Wavelet-Based Monitoring for Biosurveillance
<-- Return to the list
Date: 03-24-2006
Start Time:
1:00pm
End Time: 2:00pm
Speaker: Galit Shmueli, University of Maryland
Location: Uris 333
Abstract
Biosurveillance is the practice of monitoring disease-related data in order to detect disease outbreaks (natural and/or related to a bioterrorist attack). Biosurveillance systems have greatly improved in the last few years with respect to the types, amounts, and timeliness of data collected. They rely not only on diagnostic data but also on pre-diagnostic data. However, monitoring methods are still in their infancy.
Current biosurveillance relies on classical statistical control charts for detecting disease outbreaks. However, these are not always suitable in this context. Assumptions of normality, independence, and stationarity are typically violated in pre diagnostic data. Furthermore, outbreak signatures in such data are of unknown patterns, and therefore call for "general detectors." We describe wavelet-based methods, which make less assumptions and are suitable for detecting abnormalities of unknown form. Wavelets have been widely used for data denoising and compression, but little work exists on using them for monitoring. We discuss monitoring-based issues and illustrate them using data on U.S. military clinic visits.
A relevant working paper can be
found at
http://www.smith.umd.edu/faculty/gshmueli/biosurveillance2005.pdf
Bio
Please visit http://www.smith.umd.edu/faculty/gshmueli/