Elsevier

NeuroImage

Volume 52, Issue 4, 1 October 2010, Pages 1279-1288
NeuroImage

Localization and propagation analysis of ictal source rhythm by electrocorticography

https://doi.org/10.1016/j.neuroimage.2010.04.240Get rights and content

Abstract

The purpose of this study was to develop a novel approach for objectively estimating the locations of ictal onset zones by electrocorticography (ECoG). Conventional ECoG analyses have been performed using a 2-D space comprised of intracranial electrodes. Thus, despite the fact that ECoG data have much higher signal-to-noise ratios than electroencephalographic data, ECoG inherently requires a priori information to locate the electrodes, and thus, it is difficult to estimate the depth of epileptogenic foci using this technique. Accordingly, the authors considered that a 3-D approach is needed to determine the presence of an epileptogenic focus in the complex structure of the cortex. However, no source localization procedure has been devised to determine the location of a primary ictal source using ECoG. The authors utilized a spatiotemporal source localization technique using the first principal vectors. A directed transfer function was then employed for the time series of potential ictal sources to compute their causal inter-relationships, from which the primary sources responsible for ictal onset could be localized. Monte-Carlo simulation studies were performed to validate the feasibility and reliability of the proposed ECoG source localization technique, and the obtained results demonstrated that the mean of localization errors with a signal to white Gaussian noise ratio of 5 dB did not exceed 5 mm, even when the source was located ∼ 20 mm away from the nearest electrode. This validated ictal source localization approach was applied to a number of ictal ECoG data sets from six successfully operated epilepsy patients. The resultant 3-D ictal source locations were found to coincide with surgical resection areas and with traditional 2-D electrode-based source estimates. The authors believe that this proposed ECoG-based ictal source localization method will be found useful, especially when ictal sources are located in a deep sulcus or beyond recording planes.

Introduction

Despite recent advances in neuroimaging techniques used for the non-invasive presurgical evaluation of epilepsy, ictal electrocorticography (ECoG) is still considered the gold standard for identifying epileptogenic foci (Engel et al., 1981). However, the interpretation of ictal ECoG recordings is highly dependent on investigator experience and familiarity with the technique, due to the presence of artifacts and a variety of abnormal patterns (Arroyo et al., 1993). The conventional approaches used to determine the locations of epileptogenic foci are based on the use of intracranial electrodes, which depict the first activation on a two-dimensional sensor sheet. The first electrode site to show ictal activity in intracranial recordings often corresponds to the region of maximal epileptogenicity, and amplitude differences are manually mapped on grid figures to define the phase reversal of maximal discharges (Alarcon et al., 1997, Sutherling and Barth, 1989). Furthermore, topographic mapping using ictal ECoG discharges provides a quantitative means of identifying ictal onset zones (Otsubo et al., 2001).

The value of conventional ECoG diagnosis is controversial in both lesional and nonlesional epilepsies. In lesional epilepsy, the seizure-free rate has been reported to be 19% in 30 patients with a temporal mass lesion treated by lesionectomy alone, but 93% in patients that undergo electrophysiologically guided resection (Zooma et al., 1995). However, ECoG might not be necessary for determining surgical resection area in the case of extratemporal glioma lesions (Fried, 1995). Another study also reported good surgical outcomes (a 79% seizure-free rate) after tumor resection, regardless of electrophysiological findings (Morris et al., 1998).

In nonlesional epilepsy, invasive recordings are typically required to define the epileptogenic zone (Lüders and Comair, 2001, Quesney and Niedermeyer, 2005), and ECoG studies have demonstrated the intrinsic value of identifying resectable ictal onset zones (Henry et al., 1999, Jung et al., 1999, Paolicchi et al., 2000). However, it is still difficult to identify epileptogenic points in the sulcus precisely using subdural electrodes.

These shortcomings may arise in part from the inherent limitations of 2-D approaches. A-priori information is needed to locate ECoG electrodes, and it is difficult to estimate the depth of epileptogenic foci from subdural electrode data. Although 2-D electrocortical topographic mapping could be used to indentify ictal onset zones because they are associated with high temporal resolution and a fast sampling rate in lesional and nonlesional cases (Yoshida et al., 2007), there is a potential risk of mislocalization when the current direction is other than radial or a lesion is located some distance from electrodes. Accordingly, a 3-D approach in source space is required to avoid selection bias and to identify epileptogenic foci in the complex cortical structure. However, no ECoG-based source localization procedure has been devised for localizing primary sources.

