Revolutionizing Pediatric Neurophysiology with Magnetoencephalography

This paper considers the current and future use of magnetoencephalography (MEG) for assessing neural activity in children (birth to 18 years old), including the well-established use of SQUID (Superconducting QUantum Interference Device) MEG technology as well as the very rapidly developing Optically Pumped Magnetometry (OPM) technology. 1 A primary conclusion is that the changing landscape of pediatric neurophysiology studies foretells a revolution in electromagnetic neuroimaging. These changes will produce some discontinuity, progressing away from what once worked well enough, namely, examining neural activity at the level of the EEG or MEG sensor, but is not up to current and anticipated challenges. Given features intrinsic to MEG, including simpler mathematical models for source localization and higher-dimensional representation of neural activity, little effect of open fontanelles and sutures on infant neural measures, and reference-free neural measures, MEG will often be the preferred method for assessing neural activity in children. In particular, non-invasive, whole-brain MEG sensor data with source localization provide measures of neural activity in brain space that richly represent the structure and maturation of neural activity spanning both local and regional processes, as well as measures of connectivity within and between brain regions. Assessing neurophysiology in brain space is also essential for associating local neural activity with local brain structure (e.g., gray and white matter) and brain chemistry (e.g., magnetic resonance spectroscopy data). It is also highly likely that MEG data are more future-proof than EEG data (higher dimensionality, ease of source localization), especially for advanced source localization methods as well as advanced analysis methods yet to be developed and applied to previously collected data. The emergence of OPM-based MEG opens a new age of imaging for children and infants, such as translating the source localizing abilities of MEG in adults to wearable systems in young children. Looking forward, greater access to MEG and other advanced imaging technologies, the accessibility of greater computational power, and the rapid development of open-source software will combine to improve our methods and inform our research questions, all leading to a better understanding of how the human brain changes and supports behavioral development from birth to adulthood. Revolutionizing Pediatric Neurophysiology with Magnetoencephalography Edgar, J.C. a,b, Wilson, T.W. c,d, Taulu, S. e,f, Nugent, A.,C. g, Ahlfors, S.P. h, Kenet, T. i, Chen, Y. a,b, Bosseler, A.N e, Green, H.L. a, Heinrichs-Graham, E. c,d, Gaetz, W. a,b, Larson, E. e, Candelaria-Cook, F.T. j, Bell, N.M. k, Davenport, E.M. k, Howell, B. l, Evans, M. m, Ahtam, B. n, Huang, M.X. o, Ward, T.W. c, Share, B. p, Mody, M. h, Son, J.J. c, Cellier, D. q, Peatfield, N.A. r, Bardouille, T. s, Bailey, L.M. s, Grewal, S. t, Pratt, K. u, Knappe, S. v,w,x, Alem, O. v,w,x, Miller, G.A. y,z,aa, Stephen, J. j a Lurie Family Foundations MEG Imaging Center, Department of Radiology, The Children’s Hospital of Philadelphia, Philadelphia, PA, USA b Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA c Institute for Human Neuroscience, Boys Town National Research Hospital, Boys Town, NE, USA d Department of Pharmacology and Neuroscience, Creighton University, Omaha, NE, USA e Institute for Learning and Brain Sciences, University of Washington, Washington, WA, USA f Department of Physics, University of Washington, Washington, WA, USA g MEG Core Facility, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA h Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA i Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA j The Mind Research Network a Division of Lovelace Biomedical Research Institute, Albuquerque, NM, USA k Advanced Neuroscience Imaging Research (ANSIR) Laboratory, University of Texas Southwestern Medical Center, Dallas, TX, USA L Fralin Biomedical Research Institute at Virginia Tech Carilion, Roanoke, VA; Department of Human Development and Family Science, Virginia Tech, Blacksburg, VA, USA m Fralin Biomedical Research Institute at Virginia Tech Carilion, Roanoke, VA; Translational Biology, Medicine, and Health Program, Virginia Tech, Roanoke, VA, USA n Department of Pediatrics, Boston Children’s Hospital, Harvard Medical School, Boston, MA, USA o Department of Radiology, University of California, San Diego, California, USA p ANT North American, Inc., Nashville, TN, USA q Department of Cognitive Science, University of California, San Diego. La Jolla California, USA r HeyBro, Inc., Vancouver, BC, Canada s Department of Physics & Atmospheric Science, Dalhousie University, Halifax, NS, Canada t Department of Neurology, The Children’s Hospital of Philadelphia, Philadelphia, PA, USA u Tristan Technologies, Inc., San Diego, California v Paul M. Rady Department of Mechanical Engineering, University of Colorado Boulder, Boulder, Colorado, USA w FieldLine Industries, Boulder, Colorado, USA x FieldLine Medical, Boulder, Colorado, USA y Department of Psychology, University of California, Los Angeles, California, USA z Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, California, USA aa Department of Psychology and Beckman Institute, University of Illinois Urbana-Champaign, Urbana-Champaign, Illinois, USA

This paper considers the current and future use of magnetoencephalography (MEG) for assessing neural activity in children (birth to 18 years old), including the well-established use of SQUID (Superconducting QUantum Interference Device) MEG technology as well as the very rapidly developing Optically Pumped Magnetometry (OPM) technology. 1 A primary conclusion is that the changing landscape of pediatric neurophysiology studies foretells a revolution in electromagnetic neuroimaging. These changes will produce some discontinuity, progressing away from what once worked well enough, namely, examining neural activity at the level of the EEG or MEG sensor, but is not up to current and anticipated challenges. Given features intrinsic to MEG, including simpler mathematical models for source localization and higher-dimensional representation of neural activity, little effect of open fontanelles and sutures on infant neural measures, and reference-free neural measures, MEG will often be the preferred method for assessing neural activity in children. In particular, non-invasive, whole-brain MEG sensor data with source localization provide measures of neural activity in brain space that richly represent the structure and maturation of neural activity spanning both local and regional processes, as well as measures of connectivity within and between brain regions. Assessing neurophysiology in brain space is also essential for associating local neural activity with local brain structure (e.g., gray and white matter) and brain chemistry (e.g., magnetic resonance spectroscopy data). It is also highly likely that MEG data are more future-proof than EEG data (higher dimensionality, ease of source localization), especially for advanced source localization methods as well as advanced analysis methods yet to be developed and applied to previously collected data. The emergence of OPM-based MEG opens a new age of imaging for children and infants, such as translating the source localizing abilities of MEG in adults to wearable systems in young children. Looking forward, greater access to MEG and other advanced imaging technologies, the accessibility of greater computational power, and the rapid development of open-source software will combine to improve our methods and inform our research questions, all leading to a better understanding of how the human brain changes and supports behavioral development from birth to adulthood.

In the spring of 2023, 2024, and 2025, several pediatric brain imaging scientists from across North America (the co-authors on this paper) met to consider the current and future use of magnetoencephalography (MEG) for pediatric research in those under 18 years old. This paper reports on the key themes discussed at these three-day retreats, including the well-established use of SQUID (Superconducting QUantum Interference Device) MEG technology and the rapidly developing optically pumped magnetometry (OPM) approaches. In this paper we demonstrate how advanced electromagnetic brain imaging provides an understanding of neural activity in children that is more detailed than previously possible. As described throughout this paper, given its inherent advantages, MEG will often be the preferred method for studying neural activity in children. One primary goal of pediatric brain imaging studies should be to increase the precision of our measure of neural activity. Analyses of EEG and MEG sensor data that are limited to scalp-level measures are increasingly not scientifically satisfactory, as the provided results are difficult or impossible to interpret in a way that would expand our understanding of neural phenomena or open up new territory. For example, when describing the maturation of resting-state neural activity, acceptable solutions should not look only for scalp-region differences in sensor space. Rather, and as Sections 2 and 3 describe in detail, acceptable solutions will demonstrate differences in maturation via reporting on neural activity in brain space (i.e., describing maturation within and between brain areas). Similarly, acceptable solutions will usually not involve reporting on how EEG or MEG sensor activity, which almost always contains superimposed contributions from neural activity in multiple brain regions, changes across child development. The measures that best assess brain area-specific neural activity should be those that inform our understanding of neural activity in children. In most cases, these will be measures of brain activity in source space. Although for some purposes sensor-level analysis will continue to contribute in clinical and research contexts, making brain source analysis the standard in clinical and research contexts will produce some disruption and discontinuity, yet the case for it is compelling, as the rewards will certainly exceed the costs. This report is divided into five parts. Section 1 makes the case for examining neural activity in pediatric populations. Section 2 describes the benefits of MEG relative to EEG for assessing neural activity in children. Section 3 considers resting-state neural activity to provide an example of the need for data with a high information content (here referencing work demonstrating higher dimensional data for MEG than EEG) and good spatial resolution of source localization (here detailing problems associated with examining sensor measures) to improve our understanding of neural activity in children in a way that would substantially advance the field. Section 4 considers the cost of neurophysiology research, in terms of both money spent and amount of information acquired, as well as the perceived difficulty conducting multisite MEG studies. Section 5 considers the future of OPM MEG as a potentially revolutionary improvement in research and clinical practice. Section 1: Examining neural activity in children The shift in research strategy in the developmental neurophysiology literature over the last two decades, from primarily studying adult populations to more often studying pediatric populations, is occurring both at the level of individual laboratories and at the level of nations, with large pediatric brain imaging programs sponsored by governments such as the ChildBrain project in the European Union (https://cordis.europa.eu/project/id/641652), the Adolescent Brain Cognitive Development study in the United States (ABCD; https://abcdstudy.org/), the Province of Ontario Neurodevelopmental Disorders Network in Canada (POND; https://pond-network.ca), and the more recent HEALthy Brain and Child Development Study in the United States (HBCD; https://heal.nih.gov/research/infants-and-children/healthy-brain). The rationale for these endeavors is articulated in the NIMH 2023 Strategic Plan ( 1 ), in which Goal 2 notes: “Most mental illnesses first present in childhood or adolescence, yet mental illnesses are likely the late behavioral manifestations of changes that began years earlier. These early alterations may influence the course of brain and behavioral development and establish the trajectories of mental illnesses. To better understand these trajectories, we need to develop a comprehensive picture of typical and atypical brain and behavioral development across the lifespan.” The urgency to understand typical and atypical early brain development is further underlined by the emergence of the Developmental Origins of Health and Disease (DOHAD; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5856182/) field, which has established that early environmental insults increase the risk of adult neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease and has also found that these insults may originate in utero . Although we have learned an impressive amount about the brain from examining hemodynamic signals with fMRI ( 2 – 6 ), there is increasing agreement among neuroscientists that understanding brain function requires deep insight into the millisecond-level dynamics of neural function. In large part, this is because the brain processes information via electro-chemical signaling that typically occurs on the order of milliseconds. In line with NIMH’s Research Domain Criteria (RDoC) initiative ( 7 ), the NIMH 2023 Strategic Plan prioritizes studies examining the circuits involved in complex behavior using high spatial and temporal resolution techniques (Goal 1). The urgent need for studies that noninvasively assess neurophysiological function in typically developing as well as patient populations is also noted. For example, NIMH 2023 Strategic Goal 1 prioritizes, “Developing novel, age-appropriate imaging assays with higher spatial and temporal resolution for visualization and analyses of brain structure, maturation, connectivity, and function, with particular emphasis on advancing real-time measurement approaches” (Strategy 1.1.C.3) and “Conducting brain-wide analyses to determine which neural circuits drive network patterns associated with a pathology” (Strategy 1.3.A.1). Given such goals, the assessment of anatomically localized activity has significant advantages relative to raw sensor measurements, which, as discussed below, problematically include activity from many different brain areas (even a spatially focal source in the brain produces a wide-spread pattern of scalp potentials and magnetic fields; this is often referred to as “volume conduction” in EEG) ( 8 – 10 ). Among other advantages, source activity mapping facilitates the discovery of an individual’s neural fingerprint ( 11 ) as well as control and case group differences that are region-specific ( 12, 13 ). Invasive electrophysiology, utilizing electrodes placed directly in the brain or on the brain surface, provides remarkable richness of data. However, such data can only be obtained from patients undergoing surgery. The present paper only addresses non-invasive electromagnetic methods (EEG and MEG) in pediatric populations (NIMH Strategy 1.3.A.1). We focus on the use of MEG to measure brain activity, based on evidence that MEG methods have advanced to where they can uniquely contribute to our understanding of neural activity ( 14 ). Measures of interest include neural activity in the time domain (e.g., evoked components) and the time-frequency domain (e.g., inter-trial coherence, cross-frequency coupling, and total power measures). Source-space analyses in both domains offer far richer information than traditional sensor-space analyses, such as measures of functional connectivity that can be examined both locally (e.g., alpha-to-gamma phase-amplitude coupling within a specific brain area or inter-region connectivity) and globally (e.g., frequency-based connectivity and graph-theory analyses) ( 15 – 19 ). As methods for assessing neural activity continue to evolve, offering increasingly detailed insights into neural processes, MEG data are particularly well-suited for these advanced analyses. Multimodal studies are also increasingly used to assess how brain structure and brain chemistry are related to neural function. For example, pediatric studies increasingly report associations between brain structure (e.g., structural MRI) or chemistry (e.g. MRS) and brain function (e.g., EEG or MEG) ( 13, 20 – 25 ), including longitudinal studies examining the co-maturation of brain structure and function ( 26, 27 ). Such multimodal studies are crucial to understanding normative brain development (e.g., how maturational changes in white matter and neural processing speed are related) and will surely help identify brain abnormalities and inform potential treatments in patients. Multimodal assessment of brain structure and function in longitudinal pediatric samples also informs adult studies, as the longitudinal study of neurodevelopment provides a unique opportunity to determine causality in the linkages between structure and function as cognitive abilities emerge across childhood. Longitudinal (and cross-sectional) studies also provide the ability to identify age-specific neurophysiology measures, with some of the EEG and MEG phenomena investigated in adults not present in infants and young children ( 28 – 34 ). In addition, studies investigating neurodevelopmental disorders have shown divergent developmental trajectories in patients versus typically developing populations (e.g., Figures 2 and 3 in ( 35 ), Figure 4 in ( 36 ), Figure 3 in ( 37 ), and for a review ( 38 )). Multimodal studies will greatly benefit from source-space analyses as methods such as MRI and MRS yield results that are naturally in source space, meaning that a faithful correlation of results across methodologies can only take place if the MEG/EEG results are also mapped into source space. In sum, there is broad consensus in the field that if we want to understand brain activity in children, then we must understand how populations of neurons function with millisecond temporal precision (i.e., the neural timescale), locally within specific brain areas, as well as across different brain areas. As detailed below, progress in this area is best accomplished via studying brain activity in anatomically localized source space rather than in sensor-space signals. Section 2: Non-invasively assessing neural activity in children using EEG and MEG 1 Figure 1 shows examples of MEG and OPM systems. MEG and EEG are safe, quiet, and straightforward methods to non-invasively study neural function with high temporal resolution. This section considers the relative advantages of MEG and EEG with respect to measuring neurophysiological activity in pediatric populations. MEG and EEG instrumentation and analysis methods are not discussed in this paper; seminal papers and books on this topic are provided in Footnote 1. In the context of the goal of evaluating brain activity in anatomically localized source space, below we make a case for the use of MEG in pediatric studies, emphasizing advantages of MEG related to the physics and mathematics of signal generation (i.e., the so-called “forward problem” of determining how neural activity produces externally measured scalp potentials (EEG) and magnetic fields (MEG). Although both EEG and MEG signals have their origin mainly in the dendritic currents in pyramidal neurons, the magnetic field detected with MEG has several properties that simplify the interpretation of the data in terms of neural source activity in the brain. The important practical question of the availability and cost of EEG and MEG is addressed in Section 4. A premise for the case for MEG is that neural function characterized at the most fundamental level available is an essential component of mapping systems-level phenomena. To this end, MEG scientists often seek to obtain measures of neural activity in “brain space”, using source-localization methods to transform the extracranial signals to estimated spatiotemporal patterns of activity within the brain. One important reason to focus on brain space is that sensor measures typically contain brain activity from many different brain areas (the superposition of brain activity ( 