Analysis of ictal magnetoencephalography using gradient magnetic-field topography (GMFT) in patients with neocortical epilepsy

Highlights

• Gradient magnetic-field topography (GMFT) is useful for ictal magnetoencephalography (MEG).

• GMFT delineates the area of ictal MEG onset (ictal GMFT) at the gyral-unit level.

• The ictal GMFT area is concordant with the ictal onset zone of intracranial electroencephalography.

Abstract

Objective

We aimed to validate the usefulness of gradient magnetic-field topography (GMFT) for analysis of ictal magnetoencephalography (MEG) in patients with neocortical epilepsy.

Methods

We identified 13 patients presenting with an ictal event during preoperative MEG. We applied equivalent current dipole (ECD) estimation and GMFT to detect and localize the ictal MEG onset, and compared these methods with the ictal onset zone (IOZ) derived from chronic intracranial electroencephalography. The surgical resection areas and outcomes were also evaluated.

Results

GMFT detected and localized the ictal MEG onset in all patients, whereas ECD estimation showed localized ECDs in only 2. The delineation of GMFT was concordant with the IOZ at the gyral-unit level in 10 of 12 patients (83.3%). The detectability and precision of delineation of ictal MEG activity by GMFT were significantly superior to those of ECD (p < 0.05 and p < 0.01, respectively). Complete resection of the IOZ in the concordant group provided seizure freedom in 3 patients, whereas seizures remained in 9 patients who had incomplete resections.

Conclusions

Because of its higher spatial resolution, GMFT of ictal MEG is superior to conventional ECD estimation in patients with neocortical epilepsy.

Significance

Ictal MEG study is a useful tool to estimate the seizure onset in patients with neocortical epilepsy.

Introduction

Magnetoencephalography (MEG) is a powerful tool for noninvasive preoperative evaluation of intractable epilepsy. MEG provides information of the spatial distribution of epileptic foci, which corresponds well to the surgical results (Otsubo et al., 2001, Pataraia et al., 2004, Oishi et al., 2006, Knowlton et al., 2006). Analysis with the equivalent current dipole (ECD) method has been conventionally applied for interictal spikes in preoperative MEG. However, ictal events are only occasionally captured during MEG recording. Since Stefan et al. (1992) reported their analysis of ictal magnetoencephalographic activity, several reports have been published on the usefulness of ictal MEG (Shiraishi et al., 2001, Eliashiv et al., 2002, Oishi et al., 2002, Tilz et al., 2002, Assaf et al., 2003, Tanaka et al., 2004, Yoshinaga et al., 2004, Medvedovsky et al., 2012). Although these studies using ECD succeeded in localizing ictal magnetoencephalographic activity, ECD analysis of ictal MEG has been controversial due to the low signal-to-noise ratio (SNR) during ictal activity (Tanaka et al., 2004, Tanaka et al., 2009, Yagyu et al., 2010). Because ECD estimation requires enough SNR for statistically appropriate confidence, the application of ECD for ictal onset activity with low SNR is unreliable. Although reports on several spatial filters to localize the source of ictal onset have recently been published (Tanaka et al., 2009, Yagyu et al., 2010, Fujiwara et al., 2012), there are no adequate methods to validate their estimated results statistically.

Gradient magnetic-field topography (GMFT) was developed to represent the spatio-temporal dynamics of brain surface activity (Hashizume et al., 2007). Unlike ECD estimation or other spatial filters, with GMFT there is no need to solve the biomagnetic inverse problem. Moreover, because GMFT is less influenced by the SNR, it can delineate brain activity even at the early phase of epileptic spikes, which is usually not suitable for appropriate ECD localization (Shirozu et al., 2010). The spatio-temporal accuracy was proved by comparison between GMFT and cortical voltage activity simultaneously recorded with intracranial electroencephalography (EEG) (Shirozu et al., 2016). The aim of this study was to validate the usefulness of GMFT for analysis of ictal MEG. We hypothesized that GMFT is superior to ECD for analysis of ictal neuromagnetic dynamics, even during early ictal activity.