KR102557024B1 - Device and Method for Multi-functional Brain Computer Interface(BCI) - Google Patents

Abstract

The present invention relates to a multifunctional brain computer interface device. The device includes an EEG measuring unit for measuring brain waves, a memory storing a program for generating a control signal associated with the device, and a processor for executing the program stored in the memory, but the processor measures data from the user according to the execution of the program. Based on the EEG signal, feature information data is extracted to include feature information of the EEG signal, the extracted data is input to a classifier based on an artificial neural network model, and a control signal for controlling the user's movement is generated based on the classified data. and output the generated control signal.

Description

Multifunctional brain computer interface device and method {Device and Method for Multi-functional Brain Computer Interface (BCI)}

The present invention relates to a multifunctional brain computer interface device and method, and more particularly, multi-intent feature information through a process of transforming an EEG signal into a power spectrum of a frequency band by Fourier transform, and normalizing each frequency value of the power spectrum. It relates to a multifunctional brain computer interface device and method including a component for extracting data, predicting multiple intentions, and classifying the data.

Brain-computer interface technology (BCI) is a technology that allows you to operate machines at will with only your thoughts.

Recently, many companies and institutions are investing in the importance of BCI technology and looking at its positive future prospects, which is why patients who cannot move their bodies due to nerve paralysis or amputation can express their opinions or move their bodies according to their thoughts through technology. This is because it is a study that not only allows people to move or be active, but also provides an interface to express their opinions.

There are many ways to implement BCI, and there are many ways to implement BMI using the method used at the beginning of the study. As a method used in the early days of BMI research, there is a slow cortical potentials method. By using the phenomenon in which the potential of the brain wave becomes negative due to attention or concentration and the phenomenon in which the potential has a positive value when it is not, the phenomenon in which the potential slowly becomes positive or negative in one-dimensional tasks such as top and bottom method to use. At that time, it was a groundbreaking way to control a computer with thoughts, but it is rarely used now because it is slow to respond and high-dimensional distinction is impossible.

As another method, a method using sensorimotor rhythms is being actively researched. This method uses the phenomenon in which μ waves (8-12 Hz) or β-waves (13-30 Hz) increase or decrease according to the response in the primary sensorimotor cortex, and is generally used to distinguish between left and right two cases. has been

A research group in Berlin, Germany succeeded in controlling the mouse cursor with a success rate of 70 to 80% by using this method of increasing or decreasing sensorimotor waves (Benjamin Blankertz et al., 2008).

However, research using only a single feature of a brain signal has limitations in that it can only perform a single task. In real life, there are many cases in which multiple tasks are performed, such as typing while walking or selecting something while holding a cup. Therefore, a BCI algorithm that can simultaneously perform multiple tasks by utilizing various characteristics of brain signals is needed.

BCI researchers in China showed their research results on a hybrid BCI that uses the features of SSVEP and P300 to reach the target point by moving the mouse cursor.

(T. Yu et al, 2015). A study was also conducted to provide feedback on the training progress through SSVEP pattern extraction for the MI task, which the subject had a relatively difficult time using, by utilizing the features of MI and SSVEP (Steady-State Visually Evoked Potential).

As such, studies that utilize different features such as P300 or SSVEP to improve accuracy or performance compared to using a single feature or to help perform other tasks using one feature have been conducted, but multi-intent features There was no BCI study that could be extracted and multi-controlled at the same time.

[Patent Document 1] Korean Patent Publication No. 10-2019-0097146. 2019.08.20. open.

The present invention was created to solve the above problems, and an object of the present invention is to provide a multifunctional brain computer interface device and method.

The objects of the present invention are not limited to the objects mentioned above, and other objects not mentioned will be clearly understood from the description below.

In order to achieve the above objects, a multifunctional brain computer interface device according to an embodiment of the present invention is disclosed. The device includes an EEG measurement unit for measuring brain waves, a memory in which the device stores a program for generating a control signal associated with the device, and a processor for executing the program stored in the memory. Based on the measured EEG signal, extract feature information data to include multi-intent feature information of the EEG signal, input the extracted data to an artificial neural network model, and generate control signals corresponding to multiple intents through output values, It is possible to output the generated control signal.

In addition, according to an embodiment of the present invention, the processor includes a pre-processing unit that removes noise from the measured EEG signal, and a multi-intent feature extraction unit that extracts feature information data including the user's multiple intentions from the noise-removed EEG signal. , a multi-intent classification unit that classifies multiple intentions through multi-intent output values obtained by inputting the extracted data to an artificial neural network model, and a control signal generation unit that generates control signals corresponding to the classified multi-intents.

