This project aims to construct a set of standard anatomical templates for pediatric neuroimage analysis. It is designed to leverage the NIH funded MRI Study of Normal Brain Development, specifically the DTI (Diffusion Tensor Imaging) ancillary study. The overall goal is to form a set of age-group specific templates, or atlas spaces, by a process of group wise registration that will bring all of the subjects in each group into spatial agreement and create a statistical summary of the subject images. Many neuroscience projects are facilitated by the availability of standard templates, e.g., quantification of morphological differences between groups and automated labeling of structure. Most previous work in template, or atlas space, formation has used conventional MRI images, which have good contrast for the overall white matter / gray matter morphology of the brain. The more recent DTI images provide details of the white matter structure that are otherwise not visible in conventional MRI. We specifically plan to make use of the DTI images of the ancillary study in such a way that the white matter structures are brought into better agreement in the templates than would be the case if DTI images were not used.
The specific aims i nclude technology development, template formation, and dissemination of the resulting templates and software. The technology development effort includes the following: """""""" Extend a group registration method that we have previously developed to use DTI images in addition to MRI images """""""" Extend the deformation modeling capabilities of the group registration method. It is our expectation that the templates and companion software tools will facilitate pediatric neuroimage analysis in general, and particularly application projects that focus on white matter structure.
Relevance to Public Health This project aims to construct anatomical templates from medical images that summarize neuroanatomy and its variability in children. Templates of this sort facilitate research that can clarify the anatomical correlates of disorders and diseases. In the long term, research of this sort can lead to improvements in diagnosis and treatment.
| Toews, Matthew; Wells 3rd, William M (2013) Efficient and robust model-to-image alignment using 3D scale-invariant features. Med Image Anal 17:271-82 |
| Toews, Matthew; Wells 3rd, William M; Zöllei, Lilla (2012) A feature-based developmental model of the infant brain in structural MRI. Med Image Comput Comput Assist Interv 15:204-11 |
