Multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment

Abstract

The present application belongs to the technical field of laser, and particularly relates to a multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment, comprising: a data acquisition module for acquiring the skin horny layer thickness and sebum excretion rate of a patient and calculating the current impedance scattering ratio; a treatment planning module for triggering a wavelength reorganization instruction to recalculate the weight coefficient when the current impedance scattering ratio exceeds a first threshold; a wavelength optimization module for generating the weight coefficient containing the target wavelength through a multi-objective optimization algorithm; a pulse modulation module for calculating the pulse duration corresponding to each wavelength according to the weight coefficient; an energy control module for calculating the total output energy; a skin feedback module for generating a temperature gradient matrix; and a laser emission module for generating a driving current according to the total output energy and the temperature gradient matrix to emit the multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment. The present application realizes the precise treatment of seborrheic dermatitis and significantly improves the treatment effect.

Description

Technical Field

[0001] This invention belongs to the field of laser technology, and particularly relates to a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis. Background Technology

[0002] With the continuous development of optical communication technology, the application of fiber lasers in the biomedical field has gradually attracted attention. Seborrheic dermatitis is a common skin disease, and traditional treatment methods have certain limitations. Therefore, fiber lasers can be considered for treatment to achieve better results.

[0003] Existing laser treatment systems typically use single-wavelength lasers. While these can alleviate symptoms to some extent, the complexity of skin tissue makes it difficult for single-wavelength lasers to simultaneously target different tissue components. Therefore, they are insufficient in inhibiting Malassezia proliferation and regulating sebum secretion, resulting in less than ideal treatment outcomes. Furthermore, the lack of a dynamic response mechanism to changes in skin impedance during treatment makes it impossible to adjust treatment parameters according to the real-time condition of the skin, easily leading to unstable treatment results. Moreover, laser treatment with fixed pulse parameters can cause uneven energy deposition, resulting in thermal damage and affecting the safety and effectiveness of the treatment. Summary of the Invention

[0004] In view of the problems existing in the background technology, the purpose of the present invention is to provide a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis, which realizes precise treatment of seborrheic dermatitis and significantly improves the treatment effect.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Multi-wavelength pulsed fiber lasers for the treatment of seborrheic dermatitis include:

[0007] The data acquisition module is used to obtain the patient's skin stratum corneum thickness and sebum secretion rate, and simultaneously acquire skin impedance and optical scattering coefficient to calculate the current impedance scattering ratio;

[0008] The treatment planning module is used to trigger a wavelength recombination command to recalculate the weighting coefficients when the current impedance scattering ratio exceeds a first threshold.

[0009] The wavelength optimization module is used to generate weight coefficients containing the target wavelength based on the patient's skin stratum corneum thickness and sebum secretion rate through a multi-objective optimization algorithm.

[0010] The pulse modulation module is used to calculate the pulse duration corresponding to each wavelength based on the weighting coefficients.

[0011] The energy control module is used to integrate the energy of each wavelength and calculate the total output energy;

[0012] The skin feedback module is used to acquire the temperature distribution of the treatment area in real time and generate a temperature gradient matrix;

[0013] The laser emission module is used to generate a driving current based on the total output energy and temperature gradient matrix, thereby driving the laser emission and emitting a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis.

[0014] Preferably, in the data acquisition module, the skin impedance is obtained in the following way:

[0015] Step S01: Using a four-electrode bioimpedance measurement device, apply a 1mA / 50kHz AC signal to the skin to be tested;

[0016] Step S02: Separate the real part of the impedance using a phase detection circuit. and the virtual part ;

[0017] Step S03: Calculate the effective skin impedance value The sampling frequency is 100Hz;

[0018] Step S04: Transmit Z to the data acquisition module wirelessly.

[0019] Preferably, in the treatment planning module, the first threshold The methods for determining this include:

[0020] Step S11: Store critical impedance scattering ratio and clinical impedance scattering ratio data for different skin types to establish a treatment safety database;

[0021] Step S12: Based on the current impedance scattering ratio Q value and the treatment safety database, predict the risk level of the current impedance scattering ratio Q value using a support vector machine classification model;

[0022] Step S13: When the predicted risk level is not high, no action is taken; when the predicted risk level is high, the first threshold is automatically adjusted. Updated to :

[0023] ;

[0024] This is the minimum safety threshold.

