Multi-channel signal acquisition closed-loop ultrasonic stimulation system and method based on fuzzy PID (Proportion Integration Differentiation) control

Abstract

The invention discloses a multichannel signal acquisition closed-loop ultrasonic stimulation system and method based on fuzzy PID control, and belongs to the technical field of ultrasonic stimulation nerve regulation and control. Comprising a signal acquisition and preprocessing module integrated on a backpack adaptive to an experiment object, a signal processing module in communication connection with the signal acquisition and preprocessing module, and a fuzzy PID control module connected with the signal processing module through radio frequency. The ultrasonic stimulation module is connected with the fuzzy PID control module through radio frequency; and the power supply module supplies power to the whole system. According to the invention, the signal quality can be improved while portability is realized, the adaptability and flexibility of the system are enhanced, the dependence on parameter adjustment is reduced, and the noise and instability of a single-channel signal are effectively eliminated.

Description

Technical Field

[0001] The present invention belongs to the technical field of ultrasonic stimulation nerve regulation, and in particular relates to a multi-channel signal acquisition closed-loop ultrasonic stimulation system and method based on fuzzy PID control. Background Art

[0002] At present, the EEG and EMG signals collected by signal acquisition devices on the market are easily affected by noise interference, such as power line noise, motion artifacts, etc. These noises will cause the quality of physiological signals to deteriorate and affect subsequent data analysis. EEG and EMG signal acquisition equipment usually require electrodes to be attached or specific helmets or straps to be worn. These devices may cause discomfort when used for a long time, especially in daily continuous monitoring scenarios, where the comfort level is poor. In addition, most neural regulation stimulation is open-loop control. Open-loop control systems cannot be adjusted according to the actual output conditions and can only rely on preset instructions. They cannot respond to changes in the system, such as mechanical wear, temperature changes, etc. This causes the performance of the entire system to deteriorate over time.

[0003] For example, the Chinese invention patent document CN114272523B discloses a transcranial brain stimulation multi-channel data acquisition system, which collects the EEG signals of the experimental subjects through an ultrasonic stimulation module and processes and optimizes the signals using multi-channel data fusion technology. However, the invention focuses more on signal processing and data fusion, and lacks consideration of hardware portability and low power consumption. At the same time, the invention mainly relies on preset signal transmission and processing, and does not emphasize a real-time closed-loop feedback mechanism.

[0004] Another example is a portable animal transcranial ultrasound stimulation and brain electromyography wireless acquisition system disclosed in Chinese invention patent document CN114272523B, which realizes wireless acquisition and processing of brain electromyography activities of experimental subjects through ultrasound stimulation and wireless brain electromyography signal acquisition. However, the invention does not involve multi-channel signal acquisition, and cannot realize signal fusion processing in complex experiments, and the data accuracy and robustness are weak. The control system in the invention is an open-loop design, and it is impossible to adjust the ultrasound stimulation parameters according to the feedback signal, and the adaptability and flexibility are poor.

[0005] Another example is a sleep rhythm control method and system disclosed in Chinese invention patent application document CN115153586A, which collects and processes EEG and EMG signals and performs ultrasonic stimulation to adjust sleep rhythm and improve sleep quality, especially the rhythm of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. However, the defect of this invention application is that its adjustment process does not include real-time closed-loop control and cannot be dynamically adjusted according to the immediate feedback of physiological signals, which may lead to poor adaptability of the system under different experimental subjects or experimental conditions.

[0006] Another example is a closed-loop transcranial brain stimulation method and system for sleep disclosed in Chinese invention patent application document CN114404825A, which uses a closed-loop control mechanism to collect and process brain-muscle signals of experimental subjects in real time, calculate EEG phase and EMG activity, and transmit ultrasonic stimulation signals to preset brain areas, thereby adjusting the sleep state. However, the system mainly relies on single-channel collection of EEG and EMG signals, and the signal fusion and processing capabilities are relatively limited. At the same time, the system fails to provide an adaptive control method, relies on preset ultrasonic stimulation parameters, has low flexibility, and is more focused on fixed laboratory equipment, without emphasizing the portability and miniaturization of the system.

[0007] Therefore, there is a need for a multi-channel signal acquisition closed-loop ultrasonic stimulation system and method based on fuzzy PID control that is portable and can improve signal quality, enhance the adaptability and flexibility of the system. Summary of the invention

[0008] The purpose of the present invention is to provide a multi-channel signal acquisition closed-loop ultrasonic stimulation system and method based on fuzzy PID control, which is portable and can improve signal quality, enhance the adaptability and flexibility of the system, reduce dependence on parameter adjustment, and effectively eliminate the noise and instability of single-channel signals.

[0009] To achieve the above purpose, the technical solution adopted by the present invention is:

[0010] A multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control, comprising a signal acquisition and preprocessing module integrated on a backpack adapted to an experimental subject, a signal processing module communicatively connected to the signal acquisition and preprocessing module, a fuzzy PID control module connected to the signal processing module via radio frequency, an ultrasonic stimulation module connected to the fuzzy PID control module via radio frequency, and a power module and an external antenna for powering the entire system.

