Software Technology for Cancer Epidemiology and Prevention

A significant hindrance for military cancer studies is the lack of linkages between the disease outcomes database (Automated Central Tumor Registry) and the personnel databases (Air Force Personnel Center Personnel Database and Defense Manpower Data Center Database). Linking cancer and relevant personnel data is currently carried out manually, which is extremely cumbersome and time-consuming. There are two “bottlenecks” in a cancer researcher’s work flow. One is the recoding of cancer registry data. The other is the wait time for receiving personnel data. We propose to develop a Recoding Assistant to address the former, and a Data Format Previewer to address the latter. SSCI has assembled a strong team to develop the proposed software. The team includes Professors Shawn Murphy and Qing Zeng, both at Harvard Medical School, and Dr. William Salomon, an experienced physician/informatician. Professor Murphy is Director of the Research Patient Data Registry (RPDR), a multimillion-dollar effort sponsored by the Partners Research Council that gathers data from various hospital legacy systems and stores the data in one place for researchers to access. Professor Zeng and her collaborators have developed a suite of open-source NLP modules that they refer to as the Health Information Text Extraction (HITEx) tool, which uses GATE (General Architecture for Text Engineering) as a platform and can be assembled into pipelines for different tasks.