[0025] Preferably, in step S12, the training process of the support vector machine classification model includes:

[0026] Step S121: Construct a training set from clinical impedance scattering ratio data , where j is the index of the j-th training sample. This represents the impedance scattering ratio of the j-th training sample. This represents the hazard level label of the j-th sample, and the cleaned and labeled training dataset is obtained through the hazard level label.

[0027] Step S122: Use grid search method to apply the penalty parameter Search within the range, step size For each C value, train an SVM model and calculate its performance metrics on the cross-validation set to select the optimal penalty parameters;

[0028] Step S123: Based on the training dataset and the optimal penalty parameters, the SVM model is trained and validated using the 5-fold cross-validation method. The training dataset is randomly divided into 5 subsets. Four subsets are used for training each time, and the remaining subset is used for validation.

[0029] Step S124: Calculate the validation accuracy for each fold and take the average as the final validation accuracy;

[0030] Step S125: When the final validation accuracy exceeds the target validation accuracy, output and save the validated optimal SVM model.

[0031] Preferably, the decision function of the SVM model for:

[0032] ;

[0033] in, Let be the impedance scattering ratio of the k-th support vector, where k is the index of the support vector and n represents the total number of support vectors. Let Lagrange multipliers be the k-th support vector. Let be the label of the k-th support vector, and b be the bias term. For RBF kernel function, , This is the bandwidth parameter.

[0034] Preferably, in the wavelength optimization module, the multi-objective optimization algorithm specifically includes:

[0035] Step S21: Based on the skin stratum corneum thickness, sebum secretion rate, and the i-th wavelength, construct the first objective function F:

[0036] ;

[0037] in, For wavelength matching weights, For keratin attenuation weight, For sebum suppression weight, For the i-th wavelength, The characteristic absorption peak of the target tissue denoted as the characteristic value of stratum corneum thickness, e as the natural constant, d as the thickness of the stratum corneum, and s as the sebum secretion rate;

[0038] Step S22: Perform multi-objective optimization on the effects of wavelength matching degree, stratum corneum thickness and sebum secretion rate to obtain a progeny population containing elite individuals, and screen the non-dominated solution set of the Pareto front from the progeny population containing elite individuals.

[0039] Step S23: Based on the non-dominated solution set of the Pareto front and the thickness d of the stratum corneum, select the target wavelength and obtain the target wavelength and its corresponding weight coefficient.

[0040] Preferably, the specific steps for obtaining the non-dominated solution set of the Pareto front in step S22 include:

[0041] Step S221: Obtain population size and crossover probability Initialize the population, where each individual represents a set of wavelengths and their corresponding weight coefficients. Simultaneously, the weight coefficients corresponding to the i-th wavelength are... Introducing constraints during the initialization phase , H represents the total number of wavelengths;

[0042] Step S222: Adjust the crossover probability of the initial population according to the sebum secretion rate, using the adaptive crossover probability adjustment formula: , To obtain the new crossover probability, when the sebum secretion rate is greater than the target sebum secretion rate, the crossover probability is increased, resulting in the offspring population after the crossover operation.

[0043] Step S223: The elite individuals in the parent population whose first objective function F value is greater than the optimization objective threshold are directly retained in the offspring population to obtain an offspring population containing elite individuals.

[0044] Preferably, in the pulse modulation module, the pulse duration is... The method for obtaining it is as follows: , As the reference pulse width, To prevent the elimination of the zero constant, d represents the thickness of the stratum corneum, and s represents the sebum secretion rate. The weighting coefficient for the i-th wavelength;

[0045] Pulse duration Optimization methods also include:

[0046] Step S31: Establish thermal damage constraints Used to limit the energy deposition of laser in skin tissue. For the safe energy density threshold, Let be the wavelength energy corresponding to the i-th wavelength at time t. The pulse duration corresponding to the i-th wavelength;

[0047] Step S32: Obtain the pulse duration and constraint weights By incorporating constraints into the second objective function to solve for the optimal pulse duration, the second objective function is constructed. ,in, For the i-th wavelength during the pulse duration Total energy within;

[0048] Step S33: Through iterative solution, continuously adjust the pulse duration until the second objective function... When the target convergence condition is met, the optimal pulse duration satisfying the thermal damage constraint is obtained, whereby iterative solution yields... Correction formula , Let be the corrected pulse duration for the i-th wavelength. Let be the absorption coefficient at the i-th wavelength.

