The Shared Pathology Informatics Network (SPIN) is an initiative supported by the National Cancer Institute to create a multi-institutional distributed database of coded pathology reports useful for identifying specific types of human tissue available for research as well as the characteristics of disease across a large population of patients. Current members of the SPIN consortia include Harvard University, UCLA, the University of Pittsburgh and the Regenstrief Institute at the University of Indiana. The University of Pittsburgh is involved in several key SPIN initiatives, including analysis of negation patterns and concepts within pathology reports, development of de-identification algorithms, development of an auto-encoder, and load testing the distributed database and query software.

Our group is primarily responsible for implementing and loading the local SPIN database at Pitt and testing peer-to-peer communications in our local SPIN node. In addition, we have developed MagicMarkup, a simple browser for inspecting large numbers of pathology reports in a text file, searching them rapidly for particular text strings or regular expression patterns, and marking up text sequences within the reports with XML tags representing concepts or other text characteristics.