Elsevier

NeuroImage

Volume 51, Issue 2, June 2010, Pages 642-653
NeuroImage

Source localization of ictal epileptic activity investigated by high resolution EEG and validated by SEEG

https://doi.org/10.1016/j.neuroimage.2010.02.067Get rights and content

Abstract

High resolution electroencephalography (HR-EEG) combined with source localization methods has mainly been used to study interictal spikes and there have been few studies comparing source localization of scalp ictal patterns with depth EEG. To address this issue, 10 patients with four different scalp ictal patterns (ictal spikes, rhythmic activity, paroxysmal fast activity, obscured) were investigated by both HR-EEG and stereoelectroencephalography (SEEG). Sixty-four scalp-EEG sensors and a sampling rate of 1 kHz were used to record scalp ictal patterns. Five different source models (moving dipole, rotating dipole, MUSIC, LORETA, and sLORETA) were used in order to perform source localization. Seven to 10 intracerebral electrodes were implanted during SEEG investigations. For each source model, the concordance between ictal source localization and epileptogenic zone defined by SEEG was assessed. Results were considered to agree if they localized in the same sublobar area as defined by a trained epileptologist. Across the study population, the best concordance between source localization methods and SEEG (9/10) was obtained with equivalent current dipole modeling. MUSIC and LORETA had a concordance of 7/10 whereas sLORETA had a concordance of only 5/10. Four of our patients classified into different groups (ictal spikes, paroxysmal fast activity, obscured) had complete concordance between source localization methods and SEEG. A high signal to noise ratio, a short time window of analysis (< 1 s) and bandpass filtering around the frequency of rhythmic activity allowed improvement of the source localization results. A high level of agreement between source localization methods and SEEG can be obtained for ictal spike patterns and for scalp-EEG paroxysmal fact activities whereas scalp rhythmic discharges can be accurately localized but originated from seizure propagation network.

Introduction

Source localization techniques have been widely studied and validated for the study of interictal activity (Ebersole, 2000, Lantz et al., 2003, Michel et al., 2004a, Michel et al., 2004b, Gavaret et al., 2004, Gavaret et al., 2006, Gavaret et al., 2009). However, there have been few studies concerning ictal source localization (Lantz et al., 2001, Boon et al., 2002, Holmes et al., 2004, Beniczky et al., 2006, Ding et al., 2007). In addition, the relationship between the “irritative” zone (i.e. presenting interictal discharges) and epileptogenic zone remains poorly investigated (Talairach and Bancaud, 1966).

The small number of publications on this topic is largely due to the difficulty in recording epileptic seizures during high resolution electroencephalography (HR-EEG). This can be explained by the short recording time, due to demanding technical reasons (high number of sensors, high sampling rate, low impedances, comfort, etc.). Moreover, source modeling of seizures is much more complex than source modeling of spikes because of the low signal to noise ratio of some ictal pattern types, their frequent contamination by muscle artifacts, and propagation. Some studies have compared ictal surface EEG with intracranial EEG data (Assaf and Ebersole, 1997, Lantz et al., 1999, Lantz et al., 2001, Foldvary et al., 2001, Merlet and Gotman, 2001, Boon et al., 1997, Boon et al., 2002). Other authors have compared ictal surface EEG with anatomic lesions (Ding et al., 2007), epilepsy surgery (Boon et al., 1997, Lantz et al., 2001), or single photon-emission computed tomography (SPECT) (Mine et al., 1998, Beniczky et al., 2006). Due to poor temporal resolution, metabolic investigations are not suitable to study the temporal dynamics of seizures and cannot be used as a valid reference technique.

Different types of epilepsy (often temporal) and ictal patterns (mainly theta activity and spikes) were investigated in these studies, mainly using equivalent current dipole (ECD) and multiple signal classification (MUSIC) methods. Few studies (Worrell et al., 2000, Holmes et al., 2004) have used distributed source models (low resolution electromagnetic tomography (LORETA)) to localize ictal patterns. In these studies, the conditions of ictal surface EEG recordings were rarely adequate for source localization because of the low number of electrodes (30–40) or the use of a spherical head model.

Magnetoencephalography (MEG) recordings have been demonstrated to yield localizing information in epilepsy surgery candidates (Stefan et al., 2003, Fischer et al., 2005, Knowlton et al., 2006, Knowlton et al., 2009). Ictal activities have also been investigated by MEG in a few patients (Sutherling et al., 1987, Stefan et al., 1992, Eliashiv et al., 2002, Tanaka et al., 2009). In these studies, ictal magnetic source imaging was compared with the results of epilepsy surgery, intracranial recordings, and scalp EEGs. The authors concluded that the ictal magnetic source localizations were in close agreement with invasive and noninvasive preoperative studies. MEG ictal studies, like EEG ictal studies, concluded that imaging can help in the placement of invasive electrodes and in surgery planning. Despite fewer ergonomic constraints than MEG recordings (which require head immobility), EEG source localization studies of ictal activity are still rare. To date, no data have been published concerning source localization of ictal patterns in conditions of HR-EEG (at least 64 sensors and a high sampling rate) and few comparative studies exist with intracerebral EEG (Merlet and Gotman, 2001, Boon et al., 2002). Moreover, no study has compared dipolar and distributed source methods. It is also unclear what kind of scalp-EEG ictal patterns permit reliable source localization.

