3-D brain image registration using optimal morphological processing and iterative principal axis transform

The 3-D registration of magnetic resonance (MR) and positron emission tomographic (PET) images of the brain is of significant interest for various reasons. A new MR and PET brain registration procedure is presented here. The procedure is based on the iterative principal axis transform (IPAR) and the optimal shape description method based on the morphological signature transform. The morphological processing is used to improve the accuracy of the basic IPAR method. A genetic algorithm is used to select a near-optimal structuring element for MST-based registration. The results demonstrate that the method provides an accurate registration.