Ultraviolet treatment parameter generation system based on automatic recognition of skin type

Abstract

The application belongs to the technical field of ultraviolet light therapy, and discloses an ultraviolet treatment parameter generation system based on automatic identification of skin types; the system comprises a skin state sensing module, which is used for generating a skin static holographic snapshot, triggering a micro-energy pre-irradiation action on a target skin area through the skin static holographic snapshot, and obtaining a photosensitive response dynamic set; an intelligent feature analysis module, which is used for obtaining a photosensitive skin feature-phenotype joint vector, performing multi-objective deduction on the photosensitive skin feature-phenotype joint vector through a pre-constructed comprehensive evaluation criterion, and generating an optimal spectral power distribution and an optimal dose matrix; an adaptive optical control module, which is used for driving an adjustable double-band light source to work, and inversely deriving effective ultraviolet energy actually reaching a dermis layer to form a steady-state treatment driving current; and a thermal safety monitoring and defense module, which is used for controlling the running state of the adjustable double-band light source through an epidermal temperature field, and realizing individual difference treatment.

Description

Technical Field

[0001] This invention relates to the field of ultraviolet phototherapy technology, and more specifically, to an ultraviolet therapy parameter generation system based on automatic skin type recognition. Background Technology

[0002] Ultraviolet phototherapy is an effective clinical treatment for various chronic inflammatory skin diseases such as psoriasis, vitiligo, and atopic dermatitis. It uses photon energy of specific wavelengths to penetrate skin tissue, inducing T lymphocyte apoptosis, inhibiting excessive proliferation of keratinocytes, and regulating the local immune microenvironment. In clinical practice, precise dosage setting is a key factor determining efficacy and safety.

[0003] In existing technologies, the commonly used method for determining dosage mainly relies on the Fitzpatrick skin typing system. This system classifies people into types I to VI by inquiring about their response to sunlight and then uses a dosage reference table to determine the initial irradiation parameters. However, the traditional Fitzpatrick skin typing system only divides the complex photobiological characteristics of the skin into six discrete levels, which is insufficient to reflect the molecular-level differences in individual responses to ultraviolet radiation. For example, there are significant differences in the molecular level of individual skin, such as melanin concentration, heme distribution, collagen content, and photosensitivity. These differences directly determine an individual's tolerance to ultraviolet radiation, absorption efficiency, and treatment response. The six-level macro-typing system cannot capture these molecular-level differences, resulting in treatment parameters that cannot be accurately matched to the actual condition of the individual's skin. This can lead to undertreatment or overdose, seriously affecting the effectiveness and safety of treatment.

[0004] In view of this, the present invention proposes an ultraviolet therapy parameter generation system based on automatic skin type recognition to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a UV therapy parameter generation system based on automatic skin type recognition, comprising:

[0006] The skin state perception module is used to acquire relevant detection data of the target skin area and perform synchronous alignment and registration to generate a static holographic snapshot of the skin. It also triggers a micro-energy pre-irradiation action on the target skin area through the static holographic snapshot of the skin to obtain a dynamic set of photosensitivity response.

[0007] The intelligent feature analysis module is used to fuse features from the dynamic set of photosensitivity response and the static holographic snapshot of the skin to obtain the joint vector of photosensitivity skin features and phenotype. The joint vector of photosensitivity skin features and phenotype is then used to perform multi-objective deduction through a pre-constructed comprehensive evaluation criterion to generate the optimal spectral power distribution and the optimal dose matrix.

[0008] The adaptive optics control module is used to drive the tunable dual-band light source based on the optimal spectral power distribution and the optimal dose matrix, and to inversely calculate the effective ultraviolet energy that actually reaches the dermis. The effective ultraviolet energy is used to dynamically correct the driving current of the tunable dual-band light source to form a steady-state treatment driving current.

[0009] The thermal safety monitoring and defense module is used to monitor the epidermal temperature field of the target skin area when the steady-state treatment drive current is continuously output, and to control the operation status of the adjustable dual-band light source through the epidermal temperature field.

[0010] Furthermore, a static holographic snapshot of the skin is generated, including:

[0011] The reflectance spectrum and autofluorescence signal of the target skin area were collected; the reflectance spectrum was analyzed to obtain the detection data of melanin, heme and water distribution in the epidermis of the target skin area; the autofluorescence signal was inverted to obtain the detection data of collagen content and porphyrin content in the dermis of the target skin area.

[0012] The target skin area is scanned to obtain data on the thickness of the stratum corneum and the depth of inflammatory infiltration. Using the skin texture feature points of the target skin area as anchor points, the detection data are spatially registered and fused at the pixel level to generate a static holographic snapshot of the target skin area.

[0013] Furthermore, triggering micro-energy pre-irradiation of the target skin area through a static holographic snapshot of the skin includes:

[0014] The minimum erythema dose for the target skin area was estimated using data on stratum corneum thickness and inflammatory infiltration depth to determine the baseline tolerable dose for the target skin area. A preset safety factor was introduced to reduce and adjust the baseline tolerable dose to obtain the micro-energy pre-irradiation dose for the target skin area.

[0015] Based on melanin distribution detection data, abnormal melanin areas are identified, and the set of pixel spatial coordinates of the abnormal melanin areas is extracted as the target projection area. A single pulse of ultraviolet light is emitted to the target projection area by an adjustable dual-band light source with micro-energy pre-irradiation dose control, thus completing the micro-energy pre-irradiation action of the target skin area.

[0016] Furthermore, the dynamic set of photosensitive responses is obtained, including:

[0017] The instantaneous values ​​of heme concentration and autofluorescence signal intensity in the target skin area were continuously collected after the micro-energy pre-irradiation action was completed.

[0018] The instantaneous values ​​of heme concentration before and after micro-energy pre-irradiation were compared and the rate of change curve was derived. The peak rate in the rate of change curve was extracted as the vasodilation response rate. An exponential fitting analysis was performed on the decay curve of the instantaneous value of autofluorescence signal intensity to determine the fluorescence lifetime decay constant.