[0011] A further improvement of the technical solution of the present invention is that: the signal acquisition and preprocessing module is configured to acquire the EEG signals and EMG signals of the experimental subject and preprocess the acquired analog signals, and convert the analog signals into corresponding digital signals; the signal acquisition and preprocessing module includes a multi-channel EEG electrode for acquiring EEG signals, a multi-channel EMG electrode for acquiring EMG signals, a front-end amplifier connected to the multi-channel EEG electrodes and the multi-channel EMG electrodes respectively, a microelectrode AC amplifier connected to the front-end amplifier for communication, and an analog-to-digital converter connected to the microelectrode AC amplifier for communication.

[0012] The further improvement of the technical solution of the present invention is that: the multi-channel EEG electrode includes two single-channel microwire electrodes implanted in the M1 motor cortex on the skull of the experimental subject and two reference electrodes plugged into the nasal bone of the experimental subject; the multi-channel EMG electrode includes two surface EMG electrodes placed in a preset area of ​​the neck muscle of the experimental subject and two reference electrodes attached to the skin surface of the experimental subject near the bone; the front-end amplifier is configured to amplify the weak analog signal and reduce the introduction of noise, and transmit the amplified signal to the microelectrode AC amplifier; the microelectrode AC amplifier is configured to amplify and filter the EEG signal and the EMG signal, and send the processed signal to the analog-to-digital converter; the analog-to-digital converter is configured to convert the analog signal into a corresponding digital signal and transmit it to the signal processing module.

[0013] A further improvement of the technical solution of the present invention is that: the signal processing module is configured to perform multi-channel signal fusion, feature extraction, feature evaluation and selection on the EEG digital signals and the EMG digital signals preprocessed by the signal acquisition and preprocessing module, and transmit the optimized digital signals to the fuzzy PID control module via radio frequency; the signal processing module includes a microcontroller and a digital signal processor DSP that is communicatively connected to the microcontroller.

[0014] A further improvement of the technical solution of the present invention lies in that: the fuzzy PID control module is configured to perform fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module in sequence, to form new ultrasonic stimulation parameters and transmit them to the ultrasonic stimulation module via radio frequency communication; the fuzzy PID control module includes LabVIEW control software that responds to the EEG digital signal and the EMG digital signal sent by radio frequency and a fuzzy PID control algorithm.

[0015] A further improvement of the technical solution of the present invention is that the fuzzy PID control algorithm includes the following steps:

[0016] First, the EEG digital signal and the EMG digital signal are subjected to Hilbert transform and envelope extraction to obtain the actual characteristic values ​​of the EEG digital signal and the EMG digital signal. Then, the actual characteristic values ​​of the EEG digital signal and the EMG digital signal are compared with the characteristic values ​​of the EEG digital signal and the EMG digital signal preset in the LabVIEW control software. The error and error change rate are calculated in real time through the PID control algorithm. The control signal in the PID control algorithm is shown in the following formula:

[0017]

[0018] In the formula, u(t) represents the output of the controller at time t, e(t) represents the error at time t, and K p , K i , K dRepresent the gain coefficients of proportional, integral, and differential respectively; represents the cumulative sum of the error e(t), represents the rate of change of the error e(t); when the closed-loop ultrasound stimulation first induces EMG, When closed-loop ultrasound stimulation first evokes EEG, Starting from a set of initial values, the PID parameters are adjusted through fuzzy logic. Based on the error e(t) and the error change rate Δe(t), the fuzzy control system generates appropriate proportional, integral and derivative gain adjustments.

[0019] Then fuzzification is performed to convert the error e(t) and the error change rate Δe(t) into fuzzy linguistic variables "large", "small", "positive", and "negative";

[0020] After that, fuzzy reasoning is performed, based on the existing typical fuzzy rule base, and formulated according to experimental data or experience. The control output of fuzzy PID is shown as follows:

[0021]

[0022] In the formula, Indicates the initial setting of PID parameters, ΔK p , ΔK i , ΔK d represents the adjustment amount generated by the fuzzy control rule;

[0023] Finally, defuzzification is performed to convert the result of fuzzy reasoning, that is, the fuzzy K p , K i , K d Parameters are converted into specific actual values ​​through the defuzzification process for parameter adjustment of PID control.

[0024] A further improvement of the technical solution of the present invention is that: the fuzzy rule is set as follows: when the error e(t) is "large" and the error change rate Δe(t) is "large", increase K p value, reduce K i value, increase K d value; when the error e(t) is "large" and the error change rate Δe(t) is "small", keep K p value, increase K i value, keep K d value; when the error e(t) is "small" and the error change rate Δe(t) is "large", reduce K p value, reduce K i value, increase K d value; when the error e(t) is "small" and the error change rate Δe(t) is "small", further reduce Kp value, slightly reduce K i value, reduce K d When the error e(t) is negative and the error change rate Δe(t) is negative, reduce K p value, increase K i value, increase K d When the error e(t) is positive and the error change rate Δe(t) is positive, increase K p value, reduce K i value, reduce K d value.

[0025] A further improvement of the technical solution of the present invention is that: the ultrasonic stimulation module is configured to receive a control signal transmitted by a fuzzy PID control module to perform ultrasonic stimulation on an experimental object; the ultrasonic stimulation module includes a signal generator arranged on a microcontroller, a radio frequency power amplifier connected to the output end of the signal generator, an ultrasonic transducer connected to the output end of the radio frequency power amplifier, and a collimator filled with a coupling agent connected to the output end of the ultrasonic transducer.

