EMD를 이용한 EEG 기반 움직임 상상 분류

이다빛(David Lee),김재호(Jae-Ho Kim),정우혁(Woo-Hyuk Jung),이희재(Hee-Jae Lee),이상국(Sang-Goog Lee) 한국HCI학회 2014 한국HCI학회 학술대회 Vol.2014 No.2

Brain-Computer interfaces(BCIs)에서 Electroencephalogram(EEG)의 특징을 추출하는 것은 중요하다. 일반적으로 EEG의 특징 추출 방법으로는 Fast Fourier transform(FFT)과 Wavelet transform(WT)이 많이 사용되었다. 하지만 이러한 방법들은 신호가 linear하고 stationary 하다는 가정 하에 적용되었기 때문에 신호 분해시 신호의 왜곡이 생길 수 있다. 이에 본 논문은 움직임 상상 EEG 분류를 위해 Empirical Mode Decomposition(EMD)과 FFT를 이용하는 특징을 제안했다. 먼저 움직임 상상 EEG에 EMD를 적용하여 Implicit Mode Functions(IMF)를 추출 뒤, 추출된 IMFs에 FFT를 적용하여 해당 IMF의 주파수 성분을 확인하였다. 주파수 성분이 μ 대역을 포함하고 있는 IMF의 표준편차를 특징으로 사용하였다. 추출된 특징을 Support Vector Machine(SVM)의 입력으로 사용하였고 샘플의 검증을 위해 10-fold cross validation을 이용하였다. 제안하는 방법은 움직임 상상 EEG에 대해 84.50%의 분류 정확도를 보여주었다. Feature extraction of Electroencephalogram (EEG) is an important issue in brain-computer interfaces(BCIs). The most commonly used methods for feature extraction from EEGs is Fast Fourier transform(FFT) and Wavelet transform(WT). However, when signal decomposition is carried out , these methods can happens distortion of the signal because it assumes that the signal is linear and stationary. In this paper, we proposed to use Empirical Mode Decomposition(EMD) and FFT to feature for classification of movement imagery EEGs. The EMD was applied to generate Implicit Mode Functions(IMFs) from the movement imagery EEGs. The FFT was then used to identify frequency component of each IMF at generated IMFs. The standard deviation of IMF included mu rhythm was used as feature. In the classification process, we used the extracted feature as input of Support Vector Machine(SVM) and 10-fold cross-validation to verification of sample. Under the proposed method, the classification accuracy of movement imagery EEGs was found to be 84.50%.

뇌파 관련 국내 한의학 연구에 대한 고찰

변혁,이진호,정찬영,김은정,이재동,최도영,김갑성,이승덕 대한침구의학회 2010 대한침구의학회지 Vol.27 No.1

Objectives : To research the changes of electroencephalogram(EEG) signals for acupuncture stimulation and to establish the hereafter direction for the study on EEG. Methods : We reviewed the domestic papers searched by search engine of Korean Acupuncture & Moxibustion Society and Korea Institute of Oriental Medicine. Results : We have searched 31 articles in 10 journals. The 13 articles were concerned with acupuncture. 1. All articles were published after 2001. In 2007 there were 10 articles. 2. The studies dealing with the changes of EEG signals were 24, the studies dealing with correlation of EEG signals were 5, and the studies analyzing EEG with Korean medicine were 2. 3. In the studies dealing with the changes of EEG signals, the case-control studies were 9, the non case-control studies were 14, and the case study was 1. 10 studies used electro-acupuncture, 1 study used herbal acupuncture, and 2 studies used manual acupuncture. Conclusions : We need more various kinds of studies. 1. Excited condition by acupuncture stimulation may reduce α wave. 2. There may be the acupuncture point-specific variation of EEG signal patterns. 3. The number of responding channels for acupuncture stimulation may correlate with the quantity or variety of acupuncture effect.

