Cognitive neuroscience using wearable magnetometer arrays: Non-invasive assessment of language function

Abstract

Recent work has demonstrated that Optically Pumped Magnetometers (OPMs) can be utilised to create a wearable Magnetoencephalography (MEG) system that is motion robust. In this study, we use this system to map eloquent cortex using a clinically validated language lateralisation paradigm (covert verb generation: 120 trials, ∼10 min total duration) in healthy adults (n = 3). We show that it is possible to lateralise and localise language function on a case by case basis using this system. Specifically, we show that at a sensor and source level we can reliably detect a lateralising beta band (15-30 Hz) desynchronization in all subjects. This is the first study of human cognition using OPMs and not only highlights this technology's utility as tool for (developmental) cognitive neuroscience but also its potential to contribute to surgical planning via mapping of eloquent cortex, especially in young children.

Keywords: Language; MEG; Optically pumped magnetometers.

Fig. 1

The OPM setup for studying language lateralisation: real (a) and schematic (b). The subject is seated inside the magnetically shielded room wearing the scanner-cast with OPM sensors inserted into slots covering the bilateral aspects of the frontal lobes. The field nulling coils are placed either side of the subject and confer motion robustness to the system by nulling the field over a 40 × 40 × 40 cm³ volume within which head movement is tolerated.

Fig. 2

Sensor level spectral responses (a) and sensor positions (b). In (a) the greatest sensor level modulation (percentage change relative to rest) of beta band power with respect to the task is presented for each subject. The shading around each line indicates the standard error of the mean as assessed by bootstrapping the average across trials (100 bootstraps). Colour coded asterisks indicate when each subject deviated significantly from rest (p < .05, corrected). The boundary between the task and rest period are indicated by the grey dashed lines. In (b) the sensor positions for performing a language lateralisation experiment on an average subject are plotted relative to the MNI template. The sensors for each subject showing the greatest modulation corresponding to the traces in panel (a) are colour coded accordingly. Note that only 26 sensors were used but more sensors appear in this figure to depict the sensor for each subject that displays the maximum change in beta power.

Fig. 3

Language localization (a & c) and lateralisation (b). Brain Images (a and c, left and right hemisphere) of power change in the beta band for the verb generation task in three subjects. Images are corrected for multiple comparisons using FDR (q < 0.05). In Panel b the LI (Lateralisation Index) displays the relative number of left (positive) hemisphere vs. right (negative) hemisphere voxels in the inferior frontal gyrus as a function of threshold. The localization of the verb generation task is strongly lateralized to the left hemisphere in all 3 subjects as indexed by all LIs tending towards 1 (left) at high thresholds. Confidence intervals (95%) obtained by bootstrapping are indicating by shading surrounding the line plots. The lateralisation index was calculated from thresholds F = 1–15 but is displayed at F = 1–4 as the lateralisation saturates at low thresholds.

Fig. 4

Source level time-course of maximal beta power change. In (a) the maximum beta band power change (%) from baseline is plotted over the course of the task for each subject. Shading indicates standard error as assessed by bootstrapping (100 bootstraps over trials). Colour coded asterisks indicate when each subject deviate significantly from rest (p < .05, corrected). In (b) the location of these maximal beta band power changes (all left hemisphere) are indicated by the spheres imbedded in the surface render of the MNI template.