PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of epicurrLink to Publisher's site
 
Epilepsy Curr. 2017 Nov-Dec; 17(6): 361–362.
PMCID: PMC5706357

Localizing the Ictal Onset: Visualizing the Epileptogenic Target

Commentary

Source Localization of Ictal Epileptic Activity Based on High-Density Scalp EEG Data.

Nemtsas P, Birot G, Pittau F, Michel CM, Schaller K, Vulliemoz S, Kimiskidis VK, Seeck M. Epilepsia 2017;58:1027–1036. [PubMed]

OBJECTIVE: Electrical source imaging (ESI) is a well-established approach to localizing the epileptic focus in drug-resistant focal epilepsy. So far, ESI has been used primarily on interictal events. Emerging evidence suggests that ictal ESI is also feasible and potentially useful. We aimed to investigate the diagnostic accuracy of ESI on ictal events using high-density electroencephalography (EEG). METHODS: We performed ictal ESI on 14 patients (9 with temporal lobe epilepsy) admitted for presurgical evaluation who presented seizures during a long-term (≥18 h) high-density EEG recording (13 with 256 electrodes and one with 128 electrodes), and subsequently 8 of them underwent epilepsy surgery (postoperative follow-up >1 year). Artifact-free EEG epochs at ictal nset were selected for further analysis. The predominant ictal rhythm was identified and filtered (+1 Hz around the main frequency). ESI was computed for each time point using an individual head model and a distributed linear inverse solution, and the average across source localizations was localized. For validation, results were compared with the resection area and postoperative outcome. RESULTS: Ictal ESI correctly localized the epileptic seizure-onset zone in the resection area in five of six postoperatively seizure-free patients. Interictal and ictal ESI were concordant in 9 of 14 patients and partially concordant in additional 4 of 14 patients (93%). Divergent solutions were found in only one of the 14 patients (7%). SIGNIFICANCE: Ictal ESI is a promising localization technique in focal epilepsy.

Localizing epileptogenic sources in medically refractory focal-onset epilepsy is entering a new generation. A growing number of innovative diagnostic technologies and techniques used in combination for identifying epileptogenic foci are at various stages of deployment across well-established epilepsy centers. A multimodal non-invasive diagnostic armamentarium has evolved since the 1990s to facilitate visualizing the ictal-onset zone (1–3). Interictal electrical source imaging (ESI) using high-density scalp electrodes is a computationally intensive method that solves the inverse problem. That is, ESI estimates probable solutions for localizing focal interictal epileptiform activity within the cortical irritative zone.

The most probable inverse solutions, represented as so-called equivalent dipoles, are generated by patches of synchronous epileptogenic neurons oriented perpendicular to the scalp. The neuronal generators of these dipoles arise from the six-layered neocortical ribbon, or the deep three-layered allocortical hippocampal formation. In comparison, ESI is less sensitive to epileptogenic neuronal populations tangentially oriented to the scalp. Such tangential dipoles, also known as horizontal equivalent dipoles, however, can be detected to some extent with this technique. Computationally intensive algorithms can superimpose these interictal equivalent dipoles on high-resolution MRI datasets to estimate the location and orientation of the classically termed “irritative zones.” As a result, ESI can find probable solutions for cortical interictal epileptiform sources that generate often-complex scalp surface voltage topography.

A prerequisite for solving such inverse problems for interictal activity requires high-density EEG scalp electrode arrays (64–256 contacts) (4). Such electrode densities are well beyond the standard-density 10/20 electrode sets (21–27 scalp contacts) typically used in the epilepsy monitoring unit (EMU). Arguably, however, a threshold exists above which more scalp electrodes do not contribute to a greater confidence or improved resolution of determining localization of the irritative zone.

Nemtsas et al. aim to demonstrate that, although more challenging than capturing interictal epileptiform waveforms, high-density ictal ESI in an EMU setting can provide robust complementary information for identifying the ictal-onset zone. Although it is well known that the irritative zone and ictal-onset zone can overlap (5), the authors suggest that high-density ictal ESI is more relevant than relying on interictal ESI alone. Increasing the number of scalp electrodes improves the resolution of source estimation. However, maintaining high-density scalp electrode arrays is challenging in an EMU from both a technical and a patient comfort standpoint. Therefore, the yield of capturing ictal events with high-density scalp electrode nets can be low.

Conversely, solving the so-called forward problem can be potentially improved by optimizing the head model used in the software algorithm by transforming imaging information gathered from gapless high-resolution MRI datasets. Such realistic head model information can estimate with patient specificity the isotropic or uniform bulk conductivities for grey matter, white matter, and CSF.

Nemtsas et al. presented complementary ictal SPECT and PET findings along with an inverse solution comparison for interictal and ictal ESI in 14 adult patients. However, functional neuroimaging data arising from such multimodality information were not used to influence solving the forward problem of electrical currents. It must be emphasized that high-density ESI used in isolation can be insensitive to detecting the ictal-onset zone arising from tangentially oriented deep fissural, sulcal and mesial temporal regions. Although not discussed by the authors, multimodality functional neuroimaging techniques complementing ESI can potentially overcome such limitations.