닫힌 눈 (eye-closed) EEG신호를 이용한 높은 비율BCI 맞춤법 시스템

김종진,웬충하오,양다린,정완영 한국융합신호처리학회 2017 융합신호처리학회 논문지 (JISPS) Vol.18 No.2

This study aims to develop an BCI speller utilizing eye-closed and double-blinking EEG based on asynchronous mechanism. The proposed system comprised a signal processing module and a graphical user interface (virtual keyboard-VK) with 26 English characters plus a special symbol. A detected “eye-closed” event induces the “select” command, whereas a “double-blinking” (DB) event functions the “undo” command. A three-class support vector machine (SVM) classifier involving EEG signal analysis of three groups of events (“eye-open”-idle state, “eye-closed”, and “double -blinking”) is proposed. The results showed that the proposed BCI could achieve an overall accuracy of 92.6% and a spelling rate of 5 letters/min on average. Overall, this study showed an improvement of accuracy and the spelling rate resulting from in the feasibility and reliability of implementing a real-world BCI spelle 이 연구는 비동기 매커니즘을 바탕으로 닫힌 눈(eye-closed) 및 이중 블링크 (double-blinking) EEG를 사용하여 BCI를 개발하는 것을 목표한다. 제안된 시스템은 신호 처리 모듈과 그래픽 사용자 인터페이스 (VK- 가상 키보드)로 구성되어 있으며 26개의 영문자와 특수 기호로 구성됩니다. “눈 닫기”이벤트는 “선택”(select)명 령을 유발하는 반면, “이중 블링크”(DB) 이벤트는 “실행 취소”(undo) 명령에 따라 실행합니다. 3개의 이벤트 그 룹 (“열린 눈”(eye-open, “닫힌 눈” (eye-closed)및 “이중 블링크”(double-blinking)에 대한 EEG 신호 분석과 관 련된 3 등급 벡터 보조 분류 (SVM) 기계가 제안되었습니다. 결과는 제안된 BCI가 평균 92.6 %의 전체 정확도 와 5 글자 / 분의 맞춤법 비율을 달성 할 수 있음을 보여주었습니다. 전반적으로 이 연구는 실제 BCI 맞춤법을 구현하기의 실현 가능성과 신뢰성으로 인해 정확도와 철자 비율의 향상을 보여주었습니다

Relative Wavelet Energy and Wavelet Entropy Based Epileptic Brain Signals Classification

Yatindra Kumar,Mohan Lal Dewal,Radhey Shyam Anand 대한의용생체공학회 2012 Biomedical Engineering Letters (BMEL) Vol.2 No.3

Purpose Manual analysis of EEG signals by an expert is very much time consuming due to the long length of EEG recordings. The suitable computerized analysis is essentially required to differentiate among the normal, interictal and ictal (epileptic) EEGs. Methods In the present work the EEG signals are decomposed into different sub-bands using discrete wavelet transform (DWT) to obtain the detail and the approximation wavelet coefficients. The coefficients are used to calculate the quantitative values of relative wavelet energy and wavelet entropy from different data sets to select the features of EEG signals. The support vector machine (SVM), feed forward back- propagation neural network (FFBPNN), k-Nearest Neighbor Classifier (k-NN) and Decision tree classifier (DT) are used to classify the EEG signals. Results It is revealed that the accuracy between normal subjects with eyes open condition (data set A) epileptic data set E using SVM is obtained as 96.25%. Classification accuracy between the normal subjects with eye closed condition and epileptic data set E is obtained as 83.75%using k-NN classifier. Similar accuracies while discriminating the interictal data set C versus ictal data set E, and interictal data set D versus ictal data set E are obtained as 97.5% and 97.5% respectively, using a FFBPNN. These accuracies are quite higher than the earlier results published. The results are discussed quite in detail towards the last sections of the present paper. Conclusions Our experimental results demonstrate that the proposed method gives quite high statistical parameters for EEG classifications especially to classify the interictal data(C, D) and ictal data (E). These experiments indicate that the present method can be useful in analyzing and detecting the EEG signal associated with epilepsy.

Effect of temperature on attention ability based on electroencephalogram measurements

Choi, Yoorim,Kim, Minjung,Chun, Chungyoon Elsevier 2019 Building and environment Vol.147 No.-

<P><B>Abstract</B></P> <P>In this study, subjects' attention abilities in seven predicted mean vote (PMV) conditions (−3 to +3) were measured using electroencephalograms (EEGs). A total of 49 EEG recordings were performed (in seven PMV conditions for seven subjects), each lasting for 65 min. EEGs were recorded through the scalp and sorted by frequency using power spectral analysis. The best PMV condition for attention ability changed over time, from slightly cooler temperatures to higher temperatures. However, extreme PMV conditions led to poor attention ability during the experiment. The highest attention level was at PMV +1, according to our EEG analysis, but was reported as both PMV 0 and + 1 by the subjects. The lowest attention level was in higher temperature conditions (PMV +2, +3), according to subjects' own evaluations, while the lowest brain activity was measured in lower temperature conditions (PMV -2, −3).</P> <P><B>Highlights</B></P> <P> <UL> <LI> Attention ability was measured by EEG as a metric of productivity, instead of measuring productivity itself. </LI> <LI> The best temperature for attention changed over time and the perceived attention did not match the physiological response as assessed by EEG. </LI> <LI> The results provide a new method to measure productivity and a basis to develop new indoor environment controls for optimal productivity. </LI> </UL> </P>