39, 40 ), see examples in Section 3), as well as non-brain noise such as electromagnetic activity from face and neck muscles, eye-blinks, and heartbeat, and these signals are often difficult to separate. As such, analysis methods that are less susceptible to this problem are much preferable. In infants, EEG and MEG exams are relatively well tolerated (especially compared to MRI exams) and can be performed when infants are awake or asleep. The recent availability of high-density infant EEG sensor nets has greatly improved the topographic resolution of EEG and the feasibility of such recordings. However, EEG is strongly dependent on the different conductivities of various tissues (e.g., brain, cerebral spinal fluid, scalp, and skull) ( 41 – 43 ). Furthermore, distortions of scalp potentials due to open fontanels and sutures in younger infants further add to the challenges inherent in EEG. The conductivity values needed for EEG source localization are not only difficult to determine in infants but change during infant development due to skull maturation ( 41, 42 ) . In contrast, as detailed in Okada et al. ( 44 ), the magnetic field outside of the scalp produced by neuronal currents is only weakly affected by inhomogeneities and anisotropy of the electrical conductivity within the head ( 45 – 47 ), a fundamental advantage for the use of MEG in participants of all ages. The magnetic field is also relatively unaffected by the fontanels and sutures that provide paths of high electrical conductivity through unfused cranial bones ( 42, 43, 48 – 50 ). Crucially for infant clinical and research studies, and especially those with longitudinal designs, MEG is thus largely unaffected by skull maturation. These properties provide notable advantages for MEG over EEG with respect to the accuracy of the forward model and thus for source localization in general ( 51 ) , thus providing more valid mapping of such neural activity at any age, especially as the brain and skull mature early in life. In addition to this pervasive structural biophysical advantage, a second advantage of MEG over EEG is the nature of the raw sensor data. In EEG, scalp potential recordings are always differential, requiring a choice of a reference electrode and thus an indirect measure of the electrical field. In contrast, in MEG the magnetic field can be recorded directly at a single location, without requiring a reference measurement. The reference issue complicates the practical interpretation of EEG sensor waveforms. However, the choice of the reference electrode does not affect source localization analyses ( 52 ), further emphasizing the benefits of analyzing data in brain space (for discussion of issues associated with the choice of EEG reference and inference about brain function, see an editorial by Bringas Vega et al. ( 53 ) and the associated collection of papers). A third advantage of MEG over EEG is the richness of MEG data, demonstrated in a recent paper by Taulu and Larson ( 54 ). Comparing the properties of electrical and magnetic recordings, Taulu and Larson showed that there is inherently less spatial information available in EEG than in MEG data. They noted (p. 999): “…even without considering any effects caused by conductivity boundaries [and the much greater distortions they create in EEG than MEG], the little spatial information in the measured EEG signal can be understood by the fact that the electric potential is always a proportional quantity between two points”. This means that it is difficult to recover the finer spatial details of the underlying source currents. In contrast, MEG records magnetic field, comparable to the electric field rather than the electric potentials. For a whole-head recording, Taulu and Larson ( 54 ) estimated up to 80 degrees of freedom for high-density MEG versus less than 40 degrees of freedom for high-density EEG. As a real-world example, they examined an auditory evoked response recorded using a 60-channel EEG system. They showed that using just the 21 sites of the standard 10-20 EEG electrode montage they could explain more than 93% of the 60-channel auditory response variance. Thus, the standard 10-20 EEG montage yields essentially the same amount of information as the 60-channel montage. Their mathematical demonstrations indicate that whole-head MEG data contain considerably more neural information than whole-head EEG data and thus likely have higher utility (e.g., better source localization, better functional connectivity measures). Although sparse channel arrays sometimes suffice to test research hypotheses and are of clinical interest (e.g., assessing brain injury on the sports field ( 55, 56 ), for use as a predictive marker in individuals at clinical high risk for psychosis ( 57, 58 ), the sparseness of EEG data limits our ability to characterize what we know to be regionally specific, complex neural circuit processes. Many clinical uses of EEG are well established, including diagnosis of epilepsy and sleep disturbances with new applications quickly emerging such as ambulatory EEG and EEG in a home setting ( 59 – 62 ). For clinical applications, a case for a very targeted data collection and analysis strategy (e.g., focusing on a few EEG or MEG channels) can sometimes be made ( 63, 64 ). The recent renaissance in EEG was also discussed, in particular the development of mobile EEG systems (able to tolerate motion ( 65 ), portable EEG systems (small physical size, including EEG devices with wireless capability( 66 )), and transparent EEG systems (portable, motion tolerant, self-applicable, easy to wear, and unobtrusive ( 65 )). Although enabling studies outside the laboratory, these applications are usually low-density (<<64 channels; e.g. see ( 67, 68 )), thus providing low-dimensionality data, limiting suitability for source localization. As an example of cutting-edge work in this area is the use of high-density EEG infant caps in preterm infants in the Neonatal Intensive Care Unit ( 69, 70 ). A major limitation of MEG (both SQUID and OPM technologies) is that recordings in the field are not possible, precluding some essential studies. Of note is the need to distinguish between the richness of MEG versus EEG data in general from the accuracy of MEG and EEG data as a basis for localizing brain activity in specific regions. Studies have shown that MEG and EEG can have comparable accuracy of localizing a single focal source, e.g., early event-related responses in primary sensory areas ( 71 – 74 ). For bilateral auditory cortex activity, however, the biophysics and the evidence favor MEG over EEG. MEG is preferentially sensitive to superficial, tangentially-oriented neural currents such as the superior temporal gyrus (STG) auditory generators ( 40 ). This is true even in infants and children ( 30, 75 ), with the differentiation of left and right auditory cortex important given hemisphere differences in the maturation of auditory encoding processes ( 76 ). In contrast, with EEG, the orientation of those bilateral auditory sources leads to a single maximum located at Cz, depending on the reference, thus making it challenging to differentiate left- and right-hemisphere activity. Reporting on multidetermined auditory EEG midline sensors (where the auditory event-related potentials are the largest) versus left and right STG auditory generator activity is one example of a more general difference in analysis strategy between EEG (often focusing on sensor measures) and MEG (often focusing on source space). EEG papers often infer regional differences in brain activity based on analysis of groupings of spatially adjacent EEG sensor ROIs. Such an analysis strategy encourages what is often an unwarranted correspondence between a sensor ROI (e.g., frontal electrodes) and the brain area near the EEG sensors (frontal lobes). Authors should justify such assumed sensor-to-brain correspondences, demonstrating that the sensors provide valid and reliable measures of neural activity from the inferred brain region(s). Authors should also demonstrate that their sensor ROI is unbiased, demonstrating (a) across participants and ages that the same neural generator(s) contribute equally to the sensor ROI and (b) across participants and ages as well as within an individual that the homologous left- and right-hemisphere generators equally contribute to the sensor ROI. Such a sensor ROI approach will often be difficult to defend, with studies examining differences in the orientation of neural generator activity (often due to individual differences in gyrus anatomy) showing large hemisphere and between-subject differences in the orientation of the neurons giving rise to the EEG and MEG sensor measures ( 40, 77 ). Analyzing data in source space at the individual level and performing group analysis on source-level data mitigates this difficulty. When comparing MEG and EEG, of note is the different spatial sensitivity properties of MEG and EEG. For example, whereas MEG is less sensitive to source currents that are strictly radially oriented, EEG can detect neural generator sources of all orientations ( 78 ). One consequence of this is that MEG tends to favor focal and EEG extended cortical sources ( 79, 80 ). Furthermore, although both MEG and EEG are most sensitive to superficial sources, EEG is better suited than MEG to detect signals from deep structures (e.g., thalamus or brainstem). These often-noted differences between technologies, however, are simplistic. First, because of the reduced sensitivity of MEG to radially oriented sources, it is frequently assumed that MEG is insensitive not only to deep sources but to superficial gyral sources. Several studies have investigated this claim. Hillebrand and Barnes ( 78 ) found that source depth, and not orientation, was the main factor affecting the sensitivity of MEG to activity in human cortex. They also noted that, “there are thin strips (~2mm wide) of poor resolvability at the crests of gyri; however, these stripes account for only a relatively small proportion of the cortical area and are abutted by elements with a nominal tangential component that are highly resolvable due to the proximity to the sensor array”. Second, numerous studies have provided compelling examples of successful MEG differentiation of deep sources (e.g., hippocampus: ( 81 – 84 ), cerebellum ( 85, 86 ), and amygdala ( 37, 87, 88 )). Breakout Box 1 further discusses the localization accuracy of EEG and MEG. Collecting simultaneous MEG+EEG data is often useful, due to the different spatial sensitivity properties of MEG and EEG, thus providing complementary information ( 89 – 92 ). Also, as the degree of interference due to external disturbances and physiological signals of non-brain origin (eyes, heart, movement) differs for MEG and EEG, combined MEG+EEG recordings can help to dissociate signal artifacts from true brain activity. Because of these complementary properties, simultaneous recording of MEG and EEG can be highly beneficial ( 93 ). For example, interictal epileptogenic discharges can be identified in some patients with MEG only or with EEG only (e.g., ( 94, 95 ). Maximizing the information content using simultaneous MEG and EEG is important in clinical studies, such as presurgical mapping of epileptogenic as well as intact brain regions ( 96, 97 ). Combining MEG and EEG can also improve source localization (e.g., ( 98, 99 ) and may help distinguish sources ( 64, 73 ). Simultaneous MEG+EEG is expected to be particularly beneficial in exploratory studies in which new findings and phenomena are described ( 100 ). A downside of simultaneous EEG and MEG is the increased subject preparation time required for attaching EEG electrodes. Geodesic EEG nets with saline sponge electrodes ( 101 ), which can be applied very quickly, are not well suited for simultaneous use with MEG, because they typically take too much space around the scalp to fit inside the MEG helmet, although this is more feasible in pediatric populations run in MEG helmets sized for adult heads. Long preparation times can affect the vigilance and general state of the child and thus the quality of the neural data and task performance. In infants, given a very short period when the infant is awake and oriented, obtaining simultaneous EEG and MEG is very difficult. A minor disadvantage of simultaneous EEG and MEG recording is that even thin EEG electrodes under a fixed MEG helmet slightly increase the distance between the child’s head and the MEG sensors, thereby reducing the strength of the MEG signals. The choice of whether to undertake simultaneous EEG with MEG in clinical or research studies depends on the study goals. Section 3: Resting-state neural activity: An example demonstrating the need for high-dimensionality neural measures with good spatial resolution when studying brain maturation from infancy to late adolescence This section focuses on resting-state (RS) measures to demonstrate the need for brain imaging methods with high information content in order to understand how the human brain is constructed and how it matures. There is a very large EEG and MEG literature on the developmental trajectory of RS neural activity. Studies dating to the 1940s demonstrate changes in RS EEG from birth through adulthood, with age-related decreases in delta and theta activity and age-related increases in alpha, beta, and gamma activity ( 102 – 115 ). Within a frequency band, of note are regional/functional differences. For example, within the alpha band (8 to 13 Hz in adults and lower in children), there is a need to differentiate between maturation of parietal-occipital alpha rhythms modulated by closing and opening the eyes ( 110, 116, 117 ), sensorimotor alpha rhythms modulated by somatosensory input and movements (mu rhythms, ( 118, 119 ), and superior temporal gyrus alpha rhythms modulated by sound ( 118 ). Unfortunately, across decades many studies have ignored the complexity of RS neural activity, including spatial variation, temporal variation on the scale of seconds, minutes, or hours, and longitudinal trajectories. MEG studies are beginning to address this complexity head on ( 19, 120 ). As an example, Rempe et al. ( 120 ) showed brain-region and sex differences in maturation of RS neural activity in a MEG source-localization study assessing RS activity across the lifespan (433 individuals 6-84 years). Figure 2 (from Rempe et al.) shows that whereas the lower delta and theta frequencies showed a negative correlation with age, the higher alpha, beta, and delta frequencies correlated positively with age. These correlations were further probed with hierarchical regressions, which revealed significant nonlinear trajectories in several brain regions. To our knowledge, no developmental or aging EEG studies have provided RS findings with comparable spatial resolution. Recently, scientists have also called into question longstanding RS EEG and MEG findings, noting that most RS power-spectrum analyses conflate two RS brain processes: aperiodic background activity (contributing power across all frequencies if decomposed into sinusoids) co-existing with periodic oscillations (e.g., actual neural oscillations visible as peaks in the power spectrum, such as the RS dominant oscillation) ( 121 – 123 ). A growing literature demonstrates the need to distinguish these, specifically to parameterize the RS power spectrum in order to accurately characterize RS maturational changes separately for aperiodic and periodic power ( 124 – 126 ), with power values in brain space preferred to those in sensor space ( 127 ). For parameterized RS data, of note is the benefit of neurophysiological measures that are reference-free. Measurement of the exponent (slope) of the resting-state power spectrum is hypothesized to provide a non-invasive measure of the neural-circuit excitatory:inhibitory (E:I) balance ( 128 )). Valid determination of this value using EEG sensor measures is complicated by the fact that different referencing strategies, such as bipolar vs. common average, produce very different exponent estimates ( 129 ). The RS power spectrum exponent and offset parameters can be more directly measured via source-space analyses. Describing maturational change in neural activity is of growing interest, including parameterization of the RS power spectrum in children. Most RS parameterization efforts have so far parameterized the EEG sensor power spectra, which reflect activity from multiple brain regions (for discussions of the problems associated with this approach see ( 40, 77, 130 – 132 )). The mixing of activity from multiple brain sources due to volume conduction ( 8, 9 ) has been especially severe due to most studies obtaining power spectra from an average of EEG channels ( 133, 134 ) or via two or more regional clusters of EEG sensors ( 134 – 140 ). Understandably, the EEG cluster studies have produced inconsistent results: Rico-Picó et al. ( 137 ) and Schaworonkow and Voytek ( 140 ) observed statistically significant regional differences in the aperiodic measures, whereas several other studies did not observe sigificant regional differences ( 135, 138, 139, 141 ) or did not statistically assess regional differences ( 136 ). Recent studies have empirically demonstrated the difficulties interpreting sensor findings, because, due to volume conduction, sensor measures reflect activity from multiple brain areas. Schaworonkow and Nikulin ( 142 ) illustrated this with RS activity via showing how, due to volume conduction, RS alpha activity from occipital, sensorimotor, and superior temporal gyrus neural sources is mixed at the level of the EEG/MEG sensor. As examples, they showed that central sensors often include bilateral activity from all those brain regions and that at frontal EEG sensors the contribution of occipital and sensorimotor neural generators can be as high as 75%, making inferences about the activation of frontal cortex based on frontal electrodes problematic and likely inappropriate. They also noted that the contribution of activity at the sensor level from different brain regions likely changes over time (seconds, hours, days, years) and that between-subject differences in the brain areas contributing to individual sensors or sensor ROIs are to be expected. Shirzai et al. ( 143 ) used ultrahigh-density electrocorticography (activity from a 3-by-3 cm 2 patch of cortical surface) to compute the activity that would be observed at 207 EEG scalp electrodes. They concluded that, “The results from the simulated EEG showed that the µECoG activity can be detected in very far EEG electrodes as well as the close-range EEG electrodes. This finding challenges the notion of attributing the EEG channel activity to the closest cortical area, which is unfortunately a common method in EEG research,” and that, “These findings underscore the critical importance of source-level analysis methods for accurate interpretation of EEG data and suggest that channel-level approaches may fundamentally misattribute the cortical origins of observed electrical activity.” For all the reasons discussed in this Section, the difficulties associated with interpreting RS brain activity at the sensor level are many. For the field to advance, scientists and clinicians should make explicit the justification for their methods, ensuring that datasets not only enhance our understanding of brain development in health and disease but also make efficient use of the time-intensive process of collecting neurophysiology data. EEG and MEG analysis in source space, in contrast to sensor space, is designed to identify specific brain areas and distinguish between them, minimizing blurring. To our knowledge, only three pediatric MEG studies have explored brain area-specific differences in the aperiodic exponent and offset measures in source space, all reporting spatially dependent differences( 144, 145 ). For example, Vandewouw et al. ( 144 ) reported regional differences in age and aperiodic exponent and offset associations in 69 children and adults 1 to 38 years old. As shown in Figure 3, Green et al. ( 145 ) observed in 107 typically developing infants 2 to 68 months that RS aperiodic measures as well as their maturation differed across the seven brain areas examined. Adult studies have also reported regional differences in RS neural activity and demonstrate the need for spatially differentiated measures to better understand brain organization. As