A transcranial photobiomodulation pulse control method and system for tissue temperature prediction

Abstract

This invention provides a transcranial photobiomodulation pulse control method and system for tissue temperature prediction, relating to the field of adaptive pulse modulation technology. The method includes: acquiring operational data during the transcranial photobiomodulation process and preprocessing the operational data; inputting standard operational data into a preset tissue thermal state observation model to obtain the current target area tissue temperature, target area temperature rise rate, and target area cumulative absorbed dose; obtaining the target area temperature trajectory through a photothermal prediction model based on the current target area tissue temperature and candidate pulse sequences; generating a candidate pulse parameter set by combining the current target area tissue temperature, target area temperature rise rate, target area temperature trajectory, and target area cumulative absorbed dose; performing parameter consistency projection on the candidate pulse parameter set to obtain an executable pulse parameter set that satisfies pulse timing constraints; and updating the actual output optical power according to the driving light source of the executable pulse parameter set until the termination trigger condition of transcranial photobiomodulation is met.

Description

Technical Field

[0001] This invention relates to the field of adaptive pulse modulation technology, and in particular to a transcranial photobiological modulation pulse control method and system for tissue temperature prediction. Background Technology

[0002] Transcranial photobiological modulation (TPV) is a photostimulation method that uses red or near-infrared light to stimulate brain tissue through the scalp and skull. The light is absorbed and scattered in multiple layers of tissue, with some of the absorbed light energy being converted into heat. Therefore, the peak power, duty cycle, pulse width, and pulse interval of the light source not only determine the light dose to the target area but also collectively determine the transient temperature rise of the scalp and brain tissue. Individual factors such as the tissue's optical absorption coefficient, scattering coefficient, blood perfusion, ambient temperature, and the fit of the light source all affect the actual temperature distribution. Therefore, monitoring and management of tissue temperature are necessary during TPV.

[0003] Existing transcranial photobiological modulation devices typically employ continuous light, preset power pulses, or shutdown protection based on a single temperature upper limit as their primary control schemes. Temperature threshold control, as a fundamental safety measure, can shut off the light source when the temperature reaches a set upper limit to prevent tissue overheating. Pulsed light output can adjust the average light power according to preset duty cycles and frequencies, thus balancing instantaneous peak power with average heat input to some extent. These methods play a fundamental role in the safety protection and dose control of transcranial photobiological modulation, providing a feasible working framework for subsequent more refined temperature management.

[0004] However, existing control schemes mainly rely on the current temperature to make judgments, without using the rate of temperature rise to identify states where the temperature is still low but is rising rapidly. Furthermore, temperature control and cumulative dose control are often separate and lack a unified coordination mechanism. Summary of the Invention

[0005] Given that the existing control schemes mainly rely on the current temperature to make judgments, they do not utilize the temperature rise rate to identify states where the temperature is still low but is rising rapidly, and temperature control and cumulative dose control are often separate, lacking a unified coordination mechanism.

[0006] A first aspect of this invention provides a transcranial photobiological modulation pulse control method for tissue temperature prediction, comprising: S1: Acquire operational data during the transcranial photobiological modulation process, including surface temperature, ambient temperature, actual output light power, and pulse switch status; S2: Preprocess the running data to obtain standard running data; S3: Input the standard operating data into the preset tissue thermal state observation model to obtain the current target area tissue temperature, target area temperature rise rate and target area cumulative absorbed dose; S4: Generate at least one candidate pulse sequence according to the preset pulse parameter boundary, and obtain the target area temperature trajectory corresponding to each candidate pulse sequence in the prediction time domain through the photothermal prediction model based on the current target area tissue temperature, the current tissue equivalent thermal state and the candidate pulse sequence. S5: Combine the current target area tissue temperature, target area temperature rise rate, target area temperature trajectory corresponding to each candidate pulse sequence, and target area cumulative absorbed dose to generate candidate pulse parameter set; S6: Perform parameter consistency projection on the candidate pulse parameter set to obtain an executable pulse parameter set that satisfies the pulse timing constraints; S7: Update the actual output optical power according to the driving light source of the executable pulse parameter group, and return to step S1 until the termination trigger condition of transcranial photobiological modulation is met.

[0007] A second aspect of the present invention provides a transcranial photobiological modulation pulse control system for tissue temperature prediction, comprising: a processor and a memory; The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the transcranial photobiomodulation pulse control method for tissue temperature prediction as described in the first aspect.

[0008] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this embodiment of the invention, addressing the shortcomings of existing control schemes that do not utilize the temperature rise rate and separate temperature control from dose control, the target area temperature rise rate is first obtained through a tissue thermal state observation model. Then, by combining the current target area tissue temperature, the target area temperature rise rate, the target area temperature trajectory, and the cumulative absorbed dose in the target area, a candidate pulse parameter set is generated. This allows for the reduction of equivalent heat input by adjusting the pulse parameters before the current temperature reaches the hard threshold, thereby suppressing subsequent temperature overshoot. Simultaneously, the cumulative absorbed dose in the target area is incorporated as one of the state inputs into the pulse parameter decision-making process until the termination trigger condition is met, ensuring unified coordination between temperature management and dose management within the same control closed loop. Through these technical means, the problems of missing temperature rise rate and separation of temperature control and dose control in existing schemes are resolved. Attached Figure Description

[0009] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0010] Figure 1This is a schematic flowchart of a transcranial photobiological modulation pulse control method for tissue temperature prediction provided in an embodiment of the present invention.

[0011] Figure 2 This is a comparison chart of prediction errors of a temperature prediction algorithm provided in an embodiment of the present invention under multiple scenarios.

[0012] Figure 3 This is a comparison chart of temperature ripple prediction results for different schemes provided in an embodiment of the present invention.

[0013] Figure 4 This is a schematic diagram of a transcranial photobiological modulation pulse control system for tissue temperature prediction provided in an embodiment of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions 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, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] The transcranial photobiological modulation pulse control method for tissue temperature prediction provided by the present invention will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0016] Reference manual attached Figure 1 The diagram shows a flowchart of a transcranial photobiological modulation pulse control method for tissue temperature prediction provided by an embodiment of the present invention.

