KR20160135107A - Device and method for denoising noise of electroencephalogram signal using series independent component analysis - Google Patents

Abstract

A method for removing an electroencephalography noise is disclosed. According to the disclosure, the method comprises the following steps: (a) receiving an electroencephalography (EEG) signal; (c) extracting independent components from the EEG signal using an independent component analysis (ICA) method; and (d) removing the k^th noise (2<=k<=n+1) from the extracted independent components, wherein the steps (c) and (d) may be repeated n times (n is natural number equal to or greater than two).

Description

TECHNICAL FIELD [0001] The present invention relates to an EEG noise canceling apparatus and an EEG noise canceling apparatus,

An embodiment according to the concept of the present invention relates to an EEG removing device, and more particularly, to an EEG removing method using an average noise subtraction technique and an independent component analysis technique, using a simultaneous EEG (electroencephalography) -fMRI (functional magnetic resonance imaging) And an EEG noise canceling method capable of removing noise included in an EEG signal.

Non-invasive measurement of brain signals and images is possible by using MEG (magnetoencephalography) device, PET (positron emission tomography) device, fNIRS (functional) device as well as EEG device and fMRI device Near-infrared spectroscopy, near-infrared spectroscopy, and the like. However, each device has advantages and disadvantages in spatial / temporal resolution. Therefore, multimodality methods such as the simultaneous EEG-fMRI measurement method or the simultaneous EEG-fNIRS measurement method are used as a method to overcome the disadvantages of each equipment.

The EEG device is a device that measures electrical EEG activity through the electrodes attached to the scalp. It is useful for measuring the brain response generated in real time because of its excellent temporal resolution. However, since the signal measured from the electrode is the sum of the signals of many cranial nerve cells, it is difficult to grasp the signal at the correct position. On the other hand, since the fMRI device measures the blood-oxygenation-lever-dependent (BOLD) signal according to the increase of the local blood flow volume at the brain functional active site, the spatial resolution is superior to that of the EEG device, Has a drawback that it is inferior to the EEG device. Therefore, the simultaneous EEG-fMRI measurement method can simultaneously measure the EEG signal and BOLD signal generated in the brain to utilize the temporal / spatial advantage of each device and to estimate reliable brain function area and characteristics.

Disclosure of Invention Technical Problem [8] Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to provide an EEG noise canceling apparatus and EEG noise canceling method capable of effectively removing noise included in an EEG signal measured through simultaneous EEG-fMRI measurement.

The method for removing EEG noise according to an embodiment of the present invention includes the steps of (a) receiving an electroencephalography (EEG) signal, (c) using an Independent Component Analysis (ICA) (D) extracting the kth noise (2? K? N + 1) from the extracted independent components, wherein steps (c) and (d) n is a natural number of 2 or more).

The apparatus for removing EEG noise according to an embodiment of the present invention includes a signal receiving module for receiving an EEG signal simultaneously measured with an fMRI signal and an independent component analyzing unit for extracting independent components from the EEG signal using an independent component analysis technique, (N is a natural number equal to or greater than 2) times, and a second noise cancellation module for removing the kth noise (2? K? N + 1) from the extracted independent components.

The apparatus for removing EEG noise according to the embodiment of the present invention has the effect of eliminating the noise included in the EEG-fMRI signal measured through the simultaneous EEG-fMRI measurement by repeatedly performing the independent component analysis technique a plurality of times.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to more fully understand the drawings recited in the detailed description of the present invention, a detailed description of each drawing is provided.
FIG. 1 shows an EEG eliminator according to an embodiment of the present invention.
FIG. 2 is a flowchart illustrating a method for removing EEG noise using the EEG removing apparatus shown in FIG. 1. Referring to FIG.

It is to be understood that the specific structural or functional description of embodiments of the present invention disclosed herein is for illustrative purposes only and is not intended to limit the scope of the inventive concept But may be embodied in many different forms and is not limited to the embodiments set forth herein.

The embodiments according to the concept of the present invention can make various changes and can take various forms, so that the embodiments are illustrated in the drawings and described in detail herein. It should be understood, however, that it is not intended to limit the embodiments according to the concepts of the present invention to the particular forms disclosed, but includes all modifications, equivalents, or alternatives falling within the spirit and scope of the invention.

