Cincotti F, Babiloni C, Miniussi C, Carducci F, Moretti D, Salinari S, Pascual-Marqui R, Rossini P M, Babiloni F
Dipartimento di Fisiologia Umana e Farmacologia, Università di Roma La Sapienza, P.le A. Moro 5, 00185 Roma, Italy.
Methods Inf Med. 2004;43(1):114-7.
EEG scalp potential distributions recorded in humans are affected by low spatial resolution and by the dependence on the electrical reference used. High resolution EEG technologies are available to drastically increase the spatial resolution of the raw EEG. Such technologies include the computation of surface Laplacian (SL) of the recorded potentials, as well as the use of realistic head models to estimate the cortical sources via linear inverse procedure (low resolution brain electromagnetic tomography, LORETA). However, these deblurring procedures are generally used in conjunction with EEG recordings with 64-128 scalp electrodes and with realistic head models obtained via sequential magnetic resonance images (MRIs) of the subjects. Such recording setup it is not often available in the clinical context, due to both the unavailability of these technologies and the scarce compliance of the patients with them. In this study we addressed the use of SL and LORETA deblurring techniques to analyze data from a standard 10-20 system (19 electrodes) in a group of Alzheimer disease (AD) patients.
EEG data related to unilateral finger movements were gathered from 10 patients affected by AD. SL and LORETA techniques were applied for source estimation of EEG data. The use of MRIs for the construction of head models was avoided by using the quasi-realistic head model of the Brain Imaging Neurology Institute of Montreal.
A similar cortical activity estimated by the SL and LORETA techniques was observed during an identical time period of the acquired EEG data in the examined population.
The results of the present study suggest that both SL and LORETA approaches can be usefully applied in the clinical context, by using quasi-realistic head modeling and a standard 10-20 system as electrode montage (19 electrodes). These results represent a reciprocal cross-validation of the two mathematically independent techniques in a clinical environment.
人类记录的脑电图头皮电位分布受空间分辨率低以及对所用电参考的依赖性影响。高分辨率脑电图技术可大幅提高原始脑电图的空间分辨率。此类技术包括记录电位的表面拉普拉斯算子（SL）计算，以及通过线性逆过程（低分辨率脑电磁断层扫描，LORETA）使用逼真的头部模型来估计皮质源。然而，这些去模糊程序通常与使用64 - 128个头皮电极的脑电图记录以及通过受试者的序列磁共振成像（MRI）获得的逼真头部模型结合使用。由于这些技术不可用以及患者对其依从性差，这种记录设置在临床环境中并不常见。在本研究中，我们探讨了使用SL和LORETA去模糊技术来分析一组阿尔茨海默病（AD）患者标准10 - 20系统（19个电极）的数据。
从10名AD患者收集与单侧手指运动相关的脑电图数据。应用SL和LORETA技术对脑电图数据进行源估计。通过使用蒙特利尔脑成像神经学研究所的准逼真头部模型，避免了使用MRI构建头部模型。
在所检查人群中，在采集的脑电图数据的相同时间段内，观察到SL和LORETA技术估计的皮质活动相似。
本研究结果表明，通过使用准逼真头部建模和标准10 - 20系统作为电极蒙太奇（19个电极），SL和LORETA方法均可有效地应用于临床环境。这些结果代表了两种数学上独立的技术在临床环境中的相互交叉验证。