[0017] This invention provides a transcranial photobiological modulation pulse control method for tissue temperature prediction, which may include the following steps: S1: Acquire operational data during the transcranial photobiological modulation process, including surface temperature, ambient temperature, actual output light power, and pulse switching status.

[0018] Transcranial photobiological modulation is a photostimulation method that uses red or near-infrared light to act on brain tissue through the scalp and skull. Surface temperature is the contact temperature of the area where the light source contacts the skin or the non-contact surface temperature. Ambient temperature is the external ambient temperature during the transcranial photobiological modulation process. Actual output light power is the peak or average power of the light source during the on-state. Pulse switching state is a discrete signal that characterizes whether the light source is currently on or off.

[0019] Set target cumulative dose Dtar Maximum cumulative dose D max Temperature warning threshold T warn Hard temperature threshold T hard Predicted upper temperature limit T pred Temperature rise rate threshold k T Predicting safety margin M T Control cycle △t c and prediction steps N p Set upper and lower limits for peak power, switching on and off time, and constraints on the rate of change of control quantities.

[0020] It should be noted that those skilled in the art can set the temperature warning threshold, hard temperature threshold, upper limit of predicted temperature, temperature rise rate threshold, prediction safety margin, and prediction step size according to actual needs, and this invention does not limit these settings.

[0021] Traditional protection logic typically involves shutting down the system when the temperature reaches a threshold. The early intervention logic proposed in this invention is as follows: The rate of temperature rise is defined as: in, This represents the rate of temperature rise, an estimate of the first-order rate of change of the target area temperature with respect to time. T r This indicates the temperature of the target area.

[0022] when and At the same time, reduce the equivalent heat input.

[0023] in, k T Indicates the threshold for the rate of temperature rise. This indicates the hard temperature threshold.

[0024] Therefore, the present invention does not solve the simple problem of over-temperature shutdown, but rather uses temperature trends and predicted trajectories to generate an executable set of pulse parameters before the limit is exceeded.

[0025] In this embodiment of the invention, a complete operational boundary is constructed for the entire control process by establishing a parameter set including the target cumulative dose, maximum cumulative dose, temperature warning threshold, hard temperature threshold, predicted upper temperature limit, temperature rise rate threshold, predicted safety margin, control cycle, and prediction steps, as well as upper and lower limits of peak power, light-on time, light-off time, and control variable change rate constraints. Based on this, the proposed early intervention logic enables the system to proactively reduce the equivalent heat input before the temperature exceeds the limit, based on the temperature rise rate judgment, even when the target temperature is still below the hard temperature threshold. This shifts the control action from reactive shutdown to proactive predictive adjustment.

[0026] S2: Preprocess the running data to obtain standard running data.

[0027] The preprocessing process involves removing outliers, compensating for delays, performing low-pass filtering, and comparing the rationality of multiple sensors on the operating data. The standard operating data is a preprocessed effective temperature sequence used for subsequent state estimation.

[0028] In one possible implementation, S2 specifically includes sub-steps S201 to S203: S201: Perform outlier removal processing on the running data to obtain the running data after removing outliers.

[0029] Outlier removal involves marking and removing temperature measurements that exceed a reasonable range based on a preset temperature change rate threshold or statistical outlier criteria.

[0030] It should be noted that those skilled in the art can set the preset temperature change rate threshold according to actual needs, and this invention does not limit that.

[0031] S202: Perform delay compensation and low-pass filtering on the running data after removing outliers to obtain the filtered and compensated running data.

[0032] Among them, delay compensation is a timing correction for the lag in the response of the temperature sensor, and low-pass filtering is a smoothing process to suppress high-frequency noise in the temperature signal.

[0033] S203: Perform a rationality comparison and verification of the multi-channel sensor acquisition data after filtering and compensation, and use the rationality comparison and verification of the operating data as the standard operating data.

[0034] Among them, the rationality comparison and verification of multi-channel sensor data acquisition involves comparing the readings of multiple temperature sensors at the same time to determine whether the data is reasonable.

[0035] The temperature acquisition module reads the temperature of the light source-skin contact area or the temperature of the non-contact surface. Ts ( k ), ambient temperature T a ( k Actual optical power P m ( k ), pulse switch state s ( k (and a unified timestamp.)

[0036] Specifically, outlier removal, delay compensation, and low-pass filtering are performed on the temperature data, and a reasonableness comparison is made between multiple temperature sensors.

[0037] Specifically, the temperature acquisition and calibration module is used to acquire contact temperature, non-contact temperature, ambient temperature and corresponding timestamps, perform filtering, drift correction, sensor rationality verification and data validity marking, and output a reliable temperature sequence.

[0038] In this embodiment of the invention, by sequentially performing outlier removal, delay compensation, low-pass filtering, and multi-channel sensor plausibility verification on the operating data, the original temperature signal undergoes a complete quality screening and timing correction before entering the state estimation stage. Outlier removal marks and removes measured values ​​that exceed the reasonable range, preventing a single erroneous reading from interfering with subsequent state estimation. Delay compensation corrects the timing offset caused by the inherent response hysteresis of the sensors. Low-pass filtering suppresses high-frequency noise in the temperature signal. Multi-channel comparison verification ensures that the temperature sequences involved in the calculation have basic consistency at the sensing level. The standard operating data obtained after the above preprocessing provides reliable information input for subsequent tissue thermal state observation models.

[0039] S3: Input the standard operating data into the preset tissue thermal state observation model to obtain the current target area tissue temperature, target area temperature rise rate, and target area cumulative absorbed dose.

[0040] Among them, the tissue thermal state observation model is a mathematical model used to estimate the target tissue thermal state based on surface temperature and historical light source input. The target tissue temperature is the current temperature estimate of the region of interest in the transcranial photobiological regulation target area. The target temperature rise rate is the first-order rate of change of the target tissue temperature with respect to time. The cumulative absorbed dose of the target area is the time integral value of the absorbed power density in the target area from the initial time to the current time.