뇌파와 시선추적의 시계열적 분석을 통한 공간선호도의 활성화 영역 차이

최진경,김주연 한국생활과학회 2019 한국생활과학회지 Vol.28 No.5

Human beings perceive space as a stimulus to space. In the stimuli of space perception, vision causes humans to recognize space as a neurological process in the brain. In particular, commercial spaces should attract consumers' attention so that they can consume space. This research was conducted simultaneously with the EEG data measuring experiment and the Eye-tracking experiment to examine the physiological response of humans in the process of comparing and selecting space. Two images suitable for the experimenter to remodel the existing cafe space were presented with visual stimuli that could be viewed simultaneously on one screen along with an original image to measure EEG and visual data. The EEG and Visual data of the selected participants (N=24) who were suitable for data analysis was assessed. The results were as follows. First, the experimental participants were able to compare and monitor each area of the visual stimuli. The prefrontal lobes (Fp1, Fp2) and frontal lobes (F3, Fz, F4) were determined to be active. Secondly, on interests and preferences, humans had high concentration on EEG and Visual attention. Experimental subjects also showed higher concentration on EEG, although they had fewer fixation compared to interests and preferences. Thirdly, the response sample t-test was performed with the SPSS 25 Statistic Program to ensure that the EEG-Visual data had a statistical significant difference. Later, significant differences were identified for each preferred image interest and preferred area. The study also presented spatial images at the same time. This research analyzed the EEG and visual data to establish the concentration of attention to objects in physiological response to human preferred space. 인간은 공간에서 발생되는 자극으로 공간을 지각하여 받아들인다. 또한 선호하는 것에 관심을 가져 오래 바라본다. 공간을 지각하는 자극 중 시각은 인간이 공간을 인식하고 뇌의 신경처리 과정으로 공간을 느끼고 공간에서 행동하게 한다. 특히 상업공간은 소비자의 관심을 끌어 공간을 소비할 수 있도록 해야 한다. 본 연구의 뇌파신호 측정 실험과 시선추적 실험을 동시에 진행하여 공간을 비교하고 선택하는 과정에서의 인간의 생리적 반응을 살펴보았다. 실험참가자가 기존 카페공간을 리모델링하기에 적합한 이미지 2개를 원본이미지 1개와 함께 한 화면에 동시에 볼 수 있는 시각자극물 제시하여 뇌파, 시선데이터를 측정하였다. 데이터 분석에 유효한 실험 참가자(N=24)의 뇌파와 시선데이터의 분석을 진행하였다. 결과는 다음과 같다. 첫 번째, 실험 참가자가 시각자극물의 각각의 영역을 비교 주시한다. 전전두엽(Fp1, Fp2)과 전두엽(F3, Fz, F4) 영역의 활성으로 판단하는 있었다. 두 번째, 관심 및 선호하는 것에 인간은 뇌파와 주시에 대한 집중이 높았다. 비교하는 대상에도 관심 및 선호대상에 비하여 적은 주시의 횟수를 가지지만 뇌파의 집중도가 높게 나타났다. 세 번째, 뇌파-시선데이터가 통계적으로 유의한 차이를 가지는지 SPSS 25 통계분석 프로그램으로 대응표본 t-검정을 실시하였다. 각 선호된 이미지에 대한 관심 및 선호대상 영역에 대하여 유의한 차이를 확인하였다. 본 연구는 공간이미지를 동시에 제시하였다. 뇌파신호와 시각데이터를 분석하여 인간이 선호하는 공간에 대한 생리적 반응을 관심이 있는 대상에 대한 주의집중도를 살펴보았다.