The main purpose of this study was to evaluate the accuracy of ictal source localization in distinct types of scalp-EEG ictal patterns, using stereoelectroencephalography (SEEG) as a validating method. The different source localization methods were also compared.

Section snippets

Subjects

Ten drug-resistant partial epilepsy patients underwent HR-EEG recording of seizures (six patients were investigated in the Clinical Neurophysiology Department of Timone Hospital, Marseille, and four in the Neurology Department of the University Hospital, Nancy). Each patient gave his informed consent and the study was approved by the ethics committee (CPP) of Marseille.

Patients first underwent a presurgical evaluation including careful history taking, neurologic examination, video–EEG recording

Patient classification according to scalp-EEG ictal pattern

For each patient, ictal source localizations were compared with the intracerebral epileptogenic zone as determined by SEEG. In the ictal spike group (four patients), signal to noise ratios ranged from 4–8 (Table 1), and spiking frequency was < 2 Hz. In the rhythmic activity group, one patient had TLE and one had PCE; signal to noise ratios ranged from 1.5–2. Rhythmic frequencies were in the delta and theta bands. In the paroxysmal fast activity group, one patient had TLE and one had FLE; signal

Discussion

The main objective of this study was to determine the best conditions for locating the epileptogenic zone with HR-EEG in distinct types of scalp-EEG ictal patterns, using SEEG as a validating method. Simultaneous EEG–SEEG recordings would be the best validation method for this kind of study. However, surgical constraints of SEEG (asepsis, presence of guiding screws …) did not allow using a high number of scalp-EEG sensors.

Ten drug-resistant partial epilepsy patients who underwent HR-EEG and

Acknowledgments

LK was supported by France Maintenance Biomedical and the French Ministry of Health (PHRC 17-05, 2009). The authors thank Prof. J. Regis, Prof. J.C. Peragut (Marseille), and Prof. J. Auque and Dr. S. Colnat-Coulbois (Nancy) for their neurosurgical contributions. The authors also thank P. Marquis for HR-EEG and SEEG data management.

References (56)

  • G. Lantz et al.

    Frequency domain EEG source localization of ictal epileptiform activity in patients with partial complex epilepsy of temporal lobe origin

    Clin. Neurophysiol.

    (1999)
  • K. Lehnertz et al.

    Spatio-temporal dynamics of the primary epileptogenic area in temporal lobe epilepsy characterized by neuronal complexity loss

    Electroencephalogr. Clin. Neurophysiol.

    (1995)
  • P. LeVan et al.

    A system for automatic artifact removal in ictal scalp EEG based on independent component analysis and Bayesian classification

    Clin. Neurophysiol.

    (2006)
  • L. Maillard et al.

    Combined SEEG and source localisation study of temporal lobe schizencephaly and polymicrogyria

    Clin. Neurophysiol.

    (2009)
  • I. Merlet et al.

    Dipole modeling of scalp electroencephalogram epileptic discharges: correlation with intracerebral fields

    Clin. Neurophysiol.

    (2001)
  • C.M. Michel et al.

    Localization of the sources of EEG delta, theta, alpha and beta frequency bands using the FFT dipole approximation

    Electroencephalogr. Clin. Neurophysiol.

    (1992)
  • C.M. Michel et al.

    EEG source imaging

    Clin. Neurophysiol.

    (2004)
  • S. Mine et al.

    Ictal dipole source analysis based on a realistic scalp–skull–brain head model in localizing the epileptogenic zone

    Neurosci. Res.

    (2005)
  • R. Oostenveld et al.

    The five percent electrode system for high resolution EEG and ERP measurements

    Clin. Neurophysiol.

    (2001)
  • R.D. Pascual-Marqui et al.

    Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain

    Int. J. Psychophysiol.

    (1994)
  • N. Tanaka et al.

    Dynamic statistical parametric mapping for analyzing ictal magnetoencephalographic spikes in patients with intractable frontal lobe epilepsy

    Epilepsy Res.

    (2009)
  • N.J. Trujillo-Barreto et al.

    Bayesian model averaging in EEG/MEG imaging

    Neuroimage

    (2004)
  • B. Assaf et al.

    Continuous source imaging of scalp ictal rhythms in temporal lobe epilepsy

    Epilepsia

    (1997)
  • F. Bartolomei et al.

    Epileptogenicity of brain structures in human temporal lobe epilepsy: a quantified study from intracerebral EEG

    Brain

    (2008)
  • S. Beniczky et al.

    Source analysis of epileptic discharges using multiple signal classification analysis

    Neuroreport

    (2006)
  • P. Boon et al.

    Dipole modelling and intracranial EEG recording: correlation between dipole and ictal onset zone

    Acta Neurochir. (Wien)

    (1997)
  • P. Boon et al.

    Ictal source localization in presurgical patients with refractory epilepsy

    J. Clin. Neurophysiol.

    (2002)
  • G.E. Chatrian et al.

    A glossary of terms most commonly used by clinical electroencephalographers

    Electroencephalogr. Clin. Neurophysiol.

    (1974)
  • Cited by (104)

    View all citing articles on Scopus
    View full text