[0026] A multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control, using a multi-channel signal acquisition closed-loop ultrasonic stimulation system to perform multi-channel signal acquisition and closed-loop ultrasonic stimulation, comprising the following steps:

[0027] Step S1: The signal generator in the ultrasonic stimulation module enables the ultrasonic transducer according to the received control signal to perform ultrasonic stimulation on a preset area of ​​the experimental object;

[0028] Step S2: The multi-channel EEG electrodes in the signal acquisition and preprocessing module are placed in a specific preset area of ​​the cerebral cortex of the experimental subject, and the multi-channel EMG electrodes are placed in a preset area of ​​the neck muscles of the experimental subject, and the original EEG signals and EMG signals of the experimental subject are respectively acquired and preprocessed;

[0029] Step S3: The signal processing module receives the EEG signal and the EMG signal preprocessed by the signal acquisition and preprocessing module, performs multi-channel signal fusion, feature extraction, feature evaluation and selection, and transmits the optimized digital signal to the fuzzy PID control module via radio frequency;

[0030] Step S4: The fuzzy PID control module performs fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module in turn, obtains real-time PID controller parameters, forms new ultrasonic stimulation parameters, and transmits them to the ultrasonic stimulation module through radio frequency communication to accurately stimulate the designated area of ​​the experimental subject to achieve closed-loop control.

[0031] A further improvement of the technical solution of the present invention is that in step S3, the multi-channel signal fusion includes performing Kalman filtering on the data collected by a single channel and then performing window function processing on the data of the two channels to obtain fused and optimized data, which specifically includes the following steps:

[0032] Step S301: State prediction:

[0033] x k =F k x k-1 +B k u k +w k

[0034] In the formula, x k represents the state vector at time k, F k represents the state transfer matrix, B k represents the control input matrix, u k represents the control vector, w k represents the process noise, which is assumed to be Gaussian white noise;

[0035] The observation equation of the system is:

[0036] z k =H k x k +v k

[0037] In the formula, z k represents the observation vector at time k, H k represents the observation matrix, v k represents the measurement noise;

[0038] Step S302: Covariance prediction:

[0039]

[0040] In the formula, represents the error of the prediction estimate, Q k represents the process noise covariance;

[0041] Step S303: Calculate the Kalman gain:

[0042]

[0043] In the formula, K k represents the Kalman gain, R k represents the measurement noise covariance;

[0044] Step S304: Perform state update and covariance update:

[0045]

[0046] In the formula, represents the posterior state estimate at time k, P k represents the covariance matrix of the posterior state estimate, and I represents the identity matrix;

[0047] Step S305: Divide the processed signal into time windows, each window containing signal data at multiple time points T;

[0048] Step S306: In each time window, the signals of each channel are fused by taking the average value of all channels at the same time point, that is:

[0049]

[0050] In the formula, represents the fused signal, represents the signal of the i-th channel at time k, and n represents the total number of channels.

[0051] Due to the adoption of the above technical solution, the technical progress achieved by the present invention is:

[0052] The present invention is based on a multi-channel signal acquisition closed-loop ultrasonic stimulation system and method for fuzzy PID control. The fuzzy PID control adopted can achieve accurate regulation of EEG and EMG signals, adjust the output characteristic value in real time, make it quickly approach the predetermined target value, and ensure the accuracy and stability of the stimulation signal. In addition, multiple signal channels can be collected and processed at the same time, which is suitable for complex multi-point signal monitoring and stimulation, and realizes a wider range of application scenarios. Due to the portable design, it can be flexibly applied to a variety of environments, reducing the complexity and volume of the equipment while maintaining efficient signal processing and feedback response. Through signal processing methods such as fuzzy PID control and Hilbert transform, signal features can be more effectively extracted, noise interference can be reduced, thereby improving the acquisition quality and accuracy of the signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a principle block diagram of a multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control of the present invention;

[0054] Figure 2 It is a schematic diagram of the assembly of the multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control and the experimental object of the present invention;

[0055] Figure 3 It is a flow chart of a multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control of the present invention;

[0056] Figure 4It is a flow chart of multi-channel data fusion in a multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control of the present invention;

[0057] Figure 5 It is a comparison diagram of the EEG waveforms of the experimental group and the control group provided in the experimental example of the present invention;

[0058] Among them, 1. Ultrasonic stimulation module, 2. Signal processing module, 3. Power supply module, 4. Fuzzy PID control module, 5. External antenna, 6. Signal acquisition and preprocessing module. DETAILED DESCRIPTION

[0059] The present invention is further described in detail below in conjunction with embodiments:

[0060] Example 1

[0061] like Figure 1 and Figure 2 As shown, this embodiment provides a multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control, including a signal acquisition and preprocessing module 6 integrated on a backpack adapted to the experimental subject, a signal processing module 2 connected to the signal acquisition and preprocessing module 6, a fuzzy PID control module 4 connected to the signal processing module 2 via radio frequency, an ultrasonic stimulation module 1 connected to the fuzzy PID control module 4 via radio frequency, a power module 3 and an external antenna 5. Among them, the experimental subject is a rat or rabbit commonly used in experiments. The portable design of the backpack reduces the pressure on the animal, and compared with the previous tether to restrict the animal's behavior, the collected signals are more accurate and scientific.

