This is a Phase I of a fast-track STTR application submitted in response to Program Announcement PA-03-001 (""""""""Knowledge Integration across Distributed Heterogeneous Data Sources""""""""). We propose a collaborative effort between Heart Imaging Technologies, LLC (HeartIT) and Duke University's Cardiovascular Magnetic Resonance Center (DCMRC). The goal of the collaboration is to develop software supporting a worldwide research network of dynamic cardiac images based on existing world-wide web standards. Dr. Judd and his colleagues are recognized leaders in the field of cardiovascular MRI and have an established base of former collaborators and trainees currently located at different medical institutions across the world. Collaborative research, however, has been slowed by a lack of methods to share images and other data amongst these colleagues. In May 2000, Dr. Judd and others formed HeartIT to address these issues by developing image management systems based on converting images from DICOM format to web pages. Currently, HeartlT's third-generation commercial system (WebPAX(tm)) allows viewing of dynamic magnetic resonance images on any computer with a standard web browser (eg. Microsoft Explorer) and is in production clinical use at Duke University Hospital. Worldwide, a total of twelve similar HeartIT systems are operating at medical institutions such as Johns Hopkins Hospital, the Nashville Cardiovascular MRI Institute, and the Robert Bosch Hospital in Germany. These existing systems, as well as the close collaboration between Duke University and HeartIT, provide a unique opportunity to inexpensively and efficiently create a distributed network to share research data. In this STTR, we propose a major expansion of the existing WebPAX(tm) software which will allow these existing systems to be connected together to form a global network for sharing medical images based entirely on world-wide web standards. In the short term, this achievement will produce an important research resource supporting collaborative studies focused on the rapidly evolving field of cardiovascular MRI. In the longer term, however, the resulting global cardiovascular MRI network will serve as a fully-functional prototype for a generally-applicable computing architecture designed to provide worldwide integration of medical images acquired based on heterogeneous imaging modalities (MRI, echo, cath, SPECT) and equipment (GE, Siemens, Philips) distributed across distant geographic locations (Americas, Europe, Asia). In this application, we propose to create a """"""""public"""""""" WebPAX user interface which allows viewing of de-identified image data from physical locations outside Duke (Aim 1), to construct an """"""""Image Location Library"""""""" based on a secure web site with user-, group- and study-level accounts to catalog de-identified image data across multiple sites (Aim 2), and to test the resulting worldwide network of dynamic cardiac images using a real-life multicenter research study (Aim 3). Achieving these Aims will simultaneously create a powerful tool for research involving cardiovascular MRI as well as position HeartIT for an expansion to support for other imaging modalities through a Phase II SBIR/STTR.
