Image Understanding

• Aim 1: Quality Assurance and Validation
• Aim 2: Robust Image Segmentation and Registration
• Aim 3: Diffusion Data

• Develop methods for performing routine evaluation of the image processing algorithms in this project through a combination of quality assurance testing and validation
• Address the factors that prevent image segmentation and registration from achieving high-reliability in their routine use by non-experts
• Develop methods for coregistration of tensor and higher dimensional diffusion data into a common space

Connectomics

• Aim 1: Improved Voxel-Based Assessment of Fiber Integrity Using HARDI (TDF-FA)
• Aim 2: Tract-Based Statistical Analysis by Automated Clustering of Fibers
• Aim 3: Whole-Brain Connectivity Matrices
• Aim 4: Genetics of Brain Connectivity

• Advance the study of brain connectivity using diffusion imaging and its powerful extensions beyond the tensor model of diffusion (QBI, multi-shell HYDI, DSI, HARDI-TDF, Q-ball imaging, staggered HYDI, and DSI)
• Compare fiber and bundle integrity, properties and statistics across large populations
• Create NxN connectivity matrices summarizing the presence and properties of connections between all pairs of brain regions
• Develop, test and disseminate powerful new quantitative genetic approaches for discovering genetic effects on brain integrity and connectivity

Data Interpretation

• Aim 1: Develop computational tools to assist with data reduction and selection of appropriate statistical models in the multivariate analysis of imaging data
• Aim 2: Adapt multivariate statistical methods and model selection criteria so that they are suitable for use with multivariate observations confined to non-Euclidean manifolds
• Aim 3: Efficient statistical model to structural and diffusion imaging data

• Allow users to tailor analysis methods to the specific biological questions of interest
• Involve data reduction and incorporation of non-imaging derived biological measures
• Facilitate the examination of the data for errors and violations of underlying statistical assumptions
• Exploratory and Interactive data and result interpretation
• Provide a user-friendly interface for specifying, validating and applying an appropriate statistical model to structural and diffusion imaging data
• Enable multivariate statistical analyses of data in non-Euclidean manifolds