[0019] The combined vasodilation response rate and fluorescence lifetime decay constant form a dynamic set of photosensitivity responses in the target skin region.

[0020] Furthermore, the joint vector of photosensitive skin features and phenotypes is obtained, including:

[0021] Extract various types of data from the dynamic set of photosensitivity response and the static holographic snapshot of the skin and perform range standardization; then concatenate the standardized data of each dimension in sequence to obtain the joint vector of photosensitivity skin features-phenotype.

[0022] Furthermore, multi-objective extrapolation is performed on the joint vector of photosensitive skin features and phenotypes, including:

[0023] A pre-constructed comprehensive evaluation criterion is included, which contains a positive benefit term for quantifying the improvement rate of psoriasis area, a negative loss term for quantifying the erythema grade and the increase of DNA cyclobutanepyrimidine dimer, and a penalty term for quantifying the minimum effective dose.

[0024] Within the spectral bands and dose intensity range of the tunable dual-band light source, multiple sets of candidate spectral power distribution curves and dose matrices were established.

[0025] Using the joint vector of photosensitive skin features and phenotypes as constraints, a multi-objective iterative optimization of each candidate spectral power distribution curve and dose matrix is ​​performed based on a comprehensive evaluation criterion to obtain the expected index set.

[0026] Each expected index set is incorporated into the comprehensive evaluation criteria for deduction, and the comprehensive evaluation value of each group of candidate spectral power distribution curves and dose matrices is obtained; the candidate spectral power distribution curve and dose matrix with the largest comprehensive evaluation value are selected as the optimal spectral power distribution curve and the optimal dose matrix.

[0027] Furthermore, the effective ultraviolet energy actually reaching the dermis is determined, including:

[0028] The optimal spectral power distribution curve is mapped to the initial current duty cycle of the long-wave and medium-wave ultraviolet channels; the optimal dose matrix is ​​mapped to the target output power intensity.

[0029] The initial current duty cycle and the target output power intensity are pulse-width modulated to form the initial driving current and drive the adjustable dual-band light source to project ultraviolet light onto the target skin area.

[0030] The diffuse reflectance spectral signal of the target skin region is collected during the operation of the adjustable dual-band light source and the global energy is summarized to form the diffuse reflectance energy loss value; the output power of the initial driving current is subjected to energy evolution to obtain the theoretical emission energy; the loss of the diffuse reflectance energy loss value is removed from the theoretical emission energy to obtain the effective ultraviolet energy that actually reaches the dermis.

[0031] Furthermore, a steady-state therapeutic driving current is formed, including:

[0032] The stratum corneum thickness detection data in the static holographic snapshot of the skin is coupled with the preset stratum corneum light attenuation coefficient to determine the ultraviolet transmission coefficient through the stratum corneum; the transmission coefficient is used to adjust the magnification of each target output dose in the optimal dose matrix to obtain the target dose of the dermis.

[0033] The deviation between the effective ultraviolet energy and the target dose in the dermis is used as the energy deviation signal for calculation, and the calculated current compensation value is superimposed on the initial driving current to generate a steady-state therapeutic driving current.

[0034] Furthermore, monitoring the epidermal temperature field of the target skin area during continuous steady-state therapeutic drive current output includes:

[0035] Acquire each frame of thermal radiation distribution image of the target skin area under steady-state treatment driving current working state, and convert each frame of thermal radiation distribution image into an epidermal temperature field containing the temperature value of each pixel;

[0036] A full-domain scan of the epidermal temperature field is performed for each frame, and the highest temperature value obtained from the scan is determined as the peak epidermal temperature; the slope of the change of the peak epidermal temperature in the sliding time window is deduced as the epidermal temperature rise rate.

[0037] Furthermore, the operating status of the adjustable dual-band light source is controlled, including:

[0038] A preset thermal damage threshold is included, comprising an absolute safe temperature threshold and a maximum temperature rise rate threshold. The peak epidermal temperature and epidermal temperature rise rate are compared with the thermal damage threshold. If the peak epidermal temperature is greater than the absolute safe temperature threshold, or the epidermal temperature rise rate is greater than the maximum temperature rise rate threshold, it is determined that there is an immediate risk of thermal burn in the target skin area, and the steady-state treatment drive current is immediately interrupted. If the peak epidermal temperature is not greater than the absolute safe temperature threshold, and the epidermal temperature rise rate is not greater than the maximum temperature rise rate threshold, it is determined that the body is in a thermally safe state, and the steady-state treatment drive current continues to be output.

[0039] The technical effects and advantages of the ultraviolet therapy parameter generation system based on automatic skin type recognition of this invention are as follows:

[0040] 1. This invention collects reflectance spectrum, autofluorescence signal and tomographic data of the target skin area through a skin state sensing module to generate a static holographic snapshot of the skin. Through micro-energy pre-irradiation, it actively stimulates and captures the dynamic set of photosensitivity responses of skin microcirculation and fluorescence, realizing a leap from static phenotypic observation to dynamic physiological stress detection. It accurately quantifies the molecular-level differences in individual dimensions such as vasodilation response ability, melanocyte activity and proliferation ability, and immune cell activity, and truly reflects the biological tolerance limit of the skin to ultraviolet radiation. This provides accurate data support for the generation of subsequent personalized treatment parameters, avoids parameter adaptation deviation caused by incomplete feature collection from the source, and significantly improves the safety starting point of treatment.

