MEDICAL VOLUME VISUALIZATION ON GENERAL-PURPOSE PARALLEL ARCHITECTURES

Parallel volume visualization is of interest to a variety of applicati on areas since current single-processor systems fall short in interact ively rendering complex, large-sized datasets. This article presents a survey of volume-visualization methods on general-purpose parallel ar chitectures, with special attention being paid to medical imaging appl ications. First, the various approaches to volume visualization are br iefly discussed, followed by a description of relevant aspects of para llel architectures. Next, the implications of the various architecture s are illustrated on the basis of a number of existing implementations of visualization algorithms on parallel architectures and their resul ts. For parallel volume visualization, multiple instruction, multiple data (MIMD) architectures are found to be superior to single instructi on, multiple data (SIMD) architectures. The latter type suffers from a lack of performance as well as flexibility. For most applications of interactive volume visualization, including the important area of medi cal imaging, shared memory MIMD architectures are preferred over distr ibuted memory MIMD architectures. The ease of programming of shared me mory architectures allows existing algorithms to be readily implemente d without loss of performance or flexibility.