Locatelli M, De Angeli A, Leone E, Grassi B, Scarone S
Department of Biomedical and Technological Sciences, University of Milan Medical School, Istituto Scientifico San Raffaele, Italy.
Int J Neurosci. 1993 Oct;72(3-4):265-70. doi: 10.3109/00207459309024115.
Factor Analysis can extract salient features from EEG data and reduce redundancy of multi-channel computerized EEG data. A 16-channel computerized frequency analysis of background brain electrical activity during 3 functional conditions (eyes closed, eyes open and hyperventilation) was carried out in two groups, fifty healthy subjects and twenty-three schizophrenics. The power log-transformed relative values of normal subjects and schizophrenic patients were submitted to Factor Analysis and the resulting factor scores were compared. Schizophrenics showed EEG abnormalities in delta 2, theta 1 and alpha 2 bands for the first factor, accounting for the eyes closed condition, and in theta 2 and beta 2 bands for the second factor, accounting for the eyes open condition. This preliminary study demonstrates the utility of Factor Analysis in managing and comparing computerized EEG data.
因子分析可以从脑电图数据中提取显著特征，并减少多通道计算机化脑电图数据的冗余。对两组进行了16通道计算机化频率分析，其中一组为50名健康受试者，另一组为23名精神分裂症患者，分析了三种功能状态（闭眼、睁眼和过度换气）下的背景脑电活动。将正常受试者和精神分裂症患者的功率对数转换相对值进行因子分析，并比较所得的因子得分。精神分裂症患者在第一个因子的δ2、θ1和α2波段显示脑电图异常，该因子对应闭眼状态；在第二个因子的θ2和β2波段显示异常，该因子对应睁眼状态。这项初步研究证明了因子分析在管理和比较计算机化脑电图数据方面的实用性。