an example, building on a model of brain organization that postulates a dominant gradient in cortical features across sensorimotor and transmodal areas ( 146 ), Mahjoory et al. ( 147 ) used RS MEG recordings (N = 187 adults) and distributed source localization to show that the dominant periodic peak frequency decreases along a posterior-to-anterior axis, following a global hierarchy from early sensory to higher-order areas. They also showed that the spatial gradient of the peak frequency was anticorrelated with cortical thickness, in effect providing a proxy for cortical hierarchical level. This study demonstrates what is possible with respect to understanding core organizational features of the brain when adequate spatial resolution is available. With respect to pediatric clinical populations, MEG studies of children with a mild traumatic brain injury (mTBI) are of note. In two recent RS MEG studies ( 148, 149 ), differences in brain activity between children with and without a mTBI were found to be spatially specific. As an example, in a study examining adolescents with mTBI, Edgar et al. ( 148 ) found that group differences in resting-state beta-band activity were specific to superior frontal gyrus, right temporal pole, and right central sulcus. In sum, given regional differences in the maturation of infant brain structure and chemistry ( 150 – 160 ), area-specific differences in RS activity are almost certain, with age-varying associations between RS activity and brain structure and behavior very likely. High-dimensional, whole-head measures of neural activity that support source-localization analyses are best suited for such research. Section 4: Pediatric MEG: Two factors that limit progress Although the advantages of SQUID MEG in studying the developing brain are clear, two key factors limit progress: the cost of MEG and the putative difficulty of multisite MEG studies. Regarding the cost, many grant reviewers tend to perceive MEG studies as too expensive relative to EEG studies. This premise merits discussion, as it implies a false equivalence between EEG and MEG, and it does not consider the costs of MEG studies relative to the costs of other functional imaging studies, most prominently fMRI. The financial cost of operating an EEG or MEG lab is substantial, with the costs of setting up a high-end EEG lab over $300K USD for a 128-channel system with multiple EEG caps, an electrode-location digitization system, data acquisition and data analysis computers and software, and a shielded room (somewhat less without a shielded room). The initial cost of setting up an MEG lab can be 10x that of an EEG lab, consisting of an MEG system, magnetically shielded room, digitization system, data collection and analysis computers and software (although most MEG systems also include full EEG systems, valuable for clinical epilepsy cases). The higher financial costs associated with establishing a SQUID MEG lab do not end after the lab is set up, with many laboratories spending money on liquid helium (to cool the SQUIDs), perhaps hiring a full-time technologist, and paying for a yearly service contract. However, the helium operating cost has become less of a problem, as most new SQUID MEG systems incorporate an integrated helium recovery system. The lower cost of purchasing and maintaining an OPM system is discussed in Section 5. The differential financial costs of EEG and MEG research exams are also of note. EEG research exams are generally almost free, given minimal costs for EEG supplies (e.g., electrode paste or potassium chloride solution). In contrast, most MEG centers expect researchers to pay a substantial machine-time charge, sometimes as high as the cost of an MRI exam (e.g., between $300 to $700 USD an hour), to recover the cost of purchase and support of the system. In addition, if source localization of EEG or MEG is desired, this ideally includes structural MRI, adding to the cost per participant. In current practice, EEG is a much cheaper technology than MEG. Importantly but commonly ignored, the financial cost should be considered against the amount of information generated. The advantages of MEG over EEG were reviewed above. Although a low-density EEG study may be nearly cost-free to the investigator, the amount of neurophysiological information gained will be far less than that of MEG, especially if structural MRI is also obtained to accompany the MEG data. Furthermore, the high dimensionality of MEG data, combined with its superior spatial resolution, will likely to be more valuable than EEG data for future analysis methods yet to be developed. This long-term advantage is particularly interesting given the growing use of already published datasets for new studies. In sum, although the financial costs of MEG are currently well above those of EEG, they are comparable to those of fMRI research, and the clinical and research yield from MEG will often justify its higher costs, given what it offers in temporal and spatial precision. A second factor limiting progress is the perception that multisite MEG studies are difficult, based on assumptions that are no longer true or are readily addressable. Multisite studies are needed to establish large sample datasets as well as to achieve adequate representation of diverse rearing environments and of racial, ethnic, and socioeconomic variance. One barrier to such datasets is that there are relatively few pediatric MEG sites that provide recording equipment for smaller heads, as well as variability in MEG systems across sites. That technology is rapidly developing. Another barrier to multisite studies has been difficulty in processing data in a consistent way given that MEG datasets were stored using proprietary data formats requiring custom preprocessing prior to direct comparison. Fortunately, as part of the open data movement, all MEG vendors agreed to make their proprietary data formats readable for developing open-source software packages. Accordingly, the availability and growing use of open-source software that supports analysis of MEG data from different systems has removed this barrier. Importantly, differences in sensor type and sensor configurations across vendors are accounted for in the source-analysis algorithms, thus eliminating a prior barrier to multisite studies. The most widely used open-source packages include Brainstorm, MNE, and FieldTrip. Commercially available software also can read raw MEG data files from multiple vendors (e.g., Curry, BESA). These developments put MEG on the same plane as MRI, where there are a very small number of dominant hardware vendors, each with their own approach to gradient and head coils as well as proprietary sequences, yet data can readily be harmonized. Numerous techniques have been developed to harmonize multisite data, many of which are based on the ComBat algorithm ( 161, 162 ). Although there are no published reports applying harmonization algorithms to MEG data, there is ongoing work of the ENIGMA MEG working group, part of the larger ENIGMA consortium (Nugent, in preparation) that combines multisite, multi-vendor data and harmonization to achieve a large sample dataset. Large-sample (mostly adult) MEG resting-state and task datasets are also available from CamCAN (https://opendata.mrc-cbu.cam.ac.uk/projects/camcan/), OMEGA (https://www.mcgill.ca/bic/neuroinformatics/omega), and POND ((https://pond-network.ca/). Another issue for multisite studies is that the nature of raw MEG data differs for different types of sensors, for example planar vs. axial gradiometers. However, the properties of these different sensor types are well understood and accounted for in the analysis packages. Accordingly, near-identical source-analysis results have been obtained in a human phantom study of three commercially available MEG systems representing most of the vendor market ( 163 ). A general challenge for multisite MEG studies is recording-system quality in the form of external noise level, stability of sensors, external noise sources (local noise environment, mitigated by the quality of the shielded room), and data acquisition procedures. Depending on the facility setting (in a busy hospital unit vs. a dedicated wing of a building low in acoustic and electromagnetic noise by design), sites may differ in environmental noise. This concern is addressable via a variety of signal processing tools now readily available in both open-source and commercial MEG analysis packages. Differences in data acquisition procedures across sites can be managed via a standardized protocol, which has become common in multisite studies. In short, the primary challenges for multisite MEG studies are readily addressable (e.g., ( 164, 165 )). Section 5: Optically pumped magnetometers (OPMs): The next frontier There are several limitations to SQUID technology in currently available MEG systems. As discussed above, SQUID MEG laboratories are costly to set up and maintain. Another limitation is that SQUID systems use a fixed-size helmet holding the sensors, and most current MEG laboratories use systems sized for adults, generally designed to accommodate the 95 th percentile male head ( 166 ). They can accommodate smaller heads, but in children the SQUID MEG sensors can be located a considerable distance from the head surface. Because the strength of MEG signals decreases with the square of the distance from the source, on-scalp OPM sensors can provide much higher signal magnitudes than SQUID MEG systems designed for adult head sizes. SQUID MEG systems optimized for infants and toddlers have recently become available (for review see ( 167 )), such as the Artemis 123 system (Tristan Technologies, Inc.) designed to accommodate the median 3-year-old head circumference ( 168, 169 ). However, even the systems optimized for young children use a fixed-size helmet, which is not optimal for head sizes that are growing during development. Moreover, current pediatric helmets remain larger than is optimal for young infants. Thus, there remains a challenge in obtaining the optimal signal from each child across development. A related limitation of SQUID-based MEG is that the helmet is necessarily stationary. Thus, participant head movement leads to a reduction in data quality analogous to that affecting MRI measurements, and movement is a particular challenge in awake infants and young children. In modern SQUID-based MEG systems, this is addressed by continuously recording from head position indicator coils and using tools such as signal-space separation (SSS) and tSSS (temporal SSS) to adjust signals for head movement following the recording ( 170 – 174 ). Nonetheless, head movement results in changes in the distance between the sensor and source, leading to variability in signal-to-noise ratio across the data collection session and constraining the desired behavior of the participant. Emerging OPM technology provides a promising avenue with which to address these limitations of SQUID MEG. Briefly, OPM sensors are a form of compact quantum sensors that measure the magnetic fields associated with neural activity but do not require liquid helium ( 175 – 177 ). Each sensor cartridge can be placed directly on the scalp surface, minimizing the distance between source (brain) and sensor. Tremendous progress has been made in fabricating small and lightweight OPM sensors of various types 2, which has led to the development of wearable OPM MEG helmets. Whole-head OPM arrays with more than 200 channels have been demonstrated, and the newest devices are beginning to match the density of commercially available SQUID-based MEG systems. The current generation of OPM sensors enables many different implementations, including stationary helmets affixed to a scanner bed ( 178 ), wearable rigid helmets of different sizes ( 179, 180 ), and EEG-like caps ( 181, 182 ). Rigid structure helmets may be 3-D printed based on individual MRIs ( 183 ), thus enabling whole-head arrays that correspond closely to individual head shapes. However, this approach is extremely expensive and not feasible for routine studies. Given the inability of infants to fully support their own heads, OPM helmets often require some form of weight support, but this can be as simple as leaning against a caregiver. The ability to place OPM sensors directly on the surface of the scalp is a game-changer for pediatric neuroimaging. Notably, wearable OPM allows participants an impressive range of motion without seriously impacting data quality, provided the head does not move relative to the sensors ( 184, 185 ). This is particularly valuable in pediatric neuroimaging, where young children often struggle to remain still throughout a testing session. Indeed, recent work using wearable OPM has demonstrated success in testing children and toddlers ( 144, 180, 186, 187 ) and infants as young as 1 month of age ( 181 ). For example, Figure 4 (adapted from Safar et al. ( 187 )) shows neural responses to faces in a group of children aged 3-5, including age-related changes in region-specific evoked responses (left and middle panels) and whole-brain functional connectivity (right panel). This study demonstrates the suitability of OPM for capturing complex neuronal responses to stimuli in pediatric populations. The benefits of on-scalp OPM sensors also bring challenges. Knowing the precise location and orientation of each MEG sensor is required for accurate source reconstruction. The advantage of a fixed array (as in SQUID MEG) is that the sensor positions and orientations can be determined once and remain valid throughout the life of the system. The adaptability of OPM arrays to differing head shapes poses a technical consideration - if the sensor locations and/or orientations are allowed to differs across participants, their positions must be determined for each session. One way this problem can be mitigated is by using rigid helmets of varying (e.g., age-appropriate) sizes that fix the OPM sensor positions and orientations (e.g., ( 179, 180 )). There are also helmets with adjustable sensors that can be pushed into the helmet until they meet the subject’s scalp (e.g., ( 178 )). These helmets may incorporate measurement coils that allow the position of each sensor to be detected, or external coil arrays may be incorporated into the helmet for calibration. Finally, there are mathematical algorithms that can be used to determine the positions and orientations of OPM sensors with the help of recorded calibration data. For example, Iivanainen et al. ( 188 ) demonstrated a method that utilizes external field compensation coils and fluxgate measurements to fit the relevant coefficients of a vector spherical harmonic expansion in the area of the OPM array. These coefficients then served as a signal model against which the sensor parameters were fitted. This additional time for calibration is a consideration for pediatric studies. Another challenge is the fact that, in addition to sensor location geometry, in some types of OPMs sensor-specific scalar calibration factors, or gains, may change during a recording session due to the phenomenon of cross-axis projection error ( 189 ). To mitigate this issue, careful suppression of the background interference field is necessary. The most common approach to address this problem is by adding active noise cancellation coils to the magnetically shielded room ( 190, 191 ). One solution is to maintain a ~1 m 3 cube around the participant in which the ambient magnetic field has been minimized using nulling coils, although this limits participant movement to the 1 m 3 cube. Figure 5 show s a schematic of such a setup as detailed in Hill et al. ( 192 ). More recent approaches have incorporated nulling coils on the internal walls of the MSR itself, enabling a larger spatial extent of magnetic field control (e.g., ( 193 ). Such developments, alongside novel mobile recording platforms ( 194 ) have shown promise in allowing movement throughout the magnetically shielded room ( 195 ), though more work in this realm is needed before establishing OPM as a truly ambulatory technology (see also methods in ( 175, 196 )). Finally, improvements in the technical performance of the OPM sensors themselves (e.g., noise floor, dynamic range), which do not yet match that of SQUID MEG sensors, will be necessary if OPM technology is to be as useful as current SQUID MEG. Despite its higher noise floor compared to SQUIDs, OPM MEG holds the promise of increased signal-to-noise ratio (SNR) and spatial precision due to closer proximity to the brain, especially in infants and children with smaller head sizes ( 196 ). Compared with conventional SQUID MEG, on-head OPM can deliver equivalent or superior sensitivity to region-specific evoked and oscillatory responses (e.g., ( 197 )). Moreover, owing to the potentially flexible shape of OPM arrays, judicious placement of sensors to target a particular area can further improve source-localized SNR ( 198 ). Rapid technical advances in OPM sensor systems, noise rejection algorithms, and source models are ongoing. Given the tradeoffs between OPM (i.e., higher noise floor but sensors on the scalp) and SQUID MEG (i.e., more sensitive to tiny magnetic field changes but sensors are more distant from the scalp), direct comparisons are needed. The cost of an OPM system largely depends on the price of a magnetically shielded room and the number of OPM sensors. Like SQUID MEG, OPM recordings will require a magnetically shielded room (several hundred thousand dollars), although smaller rooms and even person-sized enclosures are now available at a reduced cost. Whereas at present the cost of an OPM sensor (providing one to three orthogonal channels at one scalp location) is in the range of $7K to $10K USD, with yet unknown lifespan, it is anticipated that sensor cost will decrease dramatically as volume increases and the technology matures. Because OPM systems are modular more sensors can be added later, the barrier of entry to using OPM MEG has dropped dramatically. Scientists considering purchasing an OPM system at the present time still need to consider dedicated staff, with the availability of a commercially available plug-and-play OPM system requiring little ongoing onsite support several years in the future. The full cost of an OPM system for a usable lifetime equivalent to that of a SQUID MEG system (decades in most cases) remains to be determined. However, as OPM technology is inherently different than SQUID systems, other factors such as simplified reparability and parts replacements, and appropriate service contracts can ensure system uptime of OPM MEG systems. At present, the OPM arena is diverse and evolving rapidly, with some laboratories focused on collecting dedicated research data and others more active in testing and validating OPM sensors and associated hardware, as well as optimizing OPM data collection and analysis methods. With respect to the clinical use of OPMs (e.g., identifying seizure onset zone(s)), of note is a 2023 commentary from the American Clinical MEG Society (ACMEGS) concluding that, “Based on the engineering obstacles and the clinical tradeoffs to be resolved, the assessment of experts suggests that there will most likely be another decade relying solely on ”frozen SQUIDs” in the clinical MEG field” ( 199 ). This assessment, however, may be overly pessimistic. OPM studies examining brain activity in patients with epilepsy show promising results in adult ( 200, 201 ) and pediatric populations ( 202, 203 ). Notably, in 2024 China approved the use of OPM systems for clinical work. Several hospitals in China are using OPM systems to plan surgery for epilepsy and brain tumors. It is hoped that studies from China systematically comparing SQUID and OPM clinical results will be forthcoming. Despite the above hurdles, many scientists are devoting considerable time to developing OPM technology, on their own or through partnering with industry. The ability to place sensors directly on the scalp of participants may revolutionize pediatric as well as adult MEG. Breakout Box 2 considers the use of OPM sensors to examine neural activity in more realistic situations than viewing stimuli on a computer screen, such as obtaining brain measures when two humans are interacting (hyperscanning). The ability to use OPMs to obtain measures of neural activity in parts of the body other than the head is also briefly discussed.

