An infrared physiotherapy wavelength intelligent regulation method and system based on physiological feedback

Abstract

The application discloses an infrared physiotherapy wavelength intelligent regulation method and system based on physiological feedback, aiming to solve the problem that the prior art cannot dynamically regulate infrared spectrum based on deep tissue efficacy feedback; the method comprises the following steps: acquiring a deep tissue blood oxygen saturation signal in real time in infrared physiotherapy, and extracting an efficacy characteristic index; inputting the index into a pre-trained efficacy-parameter mapping model to generate infrared light source regulation parameters for guiding the efficacy index to an ideal interval; and adjusting the energy ratio of the multi-waveband of the infrared physiotherapy light source in real time based on the parameters to form a closed-loop regulation. The system comprises a physiological signal acquisition module, a control and processing module and an infrared light source driving and output module. The application first realizes multi-wavelength collaborative dynamic regulation driven by an AI model with deep tissue oxygenation as feedback, actively avoids the risk of heat damage under model predictive control, and significantly improves the targeting and safety of infrared physiotherapy.

Description

Technical Field

[0001] This invention relates to infrared physiotherapy and intelligent control technology, specifically to a method and system for dynamically adjusting the wavelength spectrum and energy ratio of an infrared light source based on physiological feedback signals, particularly the blood oxygen saturation of deep tissues. Background Technology

[0002] Infrared therapy, as a non-invasive physical therapy method, is widely used in pain relief, inflammation reduction, and tissue repair. Its basic principle is to use infrared light of specific wavelengths to produce thermal and photobiological regulatory effects on biological tissues, such as activating the mitochondrial respiratory chain, promoting nitric oxide release, and improving local microcirculation.

[0003] However, existing infrared therapy devices generally employ a one-size-fits-all treatment model, failing to intelligently adjust to individual, real-time, and deep-seated therapeutic responses. Conventional red / infrared light therapy devices can only control power thresholds based on superficial tissue temperature, aiming at a singular, passive prevention of burns rather than actively pursuing deep tissue repair. Some solutions select one wavelength from multiple preset wavelengths based on tissue optical properties, but this selection is based on the tissue's static physical properties and cannot dynamically respond to the continuous changes in the tissue's physiological state during treatment.

[0004] The common shortcomings of these existing technologies are that they either rely on simple on / off or power control based on a single, superficial safety signal (such as skin temperature) or on discrete wavelength selection based on static tissue physical characteristics. None of them can perceive the true therapeutic effect of infrared therapy on deep tissues in real time and dynamically, nor can they intelligently and continuously adjust the core treatment parameters (the spectral combination and energy ratio of infrared wavelengths) based on therapeutic feedback. This makes the treatment process resemble open-loop blind firing, with optimal therapeutic effects heavily dependent on the operator's experience, resulting in extremely low levels of individualization and precision. Furthermore, they lack mechanisms for predictive and smooth prevention of potential tissue damage, and their safety assurance mechanisms are rudimentary. Summary of the Invention

[0005] This invention overcomes the shortcomings of existing technologies and provides a method and system for intelligent control of infrared therapy wavelength based on physiological feedback, so as to solve the problem that existing technologies cannot intelligently and dynamically control the infrared spectrum based on the therapeutic feedback of deep tissues.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for intelligent control of infrared physiotherapy wavelength based on physiological feedback, comprising the following steps: During infrared physiotherapy, the blood oxygen saturation signal of the deep tissues of the treated target is acquired in real time; The blood oxygen saturation signal of the deep tissue is processed to extract therapeutic characteristic indicators that reflect the current therapeutic level. The therapeutic effect characteristic index is used as input and provided to a pre-trained therapeutic effect-parameter mapping model. The model is configured to output infrared light source control parameters based on the input to guide the therapeutic effect characteristic index to a preset ideal therapeutic effect range. Based on the infrared light source control parameters, the output energy of at least one wavelength of the infrared therapy light source is adjusted in real time to form a closed-loop control.

[0007] Furthermore, the infrared light source control parameters are specifically the energy ratio of at least two different peak wavelength bands; the real-time adjustment of the output energy of at least one wavelength of the infrared therapy light source is specifically to coordinately adjust the output power of the at least two different peak wavelength bands to change the composition of the overall emission spectrum.

[0008] Furthermore, the at least two different peak wavelength bands include a main therapeutic band with a first penetration depth to deep tissues and an auxiliary therapeutic band with a second penetration depth to superficial tissues, wherein the first penetration depth is greater than the second penetration depth; the efficacy-parameter mapping model is configured to output a control parameter that prioritizes increasing the energy ratio of the main therapeutic band when the efficacy characteristic index indicates that the efficacy has not reached the ideal efficacy range.

