A large volume of image data, to be shared by multiple investigators, will be generated in this program project. These image data, which include MRI and PET of multiple probes and some of which are repeated at different time points, need to be orderly organized, reliably archived, and conveniently accessible by investigators. Furthermore, image processing and analysis procedures need to be performed reliably and consistently among different projects. The goal of this core is thus to provide image handling and analysis resources to support the investigators of the three projects of this Program Project Grant to facilitate their scientific investigations. The Image Analysis Core has three specific aims.
Aim 1. To organize image data orderly to allow convenient investigator access and to control image quality and integrity of images.
Aim 2. To streamline and to unify image processing, and to extract biologically relevant information from images.
Aim 3. To integrate the extracted information and to perform statistical analysis and evaluation. An Internet-based image database system will be set up and maintained to address Aim 1.
Aims 2 and 3 encompass many image and data analysis operations. Images of multiple modalities for each subject will be co-registered, and parametric images (with direct biological/functional meaning) will be generated from PET studies. Afterward, two rriajor analyses will be performed. One is to align images of different subjects to a common reference set, from which both ROI analysis and pixel-based analysis will be performed over the entire 3D volume of the brain. The second major analysis is to map the brain hemispheric cortical surface of different subjects to a reference set. From the derived mapping, cortical measures that include gray matter thickness and PET-derived functional activities can be determined for the brain cortical surface. Statistical analyses will be performed in close interaction with investigators of the three research projects. For image data of animal studies, a separate image database, similar to the one for images of human subjects, will be set up. The procedures of image co-registration and the generation of biological/functional images will parallel those for human studies.
Showing the most recent 10 out of 52 publications