This manuscript focuses on the present and future applications of MEG in pediatric research, emphasizing the practical side of data collection. The advantages and challenges associated with pediatric MEG were discussed from multiple angles, including biophysical, neurophysiological, clinical, and economic aspects. Given the advantages reviewed above of MEG over EEG (simpler source localization mathematical models, higher dimensional data, little effect of open fontanelle and sutures on infant neural measures, and reference-free neural measures), MEG is often the preferred method for assessing neural activity in children in both clinical and research contexts, including adding substantial value in multimodal imaging. The high dimensionality of MEG data also makes it better suited than EEG to complex spatially resolved analysis methods which are under active development. The promising potential of OPM technology to advance our understanding of neural function in children was also reviewed. Optimizing information acquisition is critical for understanding typical and atypical brain development, and rapid advancements in artificial intelligence are expected to benefit and to benefit from the use of rich MEG datasets and to increase the utility of high temporal and spatial resolution methods such as MEG for improving outcomes across the lifespan. Footnote 1 For readers with an interest in EEG and MEG instrumentation and technology, Orrison et al. ( 39 ) provided a detailed but accessible review. Schomer and Lopes da Silva ( 109 ) and Nunez and Srinivasan ( 8 ) provided more extensive reviews of EEG. See also ( 45 – 47, 51, 204, 205 ). For reviews of the use of MEG to study clinical disorders, see ( 205 – 207 ). Readers are also referred to papers discussing best practices for data collection, analysis, and reporting for EEG ( 208 – 210 ) and MEG ( 209 – 212 ). Finally, for readers interested in OPM instrumentation and methods, several papers are of note ( 175, 196, 213 ); and see references in Section 5). Footnote 2 OPM sensors have a chamber containing a vapor of atoms such as rubidium, cesium or helium. These atoms can be spin-polarized with light which makes them sensitive to small magnetic fields. In the presence of a magnetic field, the spin-polarization changes and this change can be detected by monitoring the fluctuations of light passing through the atomic vapor. There are a variety of implementations of OPMs (for an overview, see ( 214 )), but the type best suited for MEG is the zero-field magnetometer ( 215 ). These sensors are among the most sensitive detectors and lend themselves to miniaturization ( 216, 217 ). Currently, there are several OPM MEG vendors with their unique implementations and this has resulted in different characteristics that impact data quality and system usability. Thus, care should be exercised to not generalize a specific implementation as a feature or limitation of all OPM MEG systems. Figure 1 Photos of MEG systems: (A) Left: SQUID MEG Artemis 123-channel system (Tristan Technologies, Inc.); Right SQUID MEG 275-channel system (CTF MEG Neuro Innovations, Inc.), (B) Left: OPM MEG 128-channel (FieldLine Medical, Inc.); Right: OPM MEG 80-channel system (QuSpin, Colorado, USA) mounted in a rigid wearable helmet (Cerca Magnetics Ltd. Nottingham, UK), Cerca image from Rhodes et al. ( 180 ) and used with permission. Figure 2 Scatterplots displaying the peak values of the significant main effect of age and power with linear (gray) and quadratic (color) fits for each significant age effect. Asterisks indicate that the quadratic fit is significantly better than the linear. The box on each brain represents the peak of each cluster from which data were extracted. The dots on the scatterplots represent individual participants, with the colored line showing the quadratic fit and the shaded portion representing SE. The color distributions correspond to the canonical frequency band shown (delta-purple, theta-blue, alpha-green, beta-orange, and gamma-red). Absolute power in pseudo neural activity index (PNAI) units is plotted on the y axis, and age is plotted on the x axis. Left-hemisphere clusters are in the left column and right-hemisphere clusters are in the right column. Image from Rempe et al. ( 120 ) and used with permission. Figure 3 Plots show the relationship between age and the resting-state aperiodic exponent for the seven examined brain regions. The red lines indicate the statistically significant differences between the parietal-occipital region and the other brain regions for exponent and age associations. Image from Green et al. ( 145 ), this manuscript under review. Figure 4

from a recent OPM study in which children aged 3-5 saw images of emotional (happy and angry) faces. Left panel: the M170 evoked response to faces in the right fusiform gyrus, in different age groups. Middle panel: M710 amplitude, averaged across left and right fusiform gyri, as a function of age. Right panel: mean (across all participants) functional connectivity in response to face presentation relative to a preceding baseline period. Image from Safar et al. ( 187 ) and used with permission. Figure 5 Schematic illustrates OPM recording within a low magnetic field environment between biplanar nulling coils; these in turn are situated inside a magnetically shielded room. DAQ: data acquisition system. Figure adapted from Hill et al. ( 192 ) and used with permission. Figure 6 Example of a hyperscanning experiment. (a) Two participants engaged in an alternating hand-touching task while their brain activity is recorded with wearable OPM. (b) Source-localized beta-band modulation (thresholded to 80% of the maximum value) on odd trials (red: when participant 1 is active) and even trials (blue: participant 2 is active). Figure from ( 218 ) and used with permission. Breakout Box 1: Localization accuracy of EEG and MEG In many contexts, the desired location accuracy may be beyond what EEG or MEG can achieve, but that is often not the case (for additional discussions see ( 219, 220 )). Skull phantom work using 32 known dipoles found average spatial localization error of just 3 mm with 122-channel, 61-site MEG and 7–8 mm with 64-channel EEG, the former largely due to a 2.26-mm error in sensor location quantification ( 72 ). These studies used SQUID MEG, and OPM MEG may provide even better accuracy. For many goals, relative accuracy suffices. As an example, the relative locations of the representations of the fingers of one hand in primary motor cortex can easily be distinguished with a small number of scalp EEG electrodes, most placed over the contralateral motor strip, a case in which the experimenter would know in advance where to place the sensors. Second, the approximate locations of distant sources may suffice, such as assessing nodes in a frontal-parietal network. Third, some sources are conceived as patches or sheets of cortex, rather than dipoles, in which case millimeter accuracy may not be important. Fourth, distinguishing time courses of sources or their sensitivity to different experimental conditions may be more important than determining their precise spatial location. The common belief that the inverse problem precludes meaningful electromagnetic source localization is unjustified. Ultrasound imaging faces the inverse problem ( 221 ) without dismissal of the medical images it produces. The number of EEG or MEG sensors and their topographic sampling density will often be less important than the overall spatial coverage (especially for deeper sources), with sensor placement evaluated based on the anticipated source strength, location (including depth), orientation, and areal extent of the neural generator(s). Source-analysis algorithms differ considerably, with differing sensitivities to the number and locations of sensors. Source models may involve one-axis dipoles, three-axis dipoles (what BESA software calls regional sources), or cortical sheets. Source analysis may rely on sets of pre-specified targets, selected on the basis of a standard atlas or subject-specific sMRI or fMRI information, rather than asking the algorithm to model location in addition to orientation and/or time course. Methods exist to produce distributions of candidate source locations (e.g., ( 222, 223 )). For participants of any age, no single minimum number of sensors can be specified as a prerequisite to EEG or MEG source analysis. The chosen number and placement of the sensors in a given study will affect spatial capabilities of source localization, but the temporal resolution advantages of EEG and MEG do not depend on those. For source reconstruction, although co-registration to each child’s individual sMRI is preferred (and needed for multimodal studies, for example relating neural activity to gray-matter cortical thickness), some studies have found that the use of age-matched MRI templates is often successful ( 224, 225 ). For infants or individuals with high support needs (e.g., individuals with autism who have a low IQ and are nonverbal), it can be difficult to obtain structural MRI data. As a result, researchers frequently rely on standardized brain templates or surrogate MRIs. Unfortunately, most widely used neuroimaging templates, such as MNI or FreeSurfer’s fsaverage, are based on adult anatomy and are therefore not well-suited for modeling the infant brain. Infant heads and brains differ from those of adults in terms of overall size, cortical folding, skull thickness, fontanels, and cortical thickness, as well as in relative volumes of gray matter, white matter, and cerebrospinal fluid. As discussed in Section 2, these anatomical differences lead to errors in the forward model and subsequent source localization, with EEG much more vulnerable than MEG. Although some age-specific infant templates are available, such as the UNC or dHCP atlases, they have yet to be fully integrated into the leading open-source MEG analysis packages, and they do not cover the complete range of head sizes in infancy. Moreover, even with age-matched templates (201), small errors in co-registration can produce significant localization errors due to the small size of the infant head. Thus, at present, co-registration to pediatric template MRIs can be problematic. Rhodes et al. ( 226 ) demonstrated a novel method of co-registering on-head OPM data to “pseudo-MRI” images. These images were created by warping 3D structured-lights scans (acquired from individual participants) to age-matched template MRIs. Co-registration to the pseudo-MRI images provided source localization accuracy comparable to that of individual MRI scans. This method has since been used for OPM source localization in infants and young children ( 180, 187 ). This strategy, however, does not provide the ability to assess associations between brain structure and function. Breakout Box 2: OPMs in hyperscanning studies and for use outside the brain Hyperscanning refers to simultaneous recording of neurophysiological activity from the brains of multiple individuals, whether in the same or different geographic locations. Unlike the traditional approach of recording neural responses throughout highly controlled experiments, hyperscanning can be used to capture brain activity during naturalistic, participant-led social interactions. For example, some studies have recorded simultaneous EEG from two participants engaged in conversation ( 227, 228 ), musicians playing music together ( 229, 230 ), or high school students engaged in classroom learning activities ( 231 ). Such applications—which provide unique insights into the neural mechanisms supporting social cognition—are relevant to pediatric neurophysiology, for example examining social processes in children with autism (for a review, see ( 232 )). OPM MEG is well-positioned to build on this research. In the first OPM hyperscanning study, Holmes et al. ( 218 ) recorded from two individuals while they performed interactive hand-touching and table tennis inside a magnetically shielded room with active field control. As shown in Figure 6, despite task-induced movements, the authors reported the expected modulation of neural oscillatory activity at the participant level. The insensitivity of OPM-MEG to movement is a major advantage for hyperscanning over EEG, SQUID MEG, and fMRI, all of which are extremely sensitive to movement. For example, OPM hyperscanning may be more easily performed on young children or patients who find it difficult to remain still throughout an experimental session. More broadly, the reduced constraints on participant movement open the door to novel research areas. For example, it may be possible to perform hyperscanning during non-verbal communication (e.g., sign language), and OPM may enable more naturalistic participant behavior during verbal conversations, in which speech is typically accompanied by physical gestures ( 233 ). Beyond the usual requirements of EEG hyperscanning (see ( 227 )), the only additional requirement for OPM MEG is a magnetically-shielded room that is large enough to accommodate the desired number of individuals. By contrast, hyperscanning with SQUID MEG requires either simultaneous recording at two sites (e.g., ( 234, 235 )) or two systems in the same room ( 236 ), which is rarely feasible. OPMs also present opportunities for measurement of neurophysiology outside the brain. All currently available SQUID MEG systems employ a pre-defined sensor layout, whereas OPMs may be flexibly arranged to record from other parts of the body. For example, OPMs have been used to record muscle responses in the hand following ulnar nerve stimulation ( 237 ), evoked fields from the retina following light stimulation ( 238 ), and evoked fields from the spinal cord following median nerve stimulation ( 239 ). Remarkably, the latter study source-localized evoked fields and reported that spinal sources were consistently lateralized to the side of stimulation. This demonstration of accurate spinal source localization hints at the utility of this approach for diverse diagnostic applications. OPM has even been used to record fetal neurophysiology in utero. By placing OPMs on mothers’ abdomens, studies have successfully recorded fetal cardiac rhythms ( 240 – 242 ). Batie et al. ( 240 ) were additionally able to detect cardiac rhythm abnormalities, highlighting the clinical utility of this approach. Future OPM work may substantially expand the clinical and research applications of fetal MEG. For example, SQUID MEG has been used to detect visual and auditory evoked neural responses from fetuses in utero (for a review, see ( 167 )). OPMs offer a lower-cost method to further explore activity in the developing brain. It is also worth highlighting that the flexibility of OPM MEG facilitates sensitivity to deep sources in the brain that are less accessible to SQUID MEG, owing to the rapid signal drop-off between deep brain sources and sensors at the scalp. Tierney et al. ( 243 ) were able to record task-related activity from the hippocampus using an OPM sensor mounted on a dental mold against the roof of a participant’s mouth (although SQUID MEG can successfully record hippocampal activity from the scalp, e.g. ( 81, 82, 84 )). This novel application of OPM, alongside the other applications discussed above, demonstrates that OPMs improve study of deep tissue both inside and outside the brain. Disclaimer This research was supported, in part, by the Intramural Research Program of the National Institutes of Health (NIH). The contributions of the NIH author(s) are considered Works of the United States Government. The findings and conclusions presented in this paper are those of the author(s) and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.

