A high-frequency high-field intensity magnetic stimulator

Abstract

The application discloses a high-frequency high-field magnetic stimulator, and relates to the technical field of magnetic therapy.The application comprises a power supply module, a control module, a driving module, a switching circuit and a magnetic stimulation executor array, the multi-modal sensor unit of the contact surface of the magnetic stimulation executor array is used to collect bioelectric signals, temperature data and motion parameters of the injured part of the human body in real time, the control module extracts features and generates output labels by using a pre-trained neural network model, and then queries a control database to obtain matched magnetic stimulation control parameters to generate control instructions, the driving module cooperates with the switching circuit to realize magnetic stimulation, and the power supply module adaptively adjusts output voltage according to the input current of the switching circuit; the application effectively solves the problems of parameter adjustment depending on manual operation, low treatment efficiency and lack of preventive repair in the prior art, and significantly improves the intelligent level and effectiveness of magnetic stimulation treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of magnetic therapy, more particularly to a high-frequency high-field strength magnetic stimulator. BACKGROUND

[0002] At present, high-frequency magnetic stimulation technology has been applied to the field of sports injury repair, which can produce biological effects through high-frequency magnetic fields to nurse the injured muscles, joints and other tissues after exercise.

[0003] However, the prior art still has significant defects: first, it lacks multi-modal sensing capability for sports injuries, and cannot monitor key parameters such as deformation, skin temperature and microcirculation blood flow of the injured tissue in real time, resulting in inaccurate injury positioning and evaluation; second, parameter adjustment relies on manual presetting, making it difficult to dynamically optimize the stimulation scheme according to the injury type (such as muscle strain and ligament injury) and the repair process, resulting in low treatment efficiency; third, it lacks a linkage mechanism with sports data, making it difficult to achieve preventive repair.

[0004] Therefore, how to propose a high-frequency high-field strength magnetic stimulator to realize intelligent treatment of the magnetic stimulator and improve efficiency is a problem that needs to be solved by those skilled in the art. SUMMARY

[0005] Therefore, the present application provides a high-frequency high-field strength magnetic stimulator to improve the intelligent level and effectiveness of magnetic stimulation treatment.

[0006] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0007] A high-frequency high-field strength magnetic stimulator, comprising: a power supply module, a control module, a driving module, a switching circuit and a magnetic stimulator executor array;

[0008] The contact surface of the magnetic stimulator executor array is integrated with a multi-modal sensor unit for real-time acquisition of sensor data, including bioelectric signals, temperature data and motion parameters of the injured part of the human body;

[0009] The control module generates control instructions according to the sensor data;

[0010] The driving module converts the control instructions into a high-voltage driving signal;

[0011] The switching circuit responds to the high-voltage driving voltage to control the magnetic stimulator executor array to perform magnetic stimulation;

[0012] The power supply module adaptively adjusts the power output voltage according to the input current of the switching circuit to meet the power demand of the switching circuit; the input current of the switching circuit is determined according to the high-voltage driving voltage.

[0013] Preferably, the control module generates control instructions according to the sensor data, including:

[0014] extracting features of the sensor data;

[0015] inputting the features into a pre-trained neural network model to obtain corresponding output labels;

[0016] querying a pre-stored control database according to the output labels to obtain magnetic stimulation control parameters matched with the labels;

[0017] generating control instructions based on the control parameters.

[0018] Preferably, the neural network model is one or a combination of a convolutional neural network, a deep neural network, or a recurrent neural network, used for classifying the extracted biological electrical signal features, temperature data features, and motion parameter features, and outputting corresponding injury type labels or repair stage labels.

[0019] The neural network model is supervised learning trained based on the mapping relationship between historical sensor data and magnetic stimulation parameters, and the training data includes biological electrical signal waveforms, temperature change curves, and motion parameter thresholds of different injury types.

[0020] Preferably, the driving module includes an FPGA and a GaN driver.

[0021] The FPGA generates a PWM modulation signal according to the control instructions, and the GaN driver converts the PWM modulation signal into a high-voltage driving voltage.

[0022] Preferably, the power supply module includes a rectification and filtering unit and an LLC resonant circuit. The rectification and filtering unit is used to convert alternating current power into direct current bus voltage. The LLC resonant circuit converts the direct current bus voltage into high-frequency alternating current, and outputs adjustable direct current high voltage after transformer step-up and synchronous rectification.

[0023] Preferably, the LLC resonant circuit includes a half-bridge MOSFET, a resonant network, a high-frequency transformer, and a synchronous rectifier.