The terms first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms may be named for the purpose of distinguishing one element from another, for example, without departing from the scope of the right according to the concept of the present invention, the first element may be referred to as a second element, The component may also be referred to as a first component.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between. Other expressions that describe the relationship between components, such as "between" and "between" or "neighboring to" and "directly adjacent to" should be interpreted as well.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises" or "having" and the like are used to specify that there are features, numbers, steps, operations, elements, parts or combinations thereof described herein, But do not preclude the presence or addition of one or more other features, integers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the meaning of the context in the relevant art and, unless explicitly defined herein, are to be interpreted as ideal or overly formal Do not.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings attached hereto.

FIG. 1 shows an EEG eliminator according to an embodiment of the present invention.

Referring to FIG. 1, the EEG removing apparatus 10 receives an EEG signal through a plurality of channels, and removes noise included in the EEG signal. At this time, the EEG signal can be received from the EEG device. At this time, the EEG signal may be a signal measured simultaneously with the fMRI signal through the simultaneous EEG-fMRI measurement method, but the scope of the present invention is not limited thereto. The EEG eliminator 10 may be implemented as a part of the EEG apparatus.

The brain wave noise removing apparatus 10 includes a signal receiving module 100, a first noise removing module 200, an independent component extracting module 300, a second noise removing module 400 and a control module 500.

Under the control of the control module 600, the signal receiving module 100 receives an EEG signal outputted from an EEG measuring device capable of measuring EEG. In addition, the signal receiving module 100 may receive an electrocardiogram signal output from an electrocardiogram measuring device capable of measuring an electrocardiogram (ECG) signal. According to an embodiment, the electrocardiogram signal may be measured by the EEG device and transmitted to the signal receiving module 100.

The first noise removing module 200 may remove the first noise included in the EEG signal under the control of the control module 500. Specifically, the first noise removing module 200 removes a first noise included in the EEG signal, such as magnetic resonance (MR) gradient noise, using an average artifact subtraction (AAS) can do. The average noise subtraction technique is known as a simple and effective method widely used to remove the MR gradient noise in the field of simultaneous EEG-fMRI measurement.

An MRI device that measures an fMRI signal produces a strong magnetic field change to obtain a brain image volume or a slice of volume. Due to such a change in the magnetic field, the MR gradient signal affects the EEG signal during the simultaneous EEG-fMRI measurement, and the signal is significantly larger than the biological signal. The average noise subtraction technique estimates an average value of a specific interval based on each volume or slice based on the strong reproducibility characteristic of the MR gradient changing for each volume or slice and includes MR gradient noise from the measured EEG signal And subtracts a signal having the EEG signal from the EEG signal to remove the MR gradient noise included in the EEG signal.

The independent component extraction module 300 can extract a plurality of independent components from the noise-canceled EEG signal using the independent component analysis technique (ICA) under the control of the control module 600. [ The independent component analysis technique is widely used in the field of EEG research, and it is assumed that the electrical signals measured in each channel are linearly related to the signal sources of the EEG. Independent component extraction module 300 may generate an unmixing matrix that minimizes mutual information between independent components to estimate the independent components corresponding to the signal sources. Independent components can be extracted from the EEG signal through this process.

Specifically, the independent component extraction module 300 uses the first noise, for example, the MR gradient as the input data, with the noise-canceled EEG signal X as the input data. At this time, the independent component analysis technique assumes that the input data X is a linear relationship between the signal source S and the mixing matrix A as shown in Equation 1 below.

The independent component extraction module 300 also includes a non-mixing matrix W that minimizes the amount of mutual information between the data outputs (the inverse of the non-mixing matrix and the input data) to extract the first independent components from the input data X, Can be generated. The non-mixing matrix W generated by the independent component extraction module 300 is in inverse relation with the mixing matrix A. [ The first independent components may be extracted through an inner product of the non-mixing matrix W and the input data X generated as shown in Equation 2 below.

)로부터 제2 독립 성분들을 추출할 수 있다. 상기 제2 독립 성분들을 추출하는 과정은 제1 독립 성분들을 추출하는 과정과 동일하므로 상세한 설명은 생략하기로 한다.In addition, the independent component extraction module 300 extracts the second noise-

) &Lt; / RTI &gt; The process of extracting the second independent components is the same as the process of extracting the first independent components, and a detailed description thereof will be omitted.

In this way, the independent component extraction module 300 can perform independent component analysis repeatedly n (n is a natural number of 2 or more) times.

