On-Scalp Optically Pumped Magnetometers versus Cryogenic Magnetoencephalography for Diagnostic Evaluation of Epilepsy in School-aged Children

Abstract

PEDIATRIC IMAGING D rug-resistant epilepsy occurs in one-third of patients with epilepsy (1). The main treatment to alleviate seizures in those patients is epilepsy surgery, provided that the presumed location of the epileptogenic zone (PLEZ) is focal, well localized, and does not involve functionally eloquent cortices (1). Magnetoencephalography (MEG) provides nonredundant information for the noninvasive localization of the PLEZ in patients with refractory focal epilepsy (RFE) (2,3).Cryogenic MEG systems house hundreds of superconducting quantum interference devices (SQUIDs) in a rigid, one-size-fits-all helmet (4). SQUIDs have several major limitations (4). Due to cryogenic cooling, a thermally insulated gap is required between the scalp and SQUIDs, meaning that the brain-to-sensor distance is approximately 2-5 cm in adults who fit the system well and larger in patients with small heads, such as children. Small head size increases the brain-to-sensor signal attenuation as magnetic fields decrease with the square of the distance. Pediatric SQUID-based MEG (hereafter, SQUID-MEG) systems do not fully alleviate those limitations as they restrict the use of MEG to specific age ranges (eg, infants or school-aged children) (5).Optically pumped magnetometers (OPMs) are cryogen-free magnetic field sensors. OPMs can be placed directly on the scalp to record neuromagnetic signals with an Purpose: To determine if on-scalp MEG based on optically pumped magnetometers (OPMs) alleviates the main limitations of cryogenic MEG. Materials and Methods:In this prospective single-center study conducted in a tertiary university teaching hospital, participants underwent cryogenic (102 magnetometers, 204 planar gradiometers) and on-scalp (32 OPMs) MEG. The two modalities for the detection and localization of IEDs were compared. The t test was used to compare IED amplitude and signal-to-noise ratio (SNR). Distributed source modeling was performed on OPM-based and cryogenic MEG data.Results: Five children (median age, 9.4 years [range, 5-11 years]; four girls) with self-limited idiopathic (n = 3) or refractory (n = 2) focal epilepsy were included. IEDs were identified in all five children with comparable sensor topographies for both MEG devices. IED amplitudes were 2.3 (7.2 of 3.1) to 4.6 (3.2 of 0.7) times higher (P , .001) with on-scalp MEG, and the SNR was 27% (16.7 of 13.2) to 60% (12.8 of 8.0) higher (P value range: .001-.009) with on-scalp MEG in all but one participant (P = .93), whose head movements created pronounced motion artifacts. The neural source of averaged IEDs was located at approximately 5 mm (n = 3) or higher (8.3 mm, n = 1; 15.6 mm, n = 1) between on-scalp and cryogenic MEG. Conclusion:Despite the limited number of sensors and scalp coverage, on-scalp magnetoencephalography (MEG) based on optically pumped magnetometers helped detect interictal epileptiform discharges in school-aged children with epilepsy with a higher amplitude, higher signal-to-noise ratio, and similar localization value compared with conventional cr...

Exploring the limits of MEG spatial resolution with multipolar expansions

“…The anatomical MEG estimates considered in our numerical analyses actually reflect such cases of optimal spatial resolution. This illustrates the enormous potential of scalp MEG for the clinical diagnosis of epilepsy (Feys et al, 2022;Vivekananda et al, 2020;Widjaja, 2022). Contrariwise, high-SNR activity in deep brain regions, such as hippocampal epileptic discharges, would correspond to a small functional brain sphere and thus to both poor spatial resolution and limited improvements brought by scalp MEG (notwithstanding imaginative uses of OPMs to probe deep brain activity; see Tierney et al, 2021).…”

“…The above discussion suggests a negative answer, at least for neocortical epilepsies, because anatomical MEG spatial resolution (which models the situation of epileptiform spike-wave discharges right under the neocortical convexity) already saturates on scalp to its maximal value and thus cannot improve by invasive recordings. Although this line of thoughts would benefit from experimental validation, it adds to the long list of arguments suggesting that scalp MEG based on OPMs may lead, in time, to a step change in the clinical diagnosis of neocortical epilepsy (Feys et al, 2022;Widjaja, 2022).…”

“…The development of scalp MEG based on optically-pumped magnetometers (OPMs) may lead to major advances in this regard (Boto et al, 2018). By avoiding the heavy cryogenics needed for MEG systems based on superconducting quantum interference devices (SQUIDs), this technology allows to place magnetometers closer to the scalp (from about 2-4 cm above scalp for SQUIDs to about 5 mm for OPMs) and thus leads to substantial improvements in signal focality and amplitude (Boto et al, 2016;Feys et al, 2022;Iivanainen et al, 2017Iivanainen et al, , 2020. The ensuing refinement in MEG data remains limited in practice due to the relatively low number of magnetometers in current wearable OPM systems (up to 50 in Hill et al, 2020;Boto et al, 2021) and their sensitivity to environmental noise (Boto et al, 2018;Seymour et al, 2022).…”

OMMR: Co-registration toolbox of OPM-MEG and MRI

“…Source localization for OPM-MEG investigates the neural origin of the brain and has a wide application in neuroscience ( Boto et al, 2021 ; Seymour et al, 2021b ) and clinical research ( Liang et al, 2021 ; Feys et al, 2022 ). Reliable source localization results presuppose the interference suppression technique ( Seymour et al, 2021a ), accurate co-registration of the OPM-MEG and magnetic resonance imaging (MRI) ( Zetter et al, 2018 ), and source imaging methods ( An et al, 2022a ).…”

“…). This field is currently growing, and our toolbox will be updated following this growth.Although we advocate the use of rigid helmets, there are still special requirements for using a flexible cap, especially in clinical research(Duque-Muñoz et al, 2019;Feys et al, 2022). Another advantage of a flexible cap is that it allows the sensors, scalp, and face points to be digitized at the same time(Feys et al, 2022).…”