• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제10권12호
• /
• pp.587-594
• /
• 2021

뇌파(electroencephalogram, EEG)를 이용한 진단이 확대되면서 EEG 신호를 자동으로 분류하기 위한 다양한 연구가 활발히 이루어지고 있다. 본 논문은 일반인과 뇌전증 환자에게서 추출한 EEG 신호를 효과적으로 식별할 수 있는 CNN 모델을 제안한다. CNN의 학습에 필요한 데이터를 확장하기 위하여 EEG 신호를 낮은 차원의 신호로 분할하고, 이것을 다시 여러 개의 세그먼트로 중첩 분할하여 CNN 학습에 이용한다. 이와 더불어 CNN의 성능을 개선하기 위하여 CNN 앙상블 전략을 제안한다. 공개된 Bonn 데이터세트로 실험을 수행한 결과 뇌전증 발작을 99.0% 이상의 정확도로 검출하였고, 앙상블 방식에 의해 3-클래스와 5-클래스의 EEG 분류에서 정확도가 향상되었다.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2004년도 추계학술대회 논문집
• /
• pp.26-26
• /
• 2004

• 한국산업정보학회논문지
• /
• 제18권5호
• /
• pp.15-18
• /
• 2013

본 논문은 뇌전도 신호 처리를 위한 아날로그 전단부를 제시한다. 일반적으로 뇌전도 신호는 낮은 주파수 대역에 존재하고 신호의 크기가 미약하므로 이를 처리하기 위한 아날로그 전단부는 높은 전압 이득 및 공통모드 제거비를 가져야 하며 저주파 잡음을 효과적으로 억제해야 한다. 본 논문에서 제시하는 아날로그 전단부는 가변 이득 계측 증폭기와 대역통과 필터로 구성되어 있다. 낮은 주파수의 잡음을 제거하기 위하여 주파수 chopping을 적용하였다. 본 논문의 회로는 0.18um CMOS 공정을 이용하여 제작하였으며 측정 결과 최대 60dB의 전압이득과 100dB 이상의 공통모드 제거비를 내는 것을 확인하였다.

• 대한물리의학회지
• /
• 제5권3호
• /
• pp.313-322
• /
• 2010

Purpose : This study was to evaluate the effects of cognitive rehabilitation training on the cognitive decline of dementia patients. Therefore the purpose of this study was to examine the influence of brain activation according to bio-feedback training in dementia. Methods : Ten dementia patients were recruited this study. Experiment was performed for 30min per session, five times a week through 4 week and two measurements before and after bio-feedback training. Brain activity was measured by Korea Electroencephalogram(EEG) system. Statistical analysis was used Wilcoxon signed rank test to know difference of EEG between pre and post-test in each group and Mann-Whitney U test was to know difference between experimental and control group. Results : Significant improvement of slow-alpha wave was observed following bio-feedback in experiment group. There was no significant change in experiment and control group. Conclusion : In this study, the bio-feedback training was effective in improving slow-alpha wave in dementia patients. It is suggested that bio-feedback training with dementia patients can be useful to ameliorate the cognitive decline. And it will be effective for prevention of cognitive function decline.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2000년도 춘계 학술대회 및 국제 감성공학 심포지움 논문집 Proceeding of the 2000 Spring Conference of KOSES and International Sensibility Ergonomics Symposium
• /
• pp.230-234
• /
• 2000

Thermal comfort by wearing clothes is the important element which gives influence to a clothing comfort. The thermal comfort of clothes have been evaluated by sensory test and physical property of clothes material. To evaluate a thermophysiological comfort. a new evaluation method which measures the physiological response such as electroencephalogram(EEG) is attracting the attention of many people. In the chilly environment, the EEGs in t재 kinds of thermal conditions : with and without clothes were measured. By utilizing the chaos analysis, the behavior of the obtained EEGs were quantiatively expressed in the correlation dimension. As a result, the correlation dimension of the EEGs in being thermal comfortable feeling by putting on clothes, was bigger than the correlation dimension of the EEGs in being cold and discomfort. These results suggest that chaotic analysis of EEG is effective to the quantitative evaluation of thermal esthesis.

• 대한의용생체공학회:의공학회지
• /
• 제29권4호
• /
• pp.286-294
• /
• 2008

This work investigates the characteristics of brain interactions of experienced Zen-Buddhist practitioners by obtaining multichannel EEG (electroencephalogram) data. Brain interactions were compared among three phases-40-minute meditation (M), 5-minute Chakra-focusing practice (Z) and rest with closed eyes (R). The similarity index S, developed in nonlinear dynamical system theory, was employed to measure the degree of possibly asymmetric coupling. Meditators exhibited, overall, stronger interactions among multiple cortical areas in meditation stages M and Z than in the R state. This enhancement was greater in the M stage when the meditator was accompanied by a thought-free and fully consciousness state. In the high-frequency band (>13Hz), the interdependence was also higher in both meditation stages than at baseline rest. However, the interaction strength, especially in the posterior regions, was greatest in the Z stage, which involved internal attention. Few electrode pairs were observed with significant pair-wise asymmetry in the Z state. The similarity is a possible characteristic of dense reciprocal and strong mutual interactions between multiple cortical areas during meditation - especially in the Z state in the high-frequency band. These results demonstrate that profound Zen meditation induces various dynamic states in different phases of meditation, possibly reflected by nonlinear interdependence measure.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제55권5호
• /
• pp.240-247
• /
• 2006

