ISDS Conference

This paper describes a research effort to map the literature of bioterrorism agents research worldwide using bibliographic analysis, content map analysis, and co-authorship analysis based on Medline data. The objectives of our research are to (a) identify researchers who have expertise in the research domain of bioterrorism agents from the world, (b) identify major institutions and countries where these researchers reside, and (c) identify emerging topics and trends in bioterrorism agents research.

The primary objective of this study is to assess the capability of an advanced text analytics tool that uses natural language processing techniques to extract important medical information collected as part of routine emergency room care (history, symptoms, vital signs, test results, initial diagnosis, etc.). This information will be automatically, accurately, and efficiently converted from unstructured text into use-able information, which can then be used to identify cases that are the result of a naturally occurring outbreak or bioterrorism event. This information would then be available to (1) communicate to the treating physician, and (2) message back to organizations aggregating data at a higher level, such as the Centers for Disease Control and Prevention (CDC) and the Department of Homeland Security (DHS).

This paper describes lessons learned from a regional tabletop exercise (TTX) of the National Capital Region (NCR) Syndromic Surveillance Network, from the perspective of the Maryland Department of Health and Mental Hygiene (DHMH).

We propose a novel technique for building generative models of real-valued multivariate time series data streams. Such models are of considerable utility as baseline simulators in anomaly detection systems. The proposed algorithm, based on Linear Dynamical Systems (LDS) [1], learns stable parameters efficiently while yielding more accurate results than previously known methods. The resulting model can be used to generate infinitely long sequences of realistic baselines using small samples of training data.

This paper describes how powerful detectors of adverse events manifested in multivariate series of bio-surveillance data can be learned using only a few labeled instances of such events.

This work incorporates model learning into a Bayesian framework for outbreak detection. Our method learns the spatial characteristics of each outbreak type from a small number of labeled training examples, assuming a generative outbreak model with latent center. We show that using the learned models to calculate prior probabilities for a Bayesian scan statistic significantly improves detection performance.

This paper describes a cluster of influenza-like illness (ILI) prospectively identified through emergency department (ED) syndromic surveillance (SS).

This paper describes our integrated visual analytics framework for analyzing both human emergency room data and veterinary hospital data.

Infectious disease outbreaks require rapid access to information to support a coordinated response from healthcare providers and public health officials. They need to know the size, spread, and location of the outbreak, and they also need access to models that will help them to determine the best strategy to contain the outbreak. 

There are numerous software tools for outbreak detection, and there are also surveillance systems that depend on communication between health care professionals. Most of those systems use a single type of surveillance data (e.g., syndromic, mandatory reporting, or laboratory) and focus on human surveillance.

However, there are fewer options for planning responses to outbreaks. Modeling and simulation are complex and resource-intensive. For example, EpiSims and EpiCast, developed by the National Institute of Health Models of Infectious Disease Agent Study involve large, diverse datasets and require access to high-performance computing.

Cyberenvironments are an integrated set of tools and services tailored to a specific discipline that allows the community to leverage the national cyberinfrastructure in their research and teaching. They provide data stores, computational capabilities, analysis and visualization services, and interfaces to shared instruments and sensor networks.

The National Center for Supercomputing Applications is applying the concept of cyberenvironments to infectious disease surveillance to produce INDICATOR.

Objective

This paper describes INDICATOR, a biosurveillance cyberenvironment used to analyze hospital data and generate alerts for unusual values.

If the next influenza pandemic emerges in Southeast Asia, the identification of early detection strategies in this region could enable public health officials to respond rapidly. Accurate, real-time influenza surveillance is therefore crucial. Novel approaches to the monitoring of infectious disease, especially respiratory disease, are increasingly under evaluation in an effort to avoid the cost- and timeintensive nature of active surveillance, as well as the processing time lag of traditional passive surveillance. In response to these issues, we have developed an indications and warning (I&W) taxonomy of pandemic influenza based on social disruption indicators reported in news media.

Objective

Our aim is to analyze news media for I&W of influenza to determine if the signals they create differ significantly between seasonal and pandemic influenza years.