The second noise removal module 400 may remove noise from the independent components extracted by the independent component extraction module 300 under the control of the control module 600. [ In particular, the second noise reduction module 400 may remove the second noise from the first independent components and may remove the third noise from the second independent components. That is, the second noise elimination module 400 may remove the (n + 1) -th noise from the n-th independent components.

The second noise may denote at least one of noise that greatly affects pupil motion noise and low frequency, for example, heartbeat noise. At this time, the low frequency may be a frequency belonging to 0.8 Hz to 1.2 Hz.

When the second noise removing module 400 desires to remove the corrugated noise, an ECG (electrocardiogram) signal can be utilized. That is, the second noise elimination module 400 may calculate the mutual information between the extracted independent components and remove the cardiac trough noise based on the mutual information value, which is relatively larger than other components.

)을 생성하고, 아래의 수학식 3을 이용하여 제2 노이즈가 제거된 뇌파 신호(

)를 추정할 수 있다.When the second noise removal module 400 desires to remove noise from the eye, a brain topography can be utilized and can be removed on the basis of components showing strong activation around the eye. At this time, the second noise elimination module 400 replaces the value of the column corresponding to the component related to the noise among the components of the mixing matrix A estimated from the generated non-mixing matrix W to 0,

), And the second noise-canceled EEG signal (

) Can be estimated.

)로부터 제3 노이즈를 제거할 수 있다. 상기 제3 노이즈는 잔여 MR 그레디언트 노이즈일 수 있다. In addition, the second noise elimination module 400 generates the second noise-canceled EEG signal (

The third noise can be removed. The third noise may be residual MR gradient noise.

Since the EEG signal measured during the simultaneous EEG-fMRI measurement is sensitive to the measurement target, that is, the small motion of the human body, the MR gradient noise for each section in which the measurement target moves has different characteristics. Accordingly, the average noise subtraction technique using the strong reproducibility feature of the MR gradient leaves residual MR gradient noise. The frequency of the residual MR gradient noise can be easily determined using the total number of acquired slices / one volume acquisition time (TR).

The control module 500 controls the overall operation of the brain-wave noise removing apparatus 10. That is, the control module 500 may control operations of the signal receiving module 100, the first noise removing module 200, the independent component extracting module 300, and the second noise removing module 400.

)을 통해 일회의 독립 성분 분석 기법을 적용하는 기존의 방법에 비하여 MR 그레디언트 노이즈, 심탄도 노이즈, 및 눈동자 움직임 노이즈 등의 잔여 노이즈까지 효과적으로 제거할 수 있다.As described above, according to the present invention, noise can be removed step by step through a repeated independent component analysis technique, and a new independent component estimation (n) is performed for each n (n is a natural number of 2 or more)

), Residual noise such as MR gradient noise, heart-ball noise, and pupil motion noise can be effectively removed as compared with the conventional method of applying a single independent component analysis technique.

Each of the configurations of the brain-wave noise removing apparatus 10 shown in FIG. 1 indicates that it can be divided into functions and logically, which means that each configuration is divided into separate physical devices or written in separate codes Or may be easily inferred by an average expert in the field of the present invention.

In this specification, a module may mean a functional and structural combination of hardware for carrying out the technical idea of the present invention and software for driving the hardware. For example, the module may mean a logical unit of a predetermined code and a hardware resource for executing the predetermined code, and does not necessarily mean a physically connected code or a kind of hardware.

FIG. 2 is a flowchart illustrating a method for removing EEG noise using the EEG removing apparatus shown in FIG. 1. Referring to FIG.

Referring to FIGS. 1 and 2, the signal receiving module 100 of the brain-wave noise removing apparatus 10 receives an EEG signal outputted from an EEG measuring device capable of measuring EEG (S110). At this time, the signal receiving module 100 may receive an electrocardiogram (ECG) signal, which is capable of measuring an ECG signal, or an electrocardiogram signal, output from the EEG device.

The first noise removing module 200 of the brain-wave noise removing apparatus 10 may remove the first noise included in the EEG signal, e.g., MR gradient noise, using an average noise subtraction technique (S130).

The independent component extraction module 300 of the brain-wave noise removing apparatus 10 may extract a plurality of first independent components from the first noise-canceled EEG signal using an independent component analysis technique (S150).

The second noise removing module 400 of the brain-wave noise removing apparatus 10 may remove the second noise from the extracted first independent components (S170). The second noise may denote at least one of noise that affects a low frequency, for example, heart ballistic noise and pupil movement noise.