[0062] Furthermore, the signal acquisition and preprocessing module 6 is configured to acquire the EEG signals and EMG signals of the experimental subject and perform preprocessing on the acquired analog signals, such as filtering, re-referencing, and artifact processing, and convert the analog signals into corresponding digital signals.

[0063] The signal acquisition and preprocessing module 6 includes a multi-channel EEG electrode for acquiring EEG signals, a multi-channel electromyography electrode for acquiring electromyography signals EMG, a front-end amplifier connected to the multi-channel EEG electrode and the multi-channel electromyography electrode respectively, a microelectrode AC amplifier connected to the front-end amplifier, and an analog-to-digital converter connected to the microelectrode AC amplifier, wherein the series configuration of the front-end amplifier and the microelectrode AC amplifier can significantly improve the signal-to-noise ratio of the system, effectively suppress irrelevant noise and interference, and thus improve the signal quality. The analog-to-digital converter is responsible for gain adjustment, realizing the input of multi-channel signals, and converting the amplified analog signals into digital signals for subsequent processing. The multi-channel EEG electrode includes two single-channel microwire electrodes implanted in the M1 motor cortex on the skull of the experimental subject and two reference electrodes plugged into the nasal bone of the experimental subject, the two single-channel microwire electrodes are used to record the neural electrical signals of the cerebral cortex, and the two reference electrodes are used to improve the accuracy of signal acquisition and reduce noise interference. The multi-channel electromyographic electrodes include two surface electromyographic electrodes placed in a preset area of ​​the experimental subject's neck muscles and two reference electrodes attached to the surface of the experimental subject's skin near the bones. The two surface electromyographic electrodes are used to record the electrical activity during local muscle contraction, and the two reference electrodes are used to enhance the stability of the signal and serve as reference measurement points to accurately reflect the changes in the electrical signals of skeletal muscle activity. Preferably, the acquisition frequency of the electrodes of the signal acquisition and preprocessing module 6 is set to 3 khz.

[0064] Specifically, both EEG electrodes and EMG electrodes use tungsten wire probes. The length of the implant handle is 6 mm, the width is 200 μm, the thickness is 70 μm, the electrode spacing is 200 μm thick, and the electrode layer is composed of a sputtered iridium oxide film. When collecting EEG signals, the probe needs to be inserted into the cerebral cortex through minimally invasive surgery. When collecting EMG signals, there is no need to cut the skin. The probe can be directly inserted into the target muscle through the skin, so that the probe can accurately capture the activity of a single neuron or a group of neurons. During the placement process, the exact location of the probe insertion can be determined using anatomical data or images (such as CT, MRI).

[0065] The front-end amplifier, microelectrode AC amplifier and analog-to-digital converter are used to filter, amplify and digitize the collected signals. The front-end amplifier is configured to amplify the weak analog signal to a level that can be processed by the subsequent circuit, while reducing the introduction of noise as much as possible, and transmit the amplified signal to the microelectrode AC amplifier. The microelectrode AC amplifier is configured to amplify and filter the EEG signal and the EMG signal, and send the processed signal to the analog-to-digital converter. This configuration optimizes the availability and quality of the signal. The analog-to-digital converter is configured to convert the analog signal into a corresponding digital signal, and transmit the converted digital signal to the signal processing module 2. The analog-to-digital converter can use the RHD2132 chip, which reduces the energy consumption of the circuit and reduces the overall area of ​​the system.

[0066] The signal processing module 2 is configured to perform multi-channel signal fusion, feature extraction, feature evaluation and selection on the EEG digital signal and the EMG digital signal preprocessed by the signal acquisition and preprocessing module 6, and transmit the optimized digital signal to the fuzzy PID control module 4 via radio frequency.

[0067] Specifically, the signal processing module 2 includes a microcontroller and a digital signal processor DSP that is connected to the microcontroller for communication. The digital signal processor cooperates with the microcontroller. The microcontroller is responsible for management and control tasks. It transmits real-time data to the DSP through direct memory access DMA. The DSP is responsible for high-demand computing tasks, such as multi-channel signal fusion, feature extraction, and feature evaluation and selection. The feature evaluation and selection use existing algorithms to score and classify existing features. For example, in the correlation analysis in the filtering method, if the correlation between the feature and the target variable is low, the feature can be removed. Among them, the microcontroller adopts a low-power MSP430 microcontroller, which integrates a sub-1GHz radio transceiver, and can transmit the digital signal after multi-channel signal fusion, feature extraction, feature evaluation and selection to the fuzzy PID control module 4 for further processing. In addition, the signal processing module 2 is equipped with a built-in RHD2132 chip, which is suitable for portable devices and has high signal processing accuracy.

[0068] The fuzzy PID control module 4 is configured to perform fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module 2 in turn, and obtain real-time PID controller parameters, thereby forming new ultrasonic stimulation parameters, and transmitting the new stimulation parameters to the ultrasonic stimulation module 1 through radio frequency communication, accurately stimulating the designated area of ​​the experimental subject, and realizing closed-loop control. This transmission process abandons the traditional wired connection method, and avoids the common WiFi and Bluetooth wireless transmission schemes, and adopts the low-power, high-speed CC1110 chip to ensure the efficient transmission and reliability of the parameters, and further improve the flexibility and stability of the stimulation system. The fuzzy PID control module 4 includes LabVIEW control software and fuzzy PID control algorithm that responds to the EEG digital signal and the EMG digital signal sent by radio frequency.