Many patients with epilepsy have consistent seizure onset foci, and the localization of ictal onset zones is more beneficial for surgical planning than for the mapping of complex interictal patterns (Bebin et al., 1993). The localization of epileptogenic foci has been advanced by source localization and imaging techniques based on electroencephalography (EEG) (Michel et al., 1999, Worrell et al., 2000) and magnetoencephalography (MEG) (Assaf et al., 2003), which characterize seizures in source space rather than in sensor space. Ictal source localization methods integrate characteristic information of ictal sources in space, time, and frequency domains. Some methods utilize estimations of discrete spatiotemporal dipoles to identify ictal generators (Assaf and Elbersole, 1997, Boon et al., 2002), and frequency analysis and dipole localization have also been used in combination for this purpose (Lantz et al., 1999). Another approach was devised using the distributed source model to image more distributed seizure sources (Oishi et al., 2006). The majority of previously described methods use time–frequency parameterization of measured data as an initial step. Moreover, since useful signals may be lost during parameterization and it is difficult to determine signals of interest using time–frequency presentations, a spatiotemporal source localization approach was devised that exploits time–frequency parameterization in the source domain instead of measured channels (Ding et al., 2007). Kamiński and Blinowska (1991) utilized the directed transfer function (DTF) to estimate dynamic causal interaction patterns among sources. Furthermore, the DTF technique was found to provide an excellent guide for the localization of primary ictal sources, and was also used to the channel analysis of invasive recordings to two-dimensionally identify seizure onset electrodes and seizure propagation between electrodes (Franaszczuk and Bergey, 1998, Franaszczuk et al., 1994).

Epileptic studies using intracranial electrodes are advantageous for investigating the dynamics of seizure. An example of one such study is provided by an investigation of the dual pathology between the medial and lateral parts of the temporal lobe using subdural and depth electrodes in order to characterize the activity of epileptogenic foci (Fauser and Schulze-Bonhage, 2006). In another recent analysis of seizure dynamics using multichannel ECoG, a generic change was found in the correlation structures of the data (Schindler et al., 2007), namely, that differences between the propagation times of locally synchronous ictal discharges might disrupt correlations between electrodes. Cortico-muscular dynamics during clonic seizures have also been studied (Hammer et al., 2003). In this previous study, using subdural and depth electrodes, focal clonic seizures were generated by localized polyspike-wave activity in cortical primary motor areas; furthermore, the subthalamic nucleus was not found to be an essential component for the generation of clonic seizures. However, these approaches were all based on the use of 2-D ECoG electrodes, and neuronal dynamics at regions of seizure onset could be further analyzed in more detail by identifying causal relationships between 3-D epileptogenic sources.

Source localization and the estimations of causal interactions have not been popular in ECoG, because this modality measures neuronal activity in a field relatively close to the ictal onset zone, and thus, seizure onset zone is identified based on visual comparisons of signals recorded from different electrodes. However, it is difficult to determine the depths and three-dimensional locations of ictal sources using ECoG alone. Moreover, it may not be possible to place invasive electrodes over the entire area of an epileptogenic zone, and when multiple sources are identified, it is difficult to identify the primary ictal onset source. Therefore, a source analysis procedure is required that determines the primary ictal onset sources by utilizing source localization and propagation analysis in ECoG. Source reconstruction from ECoG data is a relatively new field because only potential maps have been drawn from measurements (Fuchs et al., 2007). In the present study, we evaluated ictal source locations using a spatiotemporal source localization technique and ECoG data. To achieve this, dipole sources were localized using the first principal vectors (FINE) algorithm (Ding and He, 2006), and then causal interactions among sources were estimated to determine the primary ictal onset source. Monte-Carlo simulations were also performed to determine the feasibility of source localization using ECoG.

Section snippets

Subjects

Six subjects with medically intractable seizures were selected for this retrospective study. All subjects were evaluated for seizure surgery using various modalities, which included intracranial subdural electrode arrays for seizure localization. Three patients had lesional epilepsy, and the other three had non-lesional epilepsy. All patients were seizure-free after surgery (Engel class I) (Engel, 1987). A brief list of the patient demographic data is presented in Table 1. Intracranial

Localization errors of simulated data

Localization errors at different distances from the boundary of the electrodes set are shown in Fig. 2. Fig. 2(a) and (b) demonstrate the three-dimensional error distribution in lateral and top views, respectively, and shows that localization error increases with distance from the nearest electrode. When dipole sources are located inside and outside the perpendicular boundary of the electrodes set, localization errors with respect to depth and boundary distance are estimated, respectively. Fig.

Discussion

In the present study, we evaluated a method derived from the directed transfer function technique to identify ictal rhythm by ECoG and epileptogenic foci. A beneficial surgical outcome depends on accurate epileptogenic zone delineation, while avoiding eloquent areas. This approach combines source localization and propagation measures between ECoG ictal sources. Furthermore, dynamic patterns between ictal sources were found to distinguish primary sources of ictal activity from secondary sources

Conclusions

In this study, we originally applied a spatiotemporal source localization technique called FINE to a number of ictal ECoG data sets, and subsequent simulation studies demonstrated the feasibility of ictal source localization using ECoG. Furthermore, propagation analysis using the directed transfer function was found to differentiate ictal onset sources from propagated sources.

Acknowledgments

This research was supported by grants from the Brain Research Center of the 21st Century Frontier Research Program (2009K001280) funded by the Ministry of Education, Science and Technology, the Republic of Korea.

The authors would like to thank Dr. Antti Ahonen and Dr. Ritva Paetau for their valuable comments.

References (55)

  • B.A. Assaf et al.

    Ictal magnetoencephalography in temporal and extratemporal lobe epilepsy

    Epilepsia

    (2003)
  • L. Astolfi et al.