1. NIMH Strategic Plan for Research 2023.2. A. Bernas, E. M. Barendse, A. P. Aldenkamp, W. H. Backes, P. A. M. Hofman, M. P. H. Hendriks, R. P. C. Kessels, F. M. J. Willems, P. H. N. De With, S. Zinger, J. F. A. Jansen, Brain resting‐state networks in adolescents with high‐functioning autism: Analysis of spatial connectivity and temporal neurodynamics. Brain Behav. 8, e00878 (2018).3. A. K. Easson, A. R. McIntosh, BOLD signal variability and complexity in children and adolescents with and without autism spectrum disorder. Dev. Cogn. Neurosci. 36, 100630 (2019).4. E. Delbruck, M. Yang, A. Yassine, E. D. Grossman, Functional connectivity in ASD: Atypical pathways in brain networks supporting action observation and joint attention. Brain Res. 1706, 157–165 (2019).5. O. Kardan, S. Kaplan, M. D. Wheelock, E. Feczko, T. K. M. Day, Ó. Miranda-Domínguez, D. Meyer, A. T. Eggebrecht, L. A. Moore, S. Sung, T. A. Chamberlain, E. Earl, K. Snider, A. Graham, M. G. Berman, K. Uğurbil, E. Yacoub, J. T. Elison, C. D. Smyser, D. A. Fair, M. D. Rosenberg, Resting-state functional connectivity identifies individuals and predicts age in 8-to-26-month-olds. Dev. Cogn. Neurosci. 56, 101123 (2022).6. H. Zhang, D. Shen, W. Lin, Resting-state functional MRI studies on infant brains: A decade of gap-filling efforts. NeuroImage 185, 664–684 (2019).7. B. N. Cuthbert, G. A. Miller, C. Sanislow, U. Vaidyanathan, “The Research Domain Criteria Project: Integrative Translation for Psychopathology” in Oxford Textbook of Psychopathology, R. F. Krueger, P. H. Blaney, Eds. (Oxford University PressNew York, ed. 4, 2023; https://academic.oup.com/book/45782/chapter/400197542), pp. 78–102.8. P. L. Nunez, R. Srinivasan, Electric Fields of the Brain (Oxford University Press, 2006; https://academic.oup.com/book/2998).9. R. Srinivasan, W. R. Winter, J. Ding, P. L. Nunez, EEG and MEG coherence: Measures of functional connectivity at distinct spatial scales of neocortical dynamics. J. Neurosci. Methods 166, 41–52 (2007).10. S. Baillet, J. C. Mosher, R. M. Leahy, Electromagnetic brain mapping. IEEE Signal Process. Mag. 18, 14–30 (2001).11. J. Da Silva Castanheira, H. D. Orozco Perez, B. Misic, S. Baillet, Brief segments of neurophysiological activity enable individual differentiation. Nat. Commun. 12, 5713 (2021).12. J. Da Silva Castanheira, A. I. Wiesman, J. Y. Hansen, B. Misic, S. Baillet, J. Breitner, J. Poirier, S. Baillet, P. Bellec, V. Bohbot, M. Chakravarty, L. Collins, P. Etienne, A. Evans, S. Gauthier, R. Hoge, Y. Ituria-Medina, G. Multhaup, L.-M. Münter, N. Rajah, P. Rosa-Neto, J.-P. Soucy, E. Vachon-Presseau, S. Villeneuve, P. Amouyel, M. Appleby, N. Ashton, D. Auld, G. Ayranci, C. Bedetti, M.-L. Beland, K. Blennow, A. B. Westman, C. Cuello, M. Dadar, L.-A. Daoust, S. Das, M. Dauar-Tedeschi, L. De Beaumont, D. Dea, M. Descoteaux, M. Dufour, S. Farzin, F. Ferdinand, V. Fonov, J. Gonneaud, J. Kat, C. Kazazian, A. Labonté, M.-E. Lafaille-Magnan, M. Lalancette, J.-C. Lambert, J.-M. Leoutsakos, L. Mahar, A. Mathieu, M. McSweeney, P.-F. Meyer, J. Miron, J. Near, H. NewboldFox, N. Nilsson, P. Orban, C. Picard, A. P. Binette, J.-B. Poline, S. Rabipour, A. Salaciak, M. Settimi, S. Subramaniapillai, A. Tam, C. Tardif, L. Théroux, J. Tremblay-Mercier, S. Tullo, I. Ulku, I. Vallée, H. Zetterberg, V. Nair, J. Pruessner, P. Aisen, E. Anthal, A. Barkun, T. Beaudry, F. Benbouhoud, J. Brandt, L. Carmo, C. E. Carrier, L. Cheewakriengkrai, B. Courcot, D. Couture, S. Craft, C. Dansereau, C. Debacker, R. Desautels, S. Dubuc, G. Duclair, M. Eisenberg, R. El-Khoury, A.-M. Faubert, D. Fontaine, J. Frappier, J. Frenette, G. Gagné, V. Gervais, R. Giles, R. Gordon, C. Jack, B. Jutras, Z. Khachaturian, D. Knopman, P. Kostopoulos, F. Lapalme, T. Lee, C. Lepage, I. Leppert, C. Madjar, D. Maillet, J.-R. Maltais, S. Mathotaarachchi, G. Mayrand, D. Michaud, T. Montine, J. Morris, V. Pagé, T. Pascoal, S. Peillieux, M. Petkova, G. Pogossova, P. Rioux, M. Sager, E. F. Saint-Fort, M. Savard, R. Sperling, S. Tabrizi, P. Tariot, E. Teigner, R. Thomas, P.-J. Toussaint, M. Tuwaig, V. Venugopalan, S. Verfaillie, J. Vogel, K. Wan, S. Wang, E. Yu, I. Beaulieu-Boire, P. Blanchet, S. Bogard, M. Bouchard, S. Chouinard, F. Cicchetti, M. Cloutier, A. Dagher, S. Das, C. Degroot, A. Desautels, M. H. Dion, J. Drouin-Ouellet, A.-M. Dufresne, N. Dupré, A. Duquette, T. Durcan, L. K. Fellows, E. Fon, J.-F. Gagnon, Z. Gan-Or, A. Genge, N. Jodoin, J. Karamchandani, A.-L. Lafontaine, M. Langlois, E. Leveille, M. Lévesque, C. Melmed, O. Monchi, J. Montplaisir, M. Panisset, M. Parent, M.-T. Pham-An, J.-B. Poline, R. Postuma, E. Pourcher, T. Rao, J. Rivest, G. Rouleau, M. Sharp, V. Soland, M. Sidel, S. L. Wing Sun, A. Thiel, P. Vitali, The neurophysiological brain-fingerprint of Parkinson’s disease. eBioMedicine 105, 105201 (2024).13. J. C. Edgar, K. Heiken, Y.-H. Chen, J. D. Herrington, V. Chow, S. Liu, L. Bloy, M. Huang, J. Pandey, K. M. Cannon, S. Qasmieh, S. E. Levy, R. T. Schultz, T. P. L. Roberts, Resting-State Alpha in Autism Spectrum Disorder and Alpha Associations with Thalamic Volume. J. Autism Dev. Disord. 45, 795–804 (2015).14. S. Sadaghiani, M. J. Brookes, S. Baillet, Connectomics of human electrophysiology. NeuroImage 247, 118788 (2022).15. S. Baillet, Magnetoencephalography for brain electrophysiology and imaging. Nat. Neurosci. 20, 327–339 (2017).16. T. W. Wilson, E. Heinrichs-Graham, A. L. Proskovec, T. J. McDermott, Neuroimaging with magnetoencephalography: A dynamic view of brain pathophysiology. Transl. Res. 175, 17–36 (2016).17. M. J. Brookes, P. K. Tewarie, B. A. E. Hunt, S. E. Robson, L. E. Gascoyne, E. B. Liddle, P. F. Liddle, P. G. Morris, A multi-layer network approach to MEG connectivity analysis. NeuroImage 132, 425–438 (2016).18. M. J. Brookes, J. R. Hale, J. M. Zumer, C. M. Stevenson, S. T. Francis, G. R. Barnes, J. P. Owen, P. G. Morris, S. S. Nagarajan, Measuring functional connectivity using MEG: Methodology and comparison with fcMRI. NeuroImage 56, 1082–1104 (2011).19. B. A. E. Hunt, S. M. Wong, M. M. Vandewouw, M. J. Brookes, B. T. Dunkley, M. J. Taylor, Spatial and spectral trajectories in typical neurodevelopment from childhood to middle age. Netw. Neurosci. 3, 497–520 (2019).20. H. L. Green, M. Dipiero, S. Koppers, J. I. Berman, L. Bloy, S. Liu, E. McBride, M. Ku, L. Blaskey, E. Kuschner, M. Airey, M. Kim, K. Konka, T. P. L. Roberts, J. C. Edgar, Peak Alpha Frequency and Thalamic Structure in Children with Typical Development and Autism Spectrum Disorder. J. Autism Dev. Disord. 52, 103–112 (2022).21. J. I. Berman, L. Bloy, L. Blaskey, C. R. Jackel, J. S. Miller, J. Ross, J. C. Edgar, T. P. L. Roberts, Contributions to auditory system conduction velocity: insights with multi-modal neuroimaging and machine learning in children with ASD and XYY syndrome. Front. Psychiatry 14, 1057221 (2023).22. J. M. Stephen, D. E. Hill, F. T. Candelaria-Cook, Examining the effects of prenatal alcohol exposure on corticothalamic connectivity: A multimodal neuroimaging study in children. Dev. Cogn. Neurosci. 52, 101019 (2021).23. C. Papadelis, B. Ahtam, M. Nazarova, D. Nimec, B. Snyder, P. E. Grant, Y. Okada, Cortical Somatosensory Reorganization in Children with Spastic Cerebral Palsy: A Multimodal Neuroimaging Study. Front. Hum. Neurosci. 8 (2014).24. A. Hunold, J. Haueisen, B. Ahtam, C. Doshi, C. Harini, S. Camposano, S. K. Warfield, P. E. Grant, Y. Okada, C. Papadelis, Localization of the Epileptogenic Foci in Tuberous Sclerosis Complex: A Pediatric Case Report. Front. Hum. Neurosci. 8 (2014).25. C. Urbain, J. Sato, C. Hammill, E. G. Duerden, M. J. Taylor, Converging function, structure, and behavioural features of emotion regulation in very preterm children. Hum. Brain Mapp. 40, 3385–3397 (2019).26. F. T. Candelaria-Cook, I. Solis, M. E. Schendel, Y.-P. Wang, T. W. Wilson, V. D. Calhoun, J. M. Stephen, Developmental trajectory of MEG resting-state oscillatory activity in children and adolescents: a longitudinal reliability study. Cereb. Cortex 32, 5404–5419 (2022).27. Y. Chen, H. L. Green, J. I. Berman, M. E. Putt, K. Otten, K. Mol, M. McNamee, O. Allison, E. S. Kuschner, M. Kim, L. Bloy, S. Liu, T. Yount, T. P. L. Roberts, J. Christopher Edgar, Functional and structural maturation of auditory cortex from 2 months to 2 years old. Clin. Neurophysiol. 166, 232–243 (2024).28. J. C. Edgar, L. Blaskey, H. L. Green, K. Konka, G. Shen, M. A. Dipiero, J. I. Berman, L. Bloy, S. Liu, E. McBride, M. Ku, E. S. Kuschner, M. Airey, M. Kim, R. E. Franzen, G. A. Miller, T. P. L. Roberts, Maturation of Auditory Cortex Neural Activity in Children and Implications for Auditory Clinical Markers in Diagnosis. Front. Psychiatry 11, 584557 (2020).29. J. C. Edgar, C. L. Fisk 4Th, S. Liu, J. Pandey, J. D. Herrington, R. T. Schultz, T. P. L. Roberts, Translating Adult Electrophysiology Findings to Younger Patient Populations: Difficulty Measuring 40-Hz Auditory Steady-State Responses in Typically Developing Children and Children with Autism Spectrum Disorder. Dev. Neurosci. 38, 1–14 (2016).30. A. N. Bosseler, M. Clarke, K. Tavabi, E. D. Larson, D. S. Hippe, S. Taulu, P. K. Kuhl, Using magnetoencephalography to examine word recognition, lateralization, and future language skills in 14-month-old infants. Dev. Cogn. Neurosci. 47, 100901 (2021).31. A. N. Bosseler, A. N. Meltzoff, S. Bierer, E. Huber, J. C. Mizrahi, E. Larson, Y. Endevelt-Shapira, S. Taulu, P. K. Kuhl, Infants’ brain responses to social interaction predict future language growth. Curr. Biol. 34, 1731-1738.e3 (2024).32. A. N. Bosseler, S. Taulu, E. Pihko, J. P. Mäkelä, T. Imada, A. Ahonen, P. K. Kuhl, Theta brain rhythms index perceptual narrowing in infant speech perception. Front. Psychol. 4 (2013).33. P. K. Kuhl, R. R. Ramírez, A. Bosseler, J.-F. L. Lin, T. Imada, Infants’ brain responses to speech suggest Analysis by Synthesis. Proc. Natl. Acad. Sci. 111, 11238–11245 (2014).34. T. C. Zhao, O. Boorom, P. K. Kuhl, R. Gordon, Infants’ neural speech discrimination predicts individual differences in grammar ability at 6 years of age and their risk of developing speech-language disorders. Dev. Cogn. Neurosci. 48, 100949 (2021).35. F. Mamashli, S. Khan, H. Bharadwaj, A. Losh, S. M. Pawlyszyn, M. S. Hämäläinen, T. Kenet, Maturational trajectories of local and long‐range functional connectivity in autism during face processing. Hum. Brain Mapp. 39, 4094–4104 (2018).36. J. Alho, H. Bharadwaj, S. Khan, F. Mamashli, T. K. Perrachione, A. Losh, N. M. McGuiggan, R. M. Joseph, M. S. Hämäläinen, T. Kenet, Altered maturation and atypical cortical processing of spoken sentences in autism spectrum disorder. Prog. Neurobiol. 203, 102077 (2021).37. K. Safar, M. M. Vandewouw, M. J. Taylor, Atypical development of emotional face processing networks in autism spectrum disorder from childhood through to adulthood. Dev. Cogn. Neurosci. 51, 101003 (2021).38. N. Rhodes, J. Sato, K. Safar, K. Amorim, M. J. Taylor, M. J. Brookes, Paediatric magnetoencephalography and its role in neurodevelopmental disorders. Br. J. Radiol. 97, 1591–1601 (2024).39. W. W. Orrison, Functional Brain Imaging (Mosby, St. Louis, 1995).40. J. C. Edgar, M. X. Huang, M. P. Weisend, A. Sherwood, G. A. Miller, L. E. Adler, J. M. Cañive, Interpreting abnormality: an EEG and MEG study of P50 and the auditory paired-stimulus paradigm. Biol. Psychol. 65, 1–20 (2003).41. L. Flemming, Y. Wang, A. Caprihan, M. Eiselt, J. Haueisen, Y. Okada, Evaluation of the distortion of EEG signals caused by a hole in the skull mimicking the fontanel in the skull of human neonates. Clin. Neurophysiol. 116, 1141–1152 (2005).42. S. Lew, D. D. Sliva, M. Choe, P. E. Grant, Y. Okada, C. H. Wolters, M. S. Hämäläinen, Effects of sutures and fontanels on MEG and EEG source analysis in a realistic infant head model. NeuroImage 76, 282–293 (2013).43. Lau, S., Flemming, L., Haueisen, J., “Effect of skull discontinuities on MEG and EEG: Experimental results” (2011)vol. 13, pp. 228–232.44. Y. Okada, M. Hämäläinen, K. Pratt, A. Mascarenas, P. Miller, M. Han, J. Robles, A. Cavallini, B. Power, K. Sieng, L. Sun, S. Lew, C. Doshi, B. Ahtam, C. Dinh, L. Esch, E. Grant, A. Nummenmaa, D. Paulson, BabyMEG: A whole-head pediatric magnetoencephalography system for human brain development research. Rev. Sci. Instrum. 87, 094301 (2016).45. D. Geselowitz, On the magnetic field generated outside an inhomogeneous volume conductor by internal current sources. IEEE Trans. Magn. 6, 346–347 (1970).46. D. Geselowitz, W. Miller, Extracorporeal magnetic fields generated by internal bioelectric sources. IEEE Trans. Magn. 9, 392–398 (1973).47. R. Plonsey, Magnetic field resulting from action currents on cylindrical fibres. Med. Biol. Eng. Comput. 19, 311–315 (1981).48. D. S. Barth, W. Sutherling, J. Broffman, J. Beatty, Magnetic localization of a dipolar current source implanted in a sphere and a human cranium. Electroencephalogr. Clin. Neurophysiol. 63, 260–273 (1986).49. Y. C. Okada, A. Lahteenmäki, C. Xu, Experimental analysis of distortion of magnetoencephalography signals by the skull. Clin. Neurophysiol. 110, 230–238 (1999).50. S. Lew, M. S. Hämäläinen, S. P. Ahlfors, Y. Okada, Influence of unfused cranial bones on magnetoencephalography signals in human infants. Clin. Neurophysiol. 132, 708–719 (2021).51. M. Hämäläinen, R. Hari, R. J. Ilmoniemi, J. Knuutila, O. V. Lounasmaa, Magnetoencephalography—theory, instrumentation, and applications to noninvasive studies of the working human brain. Rev. Mod. Phys. 65, 413–497 (1993).52. “EEG source localization” in Handbook of Clinical Neurology (Elsevier, 2019; https://linkinghub.elsevier.com/retrieve/pii/B9780444640321000060)vol. 160, pp. 85–101.53. M. L. Bringas Vega, P. Nunez, J. Riera, R. Zhang, P. A. Valdes-Sosa, Editorial: Through a Glass, Darkly: The Influence of the EEG Reference on Inference About Brain Function and Disorders. Front. Neurosci. 13, 1341 (2019).54. S. Taulu, E. Larson, Unified Expression of the Quasi-Static Electromagnetic Field: Demonstration With MEG and EEG Signals. IEEE Trans. Biomed. Eng. 68, 992–1004 (2021).55. D. Hanley, L. S. Prichep, N. Badjatia, J. Bazarian, R. Chiacchierini, K. C. Curley, J. Garrett, E. Jones, R. Naunheim, B. O’Neil, J. O’Neill, D. W. Wright, J. S. Huff, A Brain Electrical Activity Electroencephalographic-Based Biomarker of Functional Impairment in Traumatic Brain Injury: A Multi-Site Validation Trial. J. Neurotrauma 35, 41–47 (2018).56. D. Hanley, L. S. Prichep, J. Bazarian, J. S. Huff, R. Naunheim, J. Garrett, E. B. Jones, D. W. Wright, J. O’Neill, N. Badjatia, D. Gandhi, K. C. Curley, R. Chiacchierini, B. O’Neil, D. C. Hack, Emergency Department Triage of Traumatic Head Injury Using a Brain Electrical Activity Biomarker: A Multisite Prospective Observational Validation Trial. Acad. Emerg. Med. 24, 617–627 (2017).57. D. H. Mathalon, S. Nicholas, B. J. Roach, T. Billah, S. Lavoie, T. Whitford, H. K. Hamilton, L. Addamo, A. Anohkin, T. Bekinschtein, A. Belger, K. Buccilli, J. Cahill, R. E. Carrión, S. Damiani, I. Dzafic, B. H. Ebdrup, I. Izyurov, J. Jarcho, R. Jenni, A. Jo, S. Kerins, C. Lee, E. A. Martin, R. Mayol-Troncoso, M. A. Niznikiewicz, M. Parvaz, O. Pogarell, J. Prieto-Montalvo, R. Rabin, D. R. Roalf, J. Rogers, D. F. Salisbury, R. Shaik, S. Shankman, M. C. Stevens, Y. N. Suen, N. C. Swann, X. Tang, J. L. Thompson, I. Tso, J. Wenzel, J. H. Zhou, J. Addington, L. Alameda, C. Arango, N. J. K. Breitborde, M. R. Broome, K. S. Cadenhead, M. E. Calkins, R. I. Castillo-Passi, E. Y. H. Chen, J. Choi, P. Conus, C. M. Corcoran, B. A. Cornblatt, C. M. Diaz-Caneja, L. M. Ellman, P. Fusar-Poli, P. A. Gaspar, C. Gerber, L. B. Glenthøj, L. E. Horton, C. L. M. Hui, J. Kambeitz, L. Kambeitz-Ilankovic, M. S. Keshavan, M. Kim, S.