[0041] In one possible implementation, S3 specifically includes sub-steps S301 to S303: S301: Update the equivalent thermal state of tissues by using a preset tissue thermal state observation model, combined with ambient temperature in standard operating data and preset individual tissue parameters.

[0042] Among them, individual tissue parameters include tissue thickness, absorption coefficient, scattering coefficient, anisotropy factor, density, specific heat capacity, thermal conductivity and blood perfusion rate, and tissue equivalent thermal state is a set of intermediate state variables describing the spatiotemporal characteristics of temperature distribution inside the tissue.

[0043] It should be noted that those skilled in the art can set the preset size of individual tissue parameters according to actual needs, and this invention does not limit this.

[0044] S302: Through the preset tissue thermal state observation model, calculate the updated tissue equivalent thermal state and output the current target area tissue temperature and target area temperature rise rate.

[0045] S303: The cumulative absorbed dose in the target area is obtained by iterative calculation based on the actual output optical power in the standard operating data.

[0046] The iterative calculation involves performing time integration on the target region's absorbed power density within each control cycle and then summing the results to the cumulative value of the previous cycle.

[0047] The tissue thermal state observation model updates the state based on surface temperature and historical light source inputs: in, x Indicates the equivalent thermal state of the organization. A ( i k )and B ( i k This indicates dependence on individual organizational parameters. i k The state transition matrix and the input matrix, i k Indicates individual organizational parameters, q ( k () represents the equivalent absorbed heat input. E Represents the perturbation input matrix. d ( k This includes disturbances such as ambient temperature and blood temperature. L Indicates the observer gain. Indicates the first k The surface or contact temperature is collected in each control cycle. This represents the surface temperature observation matrix.

[0048] Specifically, the current target area temperature is obtained based on the tissue thermal state observation model: in, T r ( k ) indicates the firstk The target area temperature is estimated by the tissue thermal state observation model for each control cycle. C r This indicates the relationship used to derive target area temperature from tissue thermal state. x ( k ) indicates the first k The equivalent thermal state of the tissue during each control cycle.

[0049] Specifically, the rate of temperature rise is estimated using a sliding window least squares method: in, v T ( k () indicates the rate of temperature rise. m Indicates a sliding window. t j This indicates the sampling time within the sliding window. This represents the average sampling time within the sliding window. This represents the target area temperature at the corresponding time. This represents the average temperature of the target area within the sliding window.

[0050] Furthermore, compared to directly subtracting adjacent samples, the above estimation can suppress the influence of sensor noise on rate determination.

[0051] For example, the cumulative absorbed dose in the target area is integraled according to the absorbed power density within the target area: in, Indicates time t Average absorbed power density within the target area V r Indicates the target volume. Oh r Indicates the target area. Indicates position r place, time t Absorption power density, dV This represents a volume integral infinitesimal element.

[0052] The cumulative absorbed dose in the target area is integrated over time according to the average absorbed power density in the target area: Furthermore, the cumulative absorbed dose in the target area is iterated over time: in, D r ( k +1) indicates the first k +1 control cycle of cumulative absorbed dose to the target area. Dr ( k ) indicates the first k The cumulative absorbed dose to the target area over each control cycle. Indicates time t Average absorbed power density within the target area.

[0053] In this embodiment of the invention, a preset tissue thermal state observation model is used to update the equivalent thermal state of the tissue by combining ambient temperature and individual tissue parameters, thereby calculating the current target area tissue temperature and the target area temperature rise rate. Simultaneously, the cumulative absorbed dose of the target area is calculated iteratively based on the actual output optical power. The tissue thermal state observation model establishes a quantitative mapping between the measurable surface temperature and the target area temperature, enabling online estimation of the target area temperature without invasive measurement. Sliding window least squares estimation suppresses the influence of sensor noise on rate determination and avoids numerical jitter caused by direct subtraction of adjacent samples. The cumulative absorbed dose of the target area is updated synchronously as a second state variable besides temperature, providing a unified state benchmark for subsequent consideration of both temperature safety and dose completion.

[0054] S4: Generate at least one candidate pulse sequence according to the preset pulse parameter boundary, and obtain the target temperature trajectory corresponding to each candidate pulse sequence in the prediction time domain through the photothermal prediction model based on the current target area tissue temperature, the current tissue equivalent thermal state and the candidate pulse sequence.

[0055] The candidate pulse sequence includes the peak power, pulse width, and pulse interval corresponding to each control cycle in the prediction time domain. The photothermal prediction model recursively infers the future temperature response based on the current equivalent thermal state of the tissue and the candidate pulse sequence. The target area temperature trajectory is a sequence of predicted values ​​of the target area tissue temperature over time in the prediction time domain, and it corresponds one-to-one with the corresponding candidate pulse sequence.

[0056] Optionally, the photothermal prediction model includes an offline high-fidelity model and an online reduced-order model.

[0057] The offline high-fidelity model, based on the light diffusion equation and the biothermal equation, is used to generate photothermal response data according to preset or calibrated tissue parameters and to identify the response parameters of the online order-reduction model. The online order-reduction model is built based on the photothermal response data output by the offline high-fidelity model and the identified parameters, and is used to recursively deduce the target area temperature trajectory in real time.

[0058] The offline high-fidelity model is a refined numerical model based on the light diffusion equation and the biothermal equation, used to identify individualized tissue parameters and generate training data. The online reduced-order model is a computationally simplified model built based on the output parameters of the offline high-fidelity model, used to recursively deduce the temperature trajectory of the target area in real time. The light diffusion equation is an optical approximation equation describing the propagation of photons in scattering-dominant tissues, and the biothermal equation is the Pennes biothermal equation describing the heat transfer in biological tissues.

[0059] For head tissue where scattering is dominant, the diffusion approximation can be used to describe the luminous flux (Phi): in, ▽ Represents the gradient operator, F Indicates luminous flux. D opt Indicates the optical diffusion coefficient. m a Indicates the absorption coefficient. S This indicates the light source item.