    Comparison of different cortical connectivity estimators for high-resolution EEG recordings

    Human Brain Mapping

    (2007)
  • C. Baumgartner et al.

    Magnetoencephalography in focal epilepsy

    Epilepsia

    (2000)
  • C. Baumgartner et al.

    Neuromagnetic recordings in temporal lobe epilepsy

    Journal of Clinical Neurophysiology

    (2000)
  • E.M. Bebin et al.

    Surgical treatment for epilepsy in cerebral tuberous sclerosis

    Epilepsia

    (1993)
  • P. Boon et al.

    Ictal source localization in presurgical patients with refractory epilepsy

    Journal of Clinical Neurophysiology

    (2002)
  • K.M. Buckley et al.

    Spatial-spectrum estimation in a location sector

    IEEE Transactions on Acoustics, Speech, & Signal Processing

    (1990)
  • L. Ding et al.

    Spatio-temporal EEG source, localization using a three-dimensional subspace FINE approach in a realistic geometry inhomogeneous head model

    IEEE Transactions of Biomedical Engineering

    (2006)
  • J. Engel

    Outcome with respect to epileptic seizures

  • J. Engel

    Surgery for seizures

    New England Journal of Medicine

    (1996)
  • J. Engel et al.

    Correlation of criteria used for localizing epileptic foci in patients considered for surgical therapy of epilepsy

    Annals of Neurology

    (1981)
  • S. Fauser et al.

    Epileptogenicity of cortical dysplasia in temporal lobe dual pathology: an electrophysiological study with invasive recordings

    Brain

    (2006)
  • P.J. Franaszczuk et al.

    Application of the directed transfer function method to mesial and lateral onset temporal lobe seizures

    Brain Topography

    (1998)
  • I. Fried

    Management of low-grade gliomas: results of resections without electrocorticography

    Clinical Neurosurgery

    (1995)
  • M. Fuchs et al.

    Development of volume conductor and source models to localize epileptic foci

    Journal of Clinical Neurophysiology

    (2007)
  • G.H. Golub et al.

    Matrix computations

    (1996)
  • J. Gotman

    Noninvasive methods for evaluating the localization and propagation of epileptic activity

    Epilepsia

    (2003)
  • Cited by (34)

    • Ictal and interictal source imaging on intracranial EEG predicts epilepsy surgery outcome in children with focal cortical dysplasia

      2020, Clinical Neurophysiology
      Citation Excerpt :

      These findings suggest that intracranial ESI improves the 3D demarcation of the SOZ and IZ by facilitating the localization of the actual epileptic generators, which underlie the interictal and ictal electrographic events observed at the contact level. Our work adds to the few previous studies, which investigated the potential of intracranial ESI (Ramantani et al., 2013; Kim et al., 2010), by showing that interictal and ictal ESI offer augmented value in assessing individualized patient prognosis compared to conventional methods used to interpret iEEG and to delineate the IZ or SOZ. Our data indicate that the proximity of the ESI-IZ to the resection has a significant effect on the patient’s outcome with the ESI presenting a higher effect than the conventional approach: for the ESI-IZ, the odds of poor postsurgical outcome were seven times higher than the odds of good outcome for every 1-cm increase in the resection distance; while for the conventionally-defined IZ, the odds of poor outcome were 2.6 times higher than the odds of good outcome.

    • Source-reconstruction of the sensorimotor network from resting-state macaque electrocorticography

      2018, NeuroImage
      Citation Excerpt :

      These findings motivate the development of volume-conductor models and source-reconstruction methods for ECoG recordings (Fuchs et al., 2007; Zhang et al., 2008; Cho et al., 2011, 2013; Kim et al., 2010; Dumpelmann et al., 2012; Ramantani et al., 2013). Currently, however, source-reconstruction of ECoG recordings is relatively underdeveloped, as compared with scalp EEG and MEG recordings and existing studies have largely focused on epileptiform activity (Fuchs et al., 2007; Zhang et al., 2008; Cho et al., 2011, 2013; Kim et al., 2010; Dumpelmann et al., 2012; Ramantani et al., 2013). In this study, we investigate the feasibility of reconstructing source-level RSNs from sensor-level ECoG recordings in macaque monkey.

    • Epileptic Zone Resection for Magnetic Resonance Imaging–Negative Refractory Epilepsy Originating from the Primary Motor Cortex

      2017, World Neurosurgery
      Citation Excerpt :

      Although previous studies have demonstrated that an important factor in seizure freedom after surgery for epilepsy is the completeness of resection,1,9-11,34-36 to protect motor function, we did not resect the more medial part of the precentral gyrus. In our opinion, another reason which contributed to the poor seizure outcome in some patients is that the seizure focus was located at the deep part of the sulcus.37 Therefore, the implanted electrode which did not extend to that area could not find these seizure focuses and eventually the causative lesion was not resected.

    View all citing articles on Scopus
    1

    Current address: Department of Psychology, Yonsei University, 134 Shinchon-dong Seodaemun-gu, Seoul 120-749, South Korea.

    View full text