Figure 18. Data-rich nature of volume rendering. The graph in the lower part of the figure shows how attenuation data are used to assign values to a histogram-based tissue classification consisting of deformable regions for each type of tissue included. In this case, only fat, soft tissue, vessels, and bone are assigned values, but additional classifications can be added as needed. Opacity and color assignment may vary within a given region, and the shape of the region can be manipulated to achieve different image effects. Because there is often overlap in attenuation values between different tissues, the classification regions may overlap. Thus, accurate tissue and border classification may require additional mathematical calculations that take into consideration the characteristics of neighboring data. HU = Hounsfield units.