-W. Kim, N. Koutsouleris, J. S. Kwon, K. Langbein, D. Mamah, V. A. Mittal, M. Nordentoft, G. D. Pearlson, J. Perez, D. O. Perkins, A. R. Powers, F. W. Sabb, J. Schiffman, J. L. Shah, S. M. Silverstein, S. Smesny, W. S. Stone, G. P. Strauss, R. Upthegrove, S. K. Verma, J. Wang, D. H. Wolf, T. Zhang, S. Bouix, O. Pasternak, K.-I. K. Cho, M. J. Coleman, D. Dwyer, A. Nunez, Z. Tamayo, S. J. Wood, R. S. Kahn, J. M. Kane, P. D. McGorry, C. E. Bearden, B. Nelson, S. W. Woods, M. E. Shenton, Accelerating Medicines Partnership® Schizophrenia Program, G. A. Light, The electroencephalography protocol for the Accelerating Medicines Partnership® Schizophrenia Program: Reliability and stability of measures. Schizophrenia 11, 85 (2025).58. Y. Hou, H. Xia, T. He, B. Zhang, G. Qiu, A. Chen, N2 Responses in Youths With Psychosis Risk Syndrome and Their Association With Clinical Outcomes: A Cohort Follow-Up Study Based on the Three-Stimulus Visual Oddball Paradigm. Am. J. Psychiatry 181, 330–341 (2024).59. M. N. Van Stigt, E. A. Groenendijk, L. C. C. Van Meenen, A. A. G. A. Van De Munckhof, M. Theunissen, G. Franschman, M. D. Smeekes, J. A. F. Van Grondelle, G. Geuzebroek, A. Siegers, M. C. Visser, S. M. Van Schaik, P. H. A. Halkes, C. B. L. M. Majoie, Y. B. W. E. M. Roos, J. H. T. M. Koelman, M. S. Koopman, H. A. Marquering, W. V. Potters, J. M. Coutinho, Prehospital Detection of Large Vessel Occlusion Stroke With EEG: Results of the ELECTRA-STROKE Study. Neurology 101 (2023).60. A. Laguna, S. Pusil, I. Acero-Pousa, J. A. Zegarra-Valdivia, A. L. Paltrinieri, À. Bazán, P. Piras, C. Palomares I Perera, O. Garcia-Algar, S. Orlandi, How can cry acoustics associate newborns’ distress levels with neurophysiological and behavioral signals? Front. Neurosci. 17, 1266873 (2023).61. A. Biondi, V. Santoro, P. F. Viana, P. Laiou, D. K. Pal, E. Bruno, M. P. Richardson, Noninvasive mobile EEG as a tool for seizure monitoring and management: A systematic review. Epilepsia 63, 1041–1063 (2022).62. A. Lau-Zhu, M. P. H. Lau, G. McLoughlin, Mobile EEG in research on neurodevelopmental disorders: Opportunities and challenges. Dev. Cogn. Neurosci. 36, 100635 (2019).63. J. C. Edgar, C. L. Fisk, Y. Chen, B. Stone‐Howell, S. Liu, M. A. Hunter, M. Huang, J. Bustillo, J. M. Cañive, G. A. Miller, Identifying auditory cortex encoding abnormalities in schizophrenia: The utility of low‐frequency versus 40 Hz steady‐state measures. Psychophysiology 55, e13074 (2018).64. M. X. Huang, J. C. Edgar, R. J. Thoma, F. M. Hanlon, S. N. Moses, R. R. Lee, K. M. Paulson, M. P. Weisend, J. G. Irwin, J. R. Bustillo, L. E. Adler, G. A. Miller, J. M. Canive, Predicting EEG responses using MEG sources in superior temporal gyrus reveals source asynchrony in patients with schizophrenia. Clin. Neurophysiol. 114, 835–850 (2003).65. M. G. Bleichner, S. Debener, Concealed, Unobtrusive Ear-Centered EEG Acquisition: cEEGrids for Transparent EEG. Front. Hum. Neurosci. 11 (2017).66. S. Debener, F. Minow, R. Emkes, K. Gandras, M. De Vos, How about taking a low‐cost, small, and wireless EEG for a walk? Psychophysiology 49, 1617–1621 (2012).67. G. Niso, E. Romero, J. T. Moreau, A. Araujo, L. R. Krol, Wireless EEG: A survey of systems and studies. NeuroImage 269, 119774 (2023).68. N. S. J. Jacobsen, D. Kristanto, S. Welp, Y. C. Inceler, S. Debener, Preprocessing choices for P3 analyses with mobile EEG : A systematic literature review and interactive exploration. Psychophysiology 62, e14743 (2025).69. C. P. Kjeldsen, M. L. Neel, A. Jeanvoine, N. L. Maitre, Investigation of mothers’ elicited infant-directed speech and singing for preterm infants. Pediatr. Res. 97, 2367–2375 (2025).70. N. L. Maitre, A. P. Key, J. C. Slaughter, P. J. Yoder, M. L. Neel, C. Richard, M. T. Wallace, M. M. Murray, Neonatal Multisensory Processing in Preterm and Term Infants Predicts Sensory Reactivity and Internalizing Tendencies in Early Childhood. Brain Topogr. 33, 586–599 (2020).71. D. Cohen, B. N. Cuffin, K. Yunokuchi, R. Maniewski, C. Purcell, G. R. Cosgrove, J. Ives, J. G. Kennedy, D. L. Schomer, MEG versus EEG localization test using implanted sources in the human brain. Ann. Neurol. 28, 811–817 (1990).72. R. M. Leahy, J. C. Mosher, M. E. Spencer, M. X. Huang, J. D. Lewine, A study of dipole localization accuracy for MEG and EEG using a human skull phantom. Electroencephalogr. Clin. Neurophysiol. 107, 159–173 (1998).73. D. Sharon, M. S. Hämäläinen, R. B. H. Tootell, E. Halgren, J. W. Belliveau, The advantage of combining MEG and EEG: Comparison to fMRI in focally stimulated visual cortex. NeuroImage 36, 1225–1235 (2007).74. S. Komssi, J. Huttunen, H. J. Aronen, R. J. Ilmoniemi, EEG minimum-norm estimation compared with MEG dipole fitting in the localization of somatosensory sources at S1. Clin. Neurophysiol. 115, 534–542 (2004).75. R. Paetau, A. Ahonen, O. Salonen, M. Sams, Auditory Evoked Magnetic Fields to Tones and Pseudowords in Healthy Children and Adults: J. Clin. Neurophysiol. 12, 177–185 (1995).76. J. C. Edgar, Pediatric brain imaging research: Brain maturation constrains study design and clinical interpretation. Psychiatry Clin. Neurosci. 75, 267–269 (2021).77. J. C. Edgar, C. L. Fisk Iv, Y. Chen, B. Stone‐Howell, M. A. Hunter, M. Huang, J. R. Bustillo, J. M. Cañive, G. A. Miller, By our bootstraps: Comparing methods for measuring auditory 40 Hz steady‐state neural activity. Psychophysiology 54, 1110–1127 (2017).78. A. Hillebrand, G. R. Barnes, A Quantitative Assessment of the Sensitivity of Whole-Head MEG to Activity in the Adult Human Cortex. NeuroImage 16, 638–650 (2002).79. J. S. Ebersole, S. M. Ebersole, Combining MEG and EEG Source Modeling in Epilepsy Evaluations. J. Clin. Neurophysiol. 27, 360–371 (2010).80. W. Freeman, Combining fMRI with EEG and MEG in order to relate patterns of brain activity to cognition. Int. J. Psychophysiol. 73, 43–52 (2009).81. F. M. Hanlon, J. M. Houck, C. J. Pyeatt, S. L. Lundy, M. J. Euler, M. P. Weisend, R. J. Thoma, J. R. Bustillo, G. A. Miller, C. D. Tesche, Bilateral hippocampal dysfunction in schizophrenia. NeuroImage 58, 1158–1168 (2011).82. C. E. Meehan, A. I. Wiesman, R. K. Spooner, M. Schantell, J. A. Eastman, T. W. Wilson, Differences in Rhythmic Neural Activity Supporting the Temporal and Spatial Cueing of Attention. Cereb. Cortex 31, 4933–4944 (2021).83. M. P. Rempe, C. C. Casagrande, C. M. Embury, Y. Arif, S. Bashford, G. M. Garrison, R. J. Glesinger, H. J. Okelberry, E. L. Keifer, G. Picci, E. Heinrichs‐Graham, T. W. Wilson, Hippocampal and cortical oscillatory dynamics reflect semantic processing and predict behavioural performance. J. Physiol. 603, 3089–3106 (2025).84. F. M. Hanlon, M. P. Weisend, R. A. Yeo, M. Huang, R. R. Lee, R. J. Thoma, S. N. Moses, K. M. Paulson, G. A. Miller, J. M. Cañive, A Specific Test of Hippocampal Deficit in Schizophrenia. Behav. Neurosci. 119, 863–875 (2005).85. J. G. Samuelsson, P. Sundaram, S. Khan, M. I. Sereno, M. S. Hämäläinen, Detectability of cerebellar activity with magnetoencephalography and electroencephalography. Hum. Brain Mapp. 41, 2357–2372 (2020).86. C. Styliadis, R. Leung, S. Özcan, E. A. Moulton, E. Pang, M. J. Taylor, C. Papadelis, Atypical spatiotemporal activation of cerebellar lobules during emotional face processing in adolescents with autism. Hum. Brain Mapp. 42, 2099–2114 (2021).87. R. C. Leung, E. W. Pang, E. Anagnostou, M. J. Taylor, Young Adults with Autism Spectrum Disorder Show Early Atypical Neural Activity during Emotional Face Processing. Front. Hum. Neurosci. 12, 57 (2018).88. K. Safar, S. M. Wong, R. C. Leung, B. T. Dunkley, M. J. Taylor, Increased Functional Connectivity During Emotional Face Processing in Children With Autism Spectrum Disorder. Front. Hum. Neurosci. 12, 408 (2018).89. D. Cohen, B. N. Cuffin, Demonstration of useful differences between magnetoencephalogram and electroencephalogram. Electroencephalogr. Clin. Neurophysiol. 56, 38–51 (1983).90. D. M. Goldenholz, S. P. Ahlfors, M. S. Hämäläinen, D. Sharon, M. Ishitobi, L. M. Vaina, S. M. Stufflebeam, Mapping the signal‐to‐noise‐ratios of cortical sources in magnetoencephalography and electroencephalography. Hum. Brain Mapp. 30, 1077–1086 (2009).91. S. P. Ahlfors, J. Han, J. W. Belliveau, M. S. Hämäläinen, Sensitivity of MEG and EEG to Source Orientation. Brain Topogr. 23, 227–232 (2010).92. M. C. Piastra, A. Nüßing, J. Vorwerk, M. Clerc, C. Engwer, C. H. Wolters, A comprehensive study on electroencephalography and magnetoencephalography sensitivity to cortical and subcortical sources. Hum. Brain Mapp. 42, 978–992 (2021).93. S. P. Ahlfors, “MEG and Multimodal Integration” in Magnetoencephalography, S. Supek, C. J. Aine, Eds. (Springer International Publishing, Cham, 2019; http://link.springer.com/10.1007/978-3-319-62657-4_7-1), pp. 1–20.94. S. Knake, E. Halgren, H. Shiraishi, K. Hara, H. M. Hamer, P. E. Grant, V. A. Carr, D. Foxe, S. Camposano, E. Busa, T. Witzel, M. S. Hämäläinen, S. P. Ahlfors, E. B. Bromfield, P. M. Black, B. F. Bourgeois, A. J. Cole, G. R. Cosgrove, B. A. Dworetzky, J. R. Madsen, P. G. Larsson, D. L. Schomer, E. A. Thiele, A. M. Dale, B. R. Rosen, S. M. Stufflebeam, The value of multichannel MEG and EEG in the presurgical evaluation of 70 epilepsy patients. Epilepsy Res. 69, 80–86 (2006).95. G. Ramantani, R. Boor, R. Paetau, N. Ille, R. Feneberg, A. Rupp, T. Boppel, M. Scherg, D. Rating, T. Bast, MEG Versus EEG: Influence of Background Activity on Interictal Spike Detection: J. Clin. Neurophysiol. 23, 498–508 (2006).96. R. Hari, S. Baillet, G. Barnes, R. Burgess, N. Forss, J. Gross, M. Hämäläinen, O. Jensen, R. Kakigi, F. Mauguière, N. Nakasato, A. Puce, G.-L. Romani, A. Schnitzler, S. Taulu, IFCN-endorsed practical guidelines for clinical magnetoencephalography (MEG). Clin. Neurophysiol. 129, 1720–1747 (2018).97. H. Stefan, E. Trinka, Magnetoencephalography (MEG): Past, current and future perspectives for improved differentiation and treatment of epilepsies. Seizure 44, 121–124 (2017).98. Ü. Aydin, J. Vorwerk, M. Dümpelmann, P. Küpper, H. Kugel, M. Heers, J. Wellmer, C. Kellinghaus, J. Haueisen, S. Rampp, H. Stefan, C. H. Wolters, Combined EEG/MEG Can Outperform Single Modality EEG or MEG Source Reconstruction in Presurgical Epilepsy Diagnosis. PLOS ONE 10, e0118753 (2015).99. M. Fuchs, M. Wagner, H.-A. Wischmann, T. Köhler, A. Theißen, R. Drenckhahn, H. Buchner, Improving source reconstructions by combining bioelectric and biomagnetic data. Electroencephalogr. Clin. Neurophysiol. 107, 93–111 (1998).100. A. J. R. T. da Silva, A. Landreth, J. Bickle, Engineering the next Revolution in Neuroscience: The New Science of Experiment Planning (Oxford University Press, Oxford, 2014).101. D. M. Tucker, Spatial sampling of head electrical fields: the geodesic sensor net. Electroencephalogr. Clin. Neurophysiol. 87, 154–163 (1993).102. F. A. Gibbs, J. R. Knott, Growth of the electrical activity of the cortex. Electroencephalogr. Clin. Neurophysiol. 1, 223–229 (1949).103. H. Penuel, F. Corbin, R. G. Bickford, Studies of the electroencephalogram of normal children: comparison of visual and automatic frequency analyses. Electroencephalogr. Clin. Neurophysiol. 7, 15–28 (1955).104. M. Matoušek, I. Petersén, Automatic evaluation of EEG background activity by means of age-dependent EEG quotients. Electroencephalogr. Clin. Neurophysiol. 35, 603–612 (1973).105. O. Eeg-Olofsson, I. Petersén, U. Selldén, The Development of the Electroencephalogram in Normal Children from the Age of 1 Through 15 Years – Paroxysmal activity. Neuropediatrics 2, 375–404 (1971).106. T. Gasser, R. Verleger, P. Bächer, L. Sroka, Development of the EEG of school-age children and adolescents. I. Analysis of band power. Electroencephalogr. Clin. Neurophysiol. 69, 91–99 (1988).107. J. C. Edgar, S. Y. Khan, L. Blaskey, V. Y. Chow, M. Rey, W. Gaetz, K. M. Cannon, J. F. Monroe, L. Cornew, S. Qasmieh, S. Liu, J. P. Welsh, S. E. Levy, T. P. L. Roberts, Neuromagnetic Oscillations Predict Evoked-Response Latency Delays and Core Language Deficits in Autism Spectrum Disorders. J. Autism Dev. Disord. 45, 395–405 (2015).108. B. J. Fisch, R. Spehlmann, Fisch and Spehlmann’s EEG Primer: Basic Principles of Digital and Analog EEG (Elsevier, Amsterdam ; New York, 3rd rev. and enl. ed., 1999).109. D. L. Schomer, F. H. Lopes Da Silva, Eds., Niedermeyer’s Electroencephalography (Oxford University Press, 2017; https://academic.oup.com/book/35515)vol. 1.110. W. Klimesch, EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res. Rev. 29, 169–195 (1999).111. L. Cragg, N. Kovacevic, A. R. McIntosh, C. Poulsen, K. Martinu, G. Leonard, T. Paus, Maturation of EEG power spectra in early adolescence: a longitudinal study: Maturation of EEG. Dev. Sci. 14, 935–943 (2011).112. I. Petersén, O. Eeg-Olofsson, The Development of the Electroencephalogram in Normal Children from the Age of 1 Through 15 Years – Non-paroxysmal activity. Neuropediatrics 2, 247–304 (1971).113. A. R. Clarke, R. J. Barry, R. McCarthy, M. Selikowitz, Age and sex effects in the EEG: development of the normal child. Clin. Neurophysiol. 112, 806–814 (2001).114. A. Katada, H. Ozaki, H. Suzuki, K. Suhara, Developmental characteristics of normal and mentally retarded children’s EEGs. Electroencephalogr. Clin. Neurophysiol. 52, 192–201 (1981).115. M. Matsuura, K. Yamamoto, H. Fukuzawa, Y. Okubo, H. Uesugi, M. Moriiwa, T. Kojima, Y. Shimazono, Age development and sex differences of various EEG elements in healthy children and adults — Quantification by a computerized wave form recognition method. Electroencephalogr. Clin. Neurophysiol. 60, 394–406 (1985).116. E. D. Adrian, B. H. C. Matthews, The Berger rhythm: potential changes from the occipital lobes in man. Brain 57, 355–385 (1934).117. H. Berger, Über das Elektrenkephalogramm des Menschen. Arch. Für Psychiatr. Nervenkrankh. 87, 527–570 (1929).118. R. Salmelin, R. Hari, Characterization of spontaneous MEG rhythms in healthy adults. Electroencephalogr. Clin. Neurophysiol. 91, 237–248 (1994).119. J. Tiihonen, M. Kajola, R. Hari, Magnetic mu rhythm in man. Neuroscience 32, 793–800 (1989).120. M. P. Rempe, L. R. Ott, G. Picci, S. H. Penhale, N. J. Christopher-Hayes, B. J. Lew, N. M. Petro, C. M. Embury, M. Schantell, H. J. Johnson, H. J. Okelberry, K. L. Losh, M. P. Willett, R. A. Losh, Y.-P. Wang, V. D. Calhoun, J. M. Stephen, E. Heinrichs-Graham, M. J. Kurz, T. W. Wilson, Spontaneous cortical dynamics from the first years to the golden years. Proc. Natl. Acad. Sci. 120, e2212776120 (2023).121. B. J. He, Scale-free brain activity: past, present, and future. Trends Cogn. Sci. 18, 480–487 (2014).122. W. J. Freeman, J. Zhai, Simulated power spectral density (PSD) of background electrocorticogram (ECoG). Cogn. Neurodyn. 3, 97–103 (2009).123. T. Donoghue, M. Haller, E. J. Peterson, P. Varma, P. Sebastian, R. Gao, T. Noto, A. H. Lara, J. D. Wallis, R. T. Knight, A. Shestyuk, B. Voytek, Parameterizing neural power spectra into periodic and aperiodic components. Nat. Neurosci. 23, 1655–1665 (2020).124. T. Donoghue, J. Dominguez, B. Voytek, Electrophysiological Frequency Band Ratio Measures Conflate Periodic and Aperiodic Neural Activity. eneuro 7, ENEURO.0192-20.2020 (2020).125. B. Ostlund, T. Donoghue, B. Anaya, K. E. Gunther, S. L. Karalunas, B. Voytek, K. E. Pérez-Edgar, Spectral parameterization for studying neurodevelopment: How and why. Dev. Cogn. Neurosci. 