阅读说明：本技术 一种脑电信号采集装置及方法 (Electroencephalogram signal acquisition device and method ) 是由 李晓 寇建阁 石岩 王娜 王一轩 任帅 于 2020-12-23 设计创作，主要内容包括：本发明公开了一种脑电信号采集装置及方法,包括：第一信号采集电极、第二信号采集电极、信号处理器和无线发射模块；第一信号采集电极、第二信号采集电极和无线发射模块分别与信号处理器电性连接；无线发射模块与脑机接口无线通信；第一信号采集电极用于采集Spike信号；第二信号采集电极用于采集ECoG信号；信号处理器用于将Spike信号和ECoG信号进行处理及转换；无线发射模块用于将处理及转换后的Spike信号和ECoG信号发送至脑机接口；脑机接口对Spike信号和ECoG信号进行训练,并建立受体不同动作下的Spike信号和ECoG信号之间的关联性。本发明既能够长期安全采集脑电信号,又能提供高信噪比。(The invention discloses an electroencephalogram signal acquisition device and method, which comprise the following steps: the device comprises a first signal acquisition electrode, a second signal acquisition electrode, a signal processor and a wireless transmitting module; the first signal acquisition electrode, the second signal acquisition electrode and the wireless transmission module are respectively electrically connected with the signal processor; the wireless transmitting module is in wireless communication with the brain-computer interface; the first signal acquisition electrode is used for acquiring Spike signals; the second signal acquisition electrode is used for acquiring an ECoG signal; the signal processor is used for processing and converting the Spike signal and the ECoG signal; the wireless transmitting module is used for transmitting the processed and converted Spike signals and ECoG signals to the brain-computer interface; the brain-computer interface trains the Spike signals and the ECoG signals and establishes the correlation between the Spike signals and the ECoG signals under different actions of the receptor. The invention can safely collect the electroencephalogram signal for a long time and can also provide a high signal-to-noise ratio.)

1. An electroencephalogram signal acquisition device, characterized by comprising: the device comprises a first signal acquisition electrode, a second signal acquisition electrode, a signal processor and a wireless transmitting module; the first signal acquisition electrode, the second signal acquisition electrode and the wireless transmission module are respectively electrically connected with the signal processor; the wireless transmitting module is in wireless communication with the brain-computer interface;

the first signal acquisition electrode is used for acquiring Spike signals; the second signal acquisition electrode is used for acquiring an ECoG signal; the signal processor is used for processing and converting the Spike signal and the ECoG signal; the wireless transmitting module is used for transmitting the processed and converted Spike signals and the ECoG signals to the brain-computer interface; the brain-computer interface trains the Spike signals and the ECoG signals and establishes the correlation between the Spike signals and the ECoG signals under different actions of a receptor.

2. The electroencephalograph signal acquisition device of claim 1, wherein the first signal acquisition electrode comprises a hydrogel substrate and a plurality of contact electrodes; the electric shock electrodes are arranged on the surface of the hydrogel substrate in an array mode.

3. The electroencephalograph signal acquisition device according to claim 2, wherein the second signal acquisition electrode comprises a ceramic substrate and a plurality of microwire electrodes; one end of the microwire electrode is embedded into the ceramic substrate and is arranged in an array; the other end of the microwire electrode is wrapped with a silicone elastomer at the periphery, and the tip of the microwire electrode is exposed outside the silicone elastomer.

4. The electroencephalogram signal acquisition device according to claim 3, wherein the ceramic substrate is laminated on the surface of the hydrogel substrate and is close to the central position of the hydrogel substrate; the contact electrode is disposed around the ceramic substrate.

5. The electroencephalogram signal acquisition device according to claim 4, further comprising at least one reference electrode; the reference electrode is positioned proximate to a boundary located on the hydrogel substrate.

6. The electroencephalogram signal acquisition device according to claim 5, wherein an insulating layer is encapsulated in the hydrogel substrate; corresponding lines are arranged on the insulating layer; one end of the circuit is correspondingly connected with each contact electrode, each microwire electrode and each reference electrode one by one to form a plurality of signal transmission channels; the other end is connected with a data interface through a wire; the data interface is electrically connected with the signal processor.

