Analysis of the electroencephalographic activity during the Necker cube reversals by means of the wavelet transform
Analysis of the electroencephalographic activity during the Necker cube reversals by means of the wavelet transform
复制标题
DOI：
10.1007/s004220050492
复制
发表时间：
1998-11-01
影响因子：
1.9
通讯作者：
Stadler, M
中科院分区：
文献类型：
--
作者：
Isoglu-Alkaç, Ü;Basar-Eroglu, C;Stadler, M
在以前的研究中，感知切换相关的电位是在观察多稳态动态反转模式时获得的，其中单个响应的平均值由受试者按下按钮触发。本方法学研究的目的是提高感知转换相关电位的信号质量，考虑到测量方法对受试者反应时间的依赖性，反应时间在一个会话期间可能会发生显着变化，导致平均事件相关电位（ERP）中的低振幅波形。为了克服这一问题，在测量的内部事件的电生理相关性，模式选择方法的基础上，小波变换（WT），提出了选择一个子集的单一的ERPs具有更均匀的lavery。九名受试者观察到一个内克尔立方体，并指示按下按钮后立即感知切换。在按下按钮前约250 ms观察到一个缓慢的低振幅正波，其额心振幅最大。在单次扫描上应用5倍频程WT之后，在每个倍频程中获得的时间-频率系数在试验中被平均。代表平均ERP的最主要特征是在按下按钮之前250和125 ms之间发生的delta（0.5-4 Hz）系数。通过对包含该属性的单次扫描的子集进行平均，观察到响应峰的锐化和显著幅度增加。
In previous studies, a perceptual switching related potential was obtained during the observation of a multistable dynamic reversal pattern, where the averaging of the single responses was triggered by subjects pressing a button. The present methodological study aims to increase the signal quality of perceptual switching related potentials considering the dependence of the measurement method on the reaction time of the subject, which may vary significantly during a session, leading to low-amplitude waveform in the averaged event-related-potential (ERP). To overcome this problem in measuring the electrophysiological correlate of an internal event, a pattern selection method based on the wavelet transform (WT) is proposed to choose a subset of single ERPs with more homogenous latencies. Nine subjects observed a Necker cube and were instructed to press the button immediately after perceptual switching. A slow, low-amplitude positive wave with frontocentral amplitude maxima was observed around 250 ms prior to the button press. After the application of a 5 octave WT on single sweeps, the time-frequency coefficients obtained in each octave were averaged across trials. The most dominant feature representing the averaged ERP was the delta (0.5-4 Hz) coefficient occurring between 250 and 125 ms before the button press. By averaging the subset of the single sweeps containing this property, a sharpening and significant amplitude increase of the response peak was observed.