Tools for resting state functional connectivity magnetic resonance imaging in the developing brain

Thesis (M.A.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you.Developmental dyslexia (DD) is a specific language-based learning disorder estimated to affect 5 - 17% of all children. Currently, DD cannot be diagnosed until the formal onset of reading instruction, limiting the availability of early intervention techniques. Neuroimaging studies suggest that functional connections between brain regions associated with language function may be disrupted in children and adults affected by DD. Since DD seems to have a strong familial basis and genetic studies have suggested a role for aberrant developmental mechanisms, these differences may be present at even younger ages. Functional connectivity magnetic resonance imaging (fcMRI) provides a safe and non-invasive technique that can reveal intrinsic fluctuations in cerebral activity, allowing for the characterization of functional brain network organization. Investigation of functional connectivity in infants necessitates methods that minimize error due to motion artifact, and provide accurate segmentation and surface-based representation of cerebral tissues for the accurate and robust assessment of intrinsic fluctuations in the blood oxygen level-dependent (BOLD) signal. In this work, functional and structural magnetic resonance (MR) images were acquired from 16 healthy infants ranging in age from 8 to 11.5 months old, including 4 infants with, and 12 infants without, a family history of DD. Tissue segmentation and 3D surface generation tools were adapted and developed from a pipeline designed for adult MR images, and optimized for our infant cohort using age-appropriate atlases. Using these methods, we generated accurate automatic segmentations of white and grey matter tissues, and 3D renderings of the white and pial surface interfaces, from brain structural MR images of 13 subjects. The methods presented here specifically addressed some of the characteristic difficulties of working with infant MR images, including low data quality, low tissue contrast, and motion artifact, providing adequate input for accurate and robust surface-based fcMRI analysis. We conclude that the use of age-specific training atlases and optimized tools allow for accurate generation of the 3D brain surface representations necessary for carrying out surface-based resting state fcMRI analysis in infants, and has the potential to be useful in other structural and volumetric MRI studies with infant cohorts