A FINITE ELEMENT METHOD FOR ELASTIC PARAMETERIZATION AND ALIGNMENT OF CORTICAL SURFACES USING SULCAL CONSTRAINTS

We present a cortical surface registration method that simultaneously aligns sulcal landmarks and parameterizes two cortical surfaces. The approach is based on coregistration of cortical surface coordinate systems so that the labeled sulcal features share the same coordinates on both cortical surfaces. We model the cortex as an elastic sheet and solve the associated Cauchy-Navier equilibrium equation subject to sulcal alignment constraints. The elastic energy is computed directly with respect to the intrinsic surface geometry and discretized using a finite element method on triangular tessellations of the two surfaces. In contrast to alternative methods for cortical alignment, the method avoids the need for an intermediate flat space for sulcal landmark matching and provides a fast, accurate and inverse-consistent surface registration and parameterization for inter-subject neuroanatomical studies