Automated cortical projection of EEG sensors: Anatomical correlation via the international 10–10 system
Introduction
Localization of brain generators in a non-invasive way has been improved with the development of high resolution EEG. This technique allows the acquisition of EEG signals with a high spatial resolution at the level of the scalp, and high temporal definition. Manufacturers of EEG supplies have recognized this, and electrode caps which enable easy placement of electrodes according to the 10–10 standard are available. Currently, more and more researchers are moving to an even higher number of channels and EEG acquisition systems with 128 channels are no longer uncommon. Several authors have shown that a high number of electrodes increases the precision of the localization of intra-cerebral generators which are at the origin of the surface signals (Lantz et al., 2003, Michel et al., 2004).
The positioning of EEG electrodes is defined according to external landmarks such as nasion, inion and pre-auriculars, but also according to the cerebral structures beneath each sensor (Jasper, 1958). A fundamental assumption of the system is that there is a reliable correlation between scalp sensor location and the underlying cerebral structure. Some studies have examined the validity of the structural correlation using cadavers (Blume et al., 1974, Jasper, 1958), X-rays (Morris et al., 1986), CT scans (Homan et al., 1987, Myslobodsky and Bar-Ziv, 1989, Myslobodsky et al., 1990) and MRI (Steinmetz et al., 1989, Jack et al., 1990, Lagerlund et al., 1993, Towle et al., 1993). An initial study (Homan et al., 1987) plotted the 10–20 positions on Brodmann's cortical map from the temporal view. These authors successfully described the cranio–cerebral structural relationships on Brodmann's plane. However, no statistical analysis was performed and their methods could not avoid a resolution gap because of the dimensional conversion from space to plane. These problems were overcome by fitting cranial and cerebral surfaces to a sphere (Lagerlund et al., 1993, Towle et al., 1993). These authors described the cortical locations that lie beneath the 10–20 electrodes (cortical projection points) as 3D coordinates. More recently, Okamoto et al. examined the cranio–cerebral correspondences for 17 healthy adults and normalized the 10–20 cortical projection points of the subjects to the standard Montreal Neurological Institute (MNI) and Talairach stereotactic coordinates (reviewed in Brett et al., 2002). Statistical analysis was performed in order to obtain their probabilistic distribution. Automated methods have also been proposed to project head-surface locations onto the cortical surface in structural images (Okamoto and Dan, 2005).
Knowledge of cortical projections of the 10–10 system has several applications in surface brain imaging such as trans-magnetic stimulation (TMS) or near-infrared spectroscopic (NIRS) imaging (Okamoto et al., 2004). These techniques use indeed the international systems of sensor positioning initially described for EEG. At the opposite of electroencephalography or magnetoencephelography which now uses source imaging techniques for localization purpose (Michel et al., 2004), TMS and NIRS, which only concern the superficial cortical surface, strongly rely on cranio-cerebral correlations.
Although numerous studies exist, the cortical projections of electrodes within the 10–10 system and their anatomical variability have never been described. The purpose of our study was therefore to examine the cranio–cerebral correlations by determining the 10–10 cortical projection points in 16 healthy subjects, describing their locations in Talairach space with statistical estimation of variability and corresponding macro-anatomical and cytoarchitectonic features.
This work could provide a reliable database of cranio–cerebral correlations to researchers and neurophysiologists when they have no structural images for their subjects.
Section snippets
Subjects
Sixteen healthy subjects (six women and 10 men, aged 20–42 years) participated in this study. All subjects were investigated in order to confirm the absence of neurological abnormalities, and all gave their informed consent. The study was approved by the ethics committee (CCPPRB) of our institution.
10–10 Sensor placement
To reproduce a high-resolution EEG situation, 64 EEG-MRI sensors were taped onto the subject's head (Fig. 1). The EEG-MRI sensors were made by combining an EEG electrode, plastic support and MRI
Cortical projections of the 10–10 sensor system
The mean stereotactic coordinates and standard deviations of the locations of the 10–10 cortical projection points based on the 16 subjects expressed in Talairach space are presented in Table 1. Concerning the spatial dispersion around the mean cortical coordinates, we calculated a grand standard deviation of approximately 4.6 mm in x, 7.1 mm in y and 7.8 mm in z. Fig. 5 presents each mean cortical projection onto a Ch2bet brain template with anatomical structures. We calculated for each
Discussion
The aim of this study was to present a 3D anatomical atlas which completes our knowledge of the current anatomical correlations with 10–20 EEG sensors. We examined the cranio–cerebral correlations and probabilistically expressed the locations of 10–10 standard positions and their cortical projection points in standard stereotactic space for brain imaging studies with anatomical considerations.
Conceptually, our approach is an extension of the work of Okamoto et al. (2004), who projected the
Acknowledgments
We thank Mr. Christian-G. Bénar for examination of the manuscript and for his helpful suggestions. We also gratefully acknowledge the participation of the individuals involved in this study. This study was supported by the company T.E.A of Nancy (France) and the Regional Council of Lorraine (France).
References (34)
- et al.
Anatomical correlates of the ten–twenty electrode placement system in infants
Electroencephalogr. Clin. Neurophysiol.
(1974) - et al.
Usefulness of focal rhythmic discharges on scalp EEG of patients with focal cortical dysplasia and intractable epilepsy
Electroencephalogr. Clin. Neurophysiol.
(1996) - et al.
Cerebral location of international 10–20 system electrode placement
Electroencephalogr. Clin. Neurophysiol.
(1987) - et al.
Virtual 10–20 measurement on MR images for inter-modal linking of transcranial and tomographic neuroimaging methods
NeuroImage
(2005) - et al.
10/20, 10/10, and 10/5 systems revisited: their validity as relative head-surface-based positioning systems
NeuroImage
(2007) - et al.
Automatic localization of new scalp-recorded EEG sensors in MRI volume
NeuroImage
(2008) - et al.
More accurate Talairach coordinates for neuroimaging using non-linear registration
NeuroImage
(2008) - et al.
Determination of 10–20 system electrode locations using magnetic resonance image scanning with markers
Electroencephalogr. Clin. Neurophysiol.
(1993) - et al.
Epileptic source localization with high density EEG: how many electrodes are needed?
Clin. Neurophysiol.
(2003) - et al.
EEG source imaging
Clin. Neurophysiol.
(2004)