Panagiotis Korfiatis PhD, Presenter: Nothing to Disclose

Timothy Kline MS, BS, Abstract Co-Author: Nothing to Disclose

Daniel J. Blezek PhD, Abstract Co-Author: Nothing to Disclose

Steve Gerhardt Langer PhD, Abstract Co-Author: Stockholder, Evidentia Health, Inc

Bradley J. Erickson MD, PhD, Abstract Co-Author: Stockholder, Evidentia Health, Inc

Clinical studies can involve hundreds to thousands of subjects, with imaging data acquired over several time points, and at multiple institutions. Efficient execution of such studies demands a single platform that enables online collaboration and data management. The ability to develop data processing in this same environment is desireable. We created PESSCARA, a Platform to Enable Sharing of Scientific Computing Advances and Research Assets.

The execution of multi-site trials demands reliable data handling, including collection, processing, and reporting, with compliant audit trails of all steps. Prior to execution of such a study, the development of a processing pipeline is required. Validation of the pipeline requires access to data sets, and cohort selection is benefitted from flexible data and metadata tagging. PESSCARA offers a flexible data management schema incorporating data management features not available in existing solutions. Currently, PESSCARA has been tested in two different studies.

PESSCARA consists of HUBzero a platform used to create web sites for scientific collaboration; a content management system (TACTIC) to store medical images and associated metadata; and a Python based computational environment. HUBzero allows people to upload content and “publish” software tools to group members. HUBzero handles datasets, analysis tools, and incorporates collaboration capabilities. TACTIC has a highly configurable workflow, exposing a unified interface that can store, manage, and retrieve imaging studies. TACTIC plus Dcm4chee and Clinical Trials Processor is a research PACS that assures deidentification, study storage, and management of data and metadata with full auditing capabilities. Python serves as a computational environment since it enables rapid algorithm prototyping. The components of PESSCARA are open source and can be offered as a cloud solution.

PESSCARA aims to provide the community with an environment suitable to deal with the requirements of medical image analysis adapted to the spirit of open and reproducible research.

Korfiatis, P, Kline, T, Blezek, D, Langer, S, Erickson, B, Platform to Enable Shared Scientific Computing And Research Assets (PESSCARA).  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14019215.html