Development of an EEG-based workload measurement method in nuclear power plants

Choi, Moon Kyoung,Lee, Seung Min,Ha, Jun Su,Seong, Poong Hyun Elsevier 2018 Annals of nuclear energy Vol.111 No.-

<P><B>Abstract</B></P> <P>The environment of main control rooms of large scale process control systems such as nuclear power plants (NPPs) has been changed from the conventional analog type to the digital type. In digitalized advanced main control rooms, human operators conduct highly cognitive work rather than physical work compared to the case of the original control rooms in NPPs. Various operating support systems (OSSs) have been developed to reduce an operator’s workload. Most representative techniques to evaluate the workload are based on subjective ratings. However, there are some limitations including the possibility of skewed results due to self-assessment of the workload and the impossibility of continuously measuring the workload due to freezing simulation for workload assessment. As opposed to subjective ratings techniques, physiological techniques can be used for objective and continuous measurements of a human operator’s mental status by sensing the physiological changes of the autonomic or central nervous system. In this study, electroencephalogram (EEG) was used to measure the operator’s mental workload because it had been proven to be sensitive to variations of mental workload in other studies, and it allows various types of analysis. Based on various research reviews on the characteristics of brainwaves, EEG-based Workload Index (EWI) was suggested and validated through experiments. As a result, EWI is concluded to be valid for measuring an operator’s mental workload and preferable to subjective techniques. Furthermore, EWI was applied to evaluate the effects of OSSs on human operators through experiments. Finally, it is expected that the results of this study can be used to measure the operator’s workload in NPPs.</P> <P><B>Highlights</B></P> <P> <UL> <LI> A human operator’s workload in nuclear power plants(NPPs) usually has been evaluated by using subjective ratings. </LI> <LI> Subjective rating techniques have several weaknesses such as dependence on the operator’s memory as well as bias. </LI> <LI> We suggested an electroencephalogram (EEG)-based workload index for measuring the workload of human operators. </LI> <LI> The suggested index was applied to evaluate the effects of operating support systems. </LI> </UL> </P>

Use of electroencephalogram and long short-term memory networks to recognize design preferences of users toward architectural design alternatives

Sunwoo Chang,Wonhyeok Dong,Hanjong Jun 한국CDE학회 2020 Journal of computational design and engineering Vol.7 No.5

In this study, we propose an electroencephalogram (EEG)-based long short-term memory networks model for recognizing user preferences toward architectural design images. An EEG is an approach that records the electrical activity in the brain, and EEG-based affection recognition is a technique used for quantitatively recognizing human emotion by analysing the recorded signals. Decision-makers’ subjective reactions toward architectural design alternatives may play a key role in the architectural planning and design stage. In this regard, the proposed model enables the quantitative recognition of their preferences and supports architects in the planning and design stages. The suggested model classifies the recorded data using a deep-learning technique. To build the model, an EEG recording experiment was conducted with 18 subjects, who were asked to select their most/least preferred images among eight images of small-housing design. Post recording, a positive and negative affect schedule questionnaire was distributed to the subjects to rate their affection. Google TensorFlow and Keras were used to structure the model. After training, precision, recall, and f1 score metrics were used to evaluate and validate the model. This model can help designers to evaluate design alternatives in terms of decision-making. Moreover, as this model uses biosignal data, which is universal to humans, architectural design processes for children, the elderly, etc., may be supported. Furthermore, a data-driven design database may be proposed in a future research for cross-validating with previous methods such as interviews and observations.