[0024] The working process of the LLC resonant circuit includes:

[0025] The half-bridge switching unit chops the direct current input into a high-frequency square wave voltage;

[0026] The resonant network filters the high-frequency square wave voltage into a base frequency quasi-sine current;

[0027] The primary side of the high-frequency transformer flows through the base frequency quasi-sine current, and the secondary side induces a quasi-sine voltage;

[0028] The synchronous rectifier rectifies the quasi-sine voltage into a direct current output.

[0029] Preferably, the LLC resonant circuit further comprises a feedback control unit.

[0030] The feedback control unit comprises a current sampling unit, an error amplifier and a PWM controller, wherein:

[0031] The current sampling unit samples the input current of the switching circuit in real time and converts it into a sampling voltage;

[0032] The error amplifier compares the sampling voltage with a reference voltage to generate an error signal;

[0033] The PWM controller adjusts the duty cycle of the half-bridge MOSFET according to the error signal, so that the output voltage of the adjustable direct current high voltage increases with the increase of the input current of the switching circuit.

[0034] Compared with the prior art, the high-frequency high-field magnetic stimulator provided by the present application comprises a power supply module, a control module, a driving module, a switching circuit and a magnetic stimulation executor array. The biomagnetic signal, temperature data and motion parameters of the human body injury site are collected in real time by the multi-modal sensor unit of the contact surface of the magnetic stimulation executor array. The control module extracts the features and generates output labels by using a pre-trained neural network model, and then queries the control database to obtain matched magnetic stimulation control parameters to generate control instructions. The driving module cooperates with the switching circuit to realize magnetic stimulation, and the power supply module adjusts the output voltage adaptively according to the input current of the switching circuit. The present application realizes the precision of injury positioning evaluation and the dynamic optimization of stimulation parameters by combining multi-modal sensing with neural networks. With the adaptive adjustment of the power supply module and the motion data linkage mechanism, the problems of manual parameter adjustment, low treatment efficiency and lack of preventive repair in the prior art are effectively solved, and the intelligent level and effectiveness of magnetic stimulation treatment are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the provided drawings.

[0036] Fig. 1 The overall structure schematic diagram of the high-frequency high-field magnetic stimulator provided by the present application is shown in the figure.

[0037] Fig. 2A schematic diagram of the LLC resonant circuit provided by the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] This invention discloses a high-frequency, high-field-strength magnetic stimulator, such as... Figs. 1-2 As shown, it includes: a power supply module, a control module, a drive module, a switching circuit, and a magnetic stimulation actuator array;

[0040] The contact surface of the magnetic stimulation actuator array integrates a multimodal sensor unit for real-time acquisition of sensor data, including bioelectrical signals, temperature data, and motion parameters of the injured area of ​​the human body.

[0041] The control module generates control commands based on sensor data, including:

[0042] Extract features from sensor data;

[0043] The features are input into a pre-trained neural network model to obtain the corresponding output labels;

[0044] The pre-stored control database is queried based on the output label to obtain the magnetic stimulation control parameters that match the label;

[0045] Control commands are generated based on control parameters.

[0046] Preferably, the neural network model is one or a combination of convolutional neural networks, deep neural networks, or recurrent neural networks, used to classify the extracted bioelectrical signal features, temperature data features, and motion parameter features, and output corresponding damage type labels or repair stage labels.

[0047] The neural network model is trained through supervised learning based on the mapping relationship between historical sensor data and magnetic stimulation parameters. The training data includes bioelectrical signal waveforms, temperature change curves, and motion parameter thresholds for different injury types.

[0048] The driver module converts control commands into high-voltage drive signals. The driver module includes an FPGA and a GaN driver.

[0049] The FPGA generates a PWM modulation signal according to the control instructions, and the GaN driver converts the PWM modulation signal into a high-voltage drive voltage.

[0050] The switch circuit controls the magnetic stimulation actuator array to perform magnetic stimulation in response to the high-voltage driving voltage.

[0051] The power supply module adaptively adjusts the power supply output voltage according to the input current of the switch circuit to meet the power requirement of the switch circuit.

[0052] The power supply module includes a rectifier filter unit and an LLC resonant circuit.

[0053] The LLC resonant circuit includes a half-bridge MOSFET, a resonant network, a high-frequency transformer, and a synchronous rectifier.

[0054] The working process of the LLC resonant circuit includes:

[0055] The half-bridge switch unit chops the DC input into a high-frequency square wave voltage.

[0056] The resonant network filters the high-frequency square wave voltage into a fundamental frequency quasi-sine current.