[0049] Preferably, the specific implementation method of step S33 iterative solution is as follows:

[0050] Step S331: Obtain the initial constraint weights and the target convergence condition , The difference between the second objective function value L and the previous second objective function value;

[0051] Step S332: Calculate the cumulative energy in each iteration. ; This represents the cumulative energy calculated in the m-th iteration. The weighting coefficient for the i-th wavelength in the m-th iteration is... The pulse duration corresponding to the i-th wavelength in the m-th iteration;

[0052] Step S333: Update the constraint weights and adjust the weights of the second objective function. The update formula is as follows: , For learning rate, These are the constraint weights in the (m+1)th iteration. Let be the constraint weights in the m-th iteration.

[0053] Preferably, in the laser emitting module, the calibration process of the driving current includes the following steps:

[0054] Step S41: Connect a detection resistor R in parallel with the load terminal of the laser diode;

[0055] Step S42: Measure the voltage drop V in real time and calculate the actual current value. ;

[0056] Step S43: Adjust the drive voltage using a PID controller to make... , This is the target current value.

[0057] The beneficial effects of this invention are:

[0058] This invention achieves dynamic combination of treatment wavelengths through a multi-objective optimization algorithm, thereby precisely adjusting the weight coefficients of each wavelength according to the individual patient's skin condition (such as stratum corneum thickness and sebum secretion rate) to generate the most suitable wavelength combination, thus realizing precise treatment of seborrheic dermatitis, significantly improving the treatment effect, and effectively solving the problem that traditional single-wavelength lasers cannot simultaneously inhibit Malassezia proliferation and regulate sebum secretion.

[0059] This invention utilizes a dynamic compensation mechanism that monitors the current impedance scattering ratio in real time. This mechanism can sense changes in skin impedance in real time and adjust treatment parameters accordingly. This establishes a safe compensation mechanism during treatment, thereby enhancing the dynamic response to changes in skin impedance and improving the safety and stability of the treatment.

[0060] This invention establishes thermal damage constraints and uses an iterative optimization method to precisely adjust the pulse duration based on real-time feedback from the skin, thereby achieving optimized pulse energy distribution. This avoids thermal damage caused by uneven energy deposition due to fixed pulse parameters, thus improving the safety and effectiveness of treatment. Attached Figure Description

[0061] Figure 1 This is a schematic diagram of the modules of the present invention;

[0062] Figure 2 This is a flowchart illustrating the architecture of the present invention. Detailed Implementation

[0063] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0064] This embodiment provides a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis, solving the problem that traditional single-wavelength lasers cannot simultaneously inhibit Malassezia proliferation and regulate sebum secretion. It first simultaneously records skin impedance and optical scattering coefficients and calculates the impedance-scattering ratio. When this ratio exceeds a first threshold, it triggers a recalculation of weighting coefficients. It receives the patient's skin stratum corneum thickness and sebum secretion rate, and generates weighting coefficients containing target wavelengths using a multi-objective optimization algorithm. Based on the weighting coefficients, it calculates the pulse duration corresponding to each wavelength and integrates the energy of each wavelength to calculate the total output energy. It acquires the temperature distribution of the treatment area in real time and generates a temperature gradient matrix. Based on the total output energy and the temperature gradient matrix, it generates a driving current to drive laser emission, achieving precise treatment of seborrheic dermatitis and significantly improving treatment efficacy.

[0065] In this embodiment, the formulas can be dimensionless during calculation to simplify the calculation.

[0066] like Figure 1 and Figure 2 As shown, a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis includes: a data acquisition module, a treatment planning module, a wavelength optimization module, a pulse modulation module, an energy control module, a skin feedback module, and a laser emission module;

[0067] The data acquisition module obtains the patient's skin stratum corneum thickness and sebum secretion rate, and simultaneously acquires skin impedance Z and optical scattering coefficient σ, calculating the current impedance scattering ratio. ;

[0068] Treatment planning module, used when the current impedance scattering ratio exceeds a first threshold When this occurs, a wavelength recombination command is triggered to recalculate the weighting coefficients;

[0069] The wavelength optimization module is used to generate weight coefficients containing the target wavelength based on the received patient's skin stratum corneum thickness d and sebum secretion rate s through a multi-objective optimization algorithm.

