The Community for Technology Leaders
2015 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (2015)
Washington, DC, USA
Nov. 9, 2015 to Nov. 12, 2015
ISBN: 978-1-4673-6798-1
pp: 1063-1069
Jay Gholap , Information Systems, University of Maryland, Baltimore County, USA
Vandana P. Janeja , Information Systems, University of Maryland, Baltimore County, USA
Yelena Yesha , Computer Science and Electrical Engineering, University of Maryland, Baltimore County, USA
Raghu Chintalapati , Ekagra Software Technologies, USA
Harsh Marwaha , Ekagra Software Technologies, USA
Kunal Modi , Ekagra Software Technologies, USA
ABSTRACT
This paper proposes a collaborative data mining technique to provide multi-level analysis from clinical trials data. Clinical trials for clinical research and drug development generate large amount of data. Due to dispersed nature of clinical trial data, it remains a challenge to harness this data for analytics. In this paper, we propose a novel method using master data management (MDM) for analyzing clinical trial data, scattered across multiple databases, through collaborative data mining. Our aim is to validate findings by collaboratively utilizing multiple data mining techniques such as classification, clustering, and association rule mining. We complement our results with the help of interactive visualizations. The paper also demonstrates use of data stratification for identifying disparities between various subgroups of clinical trial participants. Overall, our approach aims at extracting useful knowledge from clinical trial data in order to improve design of clinical trials by gaining confidence in the outcomes using multi-level analysis. We provide experimental results in drug abuse clinical trial data.
INDEX TERMS
interactive visualization, clinical trials, collaborative data mining, master data management
CITATION

J. Gholap, V. P. Janeja, Y. Yesha, R. Chintalapati, H. Marwaha and K. Modi, "Collaborative data mining for clinical trial analytics," 2015 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Washington, DC, USA, 2015, pp. 1063-1069.
doi:10.1109/BIBM.2015.7359829
235 ms
(Ver 3.3 (11022016))