[0060] Specifically, the optical diffusion coefficient and the reduced scattering coefficient are: in, Represents the reduced scattering coefficient. m s Represents the scattering coefficient. g This represents the anisotropy factor.

[0061] Specifically, the absorbed power density is: in, Indicates position r place, time t Absorption power density, m a ( r ) indicates position r The absorption coefficient at that location, F ( r ) indicates position r Luminous flux at that location Indicates time t Peak optical power, s ( t ) indicates the pulse switch state. P ref Indicates the reference power.

[0062] It should be noted that tissue temperature is calculated using the Pennes biothermal equation: in, r Indicates tissue density, c Indicates the specific heat capacity of the tissue. T Indicates tissue temperature, k Indicates the thermal conductivity of the tissue. r b Indicates blood density, cb This indicates the specific heat capacity of blood. oh b Indicates blood perfusion. T b Indicates blood temperature. Q met Indicates metabolic heat production. Q opt This represents the heat input generated by light absorption.

[0063] Furthermore, the offline high-fidelity model can be implemented using COMSOL. The online controller employs a state-space model, thermal resistance / capacity network, or surrogate model obtained by reducing the order of the high-fidelity model.

[0064] Given a candidate pulse sequence, the prediction model recursively derives the future temperature trajectory. Considering the upper bound of the model error, the prediction safety constraint is: in, Indicates the first k The first control cycle predicts the future pulse sequence based on the current state and candidate pulse sequences. i Temperature in the target area Indicates the first i The upper bound of the model error of the step. M T This indicates the predicted safety margin. T pred This indicates the upper limit of the predicted temperature. N p This indicates the number of prediction steps.

[0065] For example, the COMSOL model can employ a two-dimensional axisymmetric layered head model or a three-dimensional individualized head model. Domains for the scalp, skull, cerebrospinal fluid, gray matter, and white matter are established sequentially. An incident flux is applied to the light source region, a convective heat dissipation boundary is set on the outer surface of the skin, and a far-field temperature or a sufficiently extended computational domain is set at the deep boundary. First, the steady-state light field under unit reference power is solved, and the absorbed power density is saved as a spatial basis function. Then, the real-time pulsed power is multiplied in the transient biothermal model. The average temperature of the target area, the maximum temperature of the target area, and the cumulative dose are calculated using domain averaging, domain maximization, and global ordinary differential equations, respectively. Rule-based control is implemented through Events, or the state and control variables are exchanged with an external predictive controller via the LiveLink interface. This modeling step is used to establish and identify the control model.

[0066] Reference manual attached Figure 2 The figure shows a comparison of prediction errors of a temperature prediction algorithm provided by an embodiment of the present invention under multiple scenarios.

[0067] Specifically, Figure 2The system includes a root mean square error (RMSE) layer and a maximum absolute error layer, both expressed in units of K. Each layer contains two sets of error parameters: "target area average" and "maximum overall head error." RMSE characterizes the overall deviation between the predicted and simulated observations, while maximum absolute error characterizes the largest single-point deviation during the prediction process. Specifically, schemes B and D represent representative data from a multi-source 10Hz phase strategy, scheme E represents actual square wave data, schemes F and H represent representative data from alternating and longer switching cycles, and schemes C and G are used for supplementary comparison and verification.

[0068] Furthermore, in terms of inclusion relationships, Figure 2 The RMSE layer contains the target area average RMSE and the highest overall RMSE for schemes B to H, respectively. The maximum absolute error layer contains the target area average maximum absolute error and the highest overall maximum absolute error for schemes B to H, respectively. In terms of connectivity, each scheme's error calculation result is formed by connecting the predicted values ​​from the same temperature prediction model with the corresponding simulated observation values. The target area average prediction result is connected to the target area average error index, and the highest overall prediction result is connected to the highest overall error index. These are then summarized and compared to the RMSE layer and the maximum absolute error layer, respectively. Relevant predicted values, observed values, and residuals are used to form the prediction error result.

[0069] It should be noted that, Figure 2 By comparing the average prediction error of the target area with the highest prediction error of the entire head under different schemes, the prediction error level of the temperature prediction model can be verified, providing a basis for subsequent rolling temperature prediction and maintaining an appropriate temperature safety margin.

[0070] In this embodiment of the invention, a photothermal prediction model is used to recursively deduce the target area temperature trajectory in the predicted time domain based on the current target area tissue temperature and candidate pulse sequences. This photothermal prediction model employs a hierarchical architecture of an offline high-fidelity model and an online reduced-order model: the offline high-fidelity model is established based on the light diffusion equation and the Pennes biothermal equation, used to identify individualized tissue parameters and generate training data. The online reduced-order model is established based on the output parameters of the offline high-fidelity model and is used to recursively deduce the target area temperature trajectory in real time. This hierarchical architecture enables the controller to balance the physical accuracy of the high-fidelity model with the real-time performance of online computation under conditions of limited computing resources. The setting of prediction safety constraints ensures that the system takes into account the model error boundary and the safety margin reserved in the design when making decisions, thereby effectively reducing the probability of prediction uncertainty transforming into actual temperature risk.

[0071] S5: Combine the current target area tissue temperature, target area temperature rise rate, target area temperature trajectory corresponding to each candidate pulse sequence, and target area cumulative absorbed dose to generate candidate pulse parameter set.

[0072] Among them, the candidate pulse parameter set is a combination of peak power, pulse width and pulse interval values ​​obtained from the decision within the current control cycle.

[0073] Optionally, the candidate pulse parameter set includes peak power, pulse width, and pulse interval, where peak power, pulse width, and pulse interval are independent control variables, and the duty cycle is a derived variable calculated based on pulse width and pulse interval.

[0074] Wherein, peak power is the instantaneous output optical power during the pulse switching phase, pulse width is the switching duration of a single pulse, and pulse interval is the switching duration between adjacent pulses.

[0075] In one possible implementation, S5 specifically includes sub-steps S501 to S503: S501: Determine the remaining dose in the target area based on the difference between the cumulative absorbed dose in the target area and the target dose.