[0041] 2. This invention constructs a joint vector of photosensitive skin features and phenotypes through an intelligent feature analysis module, and performs multi-objective iterative optimization using comprehensive evaluation criteria to deduce the optimal spectral power distribution and optimal dose matrix that balances treatment benefits and tissue protection. This achieves intelligent identification of skin types and personalized deduction of treatment parameters. Through real-time diffuse reflection inversion and thermal safety monitoring by the adaptive optics control module, it can eliminate the absorption loss of the stratum corneum, directly lock the effective energy of the dermis for closed-loop correction, further improving the accuracy of treatment parameter adaptation. This effectively solves the problem of insufficient treatment or overload burns caused by individual differences, and significantly improves the effectiveness and compliance of phototherapy. Attached Figure Description

[0042] Figure 1 This is a system schematic diagram of the ultraviolet therapy parameter generation system based on automatic skin type recognition of the present invention;

[0043] Figure 2 This is a flowchart illustrating the micro-energy pre-irradiation process of the present invention;

[0044] Figure 3 This is a flowchart for determining the operating status of the adjustable dual-band light source according to the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0046] Example 1, please refer to Figure 1 , Figure 2 and Figure 3 As shown in the figure, the main design contents of the ultraviolet therapy parameter generation system based on automatic skin type recognition described in this embodiment are as follows:

[0047] The traditional Fitzpatrick classification (6 levels) is based solely on skin color and sun exposure response (such as susceptibility to sunburn and tanning) for macroscopic classification, dividing the population into types I to VI. This low-resolution macroscopic classification method is highly subjective and experience-dependent. However, individual skin differs significantly at the molecular level in terms of melanin concentration, heme distribution, collagen content, and photosensitivity. These differences determine an individual's tolerance to ultraviolet radiation, absorption efficiency, and treatment response.

[0048] For example, melanin in the epidermis acts as an endogenous ultraviolet absorber, and its content directly affects the penetration depth of photon energy into the basal layer; the distribution of heme is closely related to skin microcirculation perfusion, indirectly reflecting the metabolic activity of the inflamed area; the content and cross-linking structure of collagen in the dermis not only affect the scattering path of photons in the tissue, but also characterize the degree of skin aging and repair potential through its autofluorescence properties.

[0049] Based on this, a UV therapy parameter generation system based on automatic skin type recognition was designed, including:

[0050] The skin state perception module is used to acquire relevant detection data of the target skin area and perform synchronous alignment and registration to generate a static holographic snapshot of the skin. It also triggers a micro-energy pre-irradiation action on the target skin area through the static holographic snapshot of the skin to obtain a dynamic set of photosensitivity response.

[0051] Generate a static holographic snapshot of the skin, including:

[0052] The reflectance spectrum and autofluorescence signal of the target skin area are acquired. A multimodal acquisition ring, located inside the treatment chamber, is activated. The multispectral imaging sensor array on the ring works in conjunction with an ultraviolet excitation source. Excitation light (e.g., at 365 nm) is emitted from the ultraviolet excitation source towards the target skin area, and exposure acquisition is performed in both the visible and near-infrared bands to obtain the reflectance spectrum of the target skin area. Simultaneously, a high-sensitivity fluorescence detector positioned on the multimodal acquisition ring is triggered. Within a nanosecond-level time window after the ultraviolet excitation source is turned off, the weak fluorescence signal generated by the excitation in the target skin area is captured, obtaining an autofluorescence signal containing fluorescence intensity and attenuation characteristics.

[0053] The reflectance spectrum was analyzed to obtain data on the distribution of melanin, heme, and moisture in the epidermal layer of the target skin region. Spectral separation processing was performed on the reflectance spectrum of the target skin region, separating the mixed spectral signal into independent spectral components corresponding to different chromophores. Standard reference curves describing the absorption capacity of pure melanin, heme, and water molecules at different wavelengths were pre-stored. The correlation coefficient between each independent spectral component and each standard reference curve was calculated. The independent spectral component with the highest correlation coefficient to the pure melanin standard reference curve was identified as the melanin spectral component; the independent spectral component with the highest correlation coefficient to the heme standard reference curve was identified as the heme spectral component; and the independent spectral component with the highest correlation coefficient to the moisture standard reference curve was identified as the moisture spectral component.

[0054] The melanin spectral component, heme spectral component, and moisture spectral component are integrated separately to obtain the melanin content index, heme content index, and moisture content index of each pixel in the epidermal layer of the target skin area. Based on the spatial coordinate information of the pixel recorded by the collected reflectance spectrum, the melanin content index, heme content index, and moisture content index of all pixels are arranged according to their spatial position to generate melanin, heme, and moisture distribution detection data.

[0055] For example, by using an independent component analysis algorithm to decompose the reflectance spectrum into three independent spectral components, the correlation coefficient between the first independent spectral component and the pure melanin standard reference curve is 0.95, the correlation coefficient between the second independent spectral component and the heme standard reference curve is 0.92, and the correlation coefficient between the third independent spectral component and the moisture standard reference curve is 0.88. Then, the first, second, and third independent spectral components are respectively identified as the melanin spectral component, the heme spectral component, and the moisture spectral component.

[0056] The autofluorescence signal was inverted to obtain the detection data of collagen content and porphyrin content in the dermis of the target skin area.

[0057] The fluorescence intensity at each time point in the autofluorescence signal is arranged chronologically to generate a fluorescence intensity decay curve. Based on pre-stored collagen fluorescence lifetime characteristic parameters and porphyrin fluorescence lifetime characteristic parameters, the portion of the fluorescence intensity decay curve whose decay rate matches the collagen fluorescence lifetime characteristic parameters is identified as the collagen contribution signal. The portion of the fluorescence intensity decay curve whose decay rate matches the porphyrin fluorescence lifetime characteristic parameters is identified as the porphyrin contribution signal. The collagen contribution signal and the porphyrin contribution signal are integrated respectively to obtain the collagen content index and porphyrin content index of each pixel in the dermis of the target skin area. Based on the pixel spatial coordinate information recorded by the autofluorescence signal, the collagen content index and porphyrin content index of all pixels are arranged according to spatial position to generate collagen content detection data (used to characterize the skin's aging degree and repair potential) and porphyrin content detection data (used to characterize the skin surface microbial metabolic activity and inflammatory state).

