Methods, systems, and media for detecting the presence of an analyte

Patent metadata

Field	Value
Publication number	US-11037032-B2
Application number	US-201816148614-A
Country	US
Kind code	B2
Filing date	Oct 1, 2018
Priority date	Oct 6, 2017
Publication date	Jun 15, 2021
Grant date	Jun 15, 2021

Abstract

In accordance with some embodiments, methods, systems, and media for detecting the presence of are provided. In some embodiments, a method of detecting an analyte is provided, the method comprising: capturing an image of liquid crystals; determining one or more features based on the brightness of the pixels in the image; providing the one or more features to a trained support vector machine, wherein the support vector machine was trained using images captured of other liquid crystals when exposed to a first analyte and the other liquid crystals exposed to a second analyte; and receiving an indication from the support vector machine indicating whether the liquid crystals have been exposed to the first analyte.

First claim

What is claimed is:

1. A method for detecting the presence of an analyte, the method comprising:
capturing an image of liquid crystals;
determining a first set of one or more features based on the brightness of pixels in the image;
determining a second set of features based on the image of the liquid crystals,
wherein the second set of features comprises a plurality of outputs of a hidden layer of a trained object detection deep learning convolution neural network that was provided with values based on the pixels in the image as an input;

concurrently providing at least the first set of one or more features and the second set of features to a trained support vector machine,
wherein the support vector machine was trained using features based on images captured of other liquid crystals when exposed to a first analyte and the other liquid crystals when exposed to a second analyte; and

receiving an indication from the support vector machine indicating whether the liquid crystals have been exposed to the first analyte.

2. The method of claim 1 , wherein the second set of features are based on a color image of the liquid crystals, and wherein the first set of one or more features are based on a grayscale image of the liquid crystals.

3. The method of claim 1 , further comprising:
generating a normalized RGB image from the first image of the liquid crystals;
converting the normalized RGB image to a grayscale image;
calculating a plurality of oriented gradients using the grayscale image; and
calculating a histogram of the plurality of oriented gradients, wherein the one or more features comprises values from the histogram of the plurality of oriented gradients.

4. The method of claim 3 , wherein the normalized RGB image is generated from a portion of the image of the liquid crystals, wherein the normalized RGB image has a lower resolution than the image of the liquid crystals.

5. The method of claim 1 , wherein the first analyte is a gas phase analyte and the second analyte is a non-targeted gas phase molecule.

6. The method of claim 5 , wherein the first analyte is DMMP and the second analyte is water vapor.

7. The method of claim 6 , wherein the analyte to be detected is sarin.

8. The method of claim 1 , wherein the first analyte is a liquid phase analyte and the second analyte is a non-targeted liquid phase analyte.

9. The method of claim 8 , wherein the first analyte is a biological analyte.

10. The method of claim 1 , wherein the liquid crystals are disposed within a micro-well.

11. The method of claim 1 , wherein the liquid crystals are in contact with self-assembling monomers.

12. The method of claim 1 , wherein the liquid crystals are in contact with a polymerized target of a bioagent.

13. The method of claim 1 , wherein the liquid crystals form at least one droplet suspended in an aqueous phase.

14. The method of claim 1 , wherein the liquid crystals are disposed within a holding compartment of a substrate over which an aqueous solution is being passed.

15. The method of claim 1 , wherein the liquid crystals are doped with a chiral molecule.

16. A system for detecting the presence of an analyte, the system comprising:
an image sensor; and
a processor that is programmed to:
cause the image sensor to capture an image of liquid crystals;
convert the image of the liquid crystals to grayscale;
determine a first set of one or more features based on the brightness of pixels in the grayscale image;
determine a second set of features based on the image of the liquid crystals,
wherein the second set of features comprises a plurality of outputs of a hidden layer of a trained object detection deep learning convolution neural network that was provided with values based on the pixels in the image as an input;

concurrently provide at least the first set of one or more features and the second set of features to a trained support vector machine,
wherein the support vector machine was trained using features based on images captured of other liquid crystals when exposed to a first analyte and the other liquid crystals when exposed to a second analyte; and

receive an indication from the support vector machine indicating whether the liquid crystals have been exposed to the first analyte.

17. A non-transitory computer readable medium containing computer executable instructions that, when executed by a processor, cause the processor to perform a method for detecting the presence of an analyte, the method comprising:
capturing an image of liquid crystals;
converting the image of the liquid crystals to grayscale;
determining a first set of one or more features based on the brightness of pixels in the grayscale image;
determining a second set of features based on the image of the liquid crystals,
wherein the second set