Paediatric magnetoencephalography and its role in neurodevelopmental disorders

Abstract

Magnetoencephalography (MEG) is a non-invasive neuroimaging technique that assesses neurophysiology through the detection of the magnetic fields generated by neural currents. In this way, it is sensitive to brain activity, both in individual regions and brain-wide networks. Conventional MEG systems employ an array of sensors that must be cryogenically cooled to low temperature, in a rigid one-size-fits-all helmet. Systems are typically designed to fit adults and are therefore challenging to use for paediatric measurements. Despite this, MEG has been employed successfully in research to investigate neurodevelopmental disorders, and clinically for presurgical planning for paediatric epilepsy. Here, we review the applications of MEG in children, specifically focussing on autism spectrum disorder and attention-deficit hyperactivity disorder. Our review demonstrates the significance of MEG in furthering our understanding of these neurodevelopmental disorders, while also highlighting the limitations of current instrumentation. We also consider the future of paediatric MEG, with a focus on newly developed instrumentation based on optically pumped magnetometers (OPM-MEG). We provide a brief overview of the development of OPM-MEG systems, and how this new technology might enable investigation of brain function in very young children and infants.

Keywords: ADHD; ASD; MEG; magnetoencephalography; neurodevelopment.

Figure 1.

Types of MEG signals. The top-panel shows event-related fields: the schematic shows how time and phase locked signal fluctuations can be averaged over trials to create an evoked response. The example on the right shows a typical response to an auditory tone. From Hajizadeh A, Matysiak A, May PJC, König R. Explaining event-related fields by a mechanistic model encapsulating the anatomical structure of auditory cortex. Biol Cybern. 2019;113(3):321-345. doi: 10.1007/s00422-019-00795-9; http://creativecommons.org/licenses/by/4.0/). The middle panel shows a neural oscillatory modulation. In the schematic, we show how a single oscillatory waveform can be decomposed into its constituent frequency bands. The example on the right from Hill RM, Devasagayam J, Holmes N, et al. Using OPM-MEG in contrasting magnetic environments. Neuroimage. 2022;253:119084. doi: 10.1016/j.neuroimage.2022.119084; http://creativecommons.org/licenses/by/4.0/ shows a time-frequency spectrogram with modulation of gamma (30+ Hz) activity during a task where subject watch a visual stimulus. Blue reflects a decrease and yellow reflects an increase in amplitude relative to the baseline (here baseline is defined as −3.4 to −2.5 seconds). The bottom panel shows functional connectivity. The schematic shows how it is possible to derive waveforms from spatially separate brain regions and look for a relationship between them. The example on the right from Boto E, Hill RM, Rea M, et al. Measuring functional connectivity with wearable MEG. Neuroimage. 2021;230:117815. doi: 10.1016/j.neuroimage.2021.117815; https://creativecommons.org/licenses/by/4.0/ shows a matrix representing connectivity between a set of parcellated brain regions. The ‘glass brain’ shows the spatial signature of the strongest 200 connections. In this example, we show beta band connectivity, which is strongest in the sensorimotor system. Abbreviation: MEG = magnetoencephalography.

Figure 2.

MEG paediatric systems. By date, the SARA system with an infant participant (Image provided by SARA Lab at University of Arkansas Medical Sciences); a schematic of the BabySQUID system (reprinted from Okada Y, Pratt K, Atwood C, et al. BabySQUID: a mobile, high-resolution multichannel magnetoencephalography system for neonatal brain assessment. Rev Sci Instrum. 2006;77:024301, with the permission of AIP Publishing); the KIT infant system (left) next to the adult system (Hirata M, Ikeda T, Kikuchi M, et al. Hyperscanning MEG for understanding mother-child cerebral interactions. Front Hum Neurosci. 2014;8:118. doi: 10.3389/fnhum.2014.00118; https://creativecommons.org/licenses/by/3.0³⁵); the Artemis123 (Roberts TPL, Paulson DN, Hirschkoff E, et al. Artemis 123: development of a whole-head infant and young child MEG system. Front Hum Neurosci. 2014;8:99; doi: 10.3389/fnhum.2014.00099; https://creativecommons.org/licenses/by/3.0/³³) and the BabyMEG (top right; reprinted from Okada Y, Hämäläinen M, Pratt K, et al. BabyMEG: a whole-head pediatric magnetoencephalography system for human brain development research. Rev Sci Instrum. 2016;87(9):094301 with the permission of AIP Publishing). Abbreviations: MEG = magnetoencephalography; SARA = SQUID Array for Reproductive Assessment.

Figure 3.

Delayed auditory event related field in ASD (adapted from Roberts TPL, Matsuzaki J, Blaskey L, et al. Delayed M50/M100 evoked response component latency in minimally verbal/nonverbal children who have autism spectrum disorder. Mol Autism. 2019;10:34. doi: 10.1186/s13229-019-0283-3; https://creativecommons.org/licenses/by/4.0⁴⁰): The 3 waveforms show representative auditory evoked responses in a typically developing (TD) child (top), a verbal child with ASD who has no intellectual disability (ASD-V; middle) and minimally verbal or nonverbal child (MVNV) with ASD (bottom). The gray lines indicate the expected response at 50 ms (M50 peaks). For the representative TD child a response was observed around 71 ms, for the representative ASD-V child at 81 ms, and for the representative ASD-MVNV child at 98 ms. Abbreviations: ASD = autism spectrum disorder.

Figure 4.

Age-related differences in gamma connectivity. Brain network (left-side) representing the significant group-by-age interaction to emotional faces in the gamma-band (30-55 Hz). The scatterplot (right-side) shows that connectivity strength within this gamma network increases with age in the TD group (in red) and decreases with age in the ASD group (in blue). From Safar K, Vandewouw MM, Taylor MJ. Atypical development of emotional face processing networks in autism spectrum disorder from childhood through to adulthood. Dev Cogn Neurosci. 2021;51:101003. License: CC BY-NC-ND 4.0.

Figure 5.

OPM-MEG system with (A) an adult helmet and (B) a child helmet.