[0058] Tomographic scanning of the target skin region was used to obtain data on stratum corneum thickness and inflammatory infiltration depth. A high-frequency optical coherence tomography (OCT) scanner on a multimodal acquisition ring performed micron-level longitudinal slice scans of the target skin region, acquiring a series of depth interference signals containing changes in the refractive index of the skin tissue. From the depth interference signals, the locations of the high-brightness signal layer at the epidermal-stratum corneum interface and the signal abrupt change points at the dermal-epidermal interface were extracted. The vertical physical distance between the high-brightness signal layer locations and the signal abrupt change points was calculated to determine the stratum corneum thickness. Low-reflectance signal regions formed by inflammatory cell infiltration within the dermal papillary layer were identified from the depth interference signals; the maximum longitudinal depth of these low-reflectance signal regions extending downwards from the dermal-epidermal interface was quantified to determine the inflammatory infiltration depth.

[0059] Using skin texture feature points of the target skin region as anchor points, pixel-level spatial registration and fusion of various detection data are performed to generate a static holographic snapshot of the target skin region. Surface skin texture intersections are extracted from the analytical image of the reflectance spectrum of the target skin region as reference anchor points; identical surface skin texture intersections are extracted from the analytical images of tomographic scans and autofluorescence signals as anchor points to be registered; affine transformation parameters are calculated based on the coordinate deviation between the reference anchor points and the anchor points to be registered, and these parameters are used to uniformly map all detection data to the same pixel coordinate system to complete pixel-level spatial registration; in the unified pixel coordinate system, the aligned detection data of melanin, heme, moisture, collagen, porphyrin content, stratum corneum thickness, and inflammatory infiltration depth are superimposed as independent channels to generate a static holographic snapshot of the target skin region containing multi-dimensional feature vectors of the entire region.

[0060] Triggering micro-energy pre-irradiation of the target skin area through static holographic snapshots of the skin includes:

[0061] Minimum erythema dose (MED) for the target skin region was adaptively estimated using stratum corneum thickness and inflammatory infiltration depth data to determine the baseline tolerable dose (BTD) for that region. The baseline tolerable dose represents the maximum UV energy density that the target skin region can withstand without inducing a visible erythema reaction. The average thickness of the stratum corneum and the average depth of inflammatory infiltration within the target skin region were calculated using stratum corneum thickness and inflammatory infiltration depth data, respectively. A pre-established minimum erythema dose estimation lookup table was used for matching; this lookup table stores the baseline tolerable dose values ​​corresponding to different combinations of stratum corneum thickness and inflammatory infiltration depth.

[0062] A preset safety factor is introduced to reduce and adjust the baseline tolerable dose, thus obtaining the micro-energy pre-irradiation dose for the target skin area. Specifically, the micro-energy pre-irradiation dose for the target skin area = preset safety factor × baseline tolerable dose.

[0063] It should be explained that the safety factor is a dimensionless value less than 1 and greater than 0, used to reduce the baseline tolerable dose to a safe sub-dose pre-irradiation energy. The safety factor is set based on the skin photobiology safety window theory and a retrospective statistical analysis of a large number of phototherapy clinical cases. Furthermore, the specific value of the safety factor can be fine-tuned according to the usage scenario. For example, for routine photosensitivity assessment scenarios, the value is 0.2 to 0.3; for children or sensitive skin, it is 0.1 to 0.15; for dark skin, it is 0.3 to 0.4; and for skin in the active inflammatory phase, it is 0.1 to 0.2.

[0064] Based on melanin distribution detection data, abnormal melanin regions are identified, and the set of pixel spatial coordinates within these regions is extracted as the target projection area. The global average and global standard deviation of melanin content for all pixels within the target skin region are calculated using the melanin distribution detection data. The difference between the global average and global standard deviation is used as a threshold. Pixels within the target skin region below this threshold are scanned and marked as abnormal melanin regions. The horizontal and vertical coordinates of all pixels within the abnormal melanin regions are extracted in a unified coordinate system, and these coordinates are aggregated to generate a set of pixel spatial coordinates, which is then defined as the target projection area.

[0065] The micro-energy pre-irradiation dose is controlled by an adjustable dual-band light source that emits a single pulse of ultraviolet light onto the target projection area, thus completing the micro-energy pre-irradiation action on the target skin area.

[0066] The dynamic set of photosensitive responses is obtained, including:

[0067] The instantaneous values ​​of heme concentration and autofluorescence signal intensity in the target skin area are continuously collected after the micro-energy pre-irradiation action. A continuous monitoring time window (e.g., 30 seconds) is set, and the instantaneous values ​​of heme concentration and autofluorescence signal intensity in the target skin area are synchronously collected at a fixed sampling frequency (e.g., 10 frames per second).

[0068] The instantaneous values ​​of heme concentration before and after micro-energy pre-irradiation were compared, and the rate of change curve was derived. The peak rate in the rate of change curve was extracted as the vasodilation response rate. The baseline heme concentration value before micro-energy pre-irradiation was extracted from a static holographic snapshot of the skin. The difference between the instantaneous heme concentration value and the baseline heme concentration value at each sampling point was calculated and divided by the corresponding time interval to obtain the instantaneous rate of change at the corresponding sampling point. The instantaneous rates of change were connected in chronological order to form a rate of change curve describing the dynamic response process of blood perfusion. The rate of change curve was smoothed and filtered, and the extreme points in the rate of change curve were extracted as the peak rate, which was defined as the vasodilation response rate.

[0069] An exponential fitting analysis was performed on the decay curves of the instantaneous values ​​of autofluorescence signal intensity to determine the fluorescence lifetime decay constant. The instantaneous values ​​of autofluorescence signal intensity at each sampling point were discretized over time to construct decay curves. The decay curves were then fitted using a nonlinear least squares method, and the time constant term in the fitting function was determined and identified as the fluorescence lifetime decay constant.