7. An electroencephalogram signal acquisition method which is suitable for the electroencephalogram signal acquisition device of any one of claims 1 to 6, and is characterized by comprising the following steps:

collecting Spike signals and ECoG signals, and respectively preprocessing the Spike signals and the ECoG signals;

respectively establishing a mapping relation between the Spike signal and different actions and a mapping relation between the ECoG signal and different actions;

based on the mapping relation between the Spike signals and different actions and the mapping relation between the ECoG signals and different actions, training and identifying the frequency spectrum characteristics and the energy characteristics of the Spike signals and the ECoG signals under different frequency bands by utilizing a deep learning algorithm, and establishing the mapping relation between the Spike signals and the ECoG signals under different actions;

the ECoG signal is gradually used to replace the weak Spike signal.

8. The method of claim 7, wherein the preprocessing of the Spike signal and the ECoG signal comprises: removing low-frequency components in the Spike signals by using a high-pass filter; and carrying out filtering processing on the acquired ECoG signal by using a Kalman filtering mode.

9. The EEG signal acquisition method according to claim 7, wherein the establishing of the mapping relationship between the Spike signal and the different actions and the mapping relationship between the ECoG signal and the different actions respectively comprises:

acquiring the time point of the specific action and the time point of the Spike signal and the ECoG signal when the corresponding specific action occurs;

the two time points are corresponding, and the Spike signal and the ECoG signal are windowed according to the corresponding time points;

identifying and classifying Spike signals in the window, and separating Spike signals with different waveforms of each signal transmission channel, wherein the Spike signal of each waveform corresponds to one type of neuron release type;

counting the neuron release class number separated from each signal transmission channel by a time window of 6ms, and drawing a grating graph of neuron release conditions changing along with time;

based on the raster pattern, establishing a mapping relation between Spike signals and different actions;

carrying out data feature extraction and classification on the ECoG signal in the window;

and based on the extracted and classified data characteristics, establishing a mapping relation between the ECoG signal and different actions by using statistics.

10. The method for acquiring electroencephalogram signals according to claim 7, before training and identifying the spectral characteristics and the energy characteristics of the Spike signals and the ECoG signals under different frequency bands by using a deep learning algorithm, further comprising: and carrying out statistical analysis on the time frequency and frequency band energy of the Spike signal and the ECoG signal by using a correlation analysis algorithm, and representing different types of nerve signals in different action execution processes.

Technical Field

The invention relates to the technical field of medical equipment, in particular to an electroencephalogram signal acquisition device and method.

Background

For many years, patients with spinal cord injuries and epilepsy have suffered from a tremendous amount of exercise deficiency, and although certain research advances have been made clinically in this category of disease, there is still a large gap from true exercise recovery. In contrast, researchers have conducted research on the method in the engineering field and have achieved good effects. Among them, the brain-computer interface technology (brain-computer interface) has become a hot spot of current international research as a new man-machine interaction mode. The brain-computer interface is a channel for directly realizing the transfer of brain and external information without depending on nerve channels such as peripheral nerves, muscles and the like, and converts the information into control signals or stimulation signals to directly control external auxiliary devices or directly stimulate human body motion actuators such as motor muscles and the like, so that a patient can obtain the motor ability again. The brain-computer interface technology has remarkable achievement in the research and use fields of the reconstruction and repair of the movement function of disabled people, barrier-free man-machine interaction, the fusion of biological intelligence and artificial intelligence and the like.

