Bioelectronic tongue for food allergy detection

• US 20100222224A1
• Filed: 09/03/2009
• Published: 09/02/2010
• Est. Priority Date: 09/03/2008
• Status: Active Grant

• Current Assignee: Ian Ivar Suni, Stephanie Ann Caswell Schuckers, Yin Huang
• Original Assignee: Ian Ivar Suni, Stephanie Ann Caswell Schuckers, Yin Huang
• Inventors: Huang, Yin, Suni, Ian Ivar, Schuckers, Stephanie Ann Caswell
• Granted Patent: US 9,201,068 B2
• Time in Patent Office: Days
• US Class Current: 506/8
• CPC Class Codes: B01J 2219/00605 the compounds being directly bound or immobilised to solid supports
• CPC Class Codes: B01J 2219/00725 Peptides
• CPC Class Codes: B01J 2219/0074 Biological products
• CPC Class Codes: G01N 33/5438 Electrodes

Abstract

The present invention is directed to a method and apparatus that satisfies the need for a bioelectronic tongue for food allergy detection. The method of detecting concentration of food allergen incorporates antibodies into an electronic tongue to create a bioelectronic tongue. Additionally the method uses impedance, capacitance, and/or other related electrochemical methods for detecting analyte in complex media. Furthermore the method additionally includes methods to subtract out non-specific interactions. The method also subtracts non-specific interactions. The device/apparatus is a Bioelectronic Tongue for detecting allergen in diluted food samples. The device includes: a sensor array; an impedance or capacitance analyzer; a preprocessor; a feature extractor; a pattern recognizer; and an output device indicating an allergen concentration. In order to implement the method of detecting food allergens on a bioelectronic tongue a computer readable medium containing an executable program is used for performing the analysis of a food sample. The executable program performs the acts of: preprocessing data from an impedance analyzer; extracting a feature pattern; recognizing a pattern of features of data representing a concentration of food allergen contained is the food sample; and outputs allergen concentration data.

Claims

1. A bioelectronic tongue system comprising:

a sensor array;

an impedance or capacitance analyzer;

a preprocessor;

a feature extractor;

a pattern recognizer; and

an output device indicating an allergen concentration.

2. The system of claim 1 wherein said sensor array further comprises a plurality of locations for diluted food or other samples.

3. The preprocessor of claim 1 preprocesses baseline data.

4. The feature extractor of claim 1 maps a dimensional feature space into a projection of features ordered by variability.

5. The pattern recognizer of claim 1 comprises a training set, a test set and a validation set.

6. The output device of claim 1 is selected from the group consisting of LED display, LCD display, OCLD display, CRT display and indicator lamp.

19. The method of claim 1 wherein said outputting comprising the outputting of allergen data to a display device.

7. A method of detecting concentration of food allergen comprising the acts of:

incorporating antibodies into an electronic tongue to create a bioelectronic tongue.

8. The method of claim 7 comprising the additional acts of:

using impedance, capacitance, and/or other related electrochemical methods for detecting analyte in complex media.

9. The method of claim 8 comprising the additional acts of including methods to subtract out non-specific interactions.

10. The method of claim 7 comprising the additional acts of using methods for subtracting non-specific interactions.

11. A method for determining a concentration of a food allergen comprising the acts of:

providing a plurality of diluted test samples;

applying said test samples to a sensor array;

using an impedance analyzer to create an electrical data output based on the contents of each sample;

preprocessing said data;

extracting a feature pattern;

recognizing a pattern of features of said data representing a concentration of food allergen contained in said test samples;

outputting an allergen concentration.

12. The method of claim 11 said preprocessing comprises processing baseline data so that baseline adjustment and drift compensation may be performed.

13. The method of claim 11 wherein said feature selection comprises mapping of dimensional feature space into a projection of features ordered by variability.

14. The method of claim 11 wherein said pattern recognition comprises a training set, a test set and a validation set.

15. The method of claim 14 wherein said training set comprises the acts of creating a model and selecting thresholds.

16. The method of claim 15 wherein said training set is an iterative process where weights/models are updated to minimize classification errors.

17. The method of claim 14 wherein said test set verifies that said model will generalize to an unseen set of data.

18. The method of claim 14 wherein said validation set uses cross-validation, where a small set is withheld for validation while the remaining data is used for training/testing with said validation process is repeated many times (100×

or more) repeatedly on small subsets of performance on an unseen data set.

20. A computer readable medium containing an executable program for performing an analysis of a food sample for allergens where the program performs the acts of:

preprocessing data from an impedance analyzer;

extracting a feature pattern;

recognizing a pattern of features of data representing a concentration of food allergen contained is said food sample; and

outputting an allergen concentration.

21. The program of claim 20 wherein said preprocessing data comprises baseline data so that baseline adjustment and drift compensation may be performed.

22. The program of claim 20 wherein said feature selection comprises a mapping of dimensional feature space into a projection of features ordered by variability.

23. The program of claim 20 wherein said pattern recognition comprises a training set, a test set and a validation set.

24. The program of claim 20 wherein said training set comprises the acts of creating a model and selecting thresholds.

25. The program of claim 24 wherein said training set is an iterative process where weights/models are updated to minimize classification errors.

26. The program of claim 23 wherein said test set verifies that said model will generalize to an unseen set of data.

27. The program of claim 23 wherein said validation set uses cross-validation, where a small set is withheld for validation while the remaining data is used for training/testing with said validation process is repeated many times (100×

or more) repeatedly on small subsets of performance on an unseen data set.

28. The program of claim 20 wherein said outputting comprising the outputting of allergen data to a display device.