A workflow can be constructed to use such multimodal information to solve forward problems by seeding the ESI algorithm. That is, concordant multimodal data overlapping the ictal onset zone can act as a starting location for the ESI calculations to augment scalp surface voltage topography generated by one or more deep cortical epileptic sources. Such multimodal datasets include: [1] transient changes in focal water diffusion detected by diffusion tensor imaging (DTI) for visualizing propagation of ictal activity through white matter pathways (6, 7); [2] transient changes in perfusion using subtracted ictal SPECT (SISCOM); [3] metabolism within the ictal onset zone leveraging PET; and [4] complementary magnetic source imaging, also known as MEG, for modeling tangentially oriented epileptiform dipoles arising from deep cortical sulci and fissures.

Such complementary multimodality imaging solutions, performed independently without relying upon interdependent input, can provide a more robust visualization of the epileptogenic source. For example, transient hyperperfusion-related SISCOM signals used as a starting seed for ESI can potentially increase the sensitivity of determining the cortical origin of the ictal onset. DTI and related tractography modeling can connect the dots between the SISCOM signal and ictal ESI, therefore enhancing the sensitivity of identifying the ictal onset zone. Such an interdependent multimodality map can strengthen the confidence level for visualizing the epileptogenic source and associated neural connectivity with individual specificity.

This information can provide a rich visual understanding of how different critical epileptogenic regions, or generator nodes, can arise in the mesial temporal, lateral temporal, or extratemporal regions and be associated with potentially distant mesial temporal sclerosis. Conversely in the absence of a structural lesion, using such a multimodality workflow can transform a non-lesional focal-onset epilepsy into a virtual lesional epilepsy. The concept of an epileptogenic network with critical grey matter zones or nodes can explain the epileptogenic origin and its propagation patterns. That is, the ictal-onset zone can be better represented as pathological neuronal networks, whether micro- or macro-networks, with potentially extensive neural connectivity. Consequently, the term “epileptic focus,” which has been used throughout the epilepsy surgery literature for decades, can become a misnomer. Of course, multiple independent epileptogenic foci without any mutual relationships also exist, as in dual pathology. Lastly, epileptogenic brain regions can include pathways that connect both distant neocortical mantle, and deep subcortical modulatory territories where neuronal populations are not oriented in parallel and, therefore, are invisible to both EEG and MEG.

The hypothesis that interictal and ictal activity share the same neural origins is the impetus for including interictal MEG studies in the presurgical evaluation. However, determining the ictal onset zone by employing interictal MEG or interictal ESI may not necessarily coincide with the ictal onset zone (8). Concordance reliability measures can be potentially maximized when comparing the orientation and location of the interictal dipole models with ictal EEG patterns in combination with other diagnostic functional neurovisualization modalities. In effect, a multimodality presurgical targeting map can be generated for planning resection of the ictal onset zone by improving spatial and temporal resolution and, therefore, minimizing the numbers of intracranial electrodes necessary for pre-resective invasive monitoring.

Nemtsas et al. demonstrate that high-density ictal ESI is a promising technique in select patients for facilitating identification of meaningful surgical targets for medically refractory focal-onset epilepsy. Studies involving larger patient numbers are necessary to determine whether continued development and refinement of complementary non-invasive multimodal functional imaging techniques can meaningfully increase positive clinical outcomes, minimize morbidity, and improve cost-effectiveness following resective epilepsy surgery.

Supplementary Material

References

1. Assaf BA, Ebersole JS. Continuous source imaging of scalp ictal rhythms in temporal lobe epilepsy. Epilepsia 1997; 38: 1114– 1123. [PubMed]
2. O'Brien TJ, So EL, Mullan BP, Hauser MF, Brinkman BH, Bohnen NI, Hanson D, Cascino GD, Jack CR, Sharbough FW. Subtraction ictal SPECT co-registered to MRI improves clinical usefulness of SPECT in localizing the surgical seizure focus. Neurology 1998; 50: 445– 454. [PubMed]
3. Fuchs M, Drenckhahn R, Wischmann HA, Wagner M. An improved boundary element method for realistic colume-conductor modeling. IEEE Trans Biomed Eng 1998; 45: 980– 997. [PubMed]
4. Brodbeck V, Spinelli L, Lascano AM. Electroencephalographic source imaging: A prospective study of 152 operated epilepsy patients. Brain 2011; 134: 2887– 2897. [PMC free article] [PubMed]
5. Megevand P, Spinelli L, Genetti M. Electric source imaging of interictal activity accurately localizes the seizure onset zone. J Neurol Neurosurg Psychiatry 2014; 85: 38– 43. [PubMed]
6. Diehl B, Symms MR, Boulby PA, Salmenpera T, Wheeler-Kingshott CAM, Barker GJ, Duncan JS. Postictal diffusion tensor imaging. Epilepsy Res 2005; 65: 137– 146. [PubMed]
7. Rossi MA, Stebbins G, Murphy C, Greene D, Brinker S, Sarcu D, Tenharmsel A, Stoub T, Stein MA, Hoeppner TJ, Byrne RW, Moseley ME, Bammer RA, Bild S, Dennis J, Arnett N, Balabanov A, Bergen D, Kanner AM, Smith MC. Predicting white matter targets for direct neurostimulation therapy. Epilepsy Res 2010; 91: 176– 186. [PubMed]
8. Pillai J, Sperling MR. Interictal EEG and the diagnosis of epilepsy. Epilepsia 2006; 47 suppl 1: 14– 22. [PubMed]

Articles from Epilepsy Currents are provided here courtesy of American Epilepsy Society