Clinical use of magnetoencephalographic (MEG) data will depend upon a variety of factors, one of which is the rapid display of electrical sources that produce signals within the brain. Currently, single dipole approximations of the source location are overlaid on two-dimensional magnetic resonance images (MRI) for display of the underlying anatomy. This procedure is slow and is a crude approximation of most of the underlying neural generators. The applicants propose to improve upon the visual display of combined MEG/MRI data by collaborating with university researchers who will supply high- resolution MRI data and inverse solutions of MEG data based on finite element analysis of current density for realistic head-shapes of both normal subjects and patients sufferingfrom epilepsy, Alzherimer's disease and closed head injury. MRI images will be combined to reconstruct a three dimensional (3-D) volumetric renderingof the head.A surface extraction algorithm will be written to provide coordinates for the interior surface of the cranium, which will act as the boundary layer for the inverse solution of the MEG data. Coordinates for neural activity within the brain, obtained from the inverse solution, will then be overlaid upon the 3-D rendering of the head for visual display. A variety of menu-driven features will be provided for the rapid, interactive display of this combined MEG/MRI data, including head rotation, tissue strip-away capability, and animation of the source activity over user-selected periods of time. This graphics capability for the wedding of functional and anatomical imaging of the brain will ultimately be applicable to many imaging modalities and transportable to a variety of general-purpose workstations.