[0070] The combined vasodilation response rate and fluorescence lifetime decay constant form a dynamic set of photosensitivity responses in the target skin region.

[0071] The intelligent feature analysis module is used to fuse features from the dynamic set of photosensitivity response and the static holographic snapshot of the skin to obtain a joint vector of photosensitivity skin features and phenotypes. The joint vector of photosensitivity skin features and phenotypes is then used to perform multi-objective deduction through a pre-constructed comprehensive evaluation criterion to generate the optimal spectral power distribution and the optimal dose matrix.

[0072] The joint vector of photosensitive skin features and phenotypes is obtained, including:

[0073] Data from the dynamic set of photosensitivity response and static holographic snapshots of the skin were extracted and normalized according to their ranges. The normalized data of each dimension were then concatenated sequentially to obtain a joint vector of photosensitive skin features and phenotypes. A full-region statistical average was performed on the detection data of melanin distribution, heme distribution, moisture distribution, collagen content, porphyrin content, stratum corneum thickness, and inflammatory infiltration depth to obtain the mean values ​​for melanin content, heme content, moisture content, collagen content, porphyrin content, stratum corneum thickness, and inflammatory infiltration depth. The maximum and minimum theoretical boundary values ​​of each data type were loaded, and range normalization was performed on the vasodilation response rate, fluorescence lifetime decay constant, and each mean value to generate a standardized index mapped to the zero-to-one interval.

[0074] Create a one-dimensional array container with nine slots, and fill it with the data from each normalized process in sequence to form a joint vector of photosensitive skin features and phenotypes.

[0075] The order of these parameters is as follows: standardized vasodilation response rate, standardized fluorescence lifetime decay constant, standardized melanin content index, standardized heme content index, standardized moisture content index, standardized collagen content index, standardized porphyrin content index, standardized stratum corneum thickness index, and standardized inflammatory infiltration depth index.

[0076] Multi-objective extrapolation of the joint vector of photosensitive skin features and phenotypes includes:

[0077] A pre-constructed comprehensive evaluation criterion is included, which contains a positive benefit term for quantifying the improvement rate of psoriasis area, a negative loss term for quantifying the erythema grade and the increase of DNA cyclobutanepyrimidine dimer, and a penalty term for quantifying the minimum effective dose.

[0078] Within the spectral bands and dose intensity range of the tunable dual-band light source, multiple sets of candidate spectral power distribution curves and dose matrices are established. A uniform grid sampling method is used to generate multiple candidate wavelength combinations within the permissible spectral bands (e.g., 280 nm to 400 nm) of the long-wave and mid-wave ultraviolet channels of the tunable dual-band light source; simultaneously, multiple candidate energy density values ​​are generated within the dose intensity range of the tunable dual-band light source (e.g., 0 mJ / cm² to 500 mJ / cm²). The multiple candidate wavelength combinations and multiple candidate energy density values ​​are then cross-combined using Cartesian products to form multiple sets of candidate spectral power distribution curves and candidate dose matrices.

[0079] Using the photosensitive skin feature-phenotype joint vector as a constraint, a multi-objective iterative optimization is performed on each group of candidate spectral power distribution curves and dose matrices based on a comprehensive evaluation criterion to obtain the expected index set. With the photosensitive skin feature-phenotype joint vector as a fixed input, each candidate spectral power distribution curve and dose matrix is ​​traversed. For each group of candidate spectral power distribution curves and dose matrices, the corresponding psoriasis area improvement rate is retrieved through a pre-established efficacy prediction lookup table, serving as the positive benefit term value in the expected index set. The erythema grade and the expected DNA cyclobutanepyrimidine dimer increment are obtained through queries in the established erythema prediction lookup table and DNA damage prediction lookup table, and the weighted sum of these two values ​​serves as the negative loss term value in the expected index set. The lowest effective dose benchmark value is extracted from a static holographic snapshot of the skin, and the degree to which the dose value in the corresponding group of candidate dose matrices is lower than the lowest effective dose benchmark value is calculated, serving as the penalty term value in the expected index set.

[0080] The positive benefit term, negative loss term, and penalty term are combined to form the expected index set of the corresponding candidate spectral power distribution curve and dose matrix.

[0081] Each expected indicator set is incorporated into the comprehensive evaluation criteria for deduction, and the comprehensive evaluation value of each candidate spectral power distribution curve and dose matrix is ​​obtained. The comprehensive evaluation value of the comprehensive evaluation criteria = positive benefit term - negative loss term - penalty term.

[0082] The candidate spectral power distribution curve and dose matrix with the highest comprehensive evaluation value are selected as the optimal spectral power distribution curve and optimal dose matrix.

[0083] For example, there are three sets of candidate spectral power distribution curves and candidate dose matrices:

[0084] Group 1: Positive benefit value is 60, negative loss value is 10, and penalty value is 2.5;

[0085] Group 2: Positive benefit value is 70, negative loss value is 15, and penalty value is 0;

[0086] Group 3: Positive benefit value is 75, negative loss value is 19, and penalty value is 0.

[0087] Substituting these values ​​into the comprehensive evaluation criteria, we obtain the following comprehensive evaluation values: 47.5 for the first group, 55 for the second group, and 56 for the third group.

[0088] If the comprehensive evaluation value of the third group is found to be 56, then the candidate spectral power distribution curve and candidate dose matrix of the third group are determined as the optimal spectral power distribution curve and optimal dose matrix.

[0089] The adaptive optics control module is used to drive the tunable dual-band light source based on the optimal spectral power distribution and the optimal dose matrix, and to inversely calculate the effective ultraviolet energy that actually reaches the dermis. The effective ultraviolet energy is used to dynamically correct the driving current of the tunable dual-band light source to form a steady-state treatment driving current.

