Engineering a peer-to-peer collaboratory for tissue microarray research

This work presents the design of a prototype peer-to-peer collaboratory for imaging, analyzing, and seamlessly sharing tissue microarrays (TMA), correlated clinical data, and experimental results across a consortium of distributed clinical and research sites. The overarching goal of this project is to facilitate cooperative oncology research aimed at improved understanding of the underlying mechanisms of disease onset and progression while simultaneously providing new insight into therapy planning and treatment. Key components of the collaboratory include a specification of metadata schematics for characterizing TMA specimens and abstracting their interpretations, an framework for automated and accurate analysis of digitized TMAs and a peer-to-peer infrastructure for indexing and discovery of TMA data and metadata, and a novel, optimized decentralized search engine that supports flexible querying with search guarantees and bounded costs. Prototype implementations of the automated TMA analysis component and the storage/discovery component and their evaluations are presented.