[0070] The pulse modulation module is used to calculate the pulse duration corresponding to each wavelength based on the weighting coefficients. , , As the reference pulse width, To prevent division by zero constant, The weighting coefficient for the i-th wavelength;

[0071] The energy control module is used to integrate energy from various wavelengths. Calculate the total output energy , The wavelength energy corresponding to the i-th wavelength. The pulse duration corresponding to the i-th wavelength. Let be the absorption coefficient at the i-th wavelength;

[0072] The skin feedback module is used to acquire the temperature distribution T(x, y) of the treatment area in real time and generate the temperature gradient matrix ΔT.

[0073] Laser emitting module, used to determine the total output energy The driving current is generated by the temperature gradient matrix ΔT. , For conversion efficiency, The temperature compensation coefficient drives the laser emission, emitting a multi-wavelength pulsed fiber laser for the treatment of seborrheic dermatitis.

[0074] In this embodiment, the wavelength optimization module uses a multi-objective optimization algorithm to determine the weighting coefficients of three wavelengths (1450nm, 2940nm, and 10600nm) based on the patient's skin parameters (stratum corneum thickness *d* and sebum secretion rate *s*) to achieve the best treatment effect. For example, when the patient's stratum corneum is thick and sebum secretion rate is high, the algorithm adjusts the weighting coefficients, increasing the weight of wavelengths that have a stronger effect on deeper tissues. The pulse modulation module calculates the pulse duration for each wavelength based on the weighting coefficients, ensuring that the pulse duration adapts to different skin conditions through a specific formula. The energy control module integrates the energy of each wavelength and calculates the total output energy. The laser emission module generates a driving current based on the total energy and temperature feedback, making the laser output stable and adaptable to the real-time skin condition. The data acquisition module records skin impedance and optical scattering coefficient, calculates the impedance-to-scattering ratio, and when this ratio exceeds a threshold, the treatment planning module recalculates the wavelength weighting coefficients to achieve dynamic adjustment.

[0075] Methods for measuring skin impedance include:

[0076] Step S01: Apply a 1mA / 50kHz AC signal using a four-electrode bioimpedance measurement device;

[0077] Step S02: Separate the real part of the impedance using a phase detection circuit. and the virtual part ;

[0078] Step S03: Calculate the effective skin impedance value The sampling frequency is 100Hz.

[0079] First threshold The methods for determining this include:

[0080] Step S11: Store critical impedance scattering ratio and clinical impedance scattering ratio data for different skin types to establish a treatment safety database;

[0081] Step S12: Using the current impedance scattering ratio Q value and treatment safety database, predict the risk level of the current impedance scattering ratio Q value through a support vector machine classification model;

[0082] Step S13: When the predicted risk level is not high, no action is taken; when the predicted risk level is high, the first threshold is automatically adjusted. Updated to :

[0083] ;

[0084] This is the minimum safety threshold.

[0085] Step S12, the training process of the support vector machine classification model, includes:

[0086] Step S121: Construct a training set from clinical impedance scattering ratio data , where j is the index of the j-th training sample. Indicates the impedance scattering ratio. , representing the danger level label of the j-th sample, yields the cleaned and labeled training dataset;

[0087] Step S122: Use grid search method to apply the penalty parameter Search within the range, step size For each C value, train an SVM model and calculate its performance metrics on the cross-validation set to select the optimal penalty parameters;

[0088] Step S123: Based on the training dataset and the optimal penalty parameters, the SVM model is trained and validated using the 5-fold cross-validation method. The training dataset is randomly divided into 5 subsets. Four subsets are used for training each time, and the remaining subset is used for validation.

[0089] Step S124: Calculate the validation accuracy for each fold and take the average as the final validation accuracy;

[0090] Step S125: When the final validation accuracy exceeds the target validation accuracy, output and save the validated optimal SVM model.

[0091] In this embodiment, the treatment safety database collects a large amount of critical impedance scattering ratio (CR) data for different skin types, providing a basis for threshold determination. The support vector machine (SVM) classification model uses this data to predict the hazard level of the current CR using the RBF kernel function (σ=0.5). For example, when the CR approaches or exceeds the historical high-risk threshold, the model classifies it as high-risk. At this time, the system automatically updates... Thresholds that combine the current threshold with the minimum safe threshold reduce security risks.

[0092] The decision function of the SVM model is :

[0093] ;

[0094] in, Let be the impedance scattering ratio of the k-th support vector, where k is the index of the support vector and n represents the total number of support vectors. Let Lagrange multipliers be the k-th support vector. Let be the label of the k-th support vector, and b be the bias term. For RBF kernel function, , This is the bandwidth parameter.