[0090] The effective ultraviolet energy actually reaching the dermis is calculated, including:

[0091] The optimal spectral power distribution curve is mapped to the initial current duty cycle of the long-wave and mid-wave ultraviolet channels. The energy proportions of the long-wave ultraviolet band and the mid-wave ultraviolet band in the optimal spectral power distribution curve are analyzed. The energy proportion of the long-wave ultraviolet band is mapped to the initial current duty cycle of the long-wave ultraviolet channel in the tunable dual-band light source, and the energy proportion of the mid-wave ultraviolet band is mapped to the initial current duty cycle of the mid-wave ultraviolet channel in the tunable dual-band light source.

[0092] The optimal dose matrix is ​​mapped to the target output power intensity. The initial current duty cycle and the target output power intensity are pulse-width modulated to form the initial drive current, which drives the adjustable dual-band light source to project ultraviolet light onto the target skin area. The initial current duty cycle of the long-wave ultraviolet channel, the initial current duty cycle of the mid-wave ultraviolet channel, and the target output power intensity are input into a pulse width modulation signal generator. The pulse width modulation signal generator generates a pulse width modulation signal for the long-wave ultraviolet channel based on the initial current duty cycle of the long-wave ultraviolet channel; it generates a pulse width modulation signal for the mid-wave ultraviolet channel based on the initial current duty cycle of the mid-wave ultraviolet channel; and it generates a power control signal based on the target output power intensity. The long-wave ultraviolet channel pulse width modulation signal, the mid-wave ultraviolet channel pulse width modulation signal, and the power control signal are synthesized and output as the initial drive current.

[0093] The diffuse reflectance spectral signal of the target skin region is acquired during the operation of the adjustable dual-band light source, and the energy is summarized across the entire domain to form a diffuse reflectance energy loss value. The diffuse reflectance energy loss value characterizes the total ultraviolet energy that fails to penetrate the deep layers of the skin due to skin surface reflection and backscattering from superficial tissues. A high-sensitivity spectrometer is activated to acquire the diffuse reflectance spectral signal reflected from the surface of the target skin region in real time; the diffuse reflectance spectral signal is integrated across the entire wavelength band to calculate the total energy value of the diffuse reflectance photon flux, and this total energy value is defined as the diffuse reflectance energy loss value.

[0094] The output power of the initial drive current is analyzed to obtain the theoretical emission energy. The actual current and voltage values ​​of the initial drive current applied to the tunable dual-band light source at the current moment are read and multiplied to obtain the instantaneous electrical power. Combined with the factory photoelectric conversion efficiency parameters of the tunable dual-band light source, the instantaneous electrical power is converted into instantaneous radiated optical power. The instantaneous radiated optical power is integrated on the time axis to obtain the theoretical emission energy of the tunable dual-band light source in the current time period.

[0095] The effective ultraviolet energy reaching the dermis is obtained by subtracting the diffuse reflection energy loss from the theoretical emission energy. Effective ultraviolet energy = Theoretical emission energy - Diffuse reflection energy loss.

[0096] The formation of a steady-state therapeutic driving current includes:

[0097] The stratum corneum thickness measurement data from static holographic snapshots of the skin is coupled with a preset stratum corneum light attenuation coefficient to determine the ultraviolet (UV) transmittance coefficient penetrating the stratum corneum. Based on the Lambert-Beer exponential decay law, the transmittance coefficient is obtained by multiplying the stratum corneum light attenuation coefficient with the stratum corneum thickness measurement data and taking the negative exponent.

[0098] It should be explained that the stratum corneum light attenuation coefficient is used to describe the energy absorption rate of ultraviolet rays when passing through a unit thickness of stratum corneum. The stratum corneum light attenuation coefficient is set based on pre-established skin tissue optical properties (including the absorption coefficient, scattering coefficient, and anisotropy factor of the stratum corneum at different wavelengths). Specifically, for the main wavelength of the tunable dual-band light source output, the stratum corneum attenuation coefficient corresponding to the main wavelength is retrieved from the skin tissue optical property database.

[0099] The target dose for the dermis is obtained by adjusting the magnification of each target output dose in the optimal dose matrix using the transmission coefficient. The target output dose corresponding to each pixel in the optimal dose matrix is ​​then multiplied by the transmission coefficient to obtain the energy density value that should theoretically reach the dermis after absorption and attenuation by the stratum corneum. All calculated energy density values ​​are then summed to determine the target dose for the dermis.

[0100] The deviation between the effective ultraviolet energy and the target dose in the dermis is used as the energy deviation signal for calculation. The calculated current compensation value is then added to the initial driving current to generate a steady-state treatment driving current. The effective ultraviolet energy actually reaching the dermis is compared pixel-by-pixel with the energy density value of the corresponding pixel in the dermal target dose matrix, and the difference (effective ultraviolet energy - energy density value) is calculated. The differences of all pixels are then weighted according to spatial location to obtain the energy deviation signal. This energy deviation signal is input to a PID controller, where a weighted summation is performed based on the proportional, integral, and derivative terms of the energy deviation signal to output a current compensation value.

[0101] The thermal safety monitoring and defense module is used to monitor the epidermal temperature field of the target skin area when the steady-state treatment drive current is continuously output, and to control the operation status of the adjustable dual-band light source through the epidermal temperature field.

[0102] Monitoring the epidermal temperature field of the target skin area during continuous steady-state output of the therapeutic driving current, including:

[0103] The thermal radiation distribution image of the target skin region is acquired frame by frame under steady-state treatment driving current operation, and each frame of thermal radiation distribution image is converted into an epidermal temperature field containing the temperature values ​​of each pixel. Each pixel in each frame of thermal radiation distribution image is analyzed point by point, and the corresponding Celsius temperature value is found in a pre-established thermal radiation-temperature calibration parameter table (which stores the nonlinear mapping relationship between different gray values ​​and absolute Celsius temperature) based on the gray value of the pixel. All the analyzed Celsius temperature values ​​are rearranged according to the spatial coordinate position of the original pixels to construct an epidermal temperature field containing the temperature values ​​of each pixel in the target skin region.