[0076] The target dose is the target value of the cumulative absorbed dose in the target area preset by the transcranial photobiological modulation plan, and the remaining dose in the target area is the difference between the target dose and the current cumulative absorbed dose in the target area.

[0077] It should be noted that those skilled in the art can set the target value of the cumulative absorbed dose in the target area according to actual needs, and this invention does not limit it.

[0078] S502: Based on the current target area tissue temperature, target area temperature rise rate, and target area temperature trajectory, determine whether the load reduction triggering condition is met. If so, reduce at least one of the following: peak power, shorten pulse width, and lengthen pulse interval to reduce equivalent heat input, thus obtaining the constrained pulse parameter set. Otherwise, based on the remaining dose in the target area and the temperature safety margin, maintain or gradually restore at least one of the following: peak power, pulse width, and pulse interval within the upper and lower limit constraints and the parameter change rate constraints, thus obtaining the constrained pulse parameter set.

[0079] The load reduction trigger condition is a logical judgment condition in which the target area temperature rise rate exceeds the preset rate threshold or the predicted temperature value in the target area temperature trajectory exceeds the upper limit of the predicted temperature. The constrained pulse parameter group is a combination of peak power, pulse width and pulse interval values ​​that meet the current safety requirements after adjustment by load reduction operation or hold-and-recovery operation.

[0080] It should be noted that those skilled in the art can set the preset rate threshold according to actual needs, and this invention does not limit that.

[0081] S503: Output the constrained pulse parameter set as a candidate pulse parameter set.

[0082] This invention uses the current temperature, temperature rise rate, predicted future temperature, and cumulative dose in the target area as state inputs, and generates candidate values ​​for duty cycle, pulse width, peak power, and pulse interval through hierarchical rules or model prediction and optimization.

[0083] when At this time, the controller first reduces the average thermal input. For example, rule-based control can be written as: in, Indicates the candidate duty cycle. d ( k ) indicates the duty cycle. K v This represents the control coefficient related to the rate of temperature rise. v T ( k () indicates the rate of temperature rise. K p This represents the control coefficient related to temperature. This indicates the current maximum temperature. T ref Indicates reference temperature. sat ( ) indicates a limit.

[0084] Specifically, the peak power can be written as: in, Indicates the candidate peak power. Indicates peak power. This represents the temperature feedback coefficient.

[0085] Furthermore: in,( z ,0) indicates taking z The larger of 0.

[0086] Specifically, when the temperature margin recovers and the remaining dose is greater than zero, the output gradually recovers according to the set rate of change to avoid power jumps.

[0087] Preferably, model predictive control is used for solving: in, u Indicates the control quantity. J Indicates the optimization objective. N p Indicates the number of prediction steps. , , , These represent the weighting coefficients for the temperature deviation term, temperature rise rate deviation term, dosage deviation term, and control quantity change term, respectively. Indicates the future number i Temperature in the target area.

[0088] It should be noted that the duty cycle is not a fourth free quantity independent of the pulse width and interval, but is determined by the pulse timing: in, d Indicates duty cycle, t on The pulse width represents the duration of the switching on of a single pulse. t off This indicates the duration of light-off between adjacent pulses, i.e., the pulse interval.

[0089] Furthermore, if the controller directly generates d c , t on and t off Then a consistent projection needs to be performed.

[0090] In this embodiment of the invention, candidate pulse parameter sets are generated by using four state information items—current target area tissue temperature, target area temperature rise rate, target area temperature trajectory, and cumulative absorbed dose in the target area—as joint inputs, employing rule-based control or model predictive control. Specifically, the available pulse energy quota for subsequent pulses is determined based on the remaining dose in the target area. The current temperature, temperature rise rate, and predicted temperature trajectory are used to determine whether the load reduction triggering condition is met. If met, the peak power and pulse width are reduced in advance, and the pulse interval is extended to actively suppress subsequent overshoot caused by thermal inertia before the current temperature reaches the hard temperature threshold. If not met, the power output is maintained or gradually restored based on the remaining dose. Simultaneously, the duty cycle is explicitly defined as a derived quantity jointly determined by the pulse width and pulse interval, rather than a fourth degree of freedom determined independently, thus providing a clear physical constraint basis for subsequent consistent projection.

[0091] S6: Perform parameter consistency projection on the candidate pulse parameter set to obtain an executable pulse parameter set that satisfies the pulse timing constraints.

[0092] Among them, parameter consistency projection is a mathematical transformation that maps the peak power, pulse width, and pulse interval in the candidate pulse parameter set to the closest executable value in the feasible domain of the driver; pulse timing constraint is a physical constraint relationship in which the pulse width and pulse interval jointly determine the duty cycle; and executable pulse parameter set is the final combination of peak power, pulse width, and pulse interval that the light source driver can physically realize.

[0093] Optionally, the pulse timing constraint is that the duty cycle is determined by both the pulse width and the pulse interval.

[0094] The duty cycle is the ratio of the pulse width to the sum of the pulse width and the pulse interval.

[0095] In one possible implementation, S6 specifically includes sub-steps S601 to S603: S601: Calculate the candidate duty cycle based on the pulse width and pulse interval in the candidate pulse parameter group, combined with the pulse timing constraints.

[0096] S602: Under the conditions of satisfying pulse timing constraints, parameter upper and lower limit constraints, and parameter change rate constraints, perform parameter consistency projection on the peak power, pulse width, pulse interval, and candidate duty cycle in the candidate pulse parameter group to solve for executable peak power, executable pulse width, executable pulse interval, and executable duty cycle.

[0097] Among them, the upper and lower limit constraints of parameters are the physical allowable ranges of peak power, pulse width and pulse interval, and the parameter change rate constraint is the maximum allowable change amplitude of each pulse parameter between adjacent control cycles.

[0098] S603: Outputs the executable peak power, executable pulse width, executable pulse interval, and executable duty cycle as an executable pulse parameter set.

[0099] The executable parameter set is solved by taking the candidate control variable as input. in, u c This represents the candidate control quantity.