[0095] Multi-objective optimization algorithms specifically include:

[0096] Step S21: Based on the skin stratum corneum thickness, sebum secretion rate, and the i-th wavelength, construct the first objective function F:

[0097] ;

[0098] in, For wavelength matching weights, For keratin attenuation weight, For sebum suppression weight, For the i-th wavelength, The characteristic absorption peak of the target tissue Here, is a characteristic value of stratum corneum thickness, and e is a natural constant.

[0099] Step S22: Perform multi-objective optimization on the effects of wavelength matching degree, stratum corneum thickness and sebum secretion rate to obtain a progeny population containing elite individuals, and screen the non-dominated solution set of the Pareto front from the progeny population containing elite individuals.

[0100] Step S23: Based on the non-dominated solution set of the Pareto front and the thickness d of the stratum corneum, select the target wavelength and obtain the target wavelength and its corresponding weight coefficient.

[0101] In this embodiment, different weights are assigned to three factors: wavelength matching degree, stratum corneum thickness, and sebum secretion rate. For example, when the patient's stratum corneum is thicker, the effect of the stratum corneum attenuation weight β is more significant, prompting the algorithm to select a wavelength combination more suitable for the thick stratum corneum. Multi-objective optimization search is performed through operations such as initializing the population and setting crossover probabilities. Pareto front screening then selects the solution that performs well on multiple objectives from all possible solutions, forming the optimal wavelength combination. This allows for adjustment of the wavelength combination according to the patient's specific skin condition, improving the stability and reliability of the treatment effect, and is particularly suitable for patients with different skin types and disease severity.

[0102] The specific steps for obtaining the non-dominated solution set of the Pareto front in step S22 include:

[0103] Step S221: Obtain population size and crossover probability Initialize the population, where each individual represents a set of wavelengths and their corresponding weight coefficients. Simultaneously, the weight coefficients corresponding to the i-th wavelength are... Introducing constraints during the initialization phase , H is the total number of wavelengths, and H is the initial population that maximizes the number of individuals that meet the constraints (creating the largest possible initial population in which each individual satisfies the constraint that the sum of the weight coefficients is 1).

[0104] Step S222: Adjust the crossover probability of the initial population according to the sebum secretion rate, using the adaptive crossover probability adjustment formula: , To obtain the new crossover probability, when the sebum secretion rate is greater than the target sebum secretion rate, the crossover probability is increased, resulting in the offspring population after the crossover operation.

[0105] Step S223: The elite individuals in the parent population whose first objective function F value is greater than the optimization objective threshold are directly retained in the offspring population to obtain an offspring population containing elite individuals.

[0106] In this embodiment, for patients with high sebum secretion rates, this constraint ensures sufficient weight is allocated to wavelengths that have a stronger effect on deep sebaceous glands. The adaptive crossover rate is dynamically adjusted based on the sebum secretion rate; when s > 20 μg / cm²•h, the crossover probability is increased to promote genetic diversity and prevent premature convergence of the algorithm. The elite preservation strategy directly passes on high-performing individuals to the next generation, maintaining superior traits.

[0107] Pulse duration Optimization methods also include:

[0108] Step S31: Establish thermal damage constraints Used to limit the energy deposition of laser in skin tissue. For the safe energy density threshold, Let be the wavelength energy corresponding to the i-th wavelength at time t;

[0109] Step S32: Obtain the pulse duration and constraint weights By incorporating constraints into the second objective function to solve for the optimal pulse duration, the second objective function is constructed. ,in, For the i-th wavelength during the pulse duration Total energy within;

[0110] Step S33: Through iterative solution, continuously adjust the pulse duration until the second objective function... When the target convergence condition is met, the optimal pulse duration satisfying the thermal damage constraint is obtained, whereby iterative solution yields... Correction formula , The corrected pulse duration is the duration of the i-th wavelength.

[0111] In this embodiment, the energy deposition of the laser in the skin tissue is limited to a safe threshold Φ_max = 15 J / cm². For example, this threshold is determined based on the skin tissue's tolerance to prevent excessive energy from causing thermal damage. An objective function is constructed using the Lagrange multiplier method, comprehensively considering both energy utilization efficiency and thermal damage constraints. During the iterative solution process, the pulse duration is continuously adjusted according to a modified formula to gradually approach the optimal value.

