Abstract
The statistical modelling of 3-dimensional brain images requires to take into account both the high dimensionality of the problem as well as the spatial dependence between brain regions (in terms of anatomical structure or functionality). This talk will start with an overview of neuroimaging data analysis and how functional data analysis (which is aimed at analysing data that come in the form of functions) can represent a valid and computationally appealing approach in this field. Then, the focus will be on a specific brain imaging modality called tensor-based morphometry (TBM). TBM images exhibit (especially in diseased groups) higher values in some brain regions called lateral ventricles. More specifically, a voxelwise analysis shows a mean-variance relationship in these areas and evidence of spatially dependent skewness. We propose a model for 3-dimensional functional data where mean, variance, and shape functions vary smoothly across brain locations. The effects of age and sex are estimated on a reference population of cognitively normal subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and mapped across the whole brain. The model returns also subject-specific normative maps and indices of deviation from a healthy condition which could help to assess the individual risk of pathological degeneration or to cluster different disease groups.
Organizzatore
Angela Montanari