Method and system for bio-surveillance detection and alerting

• US 20030009239A1
• Filed: 05/15/2002
• Published: 01/09/2003
• Est. Priority Date: 03/23/2000
• Status: Active Grant

• Current Assignee: Johns Hopkins University
• Original Assignee: Johns Hopkins University
• Inventors: Newhall, Bruce K, Lombardo, Joseph S, Wojcik, Richard A, Loschen, Wayne A, Chotani, Rashid A, Burkom, Howard S, Pineda, Fernando J
• Granted Patent: US 7,249,006 B2
• Time in Patent Office: Days
• US Class Current: 700/30
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 50/80 for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

Background noise from relevant data sets, including for example over-the-counter sales data, absenteeism data, etc., is subtracted using a background estimation algorithm that outputs residual data The effects of hypothetical anomalous events, such as a bio-terrorist attack, on the relevant data sets are modeled to create replica data. The replica data may be based on input from epidemiologists and various scenario templates including information on disease manifestation and other intelligence. The residual data and the replica data are then matched using a detector. Types of detectors include for example adaptive matched-filter detectors, change detectors and Bayesian Inference Networks. An alarm is triggered if a real anomalous event similar to a hypothetical anomalous event is detected. A Geographical Information System (GIS) may be used to display data from individual zip codes.

Claims

1. A method for bio-surveillance detection and alerting, comprising the steps of:

subtracting background noise from relevant data sets using a background estimation algorithm to create residual data;

modeling the effects of a hypothetical anomalous event on said relevant data sets to create replica data;

matching said residual data with said replica data using a detector to detect a real anomalous event similar to said hypothetical anomalous event; and

triggering an alert if a real anomalous event similar to said hypothetical anomalous event is detected.

2. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said background estimation algorithm is a center-surround algorithm.

3. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said background estimation algorithm includes a Kalman filter.

4. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said detector is an adaptive matched-filter detector.

5. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said detector is a Neyman-Pearson detector.

6. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said detector is a change detector.

7. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said detector includes a Bayesian Inference Network.

8. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise information on specific human behaviors exhibited during the onset of a disease.

9. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise over-the-counter drug sales data.

10. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise absenteeism data.

11. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise emergency room admissions data.

12. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise insurance claims billing data.

13. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise animal health data.

14. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise one or more of over-the-counter drug sales data, absenteeism data, emergency room admissions data, insurance claims billing data, and animal health data.

15. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise data from at least two different data sources.

16. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said relevant data sets comprise data from at least five different data sources.

17. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said step of matching said residual data with said replica data uses more than one detector.

18. A method for bio-surveillance detection and alerting as recited in claim 1, wherein said step of modeling the effects of a hypothetical anomalous event on said relevant data sets to create replica data, exploits historical data from influenza epidemics.

19. A method for bio-surveillance detection and alerting as recited in claim 1, further comprising the step of displaying data using a Geographical Information System.

20. A method for bio-surveillance detection and alerting as recited in claim 19, further comprising the step of normalizing said data against itself.

21. A method for bio-surveillance detection and alerting as recited in claim 19, further comprising the step of inputting information from a system user concerning a region of interest and a disease of interest.

22. A method for bio-surveillance detection and alerting as recited in claim 19, wherein the health status of a population is monitored at a geographical resolution equivalent to individual zip codes.

23. A method for bio-surveillance detection and alerting as recited in claim 19, wherein the data displayed are a function of an authorized access level of a user.

24. A method for bio-surveillance detection and alerting as recited in claim 19, further comprising the step of sending an electronic message to disease control personnel in a jurisdiction where an alert has been triggered.

25. A bio-surveillance detection and alerting system, comprising:

means for subtracting background noise from relevant data sets to create residual data;

means for modeling the effects of a hypothetical anomalous event on said relevant data sets to create replica data;

means for matching said residual data with said replica data to detect a real anomalous event similar to said hypothetical anomalous event; and

means for alerting a system user about said real anomalous event.