[0069] Specifically, the fuzzy PID control algorithm includes the following steps:

[0070] First, the EEG digital signal and the EMG digital signal are subjected to Hilbert transform to obtain the envelope to obtain the actual characteristic values ​​of the EEG digital signal and the EMG digital signal. Then, the actual characteristic values ​​of the EEG digital signal and the EMG digital signal are compared with the characteristic values ​​of the EEG digital signal and the EMG digital signal preset in the LabVIEW control software. The error and the error change rate are calculated in real time by the PID control algorithm. The control signal in the PID control algorithm is shown as follows:

[0071]

[0072] In the formula, u(t) represents the output of the controller at time t, e(t) represents the error at time t, and K p , K i , K d Represent the gain coefficients of proportional, integral, and differential respectively; represents the cumulative sum of the error e(t), represents the rate of change of the error e(t); when the closed-loop ultrasound stimulation first induces EMG, When closed-loop ultrasound stimulation first evokes EEG, Starting from a set of initial values, the PID parameters are adjusted through fuzzy logic. Based on the error e(t) and the error change rate Δe(t), the fuzzy control system generates appropriate proportional, integral and derivative gain adjustments.

[0073] Then, fuzzification is performed to convert the error e(t) and the error change rate Δe(t) into fuzzy linguistic variables "large", "small", "positive", and "negative", such as dividing the error e(t) into "positive large", "positive small", etc.

[0074] After that, fuzzy reasoning is performed, based on the existing typical fuzzy rule base, and formulated according to experimental data or experience. The control output of fuzzy PID is shown as follows:

[0075]

[0076] In the formula, Indicates the initial setting of PID parameters, ΔK p , ΔK i , ΔK d represents the adjustment amount generated by the fuzzy control rule; the fuzzy rule is set as follows: when the error e(t) is "large" and the error change rate Δe(t) is "large", increase K p value, reduce K i value, increase K d value; when the error e(t) is "large" and the error change rate Δe(t) is "small", keep K p value, increase Ki value, keep K d value; when the error e(t) is "small" and the error change rate Δe(t) is "large", reduce K p value, reduce K i value, increase K d When the error e(t) is "small" and the error change rate Δe(t) is "small", further reduce K p value, slightly reduce K i value, reduce K d When the error e(t) is negative and the error change rate Δe(t) is negative, reduce K p value, increase K i value, increase K d When the error e(t) is positive and the error change rate Δe(t) is positive, increase K p value, reduce K i value, reduce K d value;

[0077] Finally, defuzzification is performed to convert the result of fuzzy reasoning, that is, the fuzzy K p , K i , K d The parameters are converted into specific actual values ​​through the defuzzification process, and then the defuzzified values ​​are used for parameter adjustment of PID control.

[0078] The advantage of fuzzy PID control is that it does not rely on an accurate mathematical model, so it has good robustness when dealing with system uncertainty and noise, and by adjusting the fuzzy rules, it can flexibly respond to different control tasks.

[0079] The ultrasonic stimulation module 1 is configured to receive the control signal transmitted by the fuzzy PID control module 4 to perform ultrasonic stimulation on the experimental object. The ultrasonic stimulation module 1 includes a signal generator arranged on a microcontroller, a radio frequency power amplifier connected to the output end of the signal generator, an ultrasonic transducer connected to the output end of the radio frequency power amplifier, and a collimator filled with coupling agent connected to the output end of the ultrasonic transducer. The microcontroller receives the control signal transmitted by the fuzzy PID control module 4, and the signal generator generates a corresponding stimulation signal according to the control signal. The stimulation signal is amplified by the radio frequency power amplifier and then transmitted to the ultrasonic transducer, so that the ultrasonic transducer changes the parameters of the next cycle of ultrasound and generates an ultrasonic stimulation corresponding to the output of the fuzzy PID control to act on the skull of the experimental object, thereby realizing closed-loop control.

[0080] Wherein, the ultrasonic transducer is attached to the skull of the experimental subject under anesthesia. After the scalp is cut, the ultrasonic transducer is attached to the motor cortex of the experimental subject with cyanoacrylate adhesive. The center frequency of the ultrasonic transducer is set to 500kHz. The Olybas ultrasonic transducer with a diameter of 31mm is selected, and the model is V301 SU. Preferably, each ultrasonic stimulation of the ultrasonic transducer lasts for 300ms, the pulse sequence pulse repetition frequency (PRF) is 1kHz, and the duty cycle (DC) is 50%. The collimator is used to limit the output of the sound wave within a certain range, so that the ultrasonic wave can stimulate the target brain area of ​​the experimental subject in a targeted manner, thereby eliminating unnecessary interference, and the coupling agent is filled inside the collimator, which can well conduct the ultrasound to the skull of the experimental subject. The fundamental frequency of the pulse signal generated by the signal generator is 0.35MHz-0.7MHz. According to the input impedance of the ultrasonic transducer and the output impedance of the power amplifier, the matching circuit is designed.