[0057] The primary side of the high-frequency transformer flows through the fundamental frequency quasi-sine current, and the secondary side induces a quasi-sine voltage.

[0058] The synchronous rectifier rectifies the quasi-sine voltage into a DC output.

[0059] Further, the LLC resonant circuit further includes a feedback control unit.

[0060] The feedback control unit includes a current sampling unit, an error amplifier, and a PWM controller.

[0061] The current sampling unit samples the input current of the switch circuit in real time and converts it into a sampling voltage.

[0062] The error amplifier compares the sampling voltage with a reference voltage to generate an error signal.

[0063] The PWM controller adjusts the duty cycle of the half-bridge MOSFET according to the error signal, so that the output voltage of the adjustable DC high voltage increases with the increase of the input current of the switch circuit.

[0064] Specifically, the contact surface of the magnetic stimulation actuator array is integrated with multi-modal sensor units such as pressure sensors, temperature sensors, and electromyography sensors, which can collect electromyography signals, skin temperature data, joint range of motion, and other motion parameters of the human body injury site in real time; the control module uses an ARM processor to sample sensor data through an ADC circuit, extract frequency characteristics of bioelectric signals, temperature change rates, and motion angle thresholds, and other characteristic parameters; the drive module is composed of an FPGA chip and a GaN driver, the FPGA generates a frequency-adjustable PWM modulation signal according to the control instruction, and the GaN driver amplifies it into a high-voltage drive signal of more than 1000V; the switching circuit uses an IGBT power module to control the coils in the magnetic stimulation actuator array to generate a high-frequency magnetic field in response to the high-voltage drive signal; the power supply module converts 220V mains power into a 310V DC bus voltage through a rectifier and filter unit, and then chops it into a 200kHz high-frequency square wave through a half-bridge MOSFET in the LLC resonant circuit, filters it into a base frequency quasi-sine current through a resonant network, and outputs a 0-3000V adjustable DC high voltage through a synchronous rectifier after being boosted by a high-frequency transformer. The power supply module monitors the input current of the switching circuit in real time through the current sampling unit, and when the current increases, the PWM controller automatically adjusts the duty cycle of the half-bridge MOSFET to make the output voltage rise synchronously to meet the power demand.

[0065] The specific process of the control module generating control instructions based on sensor data is: denoising the electromyography signal through wavelet transform, extracting time domain features such as root mean square (RMS) and mean absolute value (MAV), extracting frequency domain features such as main frequency and frequency band energy through fast Fourier transform (FFT), and simultaneously obtaining gradient changes of temperature data and real-time displacement of joint angle; in this embodiment, the above features are input into a pre-trained CNN-RNN hybrid neural network model (CNN uses ResNet structure to extract spatial features, and RNN uses LSTM unit to process time sequence features), the model outputs injury type labels (such as muscle strain, ligament injury) or repair stage labels (acute phase, subacute phase, recovery phase); according to the label, the control database (storing the mapping relationship of 2000+ clinical cases of magnetic stimulation parameters) is queried to obtain corresponding stimulation frequency (50-500Hz), pulse width (10-200us), magnetic field strength (1-3T), and other parameters; finally, the control instruction containing PWM frequency, duty cycle, and pulse sequence is generated and sent to the drive module.

[0066] In the drive module, the FPGA generates a PWM signal of 20-500 kHz according to the control instruction, the duty cycle adjustment range is 0-90%, the output voltage amplitude of the driving signal is 1200V after being amplified by the GaN driver, and the IGBT switching circuit is driven to control the on-off timing sequence of the magnetic stimulation coil; when the LLC resonant circuit of the power supply module is working, the half-bridge MOSFET is turned on alternately at a frequency of 200 kHz, the DC bus voltage is chopped into a square wave, the resonant inductance and the resonant capacitance in the resonant network form a series resonance, the square wave is filtered into a quasi-sine current with a peak current of 20A, and the quasi-sine voltage is induced on the secondary side after being boosted by a high-frequency transformer with a transformation ratio of 1:30, and then rectified into a DC high voltage by a synchronous rectification MOSFET; the feedback control unit monitors the input current of the switching circuit in real time through a 0.1Ω sampling resistor, when the current rises from 5A to 10A, the error amplifier compares the sampling voltage (0.5-1V) with the reference voltage (1V) to generate an error signal, and the PWM controller adjusts the duty cycle of the half-bridge MOSFET from 40% to 60% according to the signal, so that the output voltage rises from 1500V to 3000V, which is used to drive the coils of the magnetic stimulation actuator array to generate a high-frequency magnetic field, and the power supply voltage is adaptively adjusted according to the load current.