[0112] The specific implementation method of step S33 iterative solution is as follows:

[0113] Step S331: Obtain the initial constraint weights and the target convergence condition , The difference between the second objective function value L and the second objective function value in the previous iteration;

[0114] Step S332: Calculate the cumulative energy in each iteration. ; This represents the cumulative energy calculated in the m-th iteration. The weighting coefficient for the i-th wavelength in the m-th iteration is... The pulse duration corresponding to the i-th wavelength in the m-th iteration;

[0115] Step S333: Update the constraint weights and adjust the weights of the second objective function. The update formula is as follows: , For learning rate, These are the constraint weights in the (m+1)th iteration. Let be the constraint weights in the m-th iteration.

[0116] The calibration process for the drive current includes the following steps:

[0117] Step S41: Connect a detection resistor R in parallel with the load terminal of the laser diode;

[0118] Step S42: Measure the voltage drop V in real time and calculate the actual current value. ;

[0119] Step S43: Adjust the drive voltage using a PID controller to make... , This is the target current value.

Claims

1. A multi-wavelength pulsed fiber laser for treatment of seborrheic dermatitis, characterized in that, The method comprises the following steps: The data acquisition module is used to obtain the stratum corneum thickness and sebum excretion rate of the patient, and synchronously obtain the skin impedance and optical scattering coefficient, and calculate the current impedance scattering ratio; The treatment planning module is used to trigger the wavelength reorganization instruction to recalculate the weight coefficient when the current impedance scattering ratio exceeds the first threshold value; The wavelength optimization module is used to generate the weight coefficient containing the target wavelength through a multi-objective optimization algorithm according to the stratum corneum thickness and sebum excretion rate of the patient; The multi-objective optimization algorithm specifically comprises the following steps: Step S21: a first objective function F is constructed according to the stratum corneum thickness, the sebum excretion rate and the i-th wavelength; ； wherein, is a wavelength matching weight, is a stratum corneum attenuation weight, is a sebum suppression weight, is the ith wavelength, is a target tissue characteristic absorption peak, is a stratum corneum thickness characteristic value, e is a natural constant, d is the skin stratum corneum thickness, and s is the sebum secretion rate. Step S22: a multi-objective optimization is performed on the wavelength matching degree, the stratum corneum thickness influence and the sebum excretion rate influence to obtain a child population containing elite individuals, and a non-dominated solution set of a Pareto front is screened from the child population containing the elite individuals; Step S23: the target wavelength is screened according to the non-dominated solution set of the Pareto front and the stratum corneum thickness d, and the target wavelength and the corresponding weight coefficient are obtained; The pulse modulation module is used to calculate the pulse duration corresponding to each wavelength according to the weight coefficient; The energy control module is used to integrate the energy of each wavelength and calculate the total output energy; The skin feedback module is used to obtain the temperature distribution of the treatment area in real time and generate a temperature gradient matrix; The laser emission module is used to generate a driving current according to the total output energy and the temperature gradient matrix, drive the laser emission, and emit the multi-wavelength pulsed fiber laser for treating seborrheic dermatitis.

2. The multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment according to claim 1, wherein, In the data acquisition module, the skin impedance is obtained in the following manner: Step S01: a four-electrode bioimpedance measurement device is used to apply a 1mA / 50kHz alternating current signal to the skin to be measured; Step S02: separating the impedance real part by a phase detection circuit and imaginary part ; Step S03: Calculate the effective skin impedance value , sampling frequency 100 Hz; Step S04: Z is transmitted to the data acquisition module by a wireless transmission mode.

3. The multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment according to claim 1, wherein, In the treatment planning module, the first threshold The determination method comprises: Step S11: critical impedance scattering ratios and clinical impedance scattering ratio data of different skin types are stored to establish a treatment safety database; Step S12: based on the current impedance scattering ratio Q value and the treatment safety database, a support vector machine classification model is used to predict the risk level of the current impedance scattering ratio Q value; Step S13: When predicted as non-high risk level, no processing is done, and when predicted as high risk level, the first threshold is automatically updated to : ； is the minimum safety threshold.