이러닝 적용을 위한 뇌파기반 인지부하 측정

김준(Jun Kim),송기상(Ki-Sang Song) 한국인지과학회 2009 인지과학 Vol.20 No.2

본 연구는 이러닝 체제에서 상호작용을 개선할 수 있도록 하기 위하여 사용자의 생리적 데이터 가운데서 뇌파를 통하여 학습자의 인지부하 발생을 파악할 수 있는 지를 연구하고자 하였다. 뇌파를 통하여 인지부하 발생을 알 수 있게 된다면 실시간 이러닝 체제에서 적절한 피드백 제공에 활용될 수 있기 때문이다. 이를 위하여 EEG를 이용하여 학습자의 뇌파를 측정하면서 인지활동을 수행하는 동안 발생되는 인지부하도를 측정하였고 인지과부하를 판별할 수 있는지를 알아보았다. 뇌파 측정을 위하여 언어 관련 작업기억 능력을 측정할 수 있는 듣기회상과제를 제시하였으며, 실험을 통한 과제 정답률 및 뇌파 분석 결과는 다음과 같다. 첫째, 듣기회상과제의 정답률은 회상반응과제에서 1단계는 84.4%, 2단계는 90.6%, 3단계는 62.5%, 4단계는 56.3%를 보였으며, 통계적으로 유의한 차이가 있음을 확인하였다. 즉, 3, 4단계의 경우는 피험자들이 매우 어려움을 겪었던 단계로 인지과부하가 발생했을 것으로 보인다. 둘째, SEF-95% 지표는 1, 2단계에 비해 3, 4단계에서 더욱 높은 값을 보였으며, 이는 피험자들의 인지부하가 3, 4단계에서 높았음을 객관적으로 보여주는 근거이다. 셋째, 감마파의 상대파워는 3, 4단계에서 파워값이 급격히 올라가는 패턴을 보였으며, 통계적으로 유의한 5개의 채널(F3, F4, C4, F7, F8)을 확인하였다. 5개의 채널은 뇌의 브로카 영역(F7, F8) 주위에 위치하고 있으며, 특히 뇌맵핑 분석을 통해 확인한 결과, F8(우반구의 브로카 영역에 해당하는 위치)에서 단계별 난이도가 올라갈수록 활성화의 차이가 크게 나타났다. 넷째, 19채널에 대한 상호 상관 분석을 통해 1, 2단계에 비해 3, 4단계에서 비동기화가 증가하였다. 위의 결과를 통한 본 연구의 결론은 뇌파를 이용하여 인간이 인지활동을 수행하는 동안 인지부하도를 측정할 수 있으며, 인지과부하를 판별해 낼 수 있음을 확인하였다. This paper describes the possibility of human physiological data, especially brain-wave activity, to detect cognitive overload, a phenomenon that may occur while learner uses an e-learning system. If it is found that cognitive overload to be detectable, providing appropriate feedback to learners may be possible. To illustrate the possibility, while engaging in cognitive activities, cognitive load levels were measured by EEG (electroencephalogram) to seek detection of cognitive overload. The task given to learner was a computerized listening and recall test designed to measure working memory capacity, and the test had four progressively increasing degrees of difficulty. Eight male, right-handed, university students were asked to answer 4 sets of tests and each test took from 61 seconds to 198 seconds. A correction ratio was then calculated and EEG results analyzed. The correction ratio of listening and recall tests were 84.5%, 90.6%, 62.5% and 56.3% respectively, and the degree of difficulty had statistical significance. The data highlighted learner cognitive overload on test level of 3 and 4, the higher level tests. Second, the SEF-95% value was greater on test3 and 4 than on tests 1 and 2 indicating that tests 3 and 4 imposed greater cognitive load on participants. Third, the relative power of EEG gamma wave rapidly increased on the 3rd and 4<SUP>th</SUP> test, and signals from channel F3, F4, C4, F7, and F8 showed statistically significance. These five channels are surrounding the brain's Broca area, and from a brain mapping analysis it was found that F8, right-half of the brain area, was activated relative to the degree of difficulty. Lastly, cross relation analysis showed greater increasing in synchronization at test3 and 4<SUP>th</SUP> at test1 and 2. From these findings, it is possible to measure brain cognitive load level and cognitive over load via brain activity, which may provide atimely feedback scheme for e-learning systems.

Accurate Multiscale Permutation Entropy Analysis of Brain Rhythm to Detect Epileptic Seizure

최영석,조명석,현광민 한국지식정보기술학회 2017 한국지식정보기술학회 논문지 Vol.12 No.1

Electroencephalogram(EEG) has been a standard tool to monitor the status of the brain. For quantification of EEG, permutation entropy has been of interest due to simplicity and robustness to noise. A multiscale extension of PE, called multiscale PE(MPE), has been promising in fully describing the dynamical characteristics of EEG over multiple temporal scales. However, an imprecise estimation of MPE at large scales limits its application for analyzing of short EEG. Here, a new multiscale PE measure which aims at estimating entropy accurately is presented. By computing PE of all possible coarse-grained time series and averaging the values of PE at each scale, the resultant composite MPE (CMPE) yields improved accuracy in estimation of entropy. Thus, the CMPE measure accomplishes consistent quantification of entropy regardless of the length of data. This advantage of CMPE renders its capability for analyzing EEG signals. Through simulations with two synthetic noises, CMPE has proved its capability over MPE in terms of accuracy. Experimental results using normal, inter-ictal and ictal EEG recordings have shown that the CMPE measure has leaded an improved discrimination capability for three different neurological states (normal, inter-ictal, and ictal states) than the conventional PE family.