In addition, according to one embodiment of the present invention, the pre-processing unit includes an epoch that cuts a certain section of the measured EEG signal by a set time and a filter for filtering noise of the epoched EEG signal, and the filter is low-pass It may be any one of a filter, a high pass filter, a band pass filter, and a notch filter.

In addition, according to an embodiment of the present invention, the multi-intent feature extraction unit transforms the preprocessed EEG signals into a power spectrum of a frequency band by performing a fast Fourier transform (FFT), and normalizes each frequency value of the power spectrum. Multi-intent feature information data can be extracted using the resulting value.

In addition, according to an embodiment of the present invention, the characteristic information data is the first case in which the result value obtained through normalization determined by the processor is greater than or equal to a preset frequency value or the change in the slope of the result value obtained through normalization is greater than or equal to the preset slope value. It may be data of a section including the second case.

Also, according to an embodiment of the present invention, normalization may be performed by subtracting the average of each frequency of the power spectrum in a resting state from each frequency value of the power spectrum and then dividing by the standard deviation.

In addition, according to an embodiment of the present invention, the processor may be characterized in that it simultaneously outputs at least one or more output values obtained by inputting at least one feature information data to an artificial neural network model.

In addition, a multifunctional brain computer interface method according to an embodiment of the present invention is disclosed. The method includes an EEG measurement step of measuring an EEG, a preprocessing step of removing noise from the measured EEG signal, a multi-intent feature extraction step of extracting feature information data including multiple user intentions from the noise-removed EEG signal, and an extracted EEG signal. The multi-intent classification step of classifying the user's multiple intentions through the output values obtained by inputting data to the artificial neural network model, the control signal generation step of generating control signals corresponding to the classified multiple intentions, and the step of outputting the generated control signals. can include

In addition, according to one embodiment of the present invention, the preprocessing step includes an epoching step of cutting a certain section of the measured EEG signal by a set time and a filter step of filtering noise of the epoched EEG signals, and the filter is low-pass It may be any one of a filter, a high pass filter, a band pass filter, and a notch filter.

In addition, according to an embodiment of the present invention, in the multi-intent feature extraction step, the preprocessed EEG signals are converted into a power spectrum of a frequency band by fast Fourier transform (FFT), and each frequency value of the power spectrum is normalized. It may be a step of extracting multi-intent feature information data using the resulting value.

In addition, according to an embodiment of the present invention, the characteristic information data is the first case in which the result value obtained through normalization determined by the processor is greater than or equal to a preset frequency value or the change in the slope of the result value obtained through normalization is greater than or equal to the preset slope value. It may be data of a section including the second case.

Also, according to an embodiment of the present invention, normalization may be performed by subtracting the average of each frequency of the power spectrum in a resting state from each frequency value of the power spectrum and then dividing by the standard deviation.

Further, according to an embodiment of the present invention, the outputting may be a step characterized by simultaneously outputting at least one or more output values obtained by inputting at least one feature information data to an artificial neural network model.

In addition, according to one embodiment of the present invention, a computer-readable recording medium recording a program for performing a multi-function brain computer interface method on a computer is provided.

Specific details for achieving the above objects will become clear with reference to embodiments to be described later in detail in conjunction with the accompanying drawings.

However, the present invention is not limited to the embodiments disclosed below, and may be configured in a variety of different forms, so that the disclosure of the present invention is complete and those of ordinary skill in the art to which the present invention belongs ( It is provided hereafter to fully inform the "ordinary skilled person") of the scope of the invention.

According to one embodiment of the present invention, the multifunctional brain computer interface device can simultaneously output a plurality of characteristic information data and perform various tasks at the same time.

The effects of the present invention are not limited to the above-mentioned effects, and the potential effects expected by the technical features of the present invention will be clearly understood from the description below.

In order that the above-mentioned features of the present invention may be understood in detail, more specifically, with reference to the following embodiments, some of the embodiments are shown in the accompanying drawings. Also, like reference numbers in the drawings are intended to refer to the same or similar function throughout the various aspects. However, it should be noted that the accompanying drawings only show specific exemplary embodiments of the present invention, and are not considered to limit the scope of the present invention, and other embodiments having the same effect may be fully appreciated. let it do
1 is a block diagram schematically showing a multifunctional brain computer interface device according to an embodiment of the present invention.
Figure 2 is a block diagram schematically showing the components of the processor of the multifunction brain computer interface device according to an embodiment of the present invention.
3 is a diagram schematically illustrating a signal processing process of a processor of a multifunctional brain computer interface device according to an embodiment of the present invention.
4 is a diagram illustrating a normalization process of a multifunctional brain computer interface device according to an embodiment of the present invention.
5 is a flowchart of a multifunctional brain computer interface method according to an embodiment of the present invention.

