PROJECT SUMMARY
This project is aimed at developing a novel magnetometer for room-temperature magnetoencephalography
(MEG), a functional neuroimaging technique that allows direct imaging of human brain electrophysiology by
measurement of weak magnetic fields generated by active neurons. Compared to functional magnetic resonance
imaging, MEG is more effective in localizing and tracking brain activities thanks to its high temporal resolution.
State-of-the-art MEG employs either superconducting quantum interference device (SQUID) or microfabricated
optically pumped magnetometers (µOPMs) as their sensing elements. SQUID provides the highest sensitivity
but requires cryogenic cooling, which severely limits its portability. µOPM offers an excellent alternative with
much reduced form factor. However, it still requires thermal insulation (heating as opposed to cryogenic cooling)
and it has a rather limited bandwidth and dynamic range compared to SQUID.
The proposed new type of magnetometer is designed to greatly improve the signal strength and bandwidth,
reduce the complexity of active shielding and further decrease the minimum channel spacing. The sensing
element consists of uniformly dispersed magnetite nanoparticles that operate at room temperature. Thermal
insulation is no longer needed, and thus the sensing element can be placed as close as 1 mm to the human
scalp, increasing the signal strength. The system bandwidth is not fundamentally limited but set to be 1 kHz by
choice so that high-quality electronics can be utilized while maintaining the capability to detect all neural activities
from delta to high gamma frequency bands. The proposed system employs chip-scale Kerr frequency comb as
the light source and balanced in-line Sagnac interferometer as the optical readout. It can thus achieve a
magnetometer sensitivity of 20 fT/Hz1/2 and a gradiometer sensitivity of 5 fT/cm·Hz1/2 under a strong ambient
field of 100 µT, reducing the complexity in field-shielding and making possible a dense array of sensor heads.
The proposed research has two key innovations. First, magnetite nanoparticles will be synthesized, stabilized
in polymer matrices and fabricated into micro-optical devices. We will investigate the use of different dopant
species and surface passivation to simultaneously achieve high Verdet constant, low insertion loss, and good
long-term stability for MEG applications. Second, we will incorporate a novel chip-scale frequency comb source
to simultaneously operate an array of mm-size magnetometer sensor heads. We will take advantage of its two-
mode squeezing property for noise reduction to below the quantum limit and further enhance the detection
sensitivity of our multichannel magnetic gradiometer. At the end of the program, we will benchmark and validate
our technology by a preliminary in vivo study of two normal human subjects under auditory stimulation. The
proposed magnetometer and gradiometer would significantly improve the accuracy and portability of MEG
system, making it much more widely applicable to frontline diagnostics.