Research On Fast Volme Rendering For Medical Image Data

Fast rendering algorithms are often applied to the field of medical visualization. Maximum Intensity Projection which achieves real-time performance is used to explore internal structure of medical image data. However, the method produces artifact of depth on two-dimensional image. In order to use MIP-like methods to generate final image which includes more internal information of medical data, we need to improve accumulation manner in ray casting process.Our method summarizes Maximum Intensity Projection, Instant Volume Visualization using Maximum Intensity Difference Accumulation, Depth-Enhanced Maximum Intensity Projection, Shape-Enhanced Maximum Intensity Projection etc, which modify accumulation of ray to enhance the structure of the identifiable. As internal structures are blocked, we can’t find the exact location of the structure. Therefore, we do some research and exploration in dissertation.We present a depth-based difference accumulation algorithm for volume rendering. The depth of each sample is used to linearly adjust sample’s value. Then, accumulated opacity and color are reduced with modified intensity difference and sample depth. Surface shading is used to enhance the details of the structure’s surface. In addition, we have implemented a Qt-based graphical interface to load volume data, adjust transfer function, render image interactively.Comparing the results of using different rendering algorithms and different rendering parameters with one-dimensional linear transfer function, our algorithm is capable of displaying richer internal structure in complex medical data, and can meet the different demands of the medical data rendering by parameter adjustment. In the future, we will be committed to study feature enhancement and parameter optimization on the basis of our method.