Abstract
For transparent visualization of medical volume data acquired by CT or MRI, realization of correct sense of depth is important. Therefore, the conventional visualization methods execute depth sorting of rendering primitives along the line of sight. However, the indefiniteness of depth-sorted orders of rendering primitives, which appears, for example, for overlapping boundaries of human organs and very fine 3D inner structures of a human body, may cause the rendering artifacts. Moreover, the depth sorting requires long computation time when treating large data. To avoid these difficulties, we proposed particle-based rendering that is a kind of stochastic point-based rendering. This rendering does not require the depth sorting and thus free from the above-mentioned difficulties. Moreover, it is easy to execute fused visualization of a medical volume, slice and surfaces only by expressing each of them with small particles. Such fused visualization helps us conceive 3D inner structures of a human body.
