Interactive volume manipulation with selective rendering for improved visualization

In this paper, we describe a methodology that will allow a viewer to select "geometrical" components of a volume and both manipulate these components and/or "highlight" them. A considerable number of volumetric datasets are now available and can be obtained from imaging techniques such as magnetic resonance imaging (MRI), cryogenic slicing, CT, or laser scans. In order to be able to better visualize these models, it is often required to repose them, e.g., remove occlusion, or to highlight various components. However, it may be still desirable to keep the rest of the volume present (and not to just cut it away) to maintain a focus+context view of the data. The methodology we propose allows the user to select regions of the data based upon the shape of the volumetric model. Once selected, the regions can be moved or highlighted via an alternate rendering. Some examples of alternate renderings include selective compression (higher resolution data is used to render the selected region), selective juxtaposition (replacing that selected component with data from another dataset), or selective transfer function determination. Both the selection and the rendering are done interactively to maximize their effectiveness. We show examples of the approach using a variety of volumetric datasets.