To develop maps at multiple scales of neuronal circuits in the human brain and study the brain dynamics, there is a need for non-invasive functional brain imaging with high spatiotemporal resolution operating in natural environments. Among non-invasive functional brain imaging methods, magnetoencephalography (MEG) is the only technology that can map cortical activity down to millimeter spatial resolution with millisecond time resolution. Current cryogenic MEG systems employ superconducting quantum interference device (SQUID) magnetometers. The cryogenic operation requires sensor arrays that are rigid and fixed in a helmet, and the helmet size is optimized to fit the largest adult heads. The rigid helmet limits source-to-sensor distances to >3 cm which compromises the maximum achievable signal-to-noise ratio (SNR) and hence spatial resolution. Furthermore, due to their design, these SQUID-based MEG systems are costly and impractical for experiments in natural environments. Recent simulation studies have shown that on-scalp MEG can maximize SNR and achieve spatial resolution approaching 1 mm. Optically pumped magnetometers (OPMs) are a valid candidate for MEG sensors, as they operate above room temperature, and the sensor layout can be conformal to the scalp. The overall objective of this project is to develop a wearable, conformable, full-head coverage, 108- channel, OPM-based MEG system with unprecedented spatial resolution approaching 1 mm.
The first Aim will develop the whole-cortex 108-channel OPM array along with the supporting systems (optical, electronic, software, etc.). The OPM MEG will be installed in a magnetically shielded room so that the array can be worn and move with the subject, enabling more naturalistic study paradigms.
The second Aim will leverage the high- frequency spatial features available to the on-scalp OPMs to enhance the spatial resolution of the MEG system. Given unique subjects' head shapes, adaptive sampling of the magnetic topography (image) is essential to maximize the captured spatial frequency. Hence, information-theoretic analysis will be used to maximize the spatial resolution by optimizing the sensors locations. With the array being reconfigurable, rapid calibration techniques will be developed to determine the position of each OPM for each subject. To eliminate external magnetic noise and compensate for movement-induced distortion, physics-based models will be employed.
The final Aim cross validates the performance metrics of the new OPM MEG system with a commercial SQUID system. By measuring retinotopy in the visual cortex, spatial localization between the MEG systems will be compared. By stimulating cerebellar activity, it will be studied if the conformal OPM array can better capture activity in this difficult-to-study region of the brain. Finally, by measuring resting-state MEG, intrinsic network connectivity in the human brain will be captured. This project will provide a whole-head OPM array that improves MEG measurements for people of all head sizes (from premature infants to the largest adults) and enable new experimental paradigms with a wearable array operating in semi-natural settings.
To develop maps at multiple scales of neuronal circuits in the human brain, for large-scale monitoring of neural activity, and for clinical applications such as presurgical mapping of epileptic foci, there is a need for non- invasive functional brain imaging with high spatiotemporal resolution operating in natural environments. Magnetoencephalography, using optically pumped magnetometers, augmented with advanced signal processing methods, such as adaptive magnetic topography sampling and physics-based artifact rejection models, shows great promise in delivering a functional brain imaging system with high spatial (~ 1 mm) and millisecond temporal resolution. Magnetoencephalography with optically pumped magnetometers allows a wearable, full-head coverage array for neuroscience experiments in more natural environments.
