Demonstrating equivalence across magnetoencephalography scanner platforms using neural fingerprinting

Abstract

Optically pumped magnetometers (OPMs) have emerged as a viable alternative to superconducting quantum interference devices (SQUIDs) to measure the magnetic fields generated by brain activity. Magnetoencephalography (MEG) systems based on OPMs offer potential advantages over conventional systems. However, there remains a pressing need for techniques and studies that demonstrate equivalence between emergent OPM-based instruments and the established conventional MEG standard. Here, we seek to demonstrate similarities between a 192-channel (64-triaxial sensor) OPM-based system and a 275-channel SQUID-based system, when measuring data during a sensory task designed elicit induced oscillatory (beta-band) and evoked responses as well as subtle attentional effects. We reasoned that in addition to measuring these primary responses (which are stable across individuals), neural fingerprinting (i.e., demonstrating that an individual’s OPM-MEG data can be identified from a group, by matching to their SQUID data) would determine whether subtle but repeatable differences between individuals are also measurable across OPM and SQUID-based platforms. Our results showed (1) that primary responses are measurable using both systems with correlations of 0.96 (beta responses) and 0.97 (evoked responses) at the group level; (2) both systems captured significant attentional modulation of the beta rhythm; and (3) fingerprinting was successful in 15 of 15 people using OPM-MEG, 14 of 15 people using SQUID-MEG, and 15 of 15 people when examining cross-platform data. Overall, our results confirm that the two MEG platforms capture similar information on brain function. This is an important step towards proving equivalence between OPMs and SQUIDs, which in turn is required if OPMs are to replace SQUIDs for clinical evaluations.