[0081] The power module 3 is configured to power the entire system, including a 3.7V EMB lithium-ion battery with a capacity of 45 mAh, a low dropout regulator, an oscillation circuit, etc. Since the neural recording system requires a constant power supply of 3.3 V, a low dropout (LDO) regulator is used to reduce the voltage of the 3.7 V battery to 3.3 V. An oscillator circuit is used after the LDO circuit to provide pulse stimulation with a constant voltage amplitude of 3.3 V.

[0082] The external antenna 5 is configured to extend the transmission distance and increase the transmission rate, and to reduce the floor space in the backpack.

[0083] Example 2

[0084] like Figure 3 As shown, this embodiment provides a multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control. Using the multi-channel signal acquisition closed-loop ultrasonic stimulation system of embodiment 1 to perform multi-channel signal acquisition and closed-loop ultrasonic stimulation includes the following steps:

[0085] Step S1: The signal generator in the ultrasonic stimulation module 1 enables the ultrasonic transducer according to the received control signal to perform ultrasonic stimulation on a preset area of ​​the experimental object;

[0086] Step S2: The multi-channel EEG electrodes in the signal acquisition and preprocessing module 6 are placed in a specific preset area of ​​the cerebral cortex of the experimental subject, and the multi-channel EMG electrodes are placed in a preset area of ​​the neck muscles of the experimental subject, respectively acquiring the original EEG signals and EMG signals of the experimental subject, and preprocessing them;

[0087] Step S3: the signal processing module 2 receives the EEG signal and the EMG signal preprocessed by the signal acquisition and preprocessing module 6 and performs multi-channel signal fusion, feature extraction, feature evaluation and selection, and transmits the optimized digital signal to the fuzzy PID control module 4 via radio frequency;

[0088] Among them, Figure 4 As shown, multi-channel signal fusion includes performing Kalman filtering on the data collected by a single channel and then performing window function processing on the data of two channels to obtain fused and optimized data, which specifically includes the following steps:

[0089] Step S301: State prediction:

[0090] x k =F k x k-1 +B k u k +w k

[0091] In the formula, x k represents the state vector at time k, F k represents the state transfer matrix, B k represents the control input matrix, u k represents the control vector, w k represents the process noise, which is assumed to be Gaussian white noise;

[0092] The observation equation of the system is:

[0093] z k =H k x k +v k

[0094] In the formula, z k represents the observation vector at time k, H k represents the observation matrix, v k represents the measurement noise;

[0095] Step S302: Covariance prediction:

[0096]

[0097] In the formula, represents the error of the prediction estimate, Q k represents the process noise covariance;

[0098] Step S303: Calculate the Kalman gain:

[0099]

[0100] In the formula, K krepresents the Kalman gain, R k represents the measurement noise covariance;

[0101] Step S304: Perform state update and covariance update:

[0102]

[0103] In the formula, represents the posterior state estimate at time k, P k represents the covariance matrix of the posterior state estimate, and I represents the identity matrix;

[0104] Step S305: Divide the processed signal into time windows, each window containing signal data at multiple time points T;

[0105] Step S306: In each time window, the signals of each channel are fused by taking the average value of all channels at the same time point, that is:

[0106]

[0107] In the formula, represents the fused signal, represents the signal of the i-th channel at time k, and n represents the total number of channels;

[0108] Step S4: The fuzzy PID control module 4 performs fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module 2 in sequence to obtain real-time PID controller parameters, form new ultrasonic stimulation parameters and transmit them to the ultrasonic stimulation module 1 through radio frequency communication, so as to accurately stimulate the designated area of ​​the experimental subject and realize closed-loop control.

[0109] Experimental example

[0110] In this experimental example, 8-week-old healthy adult male rats were selected, and a total of ten were randomly divided into a control group (TUS) and an experimental group (CTUS), with five rats in each group. After anesthesia (2% isoflurane), the rats were fixed on a stereotaxic, the skin was cleaned with 0.9% saline, the scalp was cut along the midline of the skull, the subcutaneous tissue and periosteum were removed, and a tungsten wire electrode (diameter 80μm) was implanted in the motor cortex. The electrode layer was composed of a sputtered iridium oxide film for EEG signal acquisition. The ground electrode and reference electrode were implanted in the cerebellum, and a plastic sleeve was placed in the hippocampus to facilitate ultrasound transmission. The needle of the syringe was put on the electromyography acquisition electrode, and the electrode was exposed about two centimeters from the needle. The electrode was bent to one side, and then the needle was used to pierce the electrode into the neck muscle, and then the implanted electrode was pressed and the needle was pulled out. The electromyography electrode was implanted. Each ultrasonic stimulation lasted 300ms, the pulse sequence pulse repetition frequency (PRF) was 1kHz, and the duty cycle (DC) was 50%. The ultrasonic stimulation device was connected to the control system, and the signal acquisition device was connected to the test object. According to the experimental settings, set the parameters of the first ultrasound stimulation, such as frequency, power, duty cycle, etc. Turn on the ultrasound stimulation device and stimulate according to the set parameters. Under the initial conditions (do not run the fuzzy PID control algorithm and adopt single-channel acquisition), turn on the ultrasound stimulation, and the signal acquisition device will collect physiological signals (EEG, EMG) in real time. After system filtering, amplification, analog-to-digital conversion and other processing optimizations, the experimental data of the control group (TUS) are obtained.

