Linking information at the genomic level with the vast quantities of phenotypic information derived from medical imaging is a fundamental research challenge that offers tremendous potential for near-term advances in medical treatment. To process the TeraBytes of data produced by new imaging technologies and genomic analyses, major medical research institutions such as Washington University turn to High Performance Computing (HPC) or supercomputers. There is a pressing need within the Washington University's Medical research community for a HPC located on the Medical School campus. Resources are currently restricted to two HPC systems which do not offer enhanced visualization capabilities or the combination of shared and distributed memory architectures. The establishment of the proposed HPC will achieve the following aims: 1) Provide a horizontal scaling capability which can support varying numbers of users each accessing the same application software in parallel. 2) Support NIH funded research projects such as Bayesian estimation of diffusion tensor parameters which require the functionality of a large shared memory system computing environment. 3) Allow remote use of image rendering and manipulation applications by sending the graphics display of the HPC's output to users located anywhere on the research network providing efficient visualization of large datasets at high resolution in a collaborative environment. 4) Accommodate privacy and security restrictions imposed by HIPAA while removing the complexities of transporting research data off campus to access HPC systems at other facilities or institutions. This HPC environment will provide a unique combination of high performance shared memory computing, large scale cluster computing and high performance distributed visualization all within a HIPPA secure environment and is essential for a large number of current NIH-funded projects. These capabilities will enable substantial increases in productivity for existing projects covering a broad range of imaging modalities while supporting Washington University's commitment to enhancing basic and clinical research in imaging, genomics, bioinformatics and computational biology.
Showing the most recent 10 out of 43 publications
