A 3D navigational environment for specifying positron emission tomography volumes-of-interest

Abstract

Volume of interest extraction for radionuclide and anatomical measurements requires correct identification of the anatomical feature being studied. We have developed a toolset for specifying 3D volumes-of-interest (VOIs) on a multislice Positron Emission Tomography (PET) dataset. The software is particularly suited for specifying cerebral cortex VOIs that represent a particular gyrus or mid-brain structure. A registered 3D magnetic resonance image (MRI) dataset is used to provide high-resolution anatomical information, both as oblique 2D sections and as volume renderings of a segmented cortical surface. Because most clinicians can readily identify specific sulci from high-quality renderings of the cortical surface, a crucial step in quickly identifying sulci in 2D sectional data is providing a feedback mechanism between the renderings and the section data. Our toolkit provides this mechanism by calculating a full depth map and transformation matrix for volume renderings of the cortex. A region drawing environment is then possible where the position of a main drawing cursor on a 2D section can be simultaneously mirrored on the rendered views. The feedback is further enhanced by displaying the projection of the main cursor position on any number of auxiliary 2D sections oriented at orthogonal slicing angles with respect to the principal 2D section. Complete VOIs are specified by drawing a stack of 2D contours subsequently filed together to form closed triangular mesh surface models.