping unit represents the physical features of each grid cell as a set of points in the topological space, and realizes feature conversion by mapping that maintains the topological structure; the chaotic system analysis unit distinguishes between photoaging and natural aging by identifying the attractor structure in the phase space; the quantum probability field calculation unit represents the grid cell as a quantum state, and describes the non-local interaction between distant grid cells through quantum correlation measurement.

2. The skin aging grid scoring system according to claim 1, characterized in that: The data acquisition module includes: a confocal microscope unit for acquiring high-resolution images of the skin's surface microstructure; A piezoelectric sensor array unit for measuring elastic parameters of the skin surface; an image preprocessing unit, communicatively connected to the confocal microscope unit, for performing noise reduction, enhancement and standardization processing on the high-resolution image; The elastic parameter preprocessing unit is communicatively connected to the piezoelectric sensor array unit and is used for filtering and standardizing the elastic parameters.

3. The skin aging grid scoring system according to claim 2, characterized in that: The confocal microscope unit uses a 633nm laser source and a 100x objective lens to collect skin surface microstructure images with a resolution of 0.2μm; the piezoelectric sensor array unit includes N rows and M columns of piezoelectric sensors to collect skin elasticity parameters with a pressure range of 0-10N.

4. The skin aging grid scoring system according to claim 1, characterized in that: The feature extraction module includes: A fractal dimension calculation unit, used for calculating the fractal dimension characteristics of the microstructure image using a nonlinear iterative box counting method; a microcrack analysis unit for calculating the number of microcracks based on the electrical signal distribution of the piezoelectric sensor array; The elastic modulus calculation unit is used to calculate the skin elastic modulus through the electrical signal measured by the piezoelectric sensor array.

5. The skin aging grid scoring system according to claim 4, characterized in that: The fractal dimension calculation unit calculates the fractal dimension by the following steps: The skin surface area was divided into 10 μm × 10 μm grid cells; A nonlinear iterative box counting method is applied to each grid cell; Establish a linear relationship between the box center spacing and the number of grid divisions; The fractal dimension was calculated by fitting the equation of a straight line.

6. The skin aging grid scoring system according to claim 1, characterized in that: The grid scoring module includes: a scoring generating unit, configured to map the aging stress index to a scoring scale of 0-10; a visualization unit, communicatively connected to the score generating unit, and configured to generate a heat map of the aging score; A regional evaluation unit, connected to the score generation unit for calculating average score values ​​of different functional areas; The verification and tuning unit is in communication with the regional evaluation unit and is used to compare the system score with the clinical expert score and optimize the system parameters based on the comparison results.

7. The skin aging grid scoring system according to claim 6, characterized in that: The verification and tuning unit optimizes system parameters by: Establish a standard sample library covering different age groups and different aging types; The consistency index between the system score and the clinical expert score was calculated; Dynamically adjusting parameters of the physical information neural network based on the consistency index; Regular assessments of the same subjects were performed to verify the accuracy of the system's predictions of the aging process.

8. A grid-based scoring method for skin aging, using the system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Collect microstructural images and elastic parameters of the skin surface; extracting fractal dimension features from the microstructure image, and extracting the number of microcracks and elastic modulus from the elastic parameters; Inputting the fractal dimension feature, the number of microcracks and the elastic modulus into a physical information neural network; In the physical information neural network, the fractal dimension feature, the number of microcracks and the elastic modulus are mapped to a topological feature space; identifying dynamic patterns of skin aging based on the topological feature space; generating an aging stress index based on the dynamic model; A gridded skin aging score result is generated based on the aging stress index.

Images