Application of Chaotic Analysis to Electroencephalography : Preliminary Study

혼돈 이론을 이용한 뇌파 분석에 대한 기초 연구

The object of this study is to apply a chaotic signal analysis method to the EEG research, especially in the aspect of neuropsychiatry, and to get some inspection of the chaotic phenomena according to the brain sites and subjects. We have acquired 21 channel EEG data and one EOG according to the international 10-20 system and calculated the correlation dimension. The subject groups are schizophrenics, bipolar disorder, major depression and normal control. They were all awoke and eye-closed. We have found no distinctive features from our experiments except temporal regions have slightly higher correlation dimension. There is also no specific distinctions between groups. We conjecture that these results are mainly because the subjects were not well controlled. EEG dimension may change in accordance with to the age, sex, medication and the time data were selected to calculate. We have also considered some conditions for a better and more objective research of chaotic analysis to EEG research. Better conditioning and standardizing the calculation of correlation dimension is necessary for the application of the chaotic analysis to neuropsychiatry.

선형적이고 정적인 뇌파 분석의 한계를 극복하고자 비선형적이고 동적인 접근방식인 혼돈이론을 이용하여 정신분열병환자 16명, 양극성 정동장애 10명, 주요 우울증 6명 및 정상 대조군 12명을 대상으로 뇌파 분석을 시도하고, 객관적인 혼돈 연구의 도입에 있어서 고려되어야 할 문제점을 살펴보고자 하였다. 각 질환에서의 상관차원의 분포가 정상대조군보다 개인에 따른 차이가 크다는 것을 알 수 있었고, 대체적으로 측두엽 부위에서 상관차원이 크게 나타나는 경향이 있었다. 이러한 결과와 관련하여 대상환자군의 특성, 혼돈이론을 뇌파분석에 이용할 때의 주의점등에 대하여 토론하였다.