We propose to demonstrate the feasibility and utility of the use of a heterogenous collection of databases for collaborating researchers in the OSU Brain-Tumor Research Center. The key technological innovation we hypothesize that will make such an approach feasible and effective is a Heterogenous Database Translation Server (HDTS). The HDTS will be a UNIX workstation based node on a high speed fiber network (as illustrated in Figure 1) with file exchange capabilities with the various computer systems of the individual investigators and institutional computer systems. On this UNIX workstation will reside a software architecture which will support automated creation of software for the translation among the databases of interest to collaborating investigators. A related software architecture is already in place, the Integrated Chameleon Architecture (ICA), but it was designed to work with a model of data different from that typically found in the medical research environment. In order to design and implement the required new component of the ICA, the following activities are to be undertaken: 1) A content investigation of the neuro- oncology research domain, specifically as it is instantiated in the research groups at The Ohio State University and Columbus Children's Hospital collaborating on brain-tumor research; 2) Formation of a generalized computational model of related Neuro-Oncology databases, as suggested by the content investigation; 3) Formation of an intermediate form for data interchange between these Neuro-Oncology databases; 4) Implementation of the new ICA component; 5) Testing and evaluation of the new component; 6) Preparation of the component for distribution to other Brain-Tumor Research Centers. If successful, this project will have a direct and immediate impact on the collaborative sharing of data in the context of brain-tumor research at OSU, and will also serve as an exemplar for a variety of related database research projects within the national brain-tumor research community.
