Constructing 4-dimensional Infant Cortical Surface Atlases

Dr. Gang Li

Li Lab R21The first two postnatal years is an exceptionally dynamic and critical period of the structural and functional development of the human brain. The increasing availability of longitudinal infant MR images provides unprecedented opportunities for improving our limited understanding of early brain development, and gaining fundamental insights into origins and abnormal trajectories of neurodevelopmental disorders. Cortical surface-based analysis of MR images has been extensively adopted in studies of adults and older children. However, infant cortical surface atlases, encoding age-matched anatomical and other reference information in a spatial framework, are still crucially lacking for enabling precise normalization, analysis, visualization, and comparison of results across infant studies.

The goal of this project is to create 4D infant cortical surface atlases, containing a) longitudinally-consistent, age-specific population representative shape of multi-scale cortical folding, and b) parcellations reflecting the developmentally and functionally distinct regions. To ensure the longitudinal consistency of 4D atlases, we will capitalize on within-subject longitudinal constraints to establish consistent inter-subject cortical correspondences. To ensure the clarity and representativeness of cortical folding in atlases, we will develop a sparse representation method to adaptively integrate individuals’ cortical folding. Meanwhile, the dynamic cortical developmental trajectories indicate the underlying microstructural changes, and thus can better define the developmentally, microstructurally, and functionally distinct regions than the conventional sulcal-gyral landmarks. To adaptively integrate complementary information from multiple biologically-distinct cortical attributes, we will nonlinearly fuse the similarity matrices of multiple attributes’ trajectories for joint parcellation.

A Joint Segmentation and Registration Framework for Early Brain Development

Dr. Guorong Wu


The first two years of life is the most dynamic and perhaps the most critical phase of postnatal brain development. This project aims to develop an efficient computational anatomy approach to deal with the difficult tissue segmentation and registration of infant brain images in the first years of life.

Specifically, to overcome the issues of dynamic appearance changes and spatially-varied development, we propose a joint image segmentation and registration framework to simultaneously determine the tissue type in each image point and further find the deformation pathway between any two infant images at different development stages. Considering the importance of image segmentation and registration in computational anatomy area, this cutting-edge technique will be also very useful for many ongoing early brain development studies.


Normative Functional Brain Atlases During Infancy

Dr. Wei Gao, Gao Lab

Functional sub-division of the insula and the associated functional networks from neonates to 2-year olds.

There is an increasing interest in exploring the mechanisms underlying early brain functional development using the resting state fMRI (rsfMRI) technique. Such explorations are promising for the detection of early functional connectivity biomarkers that are essential for the development of early diagnosis and intervention schemes for different pediatric neurological disorders such as cerebral palsy and epilepsy. However, given the dramatic functional evolution between infancy and adulthood, there are noteworthy difficulties for early developmental studies which include both the definition of infant-appropriate regions of interest (ROIs) and the accurate interpretation of the resulting functional connectivity patterns. Therefore, the establishment of infant-specific functional brain atlases represents an urgent mission for more rapid progress in the field.

Our team has extensive experience in using rsfMRI to delineate normal brain functional connectivity development patterns during infancy and has accumulated a large sample of normal singleton infants (N=174) with longitudinal rsfMRI scans during the first two years of life. The availability of such a large-scale dataset provides us a unique opportunity to establish normative functional brain atlases during infancy. In fact, we have demonstrated the feasibility of such an endeavor and delineated the sub-regional functional segregation profile of the insula and thalamus during the first two years of life. Building on these previous experiences and leveraging the large-scale pre-existing data, the proposed study aims to establish a set of normative functional atlases for the first two years of life. Additionally, we will utilize existing behavioral data (i.e., Mullen Scores) measured at 1 and 2 years of age to evaluate the behavioral relevance and significance of the established functional atlases (Aim 1). Secondly, another equally important and unique dataset consisting of 120 dizygotic (DZ) and 88 monozygotic (MZ) twin infants scanned at the same age interval is also available and will be used to: i) independently validate the atlases established based on singleton infants; and ii) examine the associated genetic and environmental contributions to the functional atlases (Aim 2).

Upon successful completion of the proposed project, we expect that a software package containing the established infant-specific normative functional brain atlases together with their behavioral correlation, genetic association, and environmental influences, will be made freely available to the early brain development research community to foster more rapid progress in the field (Aim 3). The approach is innovative because it will be the first to apply established functional connectivity clustering techniques to infant data and create normative functional brain atlases during infancy. The proposed research is significant because it is expected to provide the research community with an unprecedented set of references to expedite future explorations of the functional mechanisms underlying both normal and abnormal early brain development.