The Basic Research of mine consists of two directions: 1) infant brain functional development, and 2) methodological development on brain dynamics and “chronnectome”.
Basic Research 1 Understanding the infant brain, especially the neonate’s brain, could help us know better about how our brain develops in such a pivotal early life stage where our cognitive abilities are emerging, rapidly growing and dramatically changing. It also helps to early detect and intervention to neurodevelopmental diseases, such as autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD). My role is to, based on the big longitudinal data collected in the Baby Connectome Project (BCP), developing methods and establishing pipelines for resting-state fMRI studies on characterizing normative developmental trajectories of early-life brain functional connectome and the predictive value for early cognitive development. This work utilizes approaches developed by an NIH grant (1U01MH110274) and the efforts of the UNC/UMN Baby Connectome Project Consortium. The data and pipeline will be shared with the public, which will significantly boost the research and fill the long-existing gaps in this field.
Basic Research 2 Characterizing brain dynamics is an interesting topic. Along this track, developing and aging is one type of long-term dynamics and brain plasticity is another type of dynamics at an intermedium scale; but what I am focusing on is the short-term dynamics happening within seconds/minutes, i.e., time-varying functional connectivity, or dynamic FC (dFC). Such dFC is believed to be important to our daily life and carry supplementary information to traditional “static” functional connectivity. For example, dFC may reflect switching attention, real-time balancing resources in the brain, the flexibility and adaptation to prepare for and respond to changing environment and, simply, a small set of “brain statuses”. Actually, brain functional dynamics could be spatially and temporally complex, and more advanced methods need to be developed to characterize such comprehensive “chronnectome”. For example, using dFC to early detect a brain disease, please refer to our Human Brain Mapping paper here. This direction is supported by the NIH Brain Initiative Project EB022880. For more advanced chronnectome measurements, please see our recent publications.
Clinical Research of mine consists of neuroimaging-based 1) brain tumor imaging analysis for diagnosis and prognosis, and 2) early detection of Alzheimer’s disease. Specifically, we are developing a series of machine learning and deep learning-based tumor grading, segmentation, and outcome prediction methods based on either brain images or microscopic images. Our goal is to ease the laborious works of radiologists and benefit the patients towards precision medicine and individualized treatment. See our MICCAI paper on using multimodal MRI to predict overall survival time for patients with high-grade gliomas. For early Alzheimer’s disease detection, we developed a series of “high-order functional connectivity (HOFC)” metrics and brain network construction methods for better diagnosis. The HOFC, calculated by “correlations’ correlation”, measures higher-level, more complex functional organization and provides substantial supplementary information to the traditional BOLD signal-based “low-order” functional connectivity. For details of several HOFC metrics, please see our review paper and software paper. We also developed an easy-to-use Matlab toolbox for neuroscientists and clinicians for them to conduct rigorous disease diagnosis based on brain networks constructed by advanced methods.