[0009] Furthermore, the therapeutic efficacy indicators include: the rate of change of the deep tissue oxygen saturation signal, the time to peak, the time to recover to the target value, or the increment of the signal relative to the individualized baseline.

[0010] Furthermore, it also includes: acquiring the skin surface temperature signal of the target being treated in real time, and extracting the temperature change rate feature based on this; the efficacy-parameter mapping model is a multi-input model, the input of which further includes the temperature change rate feature, so that when the model predicts the potential risk of thermal damage based on the temperature change rate feature, it outputs infrared light source control parameters that prioritize safety.

[0011] Furthermore, the infrared light source control parameters that prioritize safety in the output specifically include: reducing the energy ratio of the near-infrared band in the current output spectrum, increasing the energy ratio of the far-infrared band, or reducing the total output optical power.

[0012] Furthermore, the efficacy-parameter mapping model is a neural network model pre-trained based on population clinical data and is configured to be adaptively fine-tuned online based on the individual physiological response data of the treated target during individual use.

[0013] Furthermore, the real-time acquisition of deep tissue oxygen saturation signals of the treated target is achieved through near-infrared spectroscopy technology.

[0014] Furthermore, the at least two different peak wavelength bands are selected from at least two of 800-840nm, 850-890nm and 900-960nm.

[0015] Furthermore, it also includes: establishing an individual physiological response profile of the treated target, determining an individualized setting value for the ideal therapeutic range based on the individual physiological response profile, and using it as the control target of the therapeutic-parameter mapping model.

[0016] This invention also discloses an intelligent control system for infrared therapy wavelengths based on physiological feedback, comprising: The physiological signal acquisition module is used to acquire the deep tissue blood oxygen saturation signal of the treated target in real time, as well as the optional skin surface temperature signal. A control and processing module, which has a pre-trained efficacy-parameter mapping model deployed therein, is configured to perform the steps of the method of any one of claims 1 to 10 to generate infrared light source modulation parameters based on the acquired signals; An infrared light source driving and output module is used to drive an infrared light source array to adjust the output energy of at least one wavelength in real time according to the infrared light source control parameters.

[0017] Furthermore, the infrared light source array is composed of multiple light-emitting units, and the peak wavelengths of the multiple light-emitting units fall within at least two of the following bands: 800-840nm, 850-890nm, and 900-960nm; the infrared light source driving and output module independently controls the driving current of the light-emitting unit corresponding to each band.

[0018] Furthermore, the physiological signal acquisition module includes a near-infrared spectral probe and a temperature sensor. The near-infrared spectral probe and the infrared light source array are integrated in the same treatment head, and the temperature sensor is located at the interface between the treatment head and the target being treated.

[0019] The beneficial effects of this invention compared to the prior art are: 1. This invention is the first to select deep tissue blood oxygen saturation, a gold standard indicator that directly reflects tissue oxygenation and metabolism, as the control target, making treatment more targeted and effective rather than relying on experience.

[0020] 2. This invention achieves intelligent formulation of infrared spectrum by dynamically adjusting the energy ratio of multiple bands through AI model, rather than simply switching power or single wavelength, thus taking into account both deep treatment efficiency and superficial thermal comfort.

[0021] 3. This invention utilizes the time-series prediction of temperature change rate to anticipate risks in advance and smoothly adjust the spectral composition, transforming passive protection into active avoidance, which greatly improves user experience and treatment safety.

[0022] 4. The model involved in this invention can continuously learn and optimize during individual use and establish personalized physiological profiles, so that the device understands the user better and better with use, and achieves truly personalized and precise physiotherapy. Attached Figure Description

[0023] The present invention will now be further described with reference to the accompanying drawings.

[0024] Figure 1 The basic flowchart of the intelligent control method for infrared physiotherapy wavelength based on physiological feedback provided in the embodiments of the present invention is shown.

[0025] Figure 2 This is a schematic diagram of the system structure provided in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram illustrating the working principle of the efficacy-parameter mapping model in an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the present invention clearer, a further detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it.

[0028] In the following description, “penetration depth” is defined as the tissue depth reached when the light intensity decays to approximately 37% of the incident intensity at the skin surface, and it is generally positively correlated with wavelength. Example

[0029] This embodiment provides a method for intelligent control of infrared therapy wavelength based on physiological feedback, the process of which is as follows: Figure 1 , Figure 2 As shown.

[0030] Step S101, Signal Acquisition: At the start of infrared therapy and during treatment, near-infrared spectroscopy (NIRS) technology is used to monitor the blood oxygen saturation (StO2) of deep tissues in the treated target (such as muscles, joints, etc.) in real time and non-invasively. Specifically, an LED light source integrating two wavelengths of near-infrared light, 760nm and 850nm, and a photodetector probe positioned approximately 3cm away from the light source are used. After the light emitted by the light source is scattered and absorbed by the tissue, part of it is received by the detector. The control and processing module uses a modified Lambert-Beer law to calculate the concentrations of oxyhemoglobin and deoxyhemoglobin in the tissue based on the difference in attenuation between the two wavelengths, thereby obtaining the tissue blood oxygen saturation (StO2). This signal directly reflects the tissue oxygenation status and is a sensitive and direct indicator for measuring the deep therapeutic effects such as mitochondrial respiratory chain activity and microcirculation improvement under photobiological regulation.

