There are many reasons for the relatively slow proliferation of advanced medical image processing methods but a significant reason is the present paradigm for providing access: most applications are still tied to proprietary software and hardware environments that carry significant up-front costs. The ultimate intent of this work is leverage commodity computing technologies to develop an open, extensible framework for deploying medical image processing applications in the heterogeneous, networked computing environment of today. The framework will provide clinicians and researchers access to state-of-the-art image processing applications regardless of their particular computing platform or locally available computing resources connecting them with federated database resources, with high-end computing resources, or even with their colleagues in a peer-to-peer computing environment.
The aims for Phase I of this project are: (1) Demonstrate that the framework provides access to image processing applications to an extent that is largely independent of local computing resources. (2) Demonstrate that the framework is general in that the same components can be reused for deploying a wide variety of medical imaging applications. (3) Demonstrate that the framework is customizable both by third-party developers and by end-users allowing power-users to both create and deploy new applications. Work in Phase II will extend the framework and develop two-demonstration applications--computer aided diagnosis (CAD) for mammography and multimodality image fusion. The ultimate goal is to obtain key partnerships and the private equity investment necessary for commercialization, which will proceed by launching revenue-generating versions of the CAD and image fusion applications.