However, there is a technical bottleneck in the electroencephalogram signal acquisition technology, which is one of the most important links of the brain-computer interface technology. The existing electroencephalogram signal collection modes can be divided into an implantable type (invasive) and a non-implantable type (non-invasive). For implantable electrodes, there are, in turn, nerve Spike (Spike) to disrupt the dura mater, Local Field Potential (LFP), and cortical electroencephalography (ECoG) to not disrupt the dura mater. On one hand, for the electrode implanted into the dura mater, the brain tissue can generate certain immunological rejection reaction, after the electrode is inserted into the brain tissue, glial cells in the brain can gradually wrap the electrode collecting end, so that the distance between the electrode and nerve cells is increased, the electrode is insulated, the electrode impedance is increased, and the electric signal can be gradually weak until the electric signal cannot be collected; on the other hand, for the ECoG acquisition mode, the electroencephalogram signal is weak, so that the signal spatial resolution and the signal-to-noise ratio are relatively low, and the signal decoding freedom degree is not high. These problems make the development of brain-computer interface research have a bottleneck, and hinder the deep and practical development of research, so that it is an urgent need for those skilled in the art to provide an electroencephalogram signal acquisition device and method capable of acquiring safely for a long time and providing a high signal-to-noise ratio.

Disclosure of Invention

In view of this, the invention provides an electroencephalogram signal acquisition device and method, which can safely acquire electroencephalogram signals for a long time and provide a high signal-to-noise ratio.

In order to achieve the purpose, the invention adopts the following technical scheme:

an electroencephalogram signal acquisition apparatus, comprising: the device comprises a first signal acquisition electrode, a second signal acquisition electrode, a signal processor and a wireless transmitting module; the first signal acquisition electrode, the second signal acquisition electrode and the wireless transmission module are respectively electrically connected with the signal processor; the wireless transmitting module is in wireless communication with the brain-computer interface;

the first signal acquisition electrode is used for acquiring Spike signals; the second signal acquisition electrode is used for acquiring an ECoG signal; the signal processor is used for processing and converting the Spike signal and the ECoG signal; the wireless transmitting module is used for transmitting the processed and converted Spike signals and the ECoG signals to the brain-computer interface; the brain-computer interface trains the Spike signals and the ECoG signals and establishes the correlation between the Spike signals and the ECoG signals under different actions of a receptor.

Preferably, in the electroencephalogram signal acquisition device, the first signal acquisition electrode comprises a hydrogel substrate and a plurality of contact electrodes; the electric shock electrodes are arranged on the surface of the hydrogel substrate in an array mode.

Preferably, in the electroencephalogram signal acquisition device, the second signal acquisition electrode comprises a ceramic substrate and a plurality of microwire electrodes; one end of the microwire electrode is embedded into the ceramic substrate and is arranged in an array; the other end of the microwire electrode is wrapped with a silicone elastomer at the periphery, and the tip of the microwire electrode is exposed outside the silicone elastomer.

Preferably, in the electroencephalogram signal acquisition device, the ceramic substrate is laminated on the surface of the hydrogel substrate and is close to the center of the hydrogel substrate; the contact electrode is disposed around the ceramic substrate.

Preferably, in the electroencephalogram signal acquisition device, at least one reference electrode is further included; the reference electrode is positioned proximate to a boundary located on the hydrogel substrate.

Preferably, in the electroencephalogram signal acquisition device, an insulating layer is sealed in the hydrogel substrate; corresponding lines are arranged on the insulating layer; one end of the circuit is correspondingly connected with each contact electrode, each microwire electrode and each reference electrode one by one to form a plurality of signal transmission channels; the other end is connected with a data interface through a wire; the data interface is electrically connected with the signal processor.

According to the technical scheme, compared with the prior art, the electroencephalogram signal acquisition device can acquire both the nerve peak potential signal (spike) and the cortical electroencephalogram signal (ECoG), in the process of applying the acquired signals to brain-computer interaction, feature extraction and correlation machine learning training are carried out on the two signals, the connection between the two signals based on individual patients is established through long-time deep learning, and under the condition that the spike electrode signals are weak or even cannot be acquired, the trained ECoG signals can be used as the main source of control signals to be applied to external human-computer interaction, so that the working time of high signal-to-noise ratio signal acquisition of the electroencephalogram electrode is prolonged.