Since the present invention can make various changes and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail.

Various features of the invention disclosed in the claims may be better understood in consideration of the drawings and detailed description. Devices, methods, manufacturing methods, and various embodiments disclosed in the specification are provided for illustrative purposes. The disclosed structural and functional features are intended to enable a person skilled in the art to specifically implement various embodiments, and are not intended to limit the scope of the invention. The disclosed terms and phrases are intended to provide an easy-to-understand description of the various features of the disclosed invention, and are not intended to limit the scope of the invention.

In describing the present invention, if it is determined that a detailed description of related known technologies may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

The "multifunctional brain computer interface device" referred to below may be implemented as a computer or portable terminal capable of accessing a server or other terminals through a network. Here, the computer includes, for example, a laptop, desktop, laptop, etc. equipped with a web browser, and the portable terminal is, for example, a wireless communication device that ensures portability and mobility. , IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet), LTE (Long Term Evolution) communication-based terminal, smart All types of handheld-based wireless communication devices such as phones and tablet PCs may be included. In addition, “network” means a wired network such as a Local Area Network (LAN), a Wide Area Network (WAN) or a Value Added Network (VAN), a mobile radio communication network, or a satellite It can be implemented in all types of wireless networks such as communication networks.

Hereinafter, a multifunctional brain computer interface device and method according to an embodiment of the present invention will be described.

1 is a block diagram schematically showing a multifunctional brain computer interface device according to an embodiment of the present invention.

Referring to FIG. 1 , a multifunctional brain computer interface device 10 may include an EEG measuring unit 100 , a memory 200 and a processor 300 .

In one embodiment, the brain wave measuring unit 100 may be attached to a part of the user's body under the control of the processor 300 to measure the user's electroencephalogram (EEG). At this time, the brain wave measurement unit 100 measures brain waves from a plurality of channels that are in contact with a plurality of brain regions of the user. The EEG measurement unit 100 may be any one of equipment for measuring brain signals such as EEG, MEG, ECoG, and NIRS.

In one embodiment, the memory 200 may store a program for generating a control signal for controlling the operation of a robot arm device (not shown) associated with the multifunction brain computer interface device 10 . In this case, the memory 200 may collectively refer to a non-volatile storage device that maintains stored information even when power is not supplied and a volatile storage device that requires power to maintain stored information.

Figure 2 is a block diagram schematically showing the components of the processor of the multifunction brain computer interface device according to an embodiment of the present invention. 3 is a diagram schematically illustrating a signal processing process of a processor of a multifunctional brain computer interface device according to an embodiment of the present invention, and FIG. 4 is a normalization of a multifunctional brain computer interface device according to an embodiment of the present invention. A diagram showing the process.

Referring to Figures 2 to 4, the processor 300 includes one or more components for controlling the overall operation of the multi-function brain computer interface device (10). In particular, the processor 300 may include at least one processor that reads a program stored in the memory 200 and executes the program.

In addition, the processor 300 can perform various functions according to the execution of programs stored in the memory 200, and detailed modules included in the processor 300 according to each function include the pre-processing unit 310, multi-intent feature extraction It can be composed of a unit 320, a multi-intention classification unit 330, and a control signal generator 340.

That is, the processor 300 extracts feature information data to include multi-intent feature information of the EEG signals measured for each region, that is, for each channel, based on the EEG signals measured for each measurement region from the user, and extracts the extracted data. Control signals corresponding to multiple intentions may be generated through output values input to the artificial neural network model, and the generated control signals may be output.

For example, assuming a situation in which you want to move from place to place by controlling an electric wheelchair and type text on a computer at the same time, the measured EEG signal includes multiple intentions of moving to place and typing. Movement-related EEG features and typing-related EEG features are simultaneously extracted. It is possible to classify the user's multiple intentions through the classification of the output values obtained by inputting such multi-intent characteristic information data to the artificial neural network model. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto.

More specifically, the preprocessor 310 may perform signal processing corresponding to the EEG measured by the EEG measurement unit 100 . That is, the preprocessor 310 may remove noise from the EEG signal measured by the EEG measurement unit 310 . For example, the preprocessor 310 may perform noise filtering, frequency filtering, Laplacian spatial filtering, and the like on the measured EEG.