[0111] The experimental group runs the fuzzy PID control algorithm and adopts multi-channel signal acquisition. Other experimental conditions of the experimental group (CTUS) and the control group (TUS) are consistent, and the experimental data of the experimental group are obtained. Figure 5 As shown in the figure, during the experiment, by observing the response speed, overshoot and stability of the control system, the experimental results of open loop (no PID control) and closed loop control (fuzzy PID control) were compared, and the effect of closed loop control was verified. In terms of the comprehensiveness of the signal, the experimental group can ensure the overall quality of the signal, while the comprehensiveness of the signal of the control group will deteriorate once the scene is complex. In terms of the response speed, the experimental group can respond quickly. The response speed of the control group is slow, which may cause the acquisition signal to lag or be inaccurate. In terms of anti-interference ability, the experimental group can better cope with interference in the environment and automatically adjust the response speed. However, due to the lack of feedback regulation ability, the collected signal of the control group may be subject to more noise and interference, and the inhibitory effect on epilepsy lacks efficiency. Since the closed-loop ultrasound stimulation system of the experimental group is controlled by fuzzy PID, according to the experimental results, a higher quality acquisition signal is obtained, and the regulation effect is more accurate and effective.

[0112] In summary, the present invention can perform adaptive adjustment according to the physiological differences of different users or individuals through fuzzy PID control, ensure the personalized application of ultrasonic stimulation, reduce the dependence on parameter adjustment, improve the usability and adaptability of the system, and at the same time, can simultaneously collect biological signals of multiple channels, and perform signal fusion processing to eliminate the noise and instability of single-channel signals. The present invention has the characteristics of easy portability and low power consumption, and ensures that the experimental subjects can move freely and stay awake during the collection process.

[0113] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control, characterized in that: The invention comprises a signal acquisition and preprocessing module (6) integrated on a backpack adapted to an experimental subject, a signal processing module (2) communicatively connected to the signal acquisition and preprocessing module (6), a fuzzy PID control module (4) connected to the signal processing module (2) via radio frequency, an ultrasonic stimulation module (1) connected to the fuzzy PID control module (4) via radio frequency, and a power supply module (3) and an external antenna (5) for supplying power to the entire system.

2. A multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 1, characterized in that: The signal acquisition and preprocessing module (6) is configured to acquire the EEG signal and the EMG signal of the experimental subject and preprocess the acquired analog signals, and convert the analog signals into corresponding digital signals; the signal acquisition and preprocessing module (6) comprises a multi-channel EEG electrode for acquiring the EEG signal, a multi-channel EMG electrode for acquiring the EMG signal, a front-end amplifier connected to the multi-channel EEG electrode and the multi-channel EMG electrode respectively, a microelectrode AC amplifier connected to the front-end amplifier for communication, and an analog-to-digital converter connected to the microelectrode AC amplifier for communication.

3. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 2, characterized in that: The multi-channel EEG electrodes include two single-channel microwire electrodes implanted in the M1 motor cortex on the skull of the experimental subject and two reference electrodes plugged into the nasal bones of the experimental subject; The multi-channel electromyographic electrode comprises two surface electromyographic electrodes placed in a preset area of ​​the neck muscle of the experimental subject and two reference electrodes attached to the skin surface of the experimental subject near the bones; the front-end amplifier is configured to amplify the weak analog signal and reduce the introduction of noise, and transmit the amplified signal to the microelectrode AC amplifier; the microelectrode AC amplifier is configured to amplify and filter the EEG signal and the electromyographic signal, and send the processed signal to the analog-to-digital converter; The analog-to-digital converter is configured to convert the analog signal into a corresponding digital signal and transmit the digital signal to the signal processing module (2).

4. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 3, characterized in that: The signal processing module (2) is configured to perform multi-channel signal fusion, feature extraction, feature evaluation and selection on the EEG digital signal and the myoelectric digital signal preprocessed by the signal acquisition and preprocessing module (6), and transmit the optimized digital signal to the fuzzy PID control module (4) via radio frequency; the signal processing module (2) comprises a microcontroller and a digital signal processor DSP communicatively connected to the microcontroller.

5. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 4, characterized in that: The fuzzy PID control module (4) is configured to sequentially perform fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module (2), thereby forming new ultrasound stimulation parameters and transmitting them to the ultrasound stimulation module (1) via radio frequency communication; the fuzzy PID control module (4) comprises LabVIEW control software that responds to the EEG digital signal and the EMG digital signal transmitted by radio frequency, and a fuzzy PID control algorithm.

6. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 5, characterized in that: The fuzzy PID control algorithm comprises the following steps: First, the EEG digital signal and the EMG digital signal are subjected to Hilbert transform and envelope extraction to obtain the actual characteristic values ​​of the EEG digital signal and the EMG digital signal. Then, the actual characteristic values ​​of the EEG digital signal and the EMG digital signal are compared with the characteristic values ​​of the EEG digital signal and the EMG digital signal preset in the LabVIEW control software. The error and error change rate are calculated in real time through the PID control algorithm. The control signal in the PID control algorithm is shown in the following formula: In the formula, u(t) represents the output of the controller at time t, e(t) represents the error at time t, and K p , K i , K d Represent the gain coefficients of proportional, integral, and differential respectively; represents the cumulative sum of the error e(t), represents the rate of change of the error e(t); when the closed-loop ultrasound stimulation first induces EMG, When closed-loop ultrasound stimulation first evokes EEG, Starting from a set of initial values, the PID parameters are adjusted through fuzzy logic. Based on the error e(t) and the error change rate Δe(t), the fuzzy control system generates appropriate proportional, integral and derivative gain adjustments. Then, fuzzification is performed to convert the error e(t) and the error change rate Δe(t) into fuzzy linguistic variables "large", "small", "positive", and "negative". After that, fuzzy reasoning is performed, based on the existing typical fuzzy rule base, and formulated according to experimental data or experience. The control output of fuzzy PID is shown as follows: In the formula, Indicates the initial setting of PID parameters, ΔK p , ΔK i , ΔK d represents the adjustment amount generated by the fuzzy control rule; Finally, defuzzification is performed to convert the result of fuzzy reasoning, that is, the fuzzy K p , K i , K d Parameters are converted into specific actual values ​​through the defuzzification process for parameter adjustment of PID control.

7. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 6, characterized in that: The fuzzy rule is: when the error e(t) is "large" and the error change rate Δe(t) is "large", increase K p value, reduce K i value, increase K d value; when the error e(t) is "large" and the error change rate Δe(t) is "small", keep K p value, increase K i value, keep K d When the error e(t) is "small" and the error change rate Δe(t) is "large", reduce K p value, reduce K i value, increase K d When the error e(t) is "small" and the error change rate Δe(t) is "small", further reduce K p value, slightly reduce K i value, reduce K d When the error e(t) is negative and the error change rate Δe(t) is negative, reduce K p value, increase K i value, increase K d When the error e(t) is positive and the error change rate Δe(t) is positive, increase K p value, reduce K i value, reduce K d value.

8. The multi-channel signal acquisition closed-loop ultrasonic stimulation system based on fuzzy PID control according to claim 7, characterized in that: The ultrasonic stimulation module (1) is configured to receive a control signal transmitted by a fuzzy PID control module (4) to perform ultrasonic stimulation on an experimental object; the ultrasonic stimulation module (1) comprises a signal generator arranged on a microcontroller, a radio frequency power amplifier connected to an output end of the signal generator, an ultrasonic transducer connected to an output end of the radio frequency power amplifier, and a collimator filled with a coupling agent connected to an output end of the ultrasonic transducer.

9. A multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control, characterized in that: Using the multi-channel signal acquisition closed-loop ultrasound stimulation system according to any one of claims 1 to 8 to perform multi-channel signal acquisition and closed-loop ultrasound stimulation comprises the following steps: Step S1: The signal generator in the ultrasonic stimulation module (1) enables the ultrasonic transducer according to the received control signal to perform ultrasonic stimulation on a preset area of ​​the experimental object; Step S2: The multi-channel EEG electrodes in the signal acquisition and preprocessing module (6) are placed in a specific preset area of ​​the cerebral cortex of the experimental subject, and the multi-channel EMG electrodes are placed in a preset area of ​​the neck muscles of the experimental subject, respectively acquiring the original EEG signals and EMG signals of the experimental subject, and preprocessing them; Step S3: the signal processing module (2) receives the EEG signal and the EMG signal preprocessed by the signal acquisition and preprocessing module (6) and performs multi-channel signal fusion, feature extraction, feature evaluation and selection, and transmits the optimized digital signal to the fuzzy PID control module (4) via radio frequency; Step S4: The fuzzy PID control module (4) performs fuzzification processing, fuzzy inference engine processing and defuzzification processing on the EEG digital signal and the EMG digital signal optimized by the signal processing module (2) in turn, obtains real-time PID controller parameters, forms new ultrasonic stimulation parameters, and transmits them to the ultrasonic stimulation module (1) via radio frequency communication, so as to accurately stimulate the designated area of ​​the experimental subject and realize closed-loop control.

10. The multi-channel signal acquisition closed-loop ultrasonic stimulation method based on fuzzy PID control according to claim 9, characterized in that: In step S3, multi-channel signal fusion includes performing Kalman filtering on the data collected by a single channel and then performing window function processing on the data of two channels to obtain fused and optimized data, which specifically includes the following steps: Step S301: State prediction: x k =F k x k-1 +B k u k +w k In the formula, x k represents the state vector at time k, F k represents the state transfer matrix, B k represents the control input matrix, u k represents the control vector, w k represents the process noise, which is assumed to be Gaussian white noise; The observation equation of the system is: z k =H k x k +v k In the formula, z k represents the observation vector at time k, H k represents the observation matrix, v k represents the measurement noise; Step S302: Covariance prediction: In the formula, represents the error of the prediction estimate, Q k represents the process noise covariance; Step S303: Calculate the Kalman gain: In the formula, K k represents the Kalman gain, R k represents the measurement noise covariance; Step S304: Perform state update and covariance update: In the formula, represents the posterior state estimate at time k, P k represents the covariance matrix of the posterior state estimate, and I represents the identity matrix; Step S305: Divide the processed signal into time windows, each window containing signal data at multiple time points T; Step S306: In each time window, the signals of each channel are fused by taking the average value of all channels at the same time point, that is: In the formula, represents the fused signal, represents the signal of the i-th channel at time k, and n represents the total number of channels.

Images