Abstract:
|
Neuroimaging is increasingly popular in statistics, though little research has been dedicated to the brain affected by various pathologies. Cancer, stroke, multiple sclerosis, Alzheimer's disease, and traumatic brain injury produce major changes in the brain. These changes make it difficult to use standard neuroimaging tools for analysis and require new methods and a deeper understanding of the specific pathology. We will discuss the various problems that remain open, including bias field correction, lesion segmentation, lesion tracking, intensity normalization, and spatial registration. We also will discuss new tools in R that combine the strength of existent neuroimaging tools with a familiar computational environment.
|