Title | More insights into early brain development through statistical analyses of eigen-structural elements of diffusion tensor imaging using multivariate adaptive regression splines. |
Publication Type | Journal Article |
Year of Publication | 2014 |
Authors | Chen, Yasheng, Hongtu Zhu, Hongyu An, Diane Armao, Dinggang Shen, John H. Gilmore, and Weili Lin |
Journal | Brain Struct Funct |
Volume | 219 |
Issue | 2 |
Pagination | 551-69 |
Date Published | 2014 Mar |
ISSN | 1863-2661 |
Keywords | Anisotropy, Brain, Brain Mapping, Child, Preschool, Diffusion Magnetic Resonance Imaging, Diffusion Tensor Imaging, Female, Humans, Image Processing, Computer-Assisted, Infant, Infant, Newborn, Male, Nerve Fibers, Myelinated, Regression Analysis |
Abstract | The aim of this study was to characterize the maturational changes of the three eigenvalues (λ1 ≥ λ2 ≥ λ3) of diffusion tensor imaging (DTI) during early postnatal life for more insights into early brain development. In order to overcome the limitations of using presumed growth trajectories for regression analysis, we employed Multivariate Adaptive Regression Splines (MARS) to derive data-driven growth trajectories for the three eigenvalues. We further employed Generalized Estimating Equations (GEE) to carry out statistical inferences on the growth trajectories obtained with MARS. With a total of 71 longitudinal datasets acquired from 29 healthy, full-term pediatric subjects, we found that the growth velocities of the three eigenvalues were highly correlated, but significantly different from each other. This paradox suggested the existence of mechanisms coordinating the maturations of the three eigenvalues even though different physiological origins may be responsible for their temporal evolutions. Furthermore, our results revealed the limitations of using the average of λ2 and λ3 as the radial diffusivity in interpreting DTI findings during early brain development because these two eigenvalues had significantly different growth velocities even in central white matter. In addition, based upon the three eigenvalues, we have documented the growth trajectory differences between central and peripheral white matter, between anterior and posterior limbs of internal capsule, and between inferior and superior longitudinal fasciculus. Taken together, we have demonstrated that more insights into early brain maturation can be gained through analyzing eigen-structural elements of DTI. |
DOI | 10.1007/s00429-013-0517-7 |
Alternate Journal | Brain Struct Funct |
Original Publication | More insights into early brain development through statistical analyses of eigen-structural elements of diffusion tensor imaging using multivariate adaptive regression splines. |
PubMed ID | 23455648 |
PubMed Central ID | PMC3795940 |
Grant List | R01NS055754 / NS / NINDS NIH HHS / United States P01CA142538-01 / CA / NCI NIH HHS / United States R01 MH086633 / MH / NIMH NIH HHS / United States R01 AG042599 / AG / NIA NIH HHS / United States R01 MH070890 / MH / NIMH NIH HHS / United States R21 AG033387 / AG / NIA NIH HHS / United States AG033387 / AG / NIA NIH HHS / United States R01EB008374 / EB / NIBIB NIH HHS / United States R01 EB008374 / EB / NIBIB NIH HHS / United States R01 EB006733 / EB / NIBIB NIH HHS / United States 1R01EB006733 / EB / NIBIB NIH HHS / United States R01 NS055754 / NS / NINDS NIH HHS / United States RR025747-01 / RR / NCRR NIH HHS / United States R01MH070890 / MH / NIMH NIH HHS / United States R01 EB009634 / EB / NIBIB NIH HHS / United States R01 HD053000 / HD / NICHD NIH HHS / United States R01HD053000 / HD / NICHD NIH HHS / United States MH086633 / MH / NIMH NIH HHS / United States UL1 RR025747 / RR / NCRR NIH HHS / United States 1R01EB009634 / EB / NIBIB NIH HHS / United States P01 CA142538 / CA / NCI NIH HHS / United States R01 MH100217 / MH / NIMH NIH HHS / United States |
More insights into early brain development through statistical analyses of eigen-structural elements of diffusion tensor imaging using multivariate adaptive regression splines.
Project: