Gong Hui

Parallel Visualization of Visible Chinese Human with Extremely Large Datasets

Single sectional anatomy image of Visible Chinese Human datasets achieved recently in First Military Medical University is almost 127 MByte. The total datasets are estimated to be 1.1 TByte if 48 bits color image is stored. Personal computer and graph workstation with multiple CPUs are also impossible to process extremely large datasets. Therefore the challenge is to visualize the extremely large datasets efficiently. Visible Chinese Human datasets are so massive in size they require the use of parallel computing resources for effective visualization. At present a parallel visualization program has been developed based on parallelism visualization toolkit (pVTK) on high performance cluster to solve the problem. The visualization results and performance demonstrate that the parallel program we developed provides a promising solution of handling extremely large datasets