Improved Spatial Resolution in Magnetoencephalography with an Optically Pumped Magnetometer Array - Project Summary/Abstract
The fixed helmet design of commercially available magnetoencephalography (MEG) systems utilizing
superconducting quantum interference device (SQUID) magnetometers is designed to fit the 95th percentile of
head size and therefore gives suboptimal measurements of the MEG signals for most subjects, especially
children. A small head size will result in a gap between the helmet and head of several centimeters, and since
the MEG signal amplitude decays as 1/r3, where r is the distance from the neuronal source, this large gap can
result in signal attenuation by a factor of ~10. Therefore, placing the sensors on to the head will lead to
increases in signal amplitude. Additionally, if sensors are placed on or near the scalp, high spatial frequency
variations in the magnetic field will be detectable. Combining these factors, substantially improved spatial
resolution in localizing neuronal sources will be enabled. Recent developments in sensor design now makes
optically pumped magnetometers (OPMs) ideal for application to the field of MEG, and since they operate
above room temperature and can be constructed as individual sensor modules, the sensor layout can be
flexible. The long-term goal of this research is to develop a full-head MEG system based on OPMs that can
conform to any head size to give the largest possible signal while at a reduced cost compared to cryogenic
MEG. The objective of this proposal is to develop a 72-channel OPM MEG system giving partial head
coverage to demonstrate improved spatial resolution in the measurement of nearby neuronal sources within
the human brain. The system will be rapidly reconfigurable to concentrate the array coverage on an area of
interest. Our central hypothesis is that the close proximity of the OPM array will allow a new level of spatial
resolution for MEG. In Specific Aim 1, our current OPM-based MEG array with 20 channels will be expanded to
a reconfigurable 72-channel system. The reconfigurable array will accommodate varying head sizes,
particularly that of small adults and children, and the number of sensors will allow the array to be concentrated
over two sections of the brain simultaneously. In Specific Aim 2, analysis techniques specific to the
reconfigurable array will be developed. When the array is repositioned for each new subject, real-time array
calibration is required for accurate magnetic source localization and external noise suppression. In addition,
data simulation will optimize the positioning of the array and reveal the possible improvements in source
localization due to access to signals of higher spatial complexity. In Specific Aim 3, the source localization
precision between our OPM MEG array and a commercial SQUID-based MEG array will be compared. Tasks
involving auditory and visual stimulation will allow us to study spatial variation of brain activity due to changing
stimulus parameters. With the expected improvements in signal size and spatial resolution, higher fidelity MEG
measurements for people of all head sizes ranging from premature infants to the largest adults are enabled,
with broad ranging applications in neuroscience and in understanding and treating brain dysfunction.