[0067] When the magnetic stimulation actuator array contacts the damaged part, the multi-modal sensor unit synchronously collects electromyographic signal (sampling rate 1000Hz), temperature (accuracy ±0.5℃) and joint angle (resolution 0.1°) data; if the control module detects that the amplitude of the electromyographic signal suddenly decreases by 30% and the temperature rises by 1.5℃, it is determined that it is acute muscle strain, at this time the neural network outputs the "acute phase" label, and the control database matches to the corresponding stimulation parameters: frequency 100Hz, pulse width 50μs, magnetic field strength 1.2T, and the stimulation lasts for 20 minutes; if the sensor data shows that the amplitude of the electromyographic signal returns to 80% of the normal level and the temperature falls during the treatment process, the neural network updates the output "subacute phase" label, and the control module automatically increases the stimulation frequency to 200Hz and the magnetic field strength to 1.8T to promote tissue repair; at the same time, the power supply module adaptively adjusts the output voltage from 2000V to 2800V according to the change of the input current of the switching circuit from 8A to 12A, to ensure the stable output of the magnetic stimulation intensity.

[0068] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0069] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A high-frequency, high-field-strength magnetic stimulator, characterized in that, include: Power supply module, control module, drive module, switching circuit and magnetic stimulation actuator array; The contact surface of the magnetic stimulation actuator array integrates a multimodal sensor unit for real-time acquisition of sensor data, including bioelectrical signals, temperature data, and motion parameters of the injured area of ​​the human body. The control module generates control commands based on the sensor data; The drive module converts the control command into a high-voltage drive voltage. The switching circuit responds to the high-voltage drive voltage and controls the magnetic stimulation actuator array to perform magnetic stimulation. The power supply module adaptively adjusts the power supply output voltage according to the input current of the switching circuit to meet the power requirements of the switching circuit; the input current of the switching circuit is determined according to the high voltage drive voltage. The power supply module includes a rectifier and filter unit and an LLC resonant circuit. The rectifier and filter unit is used to convert AC mains power into DC bus voltage. The LLC resonant circuit converts the DC bus voltage into high-frequency AC, which is then stepped up by a transformer and synchronously rectified to output an adjustable DC high voltage. The LLC resonant circuit includes a half-bridge MOSFET, a resonant network, a high-frequency transformer, and a synchronous rectifier. The operation of the LLC resonant circuit includes: The half-bridge MOSFET chops the DC input into a high-frequency square wave voltage; The resonant network filters the high-frequency square wave voltage into a fundamental frequency quasi-sinusoidal current. The fundamental frequency quasi-sinusoidal current flows through the primary side of the high-frequency transformer, and a quasi-sinusoidal voltage is induced on the secondary side. The synchronous rectifier rectifies the quasi-sinusoidal voltage into a DC output. The LLC resonant circuit also includes a feedback control unit; The feedback control unit includes a current sampling unit, an error amplifier, and a PWM controller, wherein: The current sampling unit samples the input current of the switching circuit in real time and converts it into a sampling voltage; The error amplifier compares the sampled voltage with the reference voltage to generate an error signal; The PWM controller adjusts the duty cycle of the half-bridge MOSFET according to the error signal, so that the output voltage of the adjustable DC high voltage increases as the input current of the switching circuit increases.

2. The high-frequency, high-field-strength magnetic stimulator according to claim 1, characterized in that, The control module generates control commands based on the sensor data, including: Extract features from the sensor data; The features are input into a pre-trained neural network model to obtain the corresponding output labels; The pre-stored control database is queried based on the output label to obtain the magnetic stimulation control parameters that match the label; Control commands are generated based on the control parameters.

3. A high-frequency, high-field-strength magnetic stimulator according to claim 2, characterized in that, The neural network model is one or a combination of convolutional neural networks, deep neural networks, or recurrent neural networks, used to classify the extracted bioelectrical signal features, temperature data features, and motion parameter features, and output corresponding damage type labels or repair stage labels. The neural network model is trained through supervised learning based on the mapping relationship between historical sensor data and magnetic stimulation parameters. The training data includes bioelectrical signal waveforms, temperature change curves, and motion parameter thresholds for different injury types.

4. A high-frequency, high-field-strength magnetic stimulator according to claim 1, characterized in that, The driving module includes an FPGA and a GaN driver; The FPGA generates a PWM modulation signal according to the control command, and the GaN driver converts the PWM modulation signal into a high-voltage drive voltage.

Images