Improved on Maximum Intensity Projection

Maximum intensity projection (MIP) is one of the most common methods for the visualization of volumetric data. MIP works by traversing all viewing rays and finding the maximum data value along each of them. The main limitation of MIP is that it cannot adequately depict the spatial relationships of overlapping tissues. An approach has been proposed to eliminate this drawback: local maximum intensity projection (LMIP). However, with too low a threshold value, the first encountered local maxima are mostly noise components; with a threshold value larger than the maximum intensity among all the voxels in the 3D data, LMIP is equivalent to MIP. So the results rely on the threshold. If the threshold is low, we will not get a good result. There is no shading information in MIP and LMIP. In this paper we propose an improved local maximum intensity projection. An appropriate threshold and shading are computed in this improved method. We show that the improved method is a useful technology in volumetric dataset visualization.