Skip to Content

HPC-based computational imaging for high-resolution, quantitative magnetic resonance imaging

Brad Sutton, University of Illinois at Urbana-Champaign

Usage Details

Aaron Anderson, Brad Sutton, Curtis Johnson, Alexander Cerjanic

High-resolution, quantitative MRI techniques are powerful tools for clinicians and researchers to apply to study many problems in neurology, neurosurgery, and neuroscience. However, such techniques are hampered by excessive image reconstruction times that limit adoption into standard practice. We propose to develop fast reconstruction approaches through the incorporation of GPU acceleration and the massive parallelization enabled by HPC tools, such as the Blue Waters system. These approaches will result in efficient reconstruction of images for studies being performed at the Carle Neuroscience Institute and the Beckman Institute. The ultimate goal of this line of research is the development of solutions for real-time, on-line reconstructions based on seamless integration of the MRI scanner and Blue Waters that can be efficiently incorporated into the workflows of clinicians and brain researchers.