Cyclosarin Exposure Detection
Via Microarray Data Analysis
March 2005
Andrzej K. Brodzik, The MITRE Corporation
John Dileo, The MITRE Corporation
Andrea Jensenius, The MITRE Corporation
Taehwan Kim, The MITRE Corporation
Olivia Peters, The MITRE Corporation
ABSTRACT
The purpose of this work is to establish feasibility of cyclosarin
exposure detection via microarray data analysis and to evaluate the
sensitivity of several detection methods to exposure level. First, we
test the methods on the Golub leukemia data, and then we apply the methods
to multilevel cyclosarin exposure data. Initial results of the investigation
suggest that relatively low error rates in detection of a low dose sarin
exposure can be obtained using either Bayes classifier, neural networks,
or simple class signature matching.
» Download Paper [PDF, 261KB]
Additional Search Keywords
N/A
|