Specifically, the pre-processing unit 310 may include epoching to cut a certain section of the measured EEG signal for a set period of time and a filter to filter out noise from the epoched EEG signals. In addition, epoching may serve to cut a signal of a necessary section among EEG signals continuously measured in real time through the EEG measuring unit 100 . Epoching is for processing EEG signal data continuously measured in real time, and is to divide a certain section by a designated time. For example, an EEG signal may be epoched at 0.1 second intervals by 1 second size. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto. The preprocessing process such as epoching may be omitted or added.

In addition, the pre-processing unit 310 may include a filter for removing noise, and the filter may be any one of a low-pass filter, a high-pass filter, a band-pass filter, and a notch filter. In addition, the preprocessor 310 may include Independent Component Analysis (ICA) or Principal Component Analysis (PCA) for dimensionality reduction. For example, in order to remove noise caused by power supply of 60 Hz around the measurement environment, only a low-pass frequency band (1 to 50 Hz) may be filtered using a low pass filter. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto. The preprocessing process such as epoching may be omitted or added.

In addition, the multi-intention feature extraction unit 320 may extract feature information data including the user's multiple intentions from the EEG signal from which noise has been removed through the pre-processing unit 310 . The multi-intent feature extraction unit 320 performs at least a Fourier Transform, Power Spectrum (PS), Down Sampling, and Normalization process to extract a feature reflecting the user's intention. may contain one or more.

In addition, the feature of the output extracted through the multi-intended feature extraction unit 320 may be a normalized power spectrum value for each electrode or a down-sampled EEG signal. In addition, by obtaining the standard deviation of the power spectrum, only values greater than n times the standard deviation may be used as features.

Specifically, the multi-intent feature extractor 320 converts the EEG signals preprocessed through the preprocessor 310 into a power spectrum of a frequency band by fast Fourier transform (FFT), and converts the power spectrum into a power spectrum. Feature information data can be extracted using result values obtained by normalizing each frequency value of (PS). The brain waves of subjects who saw visual stimuli flickering at a specific frequency have high power at the same frequency as the corresponding frequency.

Here, the characteristic information data is the first case in which the result value obtained through the normalization process determined by the processor 300 is equal to or greater than a preset frequency value, or the first case in which the change in the gradient of the result value obtained through normalization is equal to or greater than the preset gradient value. It may mean data of a section including 2 cases.

In addition, the processor 300 may determine a case where a result value obtained through a normalization process is equal to or greater than a predetermined frequency value as feature information data. Alternatively, the processor 300 may determine a case where the variation in the gradient of the resultant value obtained through the normalization process is equal to or greater than a predetermined gradient value as feature information data.

In addition, normalization may be a process of subtracting an average of each frequency of the power spectrum PS in a rest state from each frequency value of the power spectrum PS, and then dividing the average by the standard deviation.

For example, it may be several Hz to several tens or hundreds of Hz to reflect the characteristics of the brain wave in the power spectrum (PS) section. For downsampling, an interval of 20 ms can be used, but an interval of several ms to hundreds of ms can be used depending on the person or situation. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto.

In one embodiment, the multi-intent classification unit 330 may classify the multi-intent through an output value obtained by inputting data extracted from the multi-intent feature extraction unit 320 to an artificial neural network model. In addition, the multi-intention classifier 330 may classify using artificial neural networks or deep learning models, and support vector machines, regression, or Kalman filters. ) can also be used to classify multiple intents.

In one embodiment, the control signal generation unit 340 may generate a control signal corresponding to the feature information data classified by the multi-intent classification unit 330 .

In addition, the processor 300 may be characterized in simultaneously outputting at least one or more output values obtained by inputting at least one feature information data to an artificial neural network model. For example, by simultaneously outputting a plurality of output values, it is possible to simultaneously output a plurality of control signals, such as moving the right arm to the right and grabbing an object with the left arm. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto.

5 is a flowchart of a multifunctional brain computer interface method according to an embodiment of the present invention.

Referring to FIG. 5, the multifunctional brain computer interface method (S500) includes an EEG measurement step (S510) of measuring an EEG signal, a preprocessing step (S520) of removing noise from the measured EEG signal, and a user's user's information from the noise-removed EEG signal. A multi-intent feature extraction step (S530) of extracting feature information data including multiple intents, a multi-intent data classification step (S540) of classifying multiple intents through the output values obtained by inputting the extracted data to the artificial neural network model (S540), It may include generating a control signal corresponding to multiple intentions (S550) and outputting the generated control signal (S560).

