CS 526 - Computer Graphics II                                           -- Yiwen Sun

Short Presentation 

In this presentation I will introduce visualization applications in biosurveillance.

---

Biosurveillance

The biosurveillance system will collect and integrate information from public health, food, agricultural, environmental monitoring and the intelligence community to provide an early warning system for an outbreak or possible bioterrorism attack, and facilitate the timely dissemination of results to appropriate decision makers, like public health officials and clinicians.

---

Data

Multi-source, multi-dimensional, time-varying data:

• Clinical results from hospitals

• Environmental Samples from EPA

• Foodborne illness data from laboratories

• Animal disease report

• School or work absenteeism

• Over-the-counter pharmacy sale data

• Emergency calls or visits

In the format of

• Unstructured text data

• Existing databases

---

Model and Tasks of Visualization

Level	Situational Awareness Model	Tasks of Visualization
Level 1	Perception of the elements imported from the environment	Represent multi-source multivariate time-varying data
Level 2	Comprehension of the current situation and early event detection	Detect and track anomalies
Level 3	Projection of future status and decision making	Help complex data analysis and decision making; Facilitate the sharing of information in real time

---

Visualization Applications

• HealthMap : A global disease alert map based on ProMED-mail System which disseminates information on outbreaks of infectious diseases and acute exposures to toxins based on media reports and reports from subscribers.

• BioSense : A national public health surveillance system hosted by Centers for Disease Control and Prevention (CDC). It is designed to improve the nation's capability for disease control and monitoring through access to real-time healthcare data from healthcare organizations, state syndromic surveillance systems, national laboratories, and other data sources.

 

• VisAware : A novel visual paradigm for situational awareness, which increases flexibility and facilitates complex visual correlations.

• Agent-based Terrorist Attack Prediction : An application that utilizes a two-level fusion architecture to detect and track disease outbreaks across public health system databases. In the first level, collected data is used to detect and track indicative bio-events using latent semantic analysis and unsupervised clustering. In the second level, clusters produced via the first are used to feed dynamic Bayesian networks which assess outbreak type and state.

Comparison of These Applications

กก	Representation			Detect Anomaly	Analysis / Projection	Collaboration (data transformation)	Timeliness
	Spatial	Multi-source	Time-varying
Health Map	Map	Same glyph	Slider	Color indicates severity	กก	mail	Latency
BioSense	Map	Separate views	Chart	Threshold in probability distribution	Analysis with events	Real-time, streaming network	Real-time
VisAware	Map	Sections on ring	Concentric ring	Line width	Visual correlation	กก	กก
ATAP	Map	Separate views	Chart	Latent semantic analysis based clustering	Dynamic Bayesian Networks to predict	Network of distributed database	Real-time

---

Future Work and Challenges

• Capability to accept large quantities of diverse information in any format, standardize the data to use with data from other sources.
• Capability to define the principal factors influencing normal data behavior.
• Collaborative approach in decision making.
• Evaluation of specificity, i.e. to assess and avoid false alerts.
• Ability to detect smaller outdoor outbreaks.
  

---

Reference

• Y. Livnat, J. Agutter, S. Moon and S. Foresti. "Visual Correlation for Situational Awareness". Proceedings of the InfoVis 2005
• Endsley, M.R., Bolte, B., & Jones, D.G. (2003) "Designing for situation awareness: An approach to human-centered design" London: Taylor & Francis
• Bolstad, C.A., Endsley, M.R., Jones, D.G., Wright, M. (2003) "Measurement of Shared Situation Awareness in the Future Objective Force" Paper Presented at the Collaborative Technology Alliances Conference: Advance Decision architecture Conference, College Park, MD
• Colleen B. Martin, SAIC contractor. "BioSense: Implementation of a National Early Event Detection and Situational Awareness System". MMWR August 26, 2005
• Blind, J.; Das, S., "Disease outbreak detection and tracking for biosurveillance: a data fusion approach," Information Fusion, 2007, 10th International Conference on
• http://en.wikipedia.org/wiki/BioWatch
  

Powerpoint Slides

---

by Yiwen Sun