SC19 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Impacting Cancer with HPC: Challenges and Opportunities

Authors: Patricia Kovatch (Icahn School of Medicine at Mount Sinai), Eric Stahlberg (Frederick National Laboratory for Cancer Research), Thomas Steinke (Zuse Institute Berlin)

Abstract: The rate of advancement in the application of AI is creating tremendous opportunities for the use of HPC in advancing research and clinical applications for cancer. Simultaneously, the importance of data and information necessary for successful AI applications to advance research efforts has also grown dramatically. The BoF will focus on opportunities in cancer research and clinical applications for HPC, emphasizing the data challenges and opportunities with AI in key applications like drug discovery and disease diagnosis. A forward focus will be pursued in data availability, model validation, sharing and adoption of developed models, and opportunities for broader collaborative efforts.

Long Description: Cancer is a disease that touches us all. As the importance of HPC and HPC technologies has grown in medical applications, underscored by several large scale initiatives including the 21st Century Cures Act, the Precision Medicine Initiative, Cancer Systems Biology Consortium and many efforts employing machine and deep learning involving large scale cancer data and high-performance computing. It is important to bring together the multiple interests as a community to increasingly share insight and expertise with a specific focus on cancer. The overall goal of the BOF is to bring together those in the HPC field who have an interest in understanding and expanding the impact and critical role for HPC can have on cancer.

The BOF for 2019 is emphasizing two primary areas with current high-levels of interest among the HPC community. The challenges around aggregating, securing and utilizing increasing volumes of information as they would be required for successful machine and deep learning applications in cancer, both in research and clinical settings. The second area of emphasis for the BOF will be in the opportunities for AI to both aid in the data challenges, but also as an opportunity for HPC to impact cancer research and clinical applications. As the number of applications of deep learning utilizing cancer data has risen dramatically, the goals of this year's BOF include bringing the communities together to discuss approaches that overcome the data challenges in cancer, share lessons learned and exciting new advances as HPC powers an increasing number of AI related applications in cancer, and lastly to bring the international community together and discuss approaches and conventions that can be used to facilitate and enable greater data sharing, as well as model sharing for transfer learning, independent validation, and broader application of training models.

The BoF is being organized to provide a framework for extending the learning and awareness about the significant role HPC has in fighting and treating cancer, both in specific domains and across many available technologies. The BOF this year will be framed with 30 minutes of brief comments and discussions by a panel of 6 leaders in the field from health, computing, academia, government, international and select commercial entities all pushing the frontiers of the field. The subsequent 30 minutes will afford a moderated discussion between the BOF attendees and the panelists, and the last 30 minutes will afford the well-received attendee question and and answer session where attendees and panelists alike raise questions of each other in a highly interactive nature.

The series of HPC and Cancer BoFs have both been well attended by participants across the HPC and scientific fields with an interest in building collaborations, gaining insight into shared challenges, and seeking new opportunities to contribute to advances in overcoming cancer.

Back to Birds of a Feather Archive Listing