The invention also provides an electroencephalogram signal acquisition method which is suitable for the electroencephalogram signal acquisition device of any one of claims 1 to 6, and comprises the following steps:

collecting Spike signals and ECoG signals, and respectively preprocessing the Spike signals and the ECoG signals;

respectively establishing a mapping relation between the Spike signal and different actions and a mapping relation between the ECoG signal and different actions;

based on the mapping relation between the Spike signals and different actions and the mapping relation between the ECoG signals and different actions, training and identifying the frequency spectrum characteristics and the energy characteristics of the Spike signals and the ECoG signals under different frequency bands by utilizing a deep learning algorithm, and establishing the mapping relation between the Spike signals and the ECoG signals under different actions;

the ECoG signal is gradually used to replace the weak Spike signal.

Preferably, in the electroencephalogram signal acquisition method, the preprocessing of the Spike signal and the ECoG signal includes: removing low-frequency components in the Spike signals by using a high-pass filter; and carrying out filtering processing on the acquired ECoG signal by using a Kalman filtering mode.

Preferably, in the electroencephalogram signal acquisition method, the respectively establishing a mapping relationship between the Spike signal and different actions and a mapping relationship between the ECoG signal and different actions includes:

acquiring the time point of the specific action and the time point of the Spike signal and the ECoG signal when the corresponding specific action occurs;

the two time points are corresponding, and the Spike signal and the ECoG signal are windowed according to the corresponding time points;

identifying and classifying Spike signals in the window, and separating Spike signals with different waveforms of each signal transmission channel, wherein the Spike signal of each waveform corresponds to one type of neuron release type;

counting the neuron release class number separated from each signal transmission channel by a time window of 6ms, and drawing a grating graph of neuron release conditions changing along with time;

based on the raster pattern, establishing a mapping relation between Spike signals and different actions;

carrying out data feature extraction and classification on the ECoG signal in the window;

and based on the extracted and classified data characteristics, establishing a mapping relation between the ECoG signal and different actions by using statistics.

Preferably, in the electroencephalogram signal acquisition method, before training and identifying the spectral features and the energy features of the Spike signal and the ECoG signal in different frequency bands by using a deep learning algorithm, the method further includes: and carrying out statistical analysis on the time frequency and frequency band energy of the Spike signal and the ECoG signal by using a correlation analysis algorithm, and representing different types of nerve signals in different action execution processes.

According to the technical scheme, compared with the prior art, the electroencephalogram signal acquisition method has the following beneficial effects that:

1. the invention realizes the simultaneous acquisition of the Spike signal and the ECoG signal, establishes the correlation between the two signals in a mode of analyzing a spectrogram and an energy chart by machine learning, and can realize the characteristic expression of the two signals to a motion plan and a motion process.

2. In the process of applying the collected signals to brain-computer interaction, the invention carries out feature extraction and correlation machine learning training on the two signals, and establishes the relation between the two signals based on the individual patient through long-time deep learning. Under the condition that spike electrode signals are weak and even cannot be acquired, ECoG signals after training can be used as main sources of control signals, the method is applied to external human-computer interaction, and the working time of high signal-to-noise ratio signal acquisition of the electroencephalogram electrode is prolonged.

3. On one hand, the invention searches the optimal ECoG signal wave band and time window by means of spike signal characteristic analysis, thereby improving the signal-to-noise ratio of the ECoG; on the other hand, because of the low signal-to-noise ratio of the ECoG signal, when the ECoG signal features are extracted, some slightly different signal features cannot be extracted.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.

FIG. 1 is a block diagram of an electroencephalogram signal acquisition device provided by the present invention;

FIG. 2 is a top view of a first signal acquisition electrode and a second signal acquisition electrode provided in accordance with the present invention;

FIG. 3 is a side view of a first signal acquisition electrode and a second signal acquisition electrode provided in accordance with the present invention;

FIG. 4 is a diagram of an application scenario of the electroencephalogram signal acquisition device provided by the present invention;

