This Ph.D.-project is about digital image processing of three-dimensional functional and anatomical image volumes of the human brain. Several approaches are investigating for improvement of the functional analysis. A problem in activation studies involving multiple subjects is the individual anatomical variation. This variation can be minimized by performing normalizing "warps" of the anatomy. Another approach investigated is noise reduction in PET activation images using Prior models for random fields. Parameter estimation in these random fields is of particular interest.
|Effective start/end date||01/02/1995 → 01/03/1998|