In one embodiment, the brain wave measuring step (S510) may be a step of measuring the user's brain wave (electroencephalogram, EEG) by attaching the brain wave measuring unit 100 to a part of the user's body under the control of the processor 300. there is.

In one embodiment, the pre-processing step (S520) may be a step of removing noise from the EEG signal measured in the EEG measurement step (S510).

In addition, the preprocessing step (S520) may include an epoching step of cutting a certain section of the measured EEG signal for a set time and a filter step of filtering noise of the epoched EEG signals, and the filter is a low-pass filter or a high-pass filter. It may be any one of a filter, a band pass filter, and a notch filter.

In addition, the epoching step may serve to cut a signal of a necessary section among EEG signals continuously measured in real time through the EEG measuring step (S510). The epoching step is for processing EEG signal data continuously measured in real time, and is to divide a certain section by a designated time. A preprocessing process such as an epoching step may be omitted or added.

In addition, the preprocessing step (S520) may include a filter for noise removal, and the filter may be any one of a low pass filter, a high pass filter, a band pass filter, and a notch filter. In addition, the preprocessing step (S520) may include Independent Component Analysis (ICA) or Principal Component Analysis (PCA) for dimensionality reduction. For example, in order to remove noise caused by power supply of 60 Hz around the measurement environment, only a low-pass frequency band (1 to 50 Hz) may be filtered using a low pass filter. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto. The preprocessing process such as epoching may be omitted or added.

In an embodiment, the multi-intent feature extraction step (S530) may be a step of extracting feature information data including the user's multiple intents from the EEG signal from which noise has been removed through the pre-processing step (S520).

In addition, in the multiple intended feature extraction step (S530), the EEG signal preprocessed in the preprocessing step (S520) is converted into a power spectrum of a frequency band by fast Fourier transform (FFT), and each frequency value of the power spectrum is normalized. It may be a step of extracting feature information data using the resulting value.

Here, the characteristic information data is the first case in which the result value obtained through the normalization process determined by the processor 300 is greater than or equal to a preset frequency value, or the second case in which the change in the slope of the result value obtained through normalization is greater than or equal to the preset slope value. It may mean data of a section including a case.

In addition, normalization may be a process of subtracting an average of each frequency of the power spectrum PS in a rest state from each frequency value of the power spectrum PS, and then dividing the average by the standard deviation.

For example, it may be several Hz to several tens or hundreds of Hz to reflect the characteristics of the brain wave in the power spectrum (PS) section. For downsampling, an interval of 20 ms can be used, but an interval of several ms to hundreds of ms can be used depending on the person or situation. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto.

In an embodiment, the multi-intent classification step (S540) may classify multiple intents through an output value obtained by inputting the data extracted in the multi-intent feature extraction step (S530) to an artificial neural network model.

In one embodiment, the control signal generation step (S550) may be a step of generating control signals corresponding to multiple intentions in the data classification step (S540).

In one embodiment, the step of outputting the control signal (S560) may be a step of outputting the control signal generated in the control signal generating step (S550). In addition, the step of outputting the control signal (S560) is a step characterized in that at least one output value obtained by inputting at least one feature information data generated in the control signal generation step (S550) to the artificial neural network model is output at the same time. can For example, by simultaneously outputting a plurality of output values, it is possible to simultaneously output a plurality of control signals, such as moving the right arm to the right and grabbing an object with the left arm. The above example is only an example for explaining the present disclosure, and the present disclosure is not limited thereto.

The multifunction brain computer interface method according to an embodiment of the present invention may be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer readable media may include computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

Although the methods and apparatus of the present invention have been described with reference to specific embodiments, some or all of their components or operations may be implemented using a computer system having a general-purpose hardware architecture.

The above description is only illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention.

Therefore, the embodiments disclosed in this specification are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the present invention is not limited by these embodiments.

The protection scope of the present invention should be interpreted according to the claims, and all technical ideas within the equivalent range should be understood to be included in the scope of the present invention.