Frequent arousals during sleep degrade the quality of sleep and result in sleep fragmentation. Visual inspection of physiological signals to detect the arousal events is cumbersome and time-consuming work. The purpose of this study is to develop an automatic algorithm to detect the arousal events. The proposed method is based on time-frequency analysis and the support vector machine classifier using single channel electroencephalogram (EEG). To extract features, first we computed 6 indices to find out the informations of a subject's sleep states. Next powers of each of 4 frequency bands were computed using spectrogram of arousal region. And finally we computed variations of power of EEG frequency to detect arousals. The performance has been assessed using polysomnographic (PSG) recordings of twenty patients with sleep apnea, snoring and excessive daytime sleepiness (EDS). We could obtain sensitivity of 79.65%, specificity of 89.52% for the data sets. We have shown that proposed method was effective for detecting the arousal events.

• 한국지능시스템학회논문지
• /
• 제12권4호
• /
• pp.366-372
• /
• 2002

뇌파(Electroencephalogram, EEG)는 뇌 신경세포가 정보를 처리하는 과정에서 발생하는 전기적인 신호를 두피 표면에서 측정한 것이다. 이러한 뇌파는 비침습적인 방법으로 전기적인 신호를 측정하며 측정시 여러가지 형태의 잡파(artifact)가 섞이기 쉽다. 이러한 잡파는 뇌의 정보처리과정에 대한 유용한 정보를 담고 있는 뇌파를 분석하는데 방해가 되므로 이를 제거하기 위한 노력이 계속되어 왔다. 그러나 본 연구에서는 보다 적극적인 방향으로 잡파가 섞인 뇌파의 특성을 분석하여 이를 통해 제어 시스템 등과 같은 시스템에 적용할 수 있는 가능성을 알아보았다. 대표적인 잡파인 eye_blinking, eye_rolling, muscle 둥이 각각 포함된 뇌파에 대해서 선형 및 비선형 분석을 실시함으로써 유의미한 특성 차이를 나타내었다.

• 한국통신학회논문지
• /
• 제39C권10호
• /
• pp.887-895
• /
• 2014

졸음운전은 전체 교통사고 원인 중 큰 비중을 차지하며 그 위험성이 음주운전보다도 크다고 알려져 있다. 따라서 운전자의 졸음을 판단하고 경고하는 시스템 개발에 대한 관심이 높아지고 있으며, 뇌파를 분석하는 것이 운전자의 피로와 졸음을 감지하는데 효과적이라는 연구결과들이 발표되었다. 본 논문은 짧은 시간에 높은 해상도를 가지는 auto-regressive 모델 기법 중 잡음에 강인한 errors-in-variables(EIV) 방법을 이용하여 특징벡터를 추출하고, 다층신경망(multilayer perceptron; MLP)에 적용하여 운전자의 상태를 각성, 천이, 졸음의 세 가지 상태로 분류하는 졸음 감지 시스템을 제안한다. 생체신호의 측정 환경에 따른 성능을 평가하기 위해 높은 진단률을 갖도록 하는 EIV차수를 결정하고, 잡음에 대한 강인성을 확인하기 위해 신호대 잡음비(signal-to-noise ratio; SNR)에 따른 성능을 선형 예측 부호화(linear predictive coding; LPC) 방법과 비교하였다. 이 결과로부터 제안한 EIV와 MLP를 결합한 졸음 감지 시스템은 기존의 LPC와 MLP를 이용한 시스템에 대해 우수한 성능을 얻을 수 있음을 확인하였다.

• 대한임상전기생리학회지
• /
• 제9권1호
• /
• pp.15-21
• /
• 2011

Purpose : This study aims to examine the effects of action observation of ski simulation on electroencephalogram (EEG). Methods : Thirty healthy subjects were randomly divided into three groups: 10 participants in an AM (actual movement) group; 10 participants in an OM (observation) group; and 10 participants in a CM (control movement) group. The EEG was measured to examine changes in EEG between action observation and actually movement at 8 areas of the scalp for one minute before, during and after action observation. Results : Relative alpha power showed statistically significant differences among groups in Cz and C4 and there were interactions among groups in the Oz area. Relative beta power showed no statistical significance among groups and in particular, there were interactions among groups in the Oz area. Conclusion : These findings show that action observation affected brain activation as in actual movement. Thus it can be expected that when movement through the brain activation is applied to patients in bed rest or those who cannot perform actually movement, it can be utilized as physiotherapy.