FIG. 5 is a flow chart of the electroencephalogram signal acquisition method provided by the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1-3, an embodiment of the present invention discloses an electroencephalogram signal acquisition apparatus, including: the device comprises a first signal acquisition electrode 1, a second signal acquisition electrode 2, a signal processor 3 and a wireless transmitting module 4; the first signal collecting electrode 1, the second signal collecting electrode 2 and the wireless transmitting module 4 are respectively electrically connected with the signal processor 3; the wireless transmitting module 4 is in wireless communication with the brain-computer interface;

the first signal acquisition electrode 1 is used for acquiring Spike signals; the second signal acquisition electrode 2 is used for acquiring an ECoG signal; the signal processor 3 is used for processing and converting the Spike signal and the ECoG signal; the wireless transmitting module 4 is used for transmitting the processed and converted Spike signal and the ECoG signal to the brain-computer interface; the brain-computer interface trains the Spike signals and the ECoG signals and establishes the correlation between the Spike signals and the ECoG signals under different actions of the receptor.

Wherein, the first signal collecting electrode 1 comprises a hydrogel substrate 11 and a plurality of contact electrodes 12; the electric contact electrodes 12 are arranged on the surface of the hydrogel substrate 11 in an array. The number of the contact electrodes 12 is 28, the diameter thereof is 1mm, and the center-to-center distance between the respective adjacent contact electrodes 12 is 3 mm. The impedance of each contact electrode 12 is controlled within 100 omega, and the ECoG signals outside the dura mater can be effectively recorded.

The second signal collecting electrode 2 includes a ceramic substrate 21 and a plurality of microwire electrodes 22; one end of the microwire electrode 22 is embedded in the ceramic substrate 21 and arranged in an array; the other end of the microwire electrode 22 is covered with silicone elastomer, and the tip thereof is exposed outside the silicone elastomer. The ceramic substrate 21 is laminated on the surface of the hydrogel substrate 11 and is close to the center of the hydrogel substrate 11; the contact electrode 12 is disposed around the ceramic substrate 21.

In this embodiment 16 microwire electrodes 22 are provided and arranged in a 4 x 4 array using platinum iridium wires, typically 20-25 μm in diameter, coated with silicone elastomer, leaving only the tips exposed for electrical conduction.

More advantageously, it also comprises at least one reference electrode 5; the reference electrode 5 is placed close to the border located on the hydrogel substrate 11.

An insulating layer is encapsulated in the hydrogel substrate 11; corresponding lines are arranged on the insulating layer; one end of the circuit is correspondingly connected with each contact electrode 12, each microwire electrode 22 and each reference electrode 5 one by one to form a plurality of signal transmission channels 6; the other end is connected with a data interface through a wire; the data interface is electrically connected with the signal processor 3

A layer of flexible polyimide material and copper foil are packaged in a hydrogel substrate 11, circuits are arranged on the flexible Polyimide (PI) insulating substrate material by utilizing a photoetching technology, and finally, all the circuits are respectively connected with all relevant electrodes and external leads to serve as signal transmission channels 6.

As shown in fig. 4, which is an application scene diagram of the electroencephalogram signal acquisition device of the present invention, the contact electrode 12 is absorbed by the negative pressure device on the skull above the motor cortex of the brain of the recipient, so as to ensure that the 4 × 4 microwire electrode 22 is partially vertically inserted into the dura mater, and the contact electrode 12 is tightly attached to the dura mater. The skull is restored by using a titanium alloy mesh, an electrode lead passage is reserved, external interfaces of each contact electrode 12 and the microwire electrode 22 are connected with an external signal processor 3 through external leads, and two electroencephalogram signals are collected in real time.

On one hand, spike signals and ECoG signals are directly processed through the signal processor 3 and are converted into control signals to carry out man-machine interaction with the outside; on the other hand, the wireless transmitting module 4 transmits the signals to the outside pc for signal processing and deep learning, the mapping relation of spike signals, ECoG signals and control signals is established through long-time training and big data calculation, and when the spike signals are weak or cannot be collected in the later period, the control signals required by outside interaction are provided by the ECoG signals.

As shown in fig. 5, an embodiment of the present invention further provides an electroencephalogram signal acquisition method, including the following steps:

and S1, acquiring the Spike signal and the ECoG signal, and respectively preprocessing the Spike signal and the ECoG signal.

