Tomberg C, Noël P, Ozaki I, Desmedt J E
Brain Research Unit, University of Brussels, Belgium.
Electroencephalogr Clin Neurophysiol. 1990 Jul-Aug;77(4):259-65. doi: 10.1016/0168-5597(90)90064-k.
The main reason for doing topographic mapping of EEG or evoked potentials is to assess regional changes in brain potentials. The use of an average reference is shown to have perverse effects in this relation, namely because it imposes on the recorded data a zero-centering effect which can reduce, eliminate or even reverse the focal changes of bit-mapped brain potentials. Concurrent studies on a true 3-shell head model suggest that such distortions of human EEG data occur because the average reference is computed from a set of (scalp) recording electrodes which do not survey the bottom half of the head volume so that the integral of scalp-recorded potentials frequently differs from zero. The results also raise the question whether the actual incidence of radial or near-radial (versus tangential) generators has been underestimated in the published data using average reference mapping.
进行脑电图（EEG）或诱发电位地形图绘制的主要原因是评估脑电位的区域变化。结果表明，使用平均参考在这方面会产生不良影响，具体而言，是因为它会给记录数据施加一种零中心化效应，这种效应可能会减少、消除甚至反转位图化脑电位的局灶性变化。对一个真实的三层头部模型进行的同步研究表明，人类EEG数据出现这种失真的原因是，平均参考是根据一组（头皮）记录电极计算得出的，这些电极并未探测到头容积的下半部分，因此头皮记录电位的积分通常不为零。这些结果还提出了一个问题，即使用平均参考映射的已发表数据中，径向或近径向（相对于切向）发生器的实际发生率是否被低估了。
