Research Project
8471703
DESCRIPTION (provided by applicant): Functional neuroimaging has led to important advances in understanding neural circuits and has emerged as an important technique in the study of psychiatric and neurological disorders such as schizophrenia, dementia, depression, and epilepsy, where anatomical imaging is negative or shows only nonspecific findings. Magnetoencephalography (MEG) is the only noninvasive functional neuroimaging technique able to directly measure neural activity with sub-centimeter spatial and millisecond temporal resolution, but its potential as a research and clinical tool has yet to be realized on a large scae due to its high acquisition and operating costs. A large portion of the cost results from the use o superconducting quantum interference device (SQUID) magnetic sensors that must be cooled with liquid helium. The cryogenic infrastructure results in a bulky MEG system that requires installation of a large magnetically shielded room to achieve acceptably low background levels and that has a fixed sensor array geometry that cannot be adjusted for head size. The goal of the proposed work is to develop a small, low-cost MEG system using an array of atomic magnetometers (AMs) as replacements for SQUIDs. AMs can achieve sensitivities comparable to SQUIDs but do not require cryogenic cooling. AMs detect magnetic fields by measuring, via laser interrogation, the interaction between a magnetic field and atoms contained within a glass cell. Recently, we developed a compact optical fiber-coupled AM and used it to detect MEG signals from human subjects. Based on these preliminary studies, we propose to develop a 36-channel array of AMs with partial-head coverage (roughly 12 cm X 12 cm) that is able detect and localize neuronal activity. In Specific Aim #1, we will design critical components of the proposed AM MEG system, including the individual AMs that serve as array elements and a person-sized magnetic shield to contain the AM-array and the human subject. Commercially available magnetic source localization software will be adapted to our array geometry. One AM will be constructed and its performance will be verified at go/no-go specifications of >100 Hz bandwidth and 20 fT/Hz1/2sensitivity. In Specific Aim #2, the AM MEG system will be constructed and its source localization accuracy will be determined by detecting an MEG phantom with the AM array geometry reconfigured to mimic a variety of head sizes. Human studies will commence only after demonstrating sub-centimeter spatial resolution. In Specific Aim #3, the AM system will be compared to a commercial SQUID MEG system by measuring the evoked response in human subjects from median nerve and auditory stimuli with both systems. A successful comparison will identify a neural source in terms of strength and location to within one standard deviation error between the two systems. The results of the proposed work are expected to clear the path for developing a full-head- coverage low-cost AM MEG system that can be adapted to accommodate a wide variety of head sizes.