4. The multi-wavelength pulsed fiber laser for treatment of seborrheic dermatitis according to claim 3, characterized in that, In step S12, the training process of the support vector machine classification model comprises the following steps: Step S121: constructing a training set for clinical impedance scatter ratio data j is an index of the jth training sample, represents an impedance scatter ratio value of the jth training sample, represents a risk level label of the jth sample, and the training data set is obtained by cleaning and labeling the training data set. Step S122: search in the range of penalty parameter with step size For each value of C, train an SVM model and calculate its performance indicator on the cross-validation set, and select the optimal penalty parameter. Step S123: according to the training data set and the optimal penalty parameter, a 5-fold cross-validation method is used to train and verify the SVM model, the training data set is randomly divided into 5 subsets, 4 subsets are used for training each time, and the remaining 1 subset is used for verification; Step S124: the verification accuracy of each fold is calculated, and the average value is taken as the final verification accuracy; Step S125: when the final verification accuracy exceeds the target verification accuracy, the verified optimal SVM model is output and saved.

5. The multi-wavelength pulsed fiber laser for treatment of seborrheic dermatitis according to claim 4, characterized in that, Decision function of the SVM model is: ； wherein, is the impedance scattering ratio value of the kth support vector, k is the index of the support vector, and n represents the total number of support vectors, is the Lagrange multiplier of the kth support vector, is the label of the kth support vector, and b is the bias term, is the RBF kernel function, , is the bandwidth parameter.

6. The multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment according to claim 1, wherein, The specific obtaining steps of the non-dominated solution set of the Pareto front in step S22 comprise the following steps: Step S221: Obtain population size and crossover probability , initialize the population, each individual in the population represents a set of wavelengths and its corresponding weight coefficient, at the same time, the weight coefficient corresponding to the i-th wavelength The initialization stage introduces a constraint condition , , H is the total number of wavelengths; Step S222: adjusting the cross probability of the initial population according to the sebum excretion rate, using an adaptive cross probability adjustment formula as , is the new cross probability, when the sebum excretion rate is greater than the target sebum excretion rate, the cross probability is increased, and a child population after the cross operation is obtained; Step S223: the elite individuals with the first objective function F value greater than the optimization target threshold in the parent population are directly reserved to the child population to obtain the child population containing the elite individuals.

7. The multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment according to claim 1, wherein, In the pulse modulation module, the pulse duration The acquisition method is: , The reference pulse width is, The zero constant is prevented, d is the skin stratum corneum thickness, and s is the sebum secretion rate, The weight coefficient of the i-th wavelength is; pulse duration The optimization method of claim 1, further comprising: Step S31: Establishing thermal damage constraint condition for limiting energy deposition of laser in skin tissue, is a safety energy density threshold, is a wavelength energy corresponding to the i-th wavelength at time t, is a pulse duration corresponding to the i-th wavelength; Step S32: Acquire pulse duration and constraint weight , the constraint condition is integrated into the second objective function to solve the optimal pulse duration, and the second objective function is constructed , wherein, is the total energy of the ith wavelength within the pulse duration ; Step S33: Adjust the pulse duration constantly by iterative solving, when the second target function reaches the target convergence condition, the optimal pulse duration satisfying the thermal damage constraint is obtained, wherein the iterative solving obtains the modified formula of the pulse duration , is the modified pulse duration of the i-th wavelength, is the absorption coefficient of the i-th wavelength.

8. The multi-wavelength pulsed fiber laser for treatment of seborrheic dermatitis according to claim 7, characterized in that, The specific implementation mode of step S33 iterative solution is as follows: Step S331: Obtain the initialization constraint weight and the target convergence condition , is the difference between the second objective function value L and the last second objective function value. Step S332: calculating the energy accumulation amount each time iteration ; is the energy accumulation amount calculated in the mth iteration, is the weight coefficient of the ith wavelength in the mth iteration, is the pulse duration corresponding to the ith wavelength in the mth iteration; Step S333: updating the constraint weight adjusts the weight of the second objective function, and the updating formula is: , is the learning rate, is the constraint weight in the m+1th iteration, is the constraint weight in the mth iteration.

9. The multi-wavelength pulsed fiber laser for seborrheic dermatitis treatment according to claim 1, wherein, In the laser emission module, the calibration process of the driving current comprises the following steps: Step S41: a detection resistor R is connected in parallel at the load end of the laser diode; Step S42: Measure the voltage drop V in real time, calculate the actual current value ; Step S43: Adjust the driving voltage by the PID controller so that , is the target current value.

Images