Mixed statistical and numerical model for sensor array detection and classification

• US 7,672,813 B2
• Filed: 12/03/2007
• Issued: 03/02/2010
• Est. Priority Date: 12/03/2007
• Status: Active Grant

• Current Assignee: Smiths Detection (Smiths Group International Holdings Ltd.)
• Original Assignee: Smiths Detection (Smiths Group International Holdings Ltd.)
• Inventors: Li, Sichu, Tabacco, Mary E.
• Primary Examiner(s): Lau; Tung S
• Assistant Examiner(s): SUN, XIUQIN
• Application Number:
  US11/987,658
• Publication Number:
  US 20090144022A1
• Time in Patent Office: 820 Days
• Field of Search: 702/189, 702/191, 367/131, 382/106
• US Class Current: 702/189
• CPC Class Codes: G16H 40/63 for local operation
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders

Abstract

A method and system for detecting and classifying biosensor and chemical sensor data includes processing data fetched from an array of sensors. A statistical analysis of the processed data is performed, to obtain statistically analyzed sensor data. A determination is made, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors. If a response signal is determined to have been generated, a numerical pattern recognition is performed on the data collected from the array of sensors, in order to categorize the data.

Claims

1. A method for detecting and classifying sensor data, comprising:

processing data collected from an array of sensors by using a data processing unit;

performing a statistical analysis of the data collected from the array of sensors, to obtain statistically analyzed sensor data by using a statistical analysis unit;

determining, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors by using said statistical analysis unit; and

performing, if a response signal is determined to have been generated, a numerical pattern recognition on the data collected from the array of sensors by using a numerical pattern recognition unit, in order to categorize the data,

wherein the step of performing a numerical pattern recognition comprises;

performing an interferent rejection pattern analysis on the sensor data;

performing a biological match pattern analysis on the sensor data;

performing an interferent rejection pattern analysis on airborne particle data; and

performing a classification pattern analysis on the sensor data by using said numerical pattern recognition unit.

2. The method according to claim 1, wherein the step of performing a statistical analysis is performed by using a principle component analysis statistical model.

3. The method according to claim 1, wherein, based on results obtained from performing a numerical pattern recognition, outputting an alarm if the data collected from the array of sensors is determined to be provided by one of a predetermined set of bioagents.

4. A method for detecting and classifying sensor data, comprising:

processing data collected from an array of sensors by using a data processing unit;

performing a statistical analysis of the data collected from the array of sensors, to obtain statistically analyzed sensor data by using a statistical analysis unit;

determining, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors by using said statistical analysis unit;

performing, if a response signal is determined to have been generated, a numerical pattern recognition on the data collected from the array of sensors, in order to categorize the data by using a numerical pattern recognition unit; and

processing airborne particle data from one or more other sensors,

wherein the step of performing a numerical pattern comprises;

performing a first interferent rejection pattern analysis on the sensor data;

performing a biological match pattern analysis on the sensor data;

performing a second interferent rejection pattern analysis on the airborne particle data; and

if the results of the three performing steps meet predetermined criteria, performing a classification pattern analysis on the sensor data by using said numerical pattern recognition unit.

5. The method according to claim 4,

wherein the step of performing a statistical analysis is performed on the data collected from the array of sensors and not on the airborne particle data.

6. The method according to claim 4, wherein the step of performing a first interferent rejection pattern analysis is performed on a combination of the statistically sensor data and on raw sensor data that has not been statistically analyzed.

7. A system for detecting and classifying sensor data, comprising:

a data processing unit configured to receive and perform data processing on data output from an array of sensors;

a statistical analysis unit configured to perform statistical analysis on the data collected from the array of sensors and data processed by the data processing unit, to obtain statistically analyzed sensor data, wherein the statistical analysis unit determines, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors; and

a numerical pattern recognition unit configured to perform a numerical pattern recognition on the data collected from the array of sensors, in order to categorize the data,

wherein the numerical pattern recognition unit comprises;

an interferent rejection pattern analysis unit configured to perform an interferent rejection pattern analysis on the sensor data;

a biological match pattern analysis unit configured to perform a biological match pattern analysis on the sensor data; and

a classification unit configured to perform a classification pattern analysis on the sensor data.

8. The system according to claim 7, wherein the numerical pattern recognition unit only performs the numerical pattern recognition when the response signal is determined to have been generated.