The Spike signal and the ECoG signal are respectively collected and stored at a sampling rate of 30kHz for subsequent data analysis.

Meanwhile, the acquired Spike signal and the ECoG signal are preprocessed,

the spike signals (S01-S16) of 16 channels are collected and preprocessed, and low-frequency components are removed by using a 300Hz high-pass filter to eliminate the influence of field potential.

And carrying out filtering processing on the acquired ECoG signal by using a Kalman filtering mode.

And S2, respectively establishing a mapping relation between the Spike signals and different actions and a mapping relation between the ECoG signals and different actions.

S21, labeling specific actions and finding out specific time points;

s22, windowing the Spike signal and the ECoG signal according to the specific time point;

when each action occurs, a time point corresponds to the action, the generation of the signal also corresponds to a time point, the two time points are corresponded, and the process can be called as labeling; the electroencephalogram signals of a period of time are respectively intercepted before and after the time point, the signals in the period of time are analyzed, the windowing is carried out, the signals in the window are determined to be meaningful signals, and the signals outside the window are nonsense signals.

S23, identifying and classifying the Spike signals in the window, and separating the Spike signals of different waveforms of each signal transmission channel, wherein the Spike signal of each waveform corresponds to one type of neuron release type;

the identification standard of Spike signals is high potential, 0 is considered to be low potential, 1 is considered to be high potential, and when the amplitude of the Spike signals is larger than a certain threshold, the signals are considered to be in an excitation state and are effective characteristics; the classification standard is that the same spike signal is identified as being emitted by the same type of cell, looking at the waveform.

S24, counting the number of neuron release classes separated from 16 signal transmission channels in a time window of 6ms (3 ms before and after action), and drawing a grating graph of neuron release conditions changing along with time; used for observing Spike release characteristics in the process of movement planning and execution.

01 combinations of different classes of spike signals, for example: 0001010010101 and 000000000010 and 0000111110000 are different three types of signals.

Spike signals of each waveform are classified into one class, specific spike signal forms can be represented through class numbers and issuing feature numbers, each form is equivalent to one issuing feature, and different issuing features can correspond to different actions.

S25, establishing a mapping relation between the Spike signal and different actions based on the raster pattern;

s26, carrying out data feature extraction and classification on the ECoG signals in the window;

and S27, based on the extracted and classified data features, using statistics to establish a mapping relation between the ECoG signal and different actions.

And S3, training and identifying the spectrum characteristics and the energy characteristics of the Spike signals and the ECoG signals under different frequency bands by using a deep learning algorithm based on the mapping relationship between the Spike signals and different actions and the mapping relationship between the ECoG signals and different actions, and establishing the mapping relationship between the Spike signals and the ECoG signals under different actions.

The energy method can be used for analyzing, each signal has energy, the energy size and distribution of the signals are different, the energy method is used for carrying out statistical analysis on the time frequency and frequency band energy of the Spike signal and the ECoG signal, and different types of nerve signals in different action execution processes are characterized. And (3) analyzing the correlation between the two nerve signals in the receptor movement planning and executing process (1s) to find out the correlation and the correlation mode.

S4, gradually replacing the weak Spike signal with the ECoG signal.

Through the assistance of spike signals, an accurate feature extraction method of low-resolution ECoG signals can be generated, the low-resolution ECoG signals can be adapted, different signal features can be recognized, and then the signals are converted into control signals.

For example, when a person performs an action, the brain sends out control instructions, each control instruction corresponds to one electroencephalogram signal, each electroencephalogram signal is classified by analyzing the characteristics of the electroencephalogram signal, and the brain is known to command organs to complete the action when the brain meets the signal next time. Therefore, the electroencephalogram signals are used for controlling the mechanical arm or other auxiliary devices.

The control signals may be applied to the exoskeleton or to external tools to assist in performing each function or to cause some action.

And acquiring an ECoG signal and a spike signal, analyzing the two signals simultaneously, analyzing corresponding specific characteristics and action planning, and sending a control signal for carrying out the action when the ECoG signal is the same.

The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.