The objective of this allocation request is to support educational users of the SimVascular open source project for cardiovascular patient-specific modeling and simulation. SimVascular (www.simvascular.org) is currently the only open source software project providing a complete pipeline from medical image data to patient-specific anatomic model construction and finite element simulation of blood flow. SimVascular is an active project, with approximately 2000 users in over 60 countries worldwide. Development of SimVascular is led jointly by Prof. Alison Marsden at Stanford, Prof. Shawn Shadden at UC Berkeley, and Dr. Nathan Wilson at Open Source Medical Software Corporation. Both PIs currently use SimVascular as their core research tool, and for education in the form of undergraduate research and class projects. SimVascular was used successfully by students in classes at Stanford and Purdue in Spring 2017. Students were given a medical image data set and each student was tasked with researching a particular cardiovascular disease, building an anatomic model of the relevant anatomy from image data, meshing and assigning boundary conditions, running a computational fluid dynamics simulation of blood flow, and presenting their findings to the class. The SimVascular project provides extensive documentation, tutorials, and example files for download on our website. Following our success in introducing SimVascular as an educational tool, we are now expanding to make this tool, and accompanying HPC resources available to students at other institutions.
We recently set up a community account and a GATEWAY for the SimVascular project. The GATEWAY allows users of SimVascular to upload their files to the GATEWAY portal site, run simulation jobs remotely on XSEDE machines (currently Comet at SDSC), and download results files when jobs are completed. This portal makes SimVascular solvers available to users who may not have local clusters or expertise to compile and manage the source code on local resources. It is particularly crucial for educational use since local clusters are often for research only, and students using SimVascular in course projects need far more computing power than on a typical student laptop.
We are now requesting an educational allocation to provide computational resources to students taking courses using SimVascular in the 2017-2018 academic year. We anticipate 2 graduate level courses will be using SimVascular in the coming year at Marquette University and Purdue University. In addition, we will be using SimVascular for undergraduate research projects at Purdue University (Craig Goergenâ€™s lab), Stanford University (Alison Marsdenâ€™s lab), UC Berkeley (Shawn Shaddenâ€™s lab) and Yale University (Jay Humphreyâ€™s lab). Software development for the SimVascular project is support by the NSF SSI program, and funding was just recently granted for a new 4-year project starting fall 2017.