54, 101073 (2022).126. W. He, T. Donoghue, P. F. Sowman, R. A. Seymour, J. Brock, S. Crain, B. Voytek, A. Hillebrand, Co-Increasing Neuronal Noise and Beta Power in the Developing Brain. Neuroscience [Preprint] (2019). https://doi.org/10.1101/839258.127. T. Donoghue, N. Schaworonkow, B. Voytek, Methodological considerations for studying neural oscillations. Eur. J. Neurosci. 55, 3502–3527 (2022).128. R. Gao, E. J. Peterson, B. Voytek, Inferring synaptic excitation/inhibition balance from field potentials. NeuroImage 158, 70–78 (2017).129. V. Shirhatti, A. Borthakur, S. Ray, Effect of Reference Scheme on Power and Phase of the Local Field Potential. Neural Comput. 28, 882–913 (2016).130. K. Hoechstetter, H. Bornfleth, D. Weckesser, N. Ille, P. Berg, M. Scherg, BESA Source Coherence: A New Method to Study Cortical Oscillatory Coupling. Brain Topogr. 16, 233–238 (2004).131. M. Scherg, P. Berg, New concepts of brain source imaging and localization. Electroencephalogr. Clin. Neurophysiol. Suppl. 46, 127–137 (1996).132. M. Scherg, T. W. Picton, Separation and identification of event-related potential components by brain electric source analysis. Electroencephalogr. Clin. Neurophysiol. Suppl. 42, 24–37 (1991).133. S. L. Karalunas, B. D. Ostlund, B. R. Alperin, M. Figuracion, H. C. Gustafsson, E. M. Deming, D. Foti, D. Antovich, J. Dude, J. Nigg, E. Sullivan, Electroencephalogram aperiodic power spectral slope can be reliably measured and predicts ADHD risk in early development. Dev. Psychobiol. 64, e22228 (2022).134. M. Tröndle, T. Popov, S. Dziemian, N. Langer, Decomposing the role of alpha oscillations during brain maturation. eLife 11, e77571 (2022).135. D. Cellier, J. Riddle, I. Petersen, K. Hwang, The development of theta and alpha neural oscillations from ages 3 to 24 years. Dev. Cogn. Neurosci. 50, 100969 (2021).136. V. Carter Leno, A. Pickles, S. Van Noordt, S. Huberty, J. Desjardins, S. J. Webb, M. Elsabbagh, 12-Month peak alpha frequency is a correlate but not a longitudinal predictor of non-verbal cognitive abilities in infants at low and high risk for autism spectrum disorder. Dev. Cogn. Neurosci. 48, 100938 (2021).137. J. Rico‐Picó, S. Moyano, Á. Conejero, Á. Hoyo, M. Á. Ballesteros‐Duperón, M. R. Rueda, Early development of electrophysiological activity: Contribution of periodic and aperiodic components of the EEG signal. Psychophysiology 60, e14360 (2023).138. J. Favaro, M. A. Colombo, E. Mikulan, S. Sartori, M. Nosadini, M. F. Pelizza, M. Rosanova, S. Sarasso, M. Massimini, I. Toldo, The maturation of aperiodic EEG activity across development reveals a progressive differentiation of wakefulness from sleep. NeuroImage 277, 120264 (2023).139. A. T. Hill, G. M. Clark, F. J. Bigelow, J. A. G. Lum, P. G. Enticott, Periodic and aperiodic neural activity displays age-dependent changes across early-to-middle childhood. Dev. Cogn. Neurosci. 54, 101076 (2022).140. N. Schaworonkow, B. Voytek, Longitudinal changes in aperiodic and periodic activity in electrophysiological recordings in the first seven months of life. Dev. Cogn. Neurosci. 47, 100895 (2021).141. M. McSweeney, S. Morales, E. A. Valadez, G. A. Buzzell, L. Yoder, W. P. Fifer, N. Pini, L. C. Shuffrey, A. J. Elliott, J. R. Isler, N. A. Fox, Age-related trends in aperiodic EEG activity and alpha oscillations during early- to middle-childhood. NeuroImage 269, 119925 (2023).142. N. Schaworonkow, V. V. Nikulin, Is sensor space analysis good enough? Spatial patterns as a tool for assessing spatial mixing of EEG/MEG rhythms. NeuroImage 253, 119093 (2022).143. S. Y. Shirazi, J. Onton, S. Makeig, Simulating Scalp EEG from Ultrahigh-Density ECoG Data Illustrates Cortex to Scalp Projection Patterns. Neuroscience [Preprint] (2025). https://doi.org/10.1101/2025.06.24.660870.144. M. M. Vandewouw, J. Sato, K. Safar, N. Rhodes, M. J. Taylor, The development of aperiodic and periodic resting-state power between early childhood and adulthood: New insights from optically pumped magnetometers. Dev. Cogn. Neurosci. 69, 101433 (2024).145. H. L. Green, J. C. Edgar, K. Mol, M. McNamee, L. Prosser, M. Kim, E. S. Kuschner, G. A. Miller, Y. Chen, Resting state periodic and aperiodic brain oscillations from birth to preschool years: Aperiodic maturity predicts developmental course. Neuroscience [Preprint] (2025). https://doi.org/10.1101/2025.05.11.653151.146. J. M. Huntenburg, P.-L. Bazin, D. S. Margulies, Large-Scale Gradients in Human Cortical Organization. Trends Cogn. Sci. 22, 21–31 (2018).147. K. Mahjoory, J.-M. Schoffelen, A. Keitel, J. Gross, The frequency gradient of human resting-state brain oscillations follows cortical hierarchies. eLife 9, e53715 (2020).148. J. C. Edgar, L. Blaskey, Y. Chen, O. E. Podolak, D. L. Murray, M. McNamee, K. Konka, J. I. Berman, T. P. L. Roberts, M. Huang, K. B. Arbogast, C. L. Master, Regionally Specific Resting-State Beta Neural Power Predicts Brain Injury and Symptom Recovery in Adolescents with Concussion: A Longitudinal Study. J. Neurotrauma, 08977151251383542 (2025).149. M.-X. Huang, A. Robb Swan, A. Angeles Quinto, J. W. Huang, B. G. De-la-Garza, C. W. Huang, J. R. Hesselink, E. D. Bigler, E. A. Wilde, J. E. Max, Resting-State Magnetoencephalography Source Imaging Pilot Study in Children with Mild Traumatic Brain Injury. J. Neurotrauma 37, 994–1001 (2020).150. R. A. I. Bethlehem, J. Seidlitz, S. R. White, J. W. Vogel, K. M. Anderson, C. Adamson, S. Adler, G. S. Alexopoulos, E. Anagnostou, A. Areces-Gonzalez, D. E. Astle, B. Auyeung, M. Ayub, J. Bae, G. Ball, S. Baron-Cohen, R. Beare, S. A. Bedford, V. Benegal, F. Beyer, J. Blangero, M. Blesa Cábez, J. P. Boardman, M. Borzage, J. F. Bosch-Bayard, N. Bourke, V. D. Calhoun, M. M. Chakravarty, C. Chen, C. Chertavian, G. Chetelat, Y. S. Chong, J. H. Cole, A. Corvin, M. Costantino, E. Courchesne, F. Crivello, V. L. Cropley, J. Crosbie, N. Crossley, M. Delarue, R. Delorme, S. Desrivieres, G. A. Devenyi, M. A. Di Biase, R. Dolan, K. A. Donald, G. Donohoe, K. Dunlop, A. D. Edwards, J. T. Elison, C. T. Ellis, J. A. Elman, L. Eyler, D. A. Fair, E. Feczko, P. C. Fletcher, P. Fonagy, C. E. Franz, L. Galan-Garcia, A. Gholipour, J. Giedd, J. H. Gilmore, D. C. Glahn, I. M. Goodyer, P. E. Grant, N. A. Groenewold, F. M. Gunning, R. E. Gur, R. C. Gur, C. F. Hammill, O. Hansson, T. Hedden, A. Heinz, R. N. Henson, K. Heuer, J. Hoare, B. Holla, A. J. Holmes, R. Holt, H. Huang, K. Im, J. Ipser, C. R. Jack, A. P. Jackowski, T. Jia, K. A. Johnson, P. B. Jones, D. T. Jones, R. S. Kahn, H. Karlsson, L. Karlsson, R. Kawashima, E. A. Kelley, S. Kern, K. W. Kim, M. G. Kitzbichler, W. S. Kremen, F. Lalonde, B. Landeau, S. Lee, J. Lerch, J. D. Lewis, J. Li, W. Liao, C. Liston, M. V. Lombardo, J. Lv, C. Lynch, T. T. Mallard, M. Marcelis, R. D. Markello, S. R. Mathias, B. Mazoyer, P. McGuire, M. J. Meaney, A. Mechelli, N. Medic, B. Misic, S. E. Morgan, D. Mothersill, J. Nigg, M. Q. W. Ong, C. Ortinau, R. Ossenkoppele, M. Ouyang, L. Palaniyappan, L. Paly, P. M. Pan, C. Pantelis, M. M. Park, T. Paus, Z. Pausova, D. Paz-Linares, A. Pichet Binette, K. Pierce, X. Qian, J. Qiu, A. Qiu, A. Raznahan, T. Rittman, A. Rodrigue, C. K. Rollins, R. Romero-Garcia, L. Ronan, M. D. Rosenberg, D. H. Rowitch, G. A. Salum, T. D. Satterthwaite, H. L. Schaare, R. J. Schachar, A. P. Schultz, G. Schumann, M. Schöll, D. Sharp, R. T. Shinohara, I. Skoog, C. D. Smyser, R. A. Sperling, D. J. Stein, A. Stolicyn, J. Suckling, G. Sullivan, Y. Taki, B. Thyreau, R. Toro, N. Traut, K. A. Tsvetanov, N. B. Turk-Browne, J. J. Tuulari, C. Tzourio, É. Vachon-Presseau, M. J. Valdes-Sosa, P. A. Valdes-Sosa, S. L. Valk, T. Van Amelsvoort, S. N. Vandekar, L. Vasung, L. W. Victoria, S. Villeneuve, A. Villringer, P. E. Vértes, K. Wagstyl, Y. S. Wang, S. K. Warfield, V. Warrier, E. Westman, M. L. Westwater, H. C. Whalley, A. V. Witte, N. Yang, B. Yeo, H. Yun, A. Zalesky, H. J. Zar, A. Zettergren, J. H. Zhou, H. Ziauddeen, A. Zugman, X. N. Zuo, 3R-BRAIN, AIBL, C. Rowe, Alzheimer’s Disease Neuroimaging Initiative, Alzheimer’s Disease Repository Without Borders Investigators, G. B. Frisoni, CALM Team, Cam-CAN, CCNP, COBRE, cVEDA, ENIGMA Developmental Brain Age Working Group, Developing Human Connectome Project, FinnBrain, Harvard Aging Brain Study, IMAGEN, KNE96, The Mayo Clinic Study of Aging, NSPN, POND, The PREVENT-AD Research Group, A. P. Binette, VETSA, E. T. Bullmore, A. F. Alexander-Bloch, Brain charts for the human lifespan. Nature 604, 525–533 (2022).151. J. H. Gilmore, W. Lin, M. W. Prastawa, C. B. Looney, Y. S. K. Vetsa, R. C. Knickmeyer, D. D. Evans, J. K. Smith, R. M. Hamer, J. A. Lieberman, G. Gerig, Regional Gray Matter Growth, Sexual Dimorphism, and Cerebral Asymmetry in the Neonatal Brain. J. Neurosci. 27, 1255–1260 (2007).152. N. Gogtay, J. N. Giedd, L. Lusk, K. M. Hayashi, D. Greenstein, A. C. Vaituzis, T. F. Nugent, D. H. Herman, L. S. Clasen, A. W. Toga, J. L. Rapoport, P. M. Thompson, Dynamic mapping of human cortical development during childhood through early adulthood. Proc. Natl. Acad. Sci. 101, 8174–8179 (2004).153. S. C. L. Deoni, D. C. Dean, J. Remer, H. Dirks, J. O’Muircheartaigh, Cortical maturation and myelination in healthy toddlers and young children. NeuroImage 115, 147–161 (2015).154. S. C. L. Deoni, E. Mercure, A. Blasi, D. Gasston, A. Thomson, M. Johnson, S. C. R. Williams, D. G. M. Murphy, Mapping infant brain myelination with magnetic resonance imaging. J. Neurosci. Off. J. Soc. Neurosci. 31, 784–791 (2011).155. M. Ouyang, J. Dubois, Q. Yu, P. Mukherjee, H. Huang, Delineation of early brain development from fetuses to infants with diffusion MRI and beyond. NeuroImage 185, 836–850 (2019).156. R. Tyzio, R. Cossart, I. Khalilov, M. Minlebaev, C. A. Hübner, A. Represa, Y. Ben-Ari, R. Khazipov, Maternal Oxytocin Triggers a Transient Inhibitory Switch in GABA Signaling in the Fetal Brain During Delivery. Science 314, 1788–1792 (2006).157. E. C. Porges, G. Jensen, B. Foster, R. A. Edden, N. A. Puts, The trajectory of cortical GABA across the lifespan, an individual participant data meta-analysis of edited MRS studies. eLife 10, e62575 (2021).158. S. K. Basu, S. Pradhan, A. J. Du Plessis, Y. Ben-Ari, C. Limperopoulos, GABA and glutamate in the preterm neonatal brain: In-vivo measurement by magnetic resonance spectroscopy. NeuroImage 238, 118215 (2021).159. Y. Ben-Ari, The GABA excitatory/inhibitory developmental sequence: A personal journey. Neuroscience 279, 187–219 (2014).160. M. Yanez Lopez, A. N. Price, N. A. J. Puts, E. J. Hughes, R. A. E. Edden, G. M. McAlonan, T. Arichi, E. De Vita, Simultaneous quantification of GABA, Glx and GSH in the neonatal human brain using magnetic resonance spectroscopy. NeuroImage 233, 117930 (2021).161. N. Maikusa, Y. Zhu, A. Uematsu, A. Yamashita, K. Saotome, N. Okada, K. Kasai, K. Okanoya, O. Yamashita, S. C. Tanaka, S. Koike, Comparison of traveling‐subject and ComBat harmonization methods for assessing structural brain characteristics. Hum. Brain Mapp. 42, 5278–5287 (2021).162. J. C. Beer, N. J. Tustison, P. A. Cook, C. Davatzikos, Y. I. Sheline, R. T. Shinohara, K. A. Linn, Longitudinal ComBat: A method for harmonizing longitudinal multi-scanner imaging data. NeuroImage 220, 117129 (2020).163. M. P. Weisend, F. M. Hanlon, R. Montaño, S. P. Ahlfors, A. C. Leuthold, D. Pantazis, J. C. Mosher, A. P. Georgopoulos, M. S. Hämäläinen, C. J. Aine, Paving the way for cross-site pooling of magnetoencephalography (MEG) data. Int. Congr. Ser. 1300, 615–618 (2007).164. D. Vaghari, R. Bruna, L. E. Hughes, D. Nesbitt, R. Tibon, J. B. Rowe, F. Maestu, R. N. Henson, A multi-site, multi-participant magnetoencephalography resting-state dataset to study dementia: The BioFIND dataset. NeuroImage 258, 119344 (2022).165. L. E. Hughes, R. N. Henson, E. Pereda, R. Bruña, D. López‐Sanz, A. J. Quinn, M. W. Woolrich, A. C. Nobre, J. B. Rowe, F. Maestú, the BioFIND Working Group, Biomagnetic biomarkers for dementia: A pilot multicentre study with a recommended methodological framework for magnetoencephalography. Alzheimers Dement. Diagn. Assess. Dis. Monit. 11, 450–462 (2019).166. K. Marinkovic, B. Cox, K. Reid, E. Halgren, Head position in the MEG helmet affects the sensitivity to anterior sources. Neurol. Clin. Neurophysiol. NCN 2004, 30 (2004).167. Y.-H. Chen, J. Saby, E. Kuschner, W. Gaetz, J. C. Edgar, T. P. L. Roberts, Magnetoencephalography and the infant brain. NeuroImage 189, 445–458 (2019).168. J. C. Edgar, R. Murray, E. S. Kuschner, K. Pratt, D. N. Paulson, J. Dell, R. Golembski, P. Lam, L. Bloy, W. Gaetz, T. P. L. Roberts, The maturation of auditory responses in infants and young children: a cross-sectional study from 6 to 59 months. Front. Neuroanat. 9 (2015).169. T. P. L. Roberts, D. N. Paulson, E. Hirschkoff, K. Pratt, A. Mascarenas, P. Miller, M. Han, J. Caffrey, C. Kincade, B. Power, R. Murray, V. Chow, C. Fisk, M. Ku, D. Chudnovskaya, J. Dell, R. Golembski, P. Lam, L. Blaskey, E. Kuschner, L. Bloy, W. Gaetz, J. C. Edgar, Artemis 123: development of a whole-head infant and young child MEG system. Front. Hum. Neurosci. 8 (2014).170. S. Taulu, R. Hari, Removal of magnetoencephalographic artifacts with temporal signal‐space separation: Demonstration with single‐trial auditory‐evoked responses. Hum. Brain Mapp. 30, 1524–1534 (2009).171. S. Taulu, M. Kajola, J. Simola, Suppression of Interference and Artifacts by the Signal Space Separation Method. Brain Topogr. 16, 269–275 (2004).172. S. Taulu, J. Simola, Spatiotemporal signal space separation method for rejecting nearby interference in MEG measurements. Phys. Med. Biol. 51, 1759–1768 (2006).173. D. T. Wehner, M. S. Hämäläinen, M. Mody, S. P. Ahlfors, Head movements of children in MEG: Quantification, effects on source estimation, and compensation. NeuroImage 40, 541–550 (2008).174. E. Larson, S. Taulu, The Importance of Properly Compensating for Head Movements During MEG Acquisition Across Different Age Groups. Brain Topogr. 30, 172–181 (2017).175. T. M. Tierney, N. Holmes, S. Mellor, J. D. López, G. Roberts, R. M. Hill, E. Boto, J. Leggett, V. Shah, M. J. Brookes, R. Bowtell, G. R. Barnes, Optically pumped magnetometers: From quantum origins to multi-channel magnetoencephalography. NeuroImage 199, 598–608 (2019).176. E. Boto, S. S. Meyer, V. Shah, O. Alem, S. Knappe, P. Kruger, T. M. Fromhold, M. Lim, P. M. Glover, P. G. Morris, R. Bowtell, G. R. Barnes, M. J. Brookes, A new generation of magnetoencephalography: Room temperature measurements using optically-pumped magnetometers. NeuroImage 149, 404–414 (2017).177. B. Zhao, L. Li, Y. Zhang, J. Tang, Y. Liu, Y. Zhai, Optically Pumped Magnetometers Recent Advances and Applications in Biomagnetism: A Review. IEEE Sens. J. 23, 18949–18962 (2023).178. T. Bardouille, V. Smith, E. Vajda, C. D. Leslie, N. Holmes, Noise Reduction and Localization Accuracy in a Mobile Magnetoencephalography System. Sensors 24, 3503 (2024).179. R. M. Hill, E. Boto, M. Rea, N. Holmes, J. Leggett, L. A. Coles, M. Papastavrou, S. K. Everton, B. A. E. Hunt, D. Sims, J. Osborne, V. Shah, R. Bowtell, M. J. Brookes, Multi-channel whole-head OPM-MEG: Helmet design and a comparison with a conventional system. NeuroImage 219, 116995 (2020).180. N. Rhodes, L. Rier, K. D. Singh, J. Sato, M. M. Vandewouw, N. Holmes, E. Boto, R. M. Hill, M. Rea, M. J. Taylor, M. J. Brookes, Measuring the neurodevelopmental trajectory of excitatory-inhibitory balance via visual gamma oscillations. Imaging Neurosci. 