[0031] Step S102, Feature Extraction: After performing a moving average filter on the acquired StO2 raw signal to remove noise such as motion artifacts, one or more quantitative indicators reflecting the dynamic changes in therapeutic efficacy are extracted. In this embodiment, the increment ΔStO2 of StO2 relative to the pre-treatment baseline level is preferably extracted as the therapeutic efficacy characteristic indicator.

[0032] Step S103, Intelligent Decision-Making: The extracted efficacy feature index ΔStO2 is input into a pre-trained efficacy-parameter mapping model. This model is built based on a deep neural network (DNN). The core task of the model is to solve an optimization problem: to find a set of infrared light source control parameters so that ΔStO2 can reach and maintain within the preset ideal efficacy range (e.g., ΔStO2 > 15%) as quickly and smoothly as possible.

[0033] Step S104, Control Execution: Based on the control parameters output by the model, adjust the output of the infrared therapy light source in real time. The control parameters here are not simply the power level, but rather the energy ratio of at least two wavebands. Example

[0034] This embodiment further optimizes the control dimensions and security mechanisms based on Embodiment 1.

[0035] The infrared light source array used contains three LEDs with different peak wavelengths: the first is 810nm (near-infrared, the main therapeutic band, which has a large penetration depth into deep tissues and can efficiently activate cytochrome c oxidase), the second is 880nm (near-infrared, the auxiliary therapeutic band, which has both a certain penetration depth and thermal effect), and the third is 940nm (near / mid-infrared territorial band, the auxiliary therapeutic band, which has shallow penetration into superficial tissues, obvious thermal effect, and is beneficial to superficial microcirculation).

[0036] At this point, the infrared light source control parameters output by the model are specifically the power ratio of these three bands, for example, expressed as P. 810 :P 880 :P 940 The system uses an independent PID (proportional-integral-derivative) control loop to adjust the PWM (pulse width modulation) duty cycle of each LED driver, thereby achieving continuous, coordinated, and smooth adjustment of the output power of each band and thus changing the composition of the overall output spectrum.

[0037] Simultaneously, the system integrates a high-precision digital temperature sensor at the point where the treatment head contacts the skin to acquire the skin surface temperature T in real time. The signal processing module calculates the rate of temperature change dT / dt. This dT / dt value serves as a safety constraint feature and, along with the efficacy characteristic index ΔStO2, is input into a multi-input efficacy-parameter mapping model built on an LSTM (Long Short-Term Memory) network.

[0038] like Figure 3 As shown, the model incorporates predictive safety logic. The LSTM network can process temporal information; when it predicts, based on the input temperature change rate sequence dT / dt(n), dT / dt(n-1)..., that the skin surface temperature T will exceed a preset safety threshold (e.g., 42℃) within the next 30 seconds, the model will prioritize implementing a safety strategy even if the current ΔStO2 does not meet the threshold. For example, the output control parameters will be adjusted by the high-penetration, high-efficiency P... 810 :P 880 :P 940 The ratio is 5:3:2, which smoothly transitions to a 2:3:5 ratio that prioritizes shallow heat dissipation and reduces deep energy deposition to facilitate heat dissipation, or the overall PWM duty cycle is reduced by 20%. This achieves an intrinsic safety upgrade from passive overheating power-off to active predictive smooth control. Example

[0039] To ensure that treatment is truly individualized, this system incorporates a personalized adaptive mechanism. When a user uses the system for the first time, it derives an initial individual model from a pre-trained population model based on group data such as age, skin type, and basic physiological state.

[0040] During and after each treatment, the system records the user's complete physiological response curves (StO2 change trajectory, temperature change trajectory) and model output instructions. This data constitutes the user's individual physiological response profile. With increased usage, the system utilizes online fine-tuning technology to adaptively optimize the user's therapy-parameter mapping model locally.

[0041] The specific fine-tuning process is as follows: After each treatment, the system compares the StO2 change curve throughout the entire treatment process with the ideal therapeutic range and calculates the mean squared error loss function. Subsequently, using the gradient descent algorithm, only the parameters of the last few fully connected network layers of the model are updated with a small learning rate (e.g., 0.0001). To prevent the model from overlearning and diverging on individual singular data, an L2 regularization term is introduced into the loss function to ensure that the fine-tuned model parameters do not deviate too far from the baseline of the population model, thus always remaining within a reasonable and safe decision range.