9. The system according to claim 7, further comprising:

a second data processing unit configured to receive and perform data processing on airborne particle data output from one or more other sensors,

wherein the statistical analysis unit performs statistical analysis on the data collected from the array of sensors and not on the airborne particle data.

10. The system according to claim 7, wherein the statistical analysis unit performs statistical analysis using a principle component analysis statistical model.

11. A system for detecting and classifying sensor data, comprising:

a first data processing unit configured to receive and perform data processing on data output from an array of sensors;

a statistical analysis unit configured to perform statistical analysis on the data collected from the array of sensors and data processed by the first data processing unit, to obtain statistically analyzed sensor data, wherein the statistical analysis unit determines, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors;

a numerical pattern recognition unit configured to perform a numerical pattern recognition on the data collected from the array of sensors, in order to categorize the data; and

a second data processing unit configured to receive and perform data processing on airborne particle data output from one or more other sensors,

wherein the statistical analysis unit performs statistical analysis on the data collected from the array of sensors and not on the airborne particle data,

wherein the step of performing a numerical pattern comprises;

a first interferent rejection pattern analysis unit configured to perform a first interferent rejection pattern analysis on the sensor data;

a biological match pattern analysis unit configured to perform a biological match pattern analysis on the sensor data;

a second interferent rejection pattern analysis unit configured to perform a second interferent rejection pattern analysis on the airborne particle data; and

a classification unit configured to perform a classification pattern analysis on the sensor data, when the results of the first and second interferent rejection pattern analyses and the biological match pattern analysis meet predetermined criteria.

12. The system according to claim 11, wherein the first interferent rejection pattern analysis unit performs the first interferent rejection pattern analysis on a combination of the statistically sensor data and on raw sensor data that has not been statistically analyzed.

13. The system according to claim 11, further comprising:

an alarm unit communicatively connected to the numerical pattern recognition unit,

wherein, based on data provided by the numerical pattern recognition unit, the alarm unit outputs an alarm if the data collected from the array of sensors is determined to be provided by one of a predetermined set of bioagents.

14. A computer readable medium embodying computer program product for detecting and classifying sensor data, the computer program product, when executed by a computer, causing the computer to perform the steps of:

fetching data from an array of sensors;

performing a statistical analysis of the data collected from the array of sensors, to obtain statistically analyzed sensor data;

determining, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors; and

performing, if a response signal is determined to have been generated, a numerical pattern recognition on the data collected from the array of sensors, in order to categorize the data,

wherein the step of performing a numerical pattern recognition comprises;

performing an interferent rejection pattern analysis on the sensor data;

performing a biological match pattern analysis on the sensor data; and

performing a classification pattern analysis on the sensor data.

15. The computer readable medium according to claim 14, further comprising:

fetching airborne particle data from one or more other sensors,

wherein the step of performing a statistical analysis is performed on the data collected from the array of sensors and not on the airborne particle data.

16. The computer readable medium according to claim 14, wherein the step of performing a statistical analysis is performed by using a principle component analysis statistical model.

17. The computer readable medium according to claim 14, wherein the statistical analysis performed in the performing step is a Principle Component Analysis.

18. A computer readable medium embodying computer program product for detecting and classifying sensor data, the computer program product, when executed by a computer, causing the computer to perform the steps of:

fetching data from an array of sensors;

performing a statistical analysis of the data collected from the array of sensors, to obtain statistically analyzed sensor data;

determining, based on the statistically analyzed sensor data, whether or not there a response signal has been generated from the array of sensors;

performing, if a response signal is determined to have been generated, a numerical pattern recognition on the data collected from the array of sensors, in order to categorize the data; and

fetching airborne particle data from one or more other sensors,

wherein the step of performing a statistical analysis is performed on the data collected from the array of sensors and not on the airborne particle data,

wherein the step of performing a numerical pattern comprises;

performing a first interferent rejection pattern analysis on the sensor data;

performing a biological match pattern analysis on the sensor data;

performing a second interferent rejection pattern analysis on the airborne particle data; and

if the results of the three performing steps meet predetermined criteria, performing a classification pattern analysis on the sensor data.

19. The computer readable medium according to claim 18, wherein the step of performing a first interferent rejection pattern analysis is performed on a combination of the statistically sensor data and on raw sensor data that has not been statistically analyzed.