3, imag_a_00527 (2025).181. P. Corvilain, V. Wens, M. Bourguignon, C. Capparini, L. Fourdin, M. Ferez, O. Feys, X. De Tiège, J. Bertels, Pushing the boundaries of MEG based on optically pumped magnetometers towards early human life. Imaging Neurosci. 3, imag_a_00489 (2025).182. L. Power, T. Bardouille, K. M. Ikeda, A. Omisade, Validation of On-Head OPM MEG for Language Laterality Assessment. Brain Topogr. 38, 8 (2025).183. E. Boto, N. Holmes, J. Leggett, G. Roberts, V. Shah, S. S. Meyer, L. D. Muñoz, K. J. Mullinger, T. M. Tierney, S. Bestmann, G. R. Barnes, R. Bowtell, M. J. Brookes, Moving magnetoencephalography towards real-world applications with a wearable system. Nature 555, 657–661 (2018).184. G. C. O’Neill, R. A. Seymour, S. Mellor, N. A. Alexander, T. M. Tierney, L. Bernachot, M. Fahimi Hnazaee, M. E. Spedden, R. C. Timms, D. Bush, S. Bestmann, M. J. Brookes, G. R. Barnes, Combining video telemetry and wearable MEG for naturalistic imaging. Imaging Neurosci. 3, imag_a_00495 (2025).185. R. A. Seymour, N. Alexander, S. Mellor, G. C. O’Neill, T. M. Tierney, G. R. Barnes, E. A. Maguire, Using OPMs to measure neural activity in standing, mobile participants. NeuroImage 244, 118604 (2021).186. L. Rier, N. Rhodes, D. O. Pakenham, E. Boto, N. Holmes, R. M. Hill, G. Reina Rivero, V. Shah, C. Doyle, J. Osborne, R. W. Bowtell, M. Taylor, M. J. Brookes, Tracking the neurodevelopmental trajectory of beta band oscillations with optically pumped magnetometer-based magnetoencephalography. eLife 13, RP94561 (2024).187. K. Safar, M. M. Vandewouw, N. Rhodes, J. Sato, M. J. Taylor, Mapping neural signatures of face processing in young children: an OPM-MEG study. Soc. Cogn. Affect. Neurosci. 20, nsaf076 (2025).188. J. Iivanainen, A. Borna, R. Zetter, T. R. Carter, J. M. Stephen, J. McKay, L. Parkkonen, S. Taulu, P. D. D. Schwindt, Calibration and Localization of Optically Pumped Magnetometers Using Electromagnetic Coils. Sensors 22, 3059 (2022).189. A. Borna, J. Iivanainen, T. R. Carter, J. McKay, S. Taulu, J. Stephen, P. D. D. Schwindt, Cross-Axis projection error in optically pumped magnetometers and its implication for magnetoencephalography systems. NeuroImage 247, 118818 (2022).190. N. Holmes, J. Leggett, E. Boto, G. Roberts, R. M. Hill, T. M. Tierney, V. Shah, G. R. Barnes, M. J. Brookes, R. Bowtell, A bi-planar coil system for nulling background magnetic fields in scalp mounted magnetoencephalography. NeuroImage 181, 760–774 (2018).191. M. Rea, N. Holmes, R. M. Hill, E. Boto, J. Leggett, L. J. Edwards, D. Woolger, E. Dawson, V. Shah, J. Osborne, R. Bowtell, M. J. Brookes, Precision magnetic field modelling and control for wearable magnetoencephalography. NeuroImage 241, 118401 (2021).192. R. M. Hill, J. Devasagayam, N. Holmes, E. Boto, V. Shah, J. Osborne, K. Safar, F. Worcester, C. Mariani, E. Dawson, D. Woolger, R. Bowtell, M. J. Taylor, M. J. Brookes, Using OPM-MEG in contrasting magnetic environments. NeuroImage 253, 119084 (2022).193. N. Holmes, M. Rea, J. Chalmers, J. Leggett, L. J. Edwards, P. Nell, S. Pink, P. Patel, J. Wood, N. Murby, D. Woolger, E. Dawson, C. Mariani, T. M. Tierney, S. Mellor, G. C. O’Neill, E. Boto, R. M. Hill, V. Shah, J. Osborne, R. Pardington, P. Fierlinger, G. R. Barnes, P. Glover, M. J. Brookes, R. Bowtell, A lightweight magnetically shielded room with active shielding. Sci. Rep. 12, 13561 (2022).194. H. Schofield, R. M. Hill, O. Feys, N. Holmes, J. Osborne, C. Doyle, D. Bobela, P. Corvilain, V. Wens, L. Rier, R. Bowtell, M. Ferez, K. J. Mullinger, S. Coleman, N. Rhodes, M. Rea, Z. Tanner, E. Boto, X. De Tiège, V. Shah, M. J. Brookes, A novel, robust, and portable platform for magnetoencephalography using optically-pumped magnetometers. Imaging Neurosci. 2, imag–2–00283 (2024).195. N. Holmes, M. Rea, R. M. Hill, J. Leggett, L. J. Edwards, P. J. Hobson, E. Boto, T. M. Tierney, L. Rier, G. R. Rivero, V. Shah, J. Osborne, T. M. Fromhold, P. Glover, M. J. Brookes, R. Bowtell, Enabling ambulatory movement in wearable magnetoencephalography with matrix coil active magnetic shielding. NeuroImage 274, 120157 (2023).196. M. J. Brookes, J. Leggett, M. Rea, R. M. Hill, N. Holmes, E. Boto, R. Bowtell, Magnetoencephalography with optically pumped magnetometers (OPM-MEG): the next generation of functional neuroimaging. Trends Neurosci. 45, 621–634 (2022).197. K. Safar, M. M. Vandewouw, J. Sato, J. Devasagayam, R. M. Hill, M. Rea, M. J. Brookes, M. J. Taylor, Using optically pumped magnetometers to replicate task-related responses in next generation magnetoencephalography. Sci. Rep. 14, 6513 (2024).198. R. M. Hill, H. Schofield, E. Boto, L. Rier, J. Osborne, C. Doyle, F. Worcester, T. Hayward, N. Holmes, R. Bowtell, V. Shah, M. J. Brookes, Optimising the sensitivity of optically-pumped magnetometer magnetoencephalography to gamma band electrophysiological activity. Imaging Neurosci. 2, imag–2–00112 (2024).199. A. I. Bagić, S. M. Bowyer, R. C. Burgess, M. E. Funke, A. Lowden, I. S. Mohamed, T. Wilson, W. Zhang, A. J. Zillgitt, J. R. Tenney, Role of optically pumped magnetometers in presurgical epilepsy evaluation: Commentary of the American Clinical Magnetoencephalography Society. Epilepsia 64, 3155–3159 (2023).200. O. Feys, V. Wens, C. Depondt, E. Rikir, N. Gaspard, W. Van Paesschen, A. Aeby, O. Bodart, E. Carrette, N. Holmes, M. Brookes, M. Ferez, P. Corvilain, X. De Tiège, On‐scalp magnetoencephalography based on optically pumped magnetometers to investigate temporal lobe epilepsy. Epilepsia 66 (2025).201. J. Ren, M. Ding, Y. Peng, C. Sun, C. Yang, S. Zhou, J. Tian, Q. Wang, Z. Li, A comparative study on the detection and localization of interictal epileptiform discharges in magnetoencephalography using optically pumped magnetometers versus superconducting quantum interference devices. NeuroImage 312, 121232 (2025).202. O. Feys, P. Corvilain, E. Labyt, M. Mahmoudzadeh, L. Routier, C. Sculier, N. Holmes, M. Brookes, S. Goldman, R. Romain, S. Mitryukovskiy, A. Palacios-Laloy, D. Schwartz, N. Betrouni, P. Derambure, F. Wallois, V. Wens, X. De Tiège, Tri-axial rubidium and helium optically pumped magnetometers for on-scalp magnetoencephalography recording of interictal epileptiform discharges: a case study. Front. Neurosci. 17, 1284262 (2023).203. O. Feys, P. Corvilain, A. Aeby, C. Sculier, N. Holmes, M. Brookes, S. Goldman, V. Wens, X. De Tiège, On-Scalp Optically Pumped Magnetometers versus Cryogenic Magnetoencephalography for Diagnostic Evaluation of Epilepsy in School-aged Children. Radiology 304, 429–434 (2022).204. R. Hari, A. Puce, MEG-EEG Primer (Oxford University Press, New York, NY, 2nd ed., 2023).205. N. Rhodes, J. Sato, K. Safar, K. Amorim, M. J. Taylor, M. J. Brookes, Paediatric magnetoencephalography and its role in neurodevelopmental disorders. Br. J. Radiol. 97, 1591–1601 (2024).206. A. A. Dmytriw, A. Hadjinicolaou, G. Ntolkeras, E. Tamilia, M. Pesce, L. F. Berto, P. E. Grant, E. Pang, B. Ahtam, Magnetoencephalography for the pediatric population, indications, acquisition and interpretation for the clinician. Neuroradiol. J. 38, 7–20 (2025).207. J. C. Edgar, A. Guha, G. A. Miller, Magnetoencephalography for Schizophrenia. Neuroimaging Clin. N. Am. 30, 205–216 (2020).208. T. W. Picton, S. Bentin, P. Berg, E. Donchin, S. A. Hillyard, R. Johnson, G. A. Miller, W. Ritter, D. S. Ruchkin, M. D. Rugg, M. J. Taylor, Guidelines for using human event-related potentials to study cognition: recording standards and publication criteria. Psychophysiology 37, 127–152 (2000).209. A. Keil, E. M. Bernat, M. X. Cohen, M. Ding, M. Fabiani, G. Gratton, E. S. Kappenman, E. Maris, K. E. Mathewson, R. T. Ward, N. Weisz, Recommendations and publication guidelines for studies using frequency domain and time‐frequency domain analyses of neural time series. Psychophysiology 59, e14052 (2022).210. A. Keil, S. Debener, G. Gratton, M. Junghöfer, E. S. Kappenman, S. J. Luck, P. Luu, G. A. Miller, C. M. Yee, Committee report: Publication guidelines and recommendations for studies using electroencephalography and magnetoencephalography. Psychophysiology 51, 1–21 (2014).211. J. Gross, S. Baillet, G. R. Barnes, R. N. Henson, A. Hillebrand, O. Jensen, K. Jerbi, V. Litvak, B. Maess, R. Oostenveld, L. Parkkonen, J. R. Taylor, V. Van Wassenhove, M. Wibral, J.-M. Schoffelen, Good practice for conducting and reporting MEG research. NeuroImage 65, 349–363 (2013).212. M. D. Clarke, A. N. Bosseler, J. C. Mizrahi, E. R. Peterson, E. Larson, A. N. Meltzoff, P. K. Kuhl, S. Taulu, Infant brain imaging using magnetoencephalography: Challenges, solutions, and best practices. Hum. Brain Mapp. 43, 3609–3619 (2022).213. O. Alem, K. J. Hughes, I. Buard, T. P. Cheung, T. Maydew, A. Griesshammer, K. Holloway, A. Park, V. Lechuga, C. Coolidge, M. Gerginov, E. Quigg, A. Seames, E. Kronberg, P. Teale, S. Knappe, An integrated full-head OPM-MEG system based on 128 zero-field sensors. Front. Neurosci. 17, 1190310 (2023).214. D. Budker, M. Romalis, Optical magnetometry. Nat. Phys. 3, 227–234 (2007).215. J. Dupont-Roc, S. Haroche, C. Cohen-Tannoudji, Detection of very weak magnetic fields (10−9gauss) by 87Rb zero-field level crossing resonances. Phys. Lett. A 28, 638–639 (1969).216. P. D. D. Schwindt, S. Knappe, V. Shah, L. Hollberg, J. Kitching, L.-A. Liew, J. Moreland, Chip-scale atomic magnetometer. Appl. Phys. Lett. 85, 6409–6411 (2004).217. V. Shah, S. Knappe, P. D. D. Schwindt, J. Kitching, Subpicotesla atomic magnetometry with a microfabricated vapour cell. Nat. Photonics 1, 649–652 (2007).218. N. Holmes, M. Rea, R. M. Hill, E. Boto, J. Leggett, L. J. Edwards, N. Rhodes, V. Shah, J. Osborne, T. M. Fromhold, P. Glover, P. R. Montague, M. J. Brookes, R. Bowtell, Naturalistic Hyperscanning with Wearable Magnetoencephalography. Sensors 23, 5454 (2023).219. M. A. Marder, G. A. Miller, The future of psychophysiology, then and now. Biol. Psychol. 189, 108792 (2024).220. G. A. Miller, T. Elbert, B. P. Sutton, W. Heller, Innovative clinical assessment technologies: Challenges and opportunities in neuroimaging. Psychol. Assess. 19, 58–73 (2007).221. S. Goudarzi, A. Basarab, H. Rivaz, Inverse Problem of Ultrasound Beamforming With Denoising-Based Regularized Solutions. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 69, 2906–2916 (2022).222. M. Huang, C. J. Aine, S. Supek, E. Best, D. Ranken, E. R. Flynn, Multi-start downhill simplex method for spatio-temporal source localization in magnetoencephalography. Electroencephalogr. Clin. Neurophysiol. Potentials Sect. 108, 32–44 (1998).223. M.-X. Huang, R. R. Lee, G. A. Miller, R. J. Thoma, F. M. Hanlon, K. M. Paulson, K. Martin, D. L. Harrington, M. P. Weisend, J. C. Edgar, J. M. Canive, A parietal–frontal network studied by somatosensory oddball MEG responses, and its cross-modal consistency. NeuroImage 28, 99–114 (2005).224. L. Douw, D. Nieboer, C. J. Stam, P. Tewarie, A. Hillebrand, Consistency of magnetoencephalographic functional connectivity and network reconstruction using a template versus native M RI for co‐registration. Hum. Brain Mapp. 39, 104–119 (2018).225. I. E. Holliday, G. R. Barnes, A. Hillebrand, K. D. Singh, Accuracy and applications of group MEG studies using cortical source locations estimated from participants’ scalp surfaces. Hum. Brain Mapp. 20, 142–147 (2003).226. N. Rhodes, L. Rier, E. Boto, R. M. Hill, M. J. Brookes, Source reconstruction without an MRI using optically pumped magnetometer-based magnetoencephalography. Imaging Neurosci. 3, IMAG.a.8 (2025).227. C. L. Douglas, A. Tremblay, A. J. Newman, A two for one special: EEG hyperscanning using a single-person EEG recording setup. MethodsX 10, 102019 (2023).228. J. C. Fachner, C. Maidhof, D. Grocke, I. Nygaard Pedersen, G. Trondalen, G. Tucek, L. O. Bonde, “Telling me not to worry…” Hyperscanning and Neural Dynamics of Emotion Processing During Guided Imagery and Music. Front. Psychol. 10, 1561 (2019).229. U. Lindenberger, S.-C. Li, W. Gruber, V. Müller, Brains swinging in concert: cortical phase synchronization while playing guitar. BMC Neurosci. 10, 22 (2009).230. V. Müller, J. Sänger, U. Lindenberger, Intra- and Inter-Brain Synchronization during Musical Improvisation on the Guitar. PLoS ONE 8, e73852 (2013).231. S. Dikker, L. Wan, I. Davidesco, L. Kaggen, M. Oostrik, J. McClintock, J. Rowland, G. Michalareas, J. J. Van Bavel, M. Ding, D. Poeppel, Brain-to-Brain Synchrony Tracks Real-World Dynamic Group Interactions in the Classroom. Curr. Biol. 27, 1375–1380 (2017).232. X. Zhou, P. C. M. Wong, Hyperscanning to explore social interaction among autistic minds. Neurosci. Biobehav. Rev. 163, 105773 (2024).233. M. Graziano, M. Gullberg, When Speech Stops, Gesture Stops: Evidence From Developmental and Crosslinguistic Comparisons. Front. Psychol. 9, 879 (2018).234. S. Ahn, H. Cho, M. Kwon, K. Kim, H. Kwon, B. S. Kim, W. S. Chang, J. W. Chang, S. C. Jun, Interbrain phase synchronization during turn‐taking verbal interaction—a hyperscanning study using simultaneous EEG/MEG. Hum. Brain Mapp. 39, 171–188 (2018).235. N. Yoneta, H. Watanabe, A. Shimojo, K. Takano, T. Saito, K. Yagyu, H. Shiraishi, K. Yokosawa, J. Boasen, Magnetoencephalography Hyperscanning Evidence of Differing Cognitive Strategies Due to Social Role During Auditory Communication. Front. Neurosci. 16, 790057 (2022).236. M. Hirata, T. Ikeda, M. Kikuchi, T. Kimura, H. Hiraishi, Y. Yoshimura, M. Asada, Hyperscanning MEG for understanding mother–child cerebral interactions. Front. Hum. Neurosci. 8 (2014).237. S. Nordenström, V. Lebedev, S. Hartwig, M. Kruse, J. Marquetand, P. Broser, T. Middelmann, Feasibility of magnetomyography with optically pumped magnetometers in a mobile magnetic shield. Sci. Rep. 14, 18960 (2024).238. B. U. Westner, J. I. Lubell, M. Jensen, S. Hokland, S. S. Dalal, Contactless measurements of retinal activity using optically pumped magnetometers. NeuroImage 243, 118528 (2021).239. L. C. Mardell, M. E. Spedden, G. C. O’Neill, T. M. Tierney, R. C. Timms, C. Zich, G. R. Barnes, S. Bestmann, Concurrent spinal and brain imaging with optically pumped magnetometers. J. Neurosci. Methods 406, 110131 (2024).240. M. Batie, S. Bitant, J. F. Strasburger, V. Shah, O. Alem, R. T. Wakai, Detection of Fetal Arrhythmia by Using Optically Pumped Magnetometers. JACC Clin. Electrophysiol. 4, 284–287 (2018).241. S. Strand, W. Lutter, J. F. Strasburger, V. Shah, O. Baffa, R. T. Wakai, Low‐Cost Fetal Magnetocardiography: A Comparison of Superconducting Quantum Interference Device and Optically Pumped Magnetometers. J. Am. Heart Assoc. 8, e013436 (2019).242. D. Wurm, P. Ewert, P. Fierlinger, R. Wakai, V. Wallner, L. Wunderl, A. Wacker-Gußmann, A Small Scale Optically Pumped Fetal Magnetocardiography System. J. Clin. Med. 12, 3380 (2023).243. T. M. Tierney, A. Levy, D. N. Barry, S. S. Meyer, Y. Shigihara, M. Everatt, S. Mellor, J. D. Lopez, S. Bestmann, N. Holmes, G. Roberts, R. M. Hill, E. Boto, J. Leggett, V. Shah, M. J. Brookes, R. Bowtell, E. A. Maguire, G. R. Barnes, Mouth magnetoencephalography: A unique perspective on the human hippocampus. NeuroImage 225, 117443 (2021). Information & Authors Information Version history Peer review timeline Published Psychophysiology Version of Record12 Apr 2026Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection Authors Metrics & Citations Metrics Article Usage 779views 268downloads Citations Download citation J. Christopher Edgar, Samu Taulu, Tony W. Wilson, et al. Revolutionizing Pediatric Neurophysiology with Magnetoencephalography. Authorea. 25 November 2025. DOI: https://doi.org/10.22541/au.176406331.15863911/v1 DOI: https://doi.org/10.22541/au.176406331.15863911/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.