[0100] Specifically, the candidate control variables are: in, This represents the executable parameter set obtained from the solution. U Represents the feasible region. u min Indicates the lower limit of the parameter. u max Indicates the upper limit of the parameter. This indicates a change in the control quantity.

[0101] Furthermore, peak power, pulse width, and pulse interval are three independent control variables, while the duty cycle is a derived variable determined jointly by pulse width and pulse interval. The final output always satisfies the physical relationship of the pulse timing. The safety monitor then checks the hard temperature, dose limit, sensor status, communication status, and model confidence level. , , In the event of a malfunction, the safety monitor overrides the normal control commands and resets the light source to zero.

[0102] Reference manual attached Figure 3 The figure shows a comparison of temperature ripple prediction results for different schemes provided in an embodiment of the present invention.

[0103] Specifically, Figure 3 This includes the target area average temperature ripple layer and the ROI maximum temperature ripple layer. The vertical axis, peak-to-peak value (μK), represents the temperature ripple amplitude within the period. The horizontal axes B, C, D, E, F, and H represent different schemes, each including COMSOL / reduced-order observations and model predictions. The temperature prediction model in the specification is used to predict the target area average temperature and the ROI maximum temperature, and the relevant data records the temperature ripple within the period.

[0104] Furthermore, in terms of inclusion relationships, both the target area average temperature ripple layer and the ROI maximum temperature ripple layer include parameters B, C, D, E, F, and H, along with their corresponding COMSOL / reduced-order observations and model predictions. In terms of connectivity, the COMSOL / reduced-order observations of each scheme are compared one-to-one with the model predictions of the same scheme. Schemes B through H are respectively connected to the target area average temperature ripple layer and the ROI maximum temperature ripple layer, thus forming two types of peak-to-peak prediction comparisons.

[0105] It should be noted that, Figure 3 By comparing the model predictions with the COMSOL / reduced-order observations, it is helpful to reflect the prediction performance of the temperature prediction model on the average temperature ripple of the target area and the highest temperature ripple of the ROI under different schemes.

[0106] In this embodiment of the invention, by performing parameter consistency projection on the candidate pulse parameter set, an executable pulse parameter set that satisfies pulse timing constraints, parameter upper and lower limit constraints, and parameter rate of change constraints is obtained. Specifically, the candidate duty cycle is calculated based on the candidate pulse width and candidate pulse interval. Then, under the premise of satisfying the timing constraint relationship between the duty cycle and the pulse width and interval, consistency projection is performed on the four parameters of peak power, pulse width, pulse interval, and duty cycle to solve for the executable parameter set closest to the candidate value within the feasible region. This projection mechanism imposes amplitude and rate of change constraints on the three independent control quantities of peak power, pulse width, and pulse interval, and determines the duty cycle based on the pulse width and pulse interval. When the system interface simultaneously receives candidate duty cycles given externally, the consistency projection coordinates and corrects the relevant parameters within the feasible region, ensuring that the final output always satisfies the physical relationship that the duty cycle equals the pulse width divided by the sum of the pulse width and the pulse interval, thereby avoiding the problem of control commands being unexecutable due to constraint conflicts between parameters.

[0107] S7: Update the actual output optical power according to the driving light source of the executable pulse parameter group, and return to step S1 until the termination trigger condition of transcranial photobiological modulation is met.

[0108] Among them, driving the light source is the operation of converting the executable pulse parameter set into a light source current or voltage driving waveform and applying it to the light source, and the termination trigger condition is the basis for determining the termination of the transcranial photobiological regulation and control cycle.

[0109] Optionally, the termination triggering conditions include the cumulative absorbed dose in the target area reaching the target dose, or the triggering of a safety shutdown condition.

[0110] Among them, the safety shutdown condition is the basis for determining the safety status that triggers the forced shutdown of the light source.

[0111] The conditions that trigger a safety shutdown include at least one of the following: temperature exceeding the limit, sensor failure, communication anomaly, and insufficient model confidence.

[0112] Among them, temperature exceeding the limit means the surface temperature exceeds the upper limit of the surface temperature safety limit or the target area tissue temperature exceeds the upper limit of the target area temperature safety limit; sensor failure means the temperature sensor or power sensor fails or has abnormal output; communication abnormality means the data exchange between the controller and the sensor or light source driver is interrupted; and insufficient model confidence means the current output confidence of the tissue thermal state observation model or photothermal prediction model is lower than the preset threshold.

[0113] Specifically, when D r ≥ D tar The stimulus ends when the time is right. If the predicted constraint is temporarily infeasible but does not trigger a hard fault, execute the minimum safe output and extend the cooling interval. Shut down immediately when a hard fault is triggered.

[0114] Furthermore, a transcranial photobiological modulation adaptive pulse control system based on tissue temperature prediction includes a temperature acquisition and calibration module, a tissue state estimation module, a future temperature prediction module, a target dose calculation module, a pulse parameter decision module, a parameter consistency projection module, an independent safety monitor, and a light source driving module. All modules interact with each other via a unified clock and control cycle.

[0115] Specifically, the temperature acquisition and calibration module is used to acquire contact temperature, non-contact temperature, ambient temperature and corresponding timestamps, perform filtering, drift correction, sensor rationality verification and data validity marking, and output a reliable temperature sequence.

[0116] Specifically, the tissue state estimation module combines reliable temperature sequences, actual pulse power, environmental perturbations, and tissue parameters to estimate the equivalent thermal state of the scalp, skull, and target area, outputting the current target area temperature, temperature rise rate, and model confidence level. This module can employ a Luenberger observer, Kalman filter, extended Kalman filter, moving time-domain estimator, or equivalent thermal resistance / capacitance network.

[0117] It should be noted that the future temperature prediction module is used to recursively deduce the future temperature trajectory based on the current thermal state and candidate pulse sequence, and outputs the maximum target area temperature, the maximum surface temperature, the expected time to reach the warning boundary, and the prediction error bound within the predicted time domain.

[0118] Furthermore, the target dose calculation module calculates the target absorbed power density based on the tissue optical model, integrates the absorbed power density over time, and outputs the current cumulative dose, remaining target dose, and dose upper limit margin. For devices with limited computing resources, the offline optical field can be converted into a spatial basis function that scales linearly with peak power.