[0104] A full-domain scan of the epidermal temperature field is performed for each frame, and the highest temperature value acquired during the scan is determined as the peak epidermal temperature. The slope of the peak epidermal temperature change within a sliding time window is derived as the epidermal temperature rise rate. The peak epidermal temperature is used to represent the "static risk of immediate thermal burns," that is, to determine whether the current temperature of the target skin area has reached the critical value that would cause protein denaturation or tissue necrosis. The epidermal temperature rise rate is used to represent the "dynamic trend of heat accumulation," that is, to preemptively cut off the risk before the temperature reaches a dangerous value, achieving predictive defense.

[0105] For example, an infrared thermal imager acquires thermal radiation distribution images at a sampling frame rate of 30 frames per second. The epidermal temperature field distribution map generated at the 10th second is scanned across the entire domain, and the highest temperature value is found to be 38.5 degrees Celsius. 38.5 degrees Celsius is determined as the peak epidermal temperature at the 10th second. The sliding time window length is set to 10 seconds, and 10 peak epidermal temperature values ​​from the 1st to the 10th second are extracted and linearly fitted. The slope of the fitted line is 0.3 degrees Celsius per minute. This 0.3 degrees Celsius per minute increase is determined as the epidermal temperature rise rate at the 10th second.

[0106] Controlling the operating status of the adjustable dual-band light source includes:

[0107] A preset thermal damage threshold includes an absolute safe temperature threshold and a maximum temperature rise rate threshold. The peak skin temperature and skin temperature rise rate are compared with the thermal damage threshold.

[0108] It should be explained that the absolute safe temperature threshold and the maximum temperature rise rate threshold are determined based on a comprehensive consideration of the physiological critical values ​​of skin thermal damage, clinical safety data, and the measurement characteristics of infrared thermal imaging technology. The absolute safe temperature threshold is generally set at 42 degrees Celsius, and the maximum temperature rise rate threshold is set at 2.0 degrees Celsius per minute.

[0109] If the peak epidermal temperature exceeds the absolute safe temperature threshold, or the epidermal temperature rise rate exceeds the maximum temperature rise rate threshold, it is determined that there is an immediate risk of thermal burn in the target skin area, and the steady-state treatment driving current is immediately interrupted.

[0110] If the peak epidermal temperature is not greater than the absolute safe temperature threshold and the epidermal temperature rise rate is not greater than the maximum temperature rise rate threshold, then it is determined to be in a thermally safe state, and the steady-state treatment drive current is continuously output.

[0111] In this embodiment, the skin state sensing module collects reflectance spectrum, autofluorescence signal, and tomographic data of the target skin area to generate a static holographic snapshot of the skin. Through micro-energy pre-irradiation, it actively stimulates and captures the dynamic set of photosensitivity responses of skin microcirculation and fluorescence, realizing a leap from static phenotypic observation to dynamic physiological stress detection. It accurately quantifies the molecular-level differences in individual dimensions such as vasodilatory response, melanocyte activity and proliferation, and immune cell activity, and truly reflects the biological tolerance limit of the skin to ultraviolet radiation. This provides accurate data support for the generation of subsequent personalized treatment parameters, avoids parameter adaptation deviations caused by incomplete feature collection from the source, and significantly improves the safety starting point of treatment.

[0112] The intelligent feature analysis module constructs a joint vector of photosensitive skin features and phenotypes, and uses a comprehensive evaluation criterion to perform multi-objective iterative optimization, generating the optimal spectral power distribution and optimal dose matrix that balances treatment benefits and tissue protection. This enables intelligent identification of skin types and personalized deduction of treatment parameters. Through real-time diffuse reflection inversion and thermal safety monitoring by the adaptive optics control module, the absorption loss of the stratum corneum can be eliminated, and the effective energy of the dermis can be directly locked for closed-loop correction, further improving the accuracy of treatment parameter adaptation. This effectively solves the problem of undertreatment or overload burns caused by individual differences, and significantly improves the effectiveness and compliance of phototherapy.

[0113] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0114] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0115] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0116] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A system for generating ultraviolet therapy parameters based on automatic skin type recognition, characterized in that, The ultraviolet therapy parameter generation system based on automatic skin type recognition includes: The skin state perception module is used to acquire relevant detection data of the target skin area and perform synchronous alignment and registration to generate a static holographic snapshot of the skin. It also triggers a micro-energy pre-irradiation action on the target skin area through the static holographic snapshot of the skin to obtain a dynamic set of photosensitivity response. The intelligent feature analysis module is used to fuse features from the dynamic set of photosensitivity response and the static holographic snapshot of the skin to obtain the joint vector of photosensitivity skin features and phenotype. The joint vector of photosensitivity skin features and phenotype is then used to perform multi-objective deduction through a pre-constructed comprehensive evaluation criterion to generate the optimal spectral power distribution and the optimal dose matrix. Multi-objective extrapolation of the joint vector of photosensitive skin features and phenotypes includes: A pre-constructed comprehensive evaluation criterion is included, which contains a positive benefit term for quantifying the improvement rate of psoriasis area, a negative loss term for quantifying the erythema grade and the increase of DNA cyclobutanepyrimidine dimer, and a penalty term for quantifying the minimum effective dose. Within the spectral bands and dose intensity range of the tunable dual-band light source, multiple sets of candidate spectral power distribution curves and dose matrices were established. Using the joint vector of photosensitive skin features and phenotypes as constraints, a multi-objective iterative optimization of each candidate spectral power distribution curve and dose matrix is ​​performed based on a comprehensive evaluation criterion to obtain the expected index set. Each expected index set is incorporated into the comprehensive evaluation criteria for deduction, and the comprehensive evaluation value of each group of candidate spectral power distribution curves and dose matrices is obtained; the candidate spectral power distribution curve and dose matrix with the largest comprehensive evaluation value are selected as the optimal spectral power distribution curve and the optimal dose matrix. The adaptive optics control module is used to drive the tunable dual-band light source based on the optimal spectral power distribution and the optimal dose matrix, and to inversely calculate the effective ultraviolet energy that actually reaches the dermis. The effective ultraviolet energy is used to dynamically correct the driving current of the tunable dual-band light source to form a steady-state treatment driving current. The thermal safety monitoring and defense module is used to monitor the epidermal temperature field of the target skin area when the steady-state treatment drive current is continuously output, and to control the operation status of the adjustable dual-band light source through the epidermal temperature field.

2. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 1, characterized in that, The generation of the static holographic snapshot of the skin includes: The reflectance spectrum and autofluorescence signal of the target skin area were collected; the reflectance spectrum was analyzed to obtain the detection data of melanin, heme and water distribution in the epidermis of the target skin area; the autofluorescence signal was inverted to obtain the detection data of collagen content and porphyrin content in the dermis of the target skin area. The target skin area is scanned to obtain data on the thickness of the stratum corneum and the depth of inflammatory infiltration. Using the skin texture feature points of the target skin area as anchor points, the detection data are spatially registered and fused at the pixel level to generate a static holographic snapshot of the target skin area.

3. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 2, characterized in that, The micro-energy pre-irradiation action of triggering the target skin area through a static holographic snapshot of the skin includes: The minimum erythema dose for the target skin area was estimated using data on stratum corneum thickness and inflammatory infiltration depth to determine the baseline tolerable dose for the target skin area. A preset safety factor was introduced to reduce and adjust the baseline tolerable dose to obtain the micro-energy pre-irradiation dose for the target skin area. Based on melanin distribution detection data, abnormal melanin areas are identified, and the set of pixel spatial coordinates of the abnormal melanin areas is extracted as the target projection area. A single pulse of ultraviolet light is emitted to the target projection area by an adjustable dual-band light source with micro-energy pre-irradiation dose control, thus completing the micro-energy pre-irradiation action of the target skin area.

4. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 3, characterized in that, The obtained dynamic set of photosensitive responses includes: The instantaneous values ​​of heme concentration and autofluorescence signal intensity in the target skin area were continuously collected after the micro-energy pre-irradiation action was completed. The instantaneous values ​​of heme concentration before and after micro-energy pre-irradiation were compared, and the rate of change curve was derived. The peak rate in the rate of change curve was extracted as the vasodilation response rate. An exponential fitting analysis was performed on the decay curve of the instantaneous value of autofluorescence signal intensity to determine the fluorescence lifetime decay constant. The combined vasodilation response rate and fluorescence lifetime decay constant form a dynamic set of photosensitivity responses in the target skin region.

5. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 4, characterized in that, The obtained photosensitive skin feature-phenotype joint vector includes: Extract various types of data from the dynamic set of photosensitivity response and the static holographic snapshot of the skin and perform range standardization; then concatenate the standardized data of each dimension in sequence to obtain the joint vector of photosensitivity skin features-phenotype.

6. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 1, characterized in that, The inverse calculation of the effective ultraviolet energy actually reaching the dermis includes: The optimal spectral power distribution curve is mapped to the initial current duty cycle of the long-wave and medium-wave ultraviolet channels; the optimal dose matrix is ​​mapped to the target output power intensity. The initial current duty cycle and the target output power intensity are pulse-width modulated to form the initial driving current and drive the adjustable dual-band light source to project ultraviolet light onto the target skin area. The diffuse reflectance spectral signal of the target skin region is collected during the operation of the adjustable dual-band light source and the global energy is summarized to form the diffuse reflectance energy loss value; the output power of the initial driving current is subjected to energy evolution to obtain the theoretical emission energy; the loss of the diffuse reflectance energy loss value is removed from the theoretical emission energy to obtain the effective ultraviolet energy that actually reaches the dermis.

7. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 6, characterized in that, The formation of the steady-state therapeutic drive current includes: The stratum corneum thickness detection data in the static holographic snapshot of the skin is coupled with the preset stratum corneum light attenuation coefficient to determine the ultraviolet transmission coefficient through the stratum corneum; the transmission coefficient is used to adjust the magnification of each target output dose in the optimal dose matrix to obtain the target dose of the dermis. The deviation between the effective ultraviolet energy and the target dose in the dermis is used as the energy deviation signal for calculation, and the calculated current compensation value is superimposed on the initial driving current to generate a steady-state therapeutic driving current.

8. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 7, characterized in that, The monitoring of the epidermal temperature field of the target skin area during continuous output of the steady-state therapeutic driving current includes: Acquire each frame of thermal radiation distribution image of the target skin area under steady-state treatment driving current working state, and convert each frame of thermal radiation distribution image into an epidermal temperature field containing the temperature value of each pixel; A full-domain scan of the epidermal temperature field is performed for each frame, and the highest temperature value obtained from the scan is determined as the peak epidermal temperature; the slope of the change of the peak epidermal temperature in the sliding time window is deduced as the epidermal temperature rise rate.

9. The ultraviolet therapy parameter generation system based on automatic skin type recognition according to claim 8, characterized in that, The control of the operating state of the adjustable dual-band light source includes: A preset thermal damage threshold is included, comprising an absolute safe temperature threshold and a maximum temperature rise rate threshold. The peak epidermal temperature and epidermal temperature rise rate are compared with the thermal damage threshold. If the peak epidermal temperature is greater than the absolute safe temperature threshold, or the epidermal temperature rise rate is greater than the maximum temperature rise rate threshold, it is determined that there is an immediate risk of thermal burn in the target skin area, and the steady-state treatment drive current is immediately interrupted. If the peak epidermal temperature is not greater than the absolute safe temperature threshold, and the epidermal temperature rise rate is not greater than the maximum temperature rise rate threshold, it is determined that the body is in a thermally safe state, and the steady-state treatment drive current continues to be output.

Images