作者: Andrew K. Knutsen , Yulin V. Chang , Cindy M. Grimm , Ly Phan , Larry A. Taber
DOI: 10.1115/1.4002430
关键词:
摘要: Folding of the cerebral cortex is a critical phase brain development in higher mammals but biomechanics folding remain incompletely understood. During folding, growth cortical surface heterogeneous and anisotropic. We developed applied new technique to measure spatial directional variations from longitudinal magnetic resonance imaging (MRI) studies single animal or human subject. MRI provides high resolution 3D image volumes at different stages development. Surface representations are obtained by segmentation these volumes. Estimation local between two times requires establishment point-to-point correspondence ("registration") surfaces measured those times. Here we present novel approach for registration which an energy function minimized solving partial differential equation on spherical surface. The includes strain-energy term due distortion "error energy" mismatch features. This algorithm, implemented with finite element method, brings features into approximate alignment while minimizing deformation regions without explicit matching criteria. method was validated application three simulated test cases characterize ferret during folding. Cortical were created data acquired vivo 14 days, 21 28 days life. Deformation gradient Lagrangian strain tensors describe kinematics over this interval. These quantitative results illuminate spatial, temporal, patterns