[0042] In this way, the model can gradually learn a user's physiological response characteristics locally without uploading the user's private data. For example, if a user's StO2 increases naturally more slowly, the model will learn this and gradually adjust its control targets or regulatory strategies for the ideal therapeutic range. The initial regulatory strategy for the next treatment will be more tailored to the individual, achieving truly personalized and precise physical therapy.

[0043] Those skilled in the art will understand that the neural network models (DNN, LSTM), specific wavelength values ​​(810nm, 880nm, 940nm), and feature indicators (ΔStO2) in the above specific embodiments are illustrative and not intended to limit the scope of protection of this invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for intelligent control of infrared therapy wavelength based on physiological feedback, characterized in that, Includes the following steps: During infrared physiotherapy, the blood oxygen saturation signal of the deep tissues of the treated target is acquired in real time; The blood oxygen saturation signal of the deep tissue is processed to extract therapeutic characteristic indicators that reflect the current therapeutic level. The therapeutic effect characteristic index is used as input and provided to a pre-trained therapeutic effect-parameter mapping model. The model is configured to output infrared light source control parameters based on the input to guide the therapeutic effect characteristic index to a preset ideal therapeutic effect range. Based on the infrared light source control parameters, the output energy of at least one wavelength of the infrared therapy light source is adjusted in real time to form a closed-loop control.

2. The method for intelligent control of infrared therapy wavelength based on physiological feedback according to claim 1, characterized in that, The infrared light source control parameters are specifically the energy ratio of at least two different peak wavelength bands; the real-time adjustment of the output energy of at least one wavelength of the infrared therapy light source is specifically to coordinately adjust the output power of the at least two different peak wavelength bands to change the composition of the overall emission spectrum.

3. The method for intelligent control of infrared physiotherapy wavelength based on physiological feedback according to claim 2, characterized in that, The at least two different peak wavelength bands include a main therapeutic band with a first penetration depth to deep tissues and an auxiliary therapeutic band with a second penetration depth to superficial tissues, wherein the first penetration depth is greater than the second penetration depth; the efficacy-parameter mapping model is configured to output a control parameter that prioritizes increasing the energy ratio of the main therapeutic band when the efficacy characteristic index indicates that the efficacy has not reached the ideal efficacy range.

4. The method for intelligent control of infrared therapy wavelength based on physiological feedback according to claim 1, characterized in that, The therapeutic efficacy indicators include: the rate of change of the deep tissue oxygen saturation signal, the time to peak, the time to recover to the target value, or the increment of the signal relative to the individualized baseline.

5. The method for intelligent control of infrared therapy wavelength based on physiological feedback according to claim 1, characterized in that, Also includes: The skin surface temperature signal of the target being treated is acquired in real time, and the temperature change rate feature is extracted based on this. The therapeutic effect-parameter mapping model is a multi-input model, the input of which further includes the temperature change rate feature, so that when the model predicts the potential risk of thermal damage based on the temperature change rate feature, it outputs infrared light source control parameters that prioritize safety.

6. The method for intelligent control of infrared therapy wavelength based on physiological feedback according to claim 5, characterized in that, The infrared light source control parameters that prioritize safety in the output include: reducing the energy ratio of the near-infrared band in the current output spectrum, increasing the energy ratio of the far-infrared band, or reducing the total output optical power.

7. The method for intelligent control of infrared physiotherapy wavelength based on physiological feedback according to claim 1, characterized in that, The efficacy-parameter mapping model is a neural network model pre-trained based on population clinical data and is configured to be adaptively fine-tuned online based on the individual physiological response data of the treated target during individual use.

8. The method for intelligent control of infrared physiotherapy wavelength based on physiological feedback according to claim 2, characterized in that, The at least two different peak wavelength bands are selected from at least two of 800-840nm, 850-890nm and 900-960nm.

9. A method for intelligent control of infrared physiotherapy wavelength based on physiological feedback according to any one of claims 1 to 8, characterized in that, It also includes: establishing an individual physiological response profile of the treated target, determining an individualized setting value for the ideal therapeutic range based on the individual physiological response profile, and using it as the control target of the therapeutic-parameter mapping model.

10. A smart infrared therapy wavelength control system based on physiological feedback, characterized in that, include: The physiological signal acquisition module is used to acquire the deep tissue blood oxygen saturation signal of the treated target in real time, as well as the optional skin surface temperature signal. A control and processing module, which has a pre-trained efficacy-parameter mapping model deployed therein, is configured to perform the steps of the method of any one of claims 1 to 10 to generate infrared light source modulation parameters based on the acquired signals; An infrared light source driving and output module is used to drive an infrared light source array to adjust the output energy of at least one wavelength in real time according to the infrared light source control parameters.

Images