[0119] For example, the pulse parameter decision module is used to generate candidate values ​​for duty cycle, pulse width, peak power, and pulse interval by taking the current temperature, temperature rise rate, predicted future temperature, and cumulative dose in the target area as state inputs, and through hierarchical rules or model prediction optimization. v T > k T When the temperature reaches the hard temperature threshold, the module reduces the duty cycle, peak power, or switching time, and can increase the switching-off interval.

[0120] Specifically, the parameter consistency projection module is used to project candidate parameters onto the driver's feasible region so that the final duty cycle, pulse width, and pulse interval satisfy the following: It also satisfies the upper and lower limits of parameters, minimum resolution, maximum rate of change, and minimum time to turn off the light.

[0121] It should be noted that the independent safety monitor is used to determine hard temperature thresholds, dose limits, sensor malfunctions, model confidence levels, communication timeouts, and watchdog status. The safety monitor has the authority to override the output of the pulse parameter decision module. When the ordinary controller is infeasible or in an abnormal state, it outputs the minimum safe power, extends the cooling interval, or immediately shuts off the light source.

[0122] Furthermore, the light source driving module is used to convert the final parameter set into a light source current or voltage driving waveform, record the actual peak power, actual switching time and pulse count, and feed the actual output back to the tissue state estimation module and the target area dose calculation module.

[0123] For example, the system may also include a parameter storage module, a log module, a human-machine interaction module, and a communication module. The parameter storage module stores the tissue model, constraint thresholds, and light source calibration parameters. The log module records the state variables, predicted variables, candidate control variables, projected control variables, and safety events for each control cycle. The human-machine interaction module displays temperature margin, dosage progress, and equipment status. The communication module enables data exchange between the controller, temperature sensor, light source driver, and host computer.

[0124] Specifically, the temperature prediction algorithm of this invention consists of two parts: offline model fitting and online temperature prediction. In the offline stage, based on the long-term temperature rise curve and short-term real pulse temperature curve obtained from COMSOL, the fast and slow thermal response characteristics of the tissue are identified. In the online stage, based on the peak power of the light source, duty cycle, light-on time, light-off time, and current surface temperature, the maximum scalp temperature, average target area temperature, and maximum ROI temperature are predicted for a subsequent period. New surface temperature measurements can be used to continuously correct the prediction, thereby reducing the impact of individual differences, environmental changes, and model errors.

[0125] Furthermore, the algorithm's raw prediction data comes from COMSOL result tables. Continuous light reference data is located in "COMSOL Table Export\Multi-source 0.2". Representative data for the multi-source 10Hz phase strategy are located in "COMSOL Table Export\Multi-source Scheme B" and "COMSOL Table Export\Multi-source Scheme D-10Hz". Real square wave data is located in "COMSOL Table Export\Multi-source Scheme E". Representative data for alternating and longer switching cycles are located in "COMSOL Table Export\Multi-source Scheme F" and "COMSOL Table Export\Multi-source Scheme H-10Hz". Scheme C and G data under the same export directory are used for supplementary comparison and verification. The tables in each directory record data such as time, temperature rise or temperature, pulse gate function, and temperature ripple within the cycle.

[0126] Specifically, the model fitting program is "fit_temperature_predictor.py". This program reads the COMSOL result table mentioned above, fits the long-term temperature rise trend and the short-period temperature ripple respectively, and outputs the prediction parameters and errors. The main parameters after fitting are saved in "Temperature Prediction Algorithm Fitting Results\temperature_predictor_parameters.json" for online use by the controller. Simulation observations, predicted values, and residuals are saved in "slow_predictions.csv". The fitting time constant, prediction error, and data applicability limitations are summarized in "Temperature Prediction Algorithm Fitting Results\Fitting Report.md".

[0127] It should be noted that during online operation, the controller first updates the temperature prediction based on the current pulse parameters, then corrects it by combining the latest surface temperature measurement results, and performs rolling predictions of future temperatures under candidate pulse parameters. When the predicted temperature approaches the preset upper limit or the temperature rise trend is too rapid, the controller reduces the peak power or duty cycle in advance, shortens the switching time, or extends the switching interval. If the measured temperature, predicted temperature, cumulative dose, or equipment status triggers a hard safety condition, the independent safety monitor will execute load reduction or shutdown.

[0128] Specifically, the above parameters belong to one embodiment of the existing COMSOL model. Some long-term temperature results from scheme 0.2 exhibit spatial consistency anomalies and were not used for fitting the main slow model. The fast periodic data for schemes F and G are sparse, and the fast ripple data for scheme H includes reduced-order estimation. Therefore, after changing the light source arrangement, target area, or tissue parameters in the actual product, the same process should be used to re-export the data, refit the parameters, and verify the prediction error, while retaining an appropriate temperature safety margin.

[0129] In this embodiment of the invention, by sending the executable pulse parameter set to the light source driver to generate a driving waveform and output pulses, and simultaneously acquiring the actual output optical power to update the actual output optical power, a complete data closed loop is formed before returning to step S1. This achieves rolling optimization execution within each control cycle—that is, only the first pulse parameter set corresponding to the current cycle is executed, and the running data is reacquired and recalculated in the next cycle, ensuring that the control decision is always based on the latest temperature measurement information. The entire control loop continues to run until the cumulative absorbed dose in the target area reaches the target dose, or any safety shutdown condition, including temperature exceeding the limit, sensor failure, communication abnormality, or insufficient model confidence, is triggered. This ensures that the control process terminates under normal operating conditions with the dose target as the guide, and under abnormal operating conditions, it is forcibly terminated with safety protection as the priority.

[0130] Reference manual attached Figure 4 The diagram shows a structural schematic of a transcranial photobiological modulation pulse control system for tissue temperature prediction provided in an embodiment of the present invention.

[0131] This invention provides a transcranial photobiological modulation pulse control system 20 for tissue temperature prediction, comprising: a processor 201 and a memory 202; The memory 202 stores programs or instructions that can run on the processor 201. When the program or instructions are executed by the processor 201, they implement the steps of the transcranial photobiological modulation pulse control method for tissue temperature prediction described above, and can achieve the same technical effect. To avoid repetition, the present invention will not elaborate further.

[0132] It should be understood that the processor 201 in this embodiment of the invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0133] It should also be understood that the memory 202 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DR RAM).

[0134] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. Wireless means can include infrared, radio, or microwave. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0135] It should be understood that, in various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0136] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein 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.

[0137] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0138] In the several embodiments provided by this invention, it should be understood that the disclosed devices, 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 a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, 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 devices or units may be electrical, mechanical, or other forms.

[0139] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0140] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0141] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. 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 protection scope of the present invention.

Claims

1. A method of transcranial photobiomodulation pulse control for tissue temperature prediction, characterized by, include: S1: Acquire operational data during the transcranial photobiological modulation process, including surface temperature, ambient temperature, actual output light power, and pulse switch status; S2: Preprocess the operating data to obtain standard operating data; S3: Input the standard operating data into the preset tissue thermal state observation model to obtain the current target area tissue temperature, target area temperature rise rate and target area cumulative absorbed dose; S4: Generate at least one candidate pulse sequence according to the preset pulse parameter boundary, and obtain the target area temperature trajectory corresponding to each candidate pulse sequence in the prediction time domain through the photothermal prediction model based on the current target area tissue temperature, the current tissue equivalent thermal state and the candidate pulse sequence. S5: Combine the current target area tissue temperature, the target area temperature rise rate, the target area temperature trajectory corresponding to each candidate pulse sequence, and the target area cumulative absorbed dose to generate a candidate pulse parameter group; S6: Perform parameter consistency projection on the candidate pulse parameter group to obtain an executable pulse parameter group that satisfies the pulse timing constraints; S7: Update the actual output optical power according to the driving light source of the executable pulse parameter group, and return to step S1 until the termination trigger condition of transcranial photobiological modulation is met.

2. The method of claim 1, wherein the method further comprises: S2 specifically includes: S201: Perform anomaly removal processing on the running data to obtain running data after removing anomalies; S202: Perform delay compensation and low-pass filtering on the running data after removing outliers to obtain the filtered and compensated running data; S203: Perform a multi-channel sensor acquisition data rationality comparison and verification on the filtered and compensated operating data, and use the operating data after rationality comparison and verification as the standard operating data.

3. The method of claim 1, wherein the method further comprises: S3 specifically includes: S301: Update the tissue equivalent thermal state by using the preset tissue thermal state observation model, combined with the ambient temperature in the standard operating data and the preset individual tissue parameters; S302: Using the preset tissue thermal state observation model, calculate the updated tissue equivalent thermal state and output the current target area tissue temperature and the target area temperature rise rate; S303: Based on the actual output optical power in the standard operating data, the cumulative absorbed dose of the target area is obtained by iterative calculation.

4. The method of claim 1, wherein the method further comprises: The photothermal prediction model includes an offline high-fidelity model and an online order-reduction model; The offline high-fidelity model is established based on the light diffusion equation and the biothermal equation. It is used to generate photothermal response data according to preset or calibrated tissue parameters and to identify the response parameters of the online reduced-order model. The online reduction model is established based on the photothermal response data and identification parameters output by the offline high-fidelity model, and is used to recursively deduce the temperature trajectory of the target area in real time.

5. The transcranial photobiological modulation pulse control method for tissue temperature prediction according to claim 1, characterized in that, The candidate pulse parameter set includes peak power, pulse width, and pulse interval. The peak power, pulse width, and pulse interval are independent control quantities, and the duty cycle is a derived quantity calculated based on the pulse width and pulse interval.

6. The transcranial photobiological modulation pulse control method for tissue temperature prediction according to claim 5, characterized in that, S5 specifically includes: S501: Determine the remaining dose in the target area based on the difference between the cumulative absorbed dose in the target area and the target dose; S502: Based on the current target area tissue temperature, the target area temperature rise rate, and the target area temperature trajectory, determine whether the load reduction triggering condition is met; if so, reduce at least one of the peak power, shorten the pulse width, and extend the pulse interval to reduce the equivalent heat input and obtain the constrained pulse parameter set; otherwise, based on the remaining dose in the target area and the temperature safety margin, maintain or gradually restore at least one of the peak power, the pulse width, and the pulse interval within the upper and lower limit constraints and the parameter change rate constraints to obtain the constrained pulse parameter set; S503: Output the constrained pulse parameter set as the candidate pulse parameter set.

7. The transcranial photobiological modulation pulse control method for tissue temperature prediction according to claim 1, characterized in that, The pulse timing constraint is that the duty cycle is determined by both the pulse width and the pulse interval.

8. The transcranial photobiological modulation pulse control method for tissue temperature prediction according to claim 6, characterized in that, S6 specifically includes: S601: Calculate the candidate duty cycle based on the pulse width and pulse interval in the candidate pulse parameter group, combined with the pulse timing constraints; S602: Under the conditions of satisfying the pulse timing constraints, parameter upper and lower limit constraints and parameter change rate constraints, perform parameter consistency projection on the peak power, pulse width, pulse interval and candidate duty cycle in the candidate pulse parameter group to solve for the executable peak power, executable pulse width, executable pulse interval and executable duty cycle; S603: Output the executable peak power, the executable pulse width, the executable pulse interval, and the executable duty cycle as the executable pulse parameter set.

9. The transcranial photobiological modulation pulse control method for tissue temperature prediction according to claim 1, characterized in that, The termination triggering conditions include the cumulative absorbed dose in the target area reaching the target dose, or the triggering of a safety shutdown condition; The triggering safety shutdown conditions include at least one of the following: temperature exceeding the limit, sensor failure, communication anomaly, and insufficient model confidence.

10. A transcranial photobiological modulation pulse control system for tissue temperature prediction, characterized in that, include: Processor and memory; The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the transcranial photobiomodulation pulse control method for tissue temperature prediction as described in any one of claims 1 to 9.

Images