10: multifunctional brain computer interface device
100: EEG measuring unit
200: memory
300: processor
310: pre-processing unit
320: multi-intent feature extraction unit
330: multi-intent classifier
340: control signal generator

Claims (14)

In the multifunctional brain computer interface device,
an EEG measurement unit measuring brain waves from a plurality of channels contacting a plurality of brain regions of the user when a user desires a plurality of tasks at the same time;
a memory storing a program for generating a control signal associated with the device by the device; and
Including a processor that executes the program stored in the memory,
According to the execution of the program, the processor,
Feature information data including all multi-intent feature information respectively corresponding to the plurality of tasks is simultaneously extracted from the EEG signal of the same time zone measured for each channel, the extracted data is input to an artificial neural network model, and the artificial neural network Generating a plurality of control signals corresponding to each output value through a plurality of output values corresponding to multiple intentions simultaneously output from the model, and outputting the generated plurality of control signals simultaneously,
The multiple intents respectively correspond to the plurality of tasks.
Multifunctional brain computer interface device.
According to claim 1,
the processor,
a pre-processor to remove noise from the measured EEG signal;
a multi-intent feature extractor extracting feature information data including multiple intentions of the user from the noise-removed brainwave signal;
a multi-intent classification unit that classifies multiple intentions through multi-intent output values obtained by inputting the extracted data to an artificial neural network model; and
Including a control signal generator for generating a control signal corresponding to the classified multiple intentions,
Multifunctional brain computer interface device.
According to claim 2,
The pre-processing unit,
Epoching (epoching) cutting a certain section of the measured EEG signal by a set time; and
A filter for filtering noise of the epoched EEG signals,
The filter is any one of a low pass filter, a high pass filter, a band pass filter and a notch filter,
Multifunctional brain computer interface device.
According to claim 2,
The multi-intent feature extraction unit,
Converting the preprocessed EEG signals into a power spectrum of a frequency band by fast Fourier transform (FFT), and extracting multi-intent feature information data using a result value obtained by normalizing each frequency value of the power spectrum ,
Multifunctional brain computer interface device.
According to claim 4,
The feature information data,
The first case in which the result value obtained through the normalization determined by the processor is greater than or equal to a preset frequency value or the second case in which the change in the slope of the result value obtained through the normalization is greater than or equal to a preset slope value Data of an interval including,
Multifunctional brain computer interface device.
According to claim 4,
The normalization is
Subtracting the average of each frequency of the power spectrum in a resting state from each frequency value of the power spectrum and then dividing by the standard deviation,
Multifunctional brain computer interface device.
delete A multifunctional brain computer interface method performed by a multifunctional brain computer interface device,
an EEG measuring step of measuring EEG from a plurality of channels contacting a plurality of brain regions of the user when the user desires a plurality of tasks at the same time;
A pre-processing step of removing noise from the EEG signal of the same time zone measured for each channel by a processor;
a multi-intent feature extraction step of simultaneously extracting, by the processor, feature information data including all of the multi-intent feature information respectively corresponding to the plurality of tasks from the noise-removed brain wave signal;
a multi-intention classification step in which the processor inputs the extracted data to an artificial neural network model and classifies multiple intentions of a user corresponding to each output value through a plurality of output values corresponding to multiple intentions simultaneously output from the artificial neural network model;
a control signal generation step of generating, by the processor, a plurality of control signals respectively corresponding to the classified multiple intentions; and
Including the step of the processor outputting the plurality of control signals generated at the same time,
The multiple intentions respectively correspond to the plurality of tasks.
A multifunctional brain computer interface method.
According to claim 8,
In the preprocessing step,
an epoching step in which the processor cuts a certain section of the measured EEG signal for a set time; and
A filter step of filtering noise of the epoched EEG signals by the processor;
The filter is any one of a low pass filter, a high pass filter, a band pass filter and a notch filter,
A multifunctional brain computer interface method.
According to claim 8,
The multiple intention feature extraction step,
The processor transforms the preprocessed EEG signals into a power spectrum of a frequency band by performing a fast Fourier transform (FFT), and normalizes each frequency value of the power spectrum to use multi-intention feature information data Which is the step of extracting
A multifunctional brain computer interface method.
According to claim 10,
The feature information data,
The first case in which the result value obtained through the normalization determined by the processor is greater than or equal to a preset frequency value or the second case in which the change in the slope of the result value obtained through the normalization is greater than or equal to a preset slope value Data of an interval including,
A multifunctional brain computer interface method.
According to claim 10,
The normalization is
Subtracting the average of each frequency of the power spectrum in a resting state from each frequency value of the power spectrum and then dividing by the standard deviation,
A multifunctional brain computer interface method.
delete A computer-readable recording medium recording a program for performing the method according to any one of claims 8 to 12 on a computer.

Images