Multimodality three-dimensional brain image registration and analysis

Registration of anatomical images such as magnetic resonance (MR) and functional images such as positron emission tomography (PET) is an essential step in multi-modality brain image analysis. I have developed a technique, called iterative principal axes registration (IPAR) for three-dimensional registration of MR and PET brain images. Using this method, partial volumes of PET images can be accurately registered with complete MR scans. I will present qualitative and quantitative results showing that IPAR is accurate and practical as a registration method for MR/PET correlation studies. When using this method processing of MR and PET data is largely automatic with very little operator interaction (and very little operator training) required. The method is fast, requiring only about 20 seconds per iteration on the SUN SPARC-2 (20 iterations are usually sufficient to obtain an accurate match). Accurate three-dimensional registration of anatomical and functional brain images can be obtained using the IPAR algorithm without fiducial or external markers.;Automated or semi-automated analysis and labeling of structural brain images, such as magnetic resonance (MR) and computed tomography (CT), is desirable for a number of reasons. Quantification of brain volumes can aid in the study of various diseases and the affect of various drug regimes. A labeled structural image, when registered with a functional image such as positron emission tomography (PET) or single photon emission computed tomography (SPECT), allows the quantification of activity in various brain subvolumes such as the major lobes. Because even low resolution scans (7.5 to 8.0 mm slices) have 15 to 17 slices in order to image the entire head of the subject, hand segmentation of these slices is a laborious process. However, because of the spatial complexity of many of the brain structures, notably the ventricles, automatic segmentation is not a simple undertaking. In order to accurately segment a structure such as the ventricles we must have a model of equal complexity to guide the segmentation. Also, we must have a model which can incorporate the variability among different subjects. The construction of these models and their use in segmentation and labeling of brain images are described in this dissertation.