Understanding brain dynamics involved in many perceptual and cognitive tasks is of particular interest to scientists in the fields of speech & hearing, as well as neuroscience at large. Furthermore, this knowledge can be combined in an engineering approach to improve our brain-computer interface (BCI) designs, as well as providing us with crucial information on computation modeling, e.g., in speech production. In this talk, I will introduce an emerging neuroimaging technique—known as magnetoencephalography (MEG)—that has just arrived at the University of Washington. I will present our ongoing work on mapping the cortical network involved in auditory attention using a multimodal imaging approach (combining MEG with electroencephalography and magnetic resonance imaging). I will also discuss why the interface between neuroscience and engineering will be the crucial juncture on which new discoveries depend.