The independent component extraction module 300 of the brain-wave noise removing apparatus 10 may extract the k-th independent components (2? K? N) from the second noise-removed brain wave signal using the independent component analysis technique (S180).

The second noise elimination module 400 of the brain-wave noise removing apparatus 10 may remove the (k + 1) -th noise from the extracted k-th independent components (S190). The k + 1 noise may be residual MR gradient noise.

In addition, since the steps S180 and S190 are repeatedly performed until the value of k becomes n, the noise can be removed in steps. That is, depending on whether there is a need for additional noise removal, the noise removal operation may be terminated or additional independent component extraction operations and additional noise removal operations may be performed.

In the above description, a method of performing noise analysis for a total of n + 1 times by performing independent component analysis n times is described, wherein n may be a natural number of 2 or more.

By repeating the independent component analysis in this manner, noise can be removed step by step, and new independent components can be extracted for each step. Therefore, according to the present invention, residual noise such as MR gradient noise, heart trajectory noise, pupil motion noise, and motion noise can be effectively removed as compared with the conventional one-time independent noise analysis method.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

10: EEG eliminator
100: Signal receiving module
200: First Noise Reduction Module
300: Independent component extraction module
400: Second Noise Reduction Module
500: control module

Claims (5)

(a) receiving an electroencephalography (EEG) signal and an electrocardiogram (ECG) signal;
(c) extracting independent components from the EEG signal using an Independent Component Analysis (ICA) technique;
(d) removing kth noise (2? k? n + 1) from the extracted independent components; And
Wherein the steps (c) to (d)

By
The independent component analysis is repeated n times (n is a natural number of 2 or more) so as to remove ballistocardiogram (BCG) noise, pupil motion noise, and residual magnetic resonance gradient noise, ) Extracts the independent components from the k-noise canceled EEG signal,
The residual MR gradient noise is determined based on a frequency value obtained by dividing the number of slices of acquired brain image volume by the volume acquisition time (repetition time)
The heartbeat noise is determined based on a value of mutual information between the ECG signal and extracted independent components,
Wherein the pupil movement noise is determined based on the degree of activation around the eye on brain topography,
A method for eliminating EEG noise.
here,

The n &lt; th &gt; noise canceled EEG signal,

Is an n-th mixing matrix,

Is an independent component estimation
The method according to claim 1,
The method of claim 1, further comprising, after step (a), (b) removing a first noise included in the EEG signal,
The method of claim 1, wherein removing the first noise comprises removing magnetic resonance (MR) gradient noise included in the EEG signal using an average artifact subtraction (AAS) technique.
3. The method of claim 2,
Wherein the EEG signal is measured simultaneously with a fMRI (functional magnetic resonance imaging) signal,
The step (d)
Replacing a value of a column corresponding to the kth noise among components of the mixing matrix with 0; And
Estimating an EEG signal from which the k &lt; th &gt; noise has been removed from a mixing matrix in which the value of the column is replaced by zero.
A method for eliminating EEG noise.
a signal receiving module for receiving an EEG signal and an electrocardiogram (ECG) signal simultaneously measured with an fMRI signal;
An independent component extraction module for performing an independent component analysis for extracting independent components from the EEG signal using an independent component analysis technique for n (n is a natural number of 2 or more) times; And
And a second noise elimination module for removing the kth noise (2 &lt; k &lt; = n + 1) from the extracted independent components,
The independent component extraction module

By
Wherein the independent component analysis is repeated n times (n is a natural number of 2 or more), extracting independent components from the k &lt; th &gt;
The second noise cancellation module removes the corrugation noise, the pupil motion noise, and the residual MR gradient noise step by step,
The residual MR gradient noise is determined based on a frequency value obtained by dividing the number of slices of the acquired brain image volume by the volume acquisition time,
The heartbeat noise is determined based on a value of mutual information between the ECG signal and extracted independent components,
Wherein the pupil motion noise is determined based on the degree of activation around the eyes on the brain topographical map,
EEG noise eliminator.
here,

The n &lt; th &gt; noise canceled EEG signal,

Is an n-th mixing matrix,

Is an independent component estimation
5. The method of claim 4,
Wherein the EEG removing device further comprises a first noise removing module for removing MR gradient noise included in the EEG signal using an average noise subtraction technique.

Images