UC San Diego

In recent times, the study of functional connectivity (FC) between spatially distinct locations yet functionally related locations in the resting brain using functional magnetic resonance imaging (fMRI) has been on the rise. However, the interpretation of such functional measures is complicated by the complex hemodynamic nature of the blood oxygenation level dependent (BOLD) signal, which can be influenced by both neural as well as vascular factors. In this work, we employed a direct measure of neuromagnetic activity, magnetoencephalography (MEG), to further validate the neural origin of caffeine-induced reductions in BOLD connectivity observed previously in our lab. Concerns regarding the performance of existing source reconstruction methods for MEG analysis motivated the development of an improved source reconstruction technique, a multi-core beamformer (MCBF), which was comprehensively tested with both simulations and neuromagnetic data. An iterative algorithm to be used in conjunction with the MCBF, allowing a solution to be obtained without any a priori knowledge about the underlying source configuration, was described and analyzed in detail. With the help of the new beamformer, the caffeine data were reexamined and the original findings were upheld. Preliminary investigation into resting-state networks by means of temporal independent component analysis (ICA) was also conducted