A User-centered Approach to Mining Health Information for Epidemiological Research and Comparative Analysis

Period of Performance: 12/14/2001 - 06/14/2002

$100K

Phase 1 SBIR

Recipient Firm

Stottler Henke Associates
1650 South Amphlett Boulevard, Suite 300
San Mateo, CA 94402
Principal Investigator

Abstract

We propose to develop a new form of data mining software application that will be uniquely capable of overcoming the challenges to the effective use of advanced data analysis algorithms in support of epidemiology and comparative analyses. In the design of MediMiner we have been motivated primarily by two potentially conflicting observations: (1) there is substantial evidence that the domain independent "data archaeology" systems available today are only useful to skilled data mining specialists; (2) it is a certainty that the objectives of our end users will change over time, as will the contents of the MDR. To address these observations we have developed an innovative approach that will enable data mining specialists (or other skilled and motivated users) the full exploratory power of an interactive integrated data mining system, but also allow the efficient construction of more targeted solutions that are better suited to the needs and capabilities of end users such as healthcare researchers. In the development of MediMiner we will build upon our integrated data mining solution called IKODA. Phase I research and development of a proof-of-concept limited prototype will lay the groundwork for the Phase II implementation of a fully functional prototype of MediMiner. The MediMiner system will be attractive to the healthcare professionals within both public and private sectors. Further, our approach is equally applicable to a great number of other domains.