back to projects

A Partnership Training Program In Breast Cancer Diagnosis:

Concept Development of the Next Generation Diagnostic Breast Imaging using Digital Image Library and Networking Techniques

Principal Investigator: Mohamed F. Chouikha, Ph.D.
Partnership Leader: S-C. Ben Lo, Ph.D.
Ahmed Jendoubi, Ph.D.
Jim Humphries

This program represents a training partnership between Howard Universit (HU) (Department of Electrical Engineering, Department of Systems and Computer Sciences, and Department of Radiology); and Georgetown University (Image Science and Information Systems, ISIS). In this partnership training program, we are training faculty and students in breast cancer imaging, digital image database library techniques and network communication strategy.

Experts from Georgetown and Howard University are participating in training through seminars, specialized tutorials and workshops. Outside distinguished speakers give talks in the lecture series. They provide the faculty investigators and students with opportunities to discuss and to broaden the knowledge in breast cancer, imaging, databases, digital library and networking.

GRID Technologies for Digital Mammography

High resolution digital mammograms and advanced algorithms must be accessible for them to be useful. To that end we have designed, and are in the process of implementing, a new end to end system that leverages our existing database and research technologies along with advances in technologies from high performance computing, the Grid, and Cyber-infrastructures. The purpose is to make real cases and our technology accessible to researchers and students over the network via an intuitive web-accessible interface that allows the user to locate data and services that are distributed and span organizational boundaries.

The prototype is designed around proven Grid and cyber-infrastructure technologies such as the Globus Grid toolkit and the San Diego Supercomputing Center's Storage Resource Broker (SRB). The initial version includes an interactive web client that will provides the ability to sign and annotate mammograms and to request and receive "consults" from radiologists, researchers, instructors, and expert systems and to automatically find additional similar cases using content based retrieval.

Additional functionality is planned to include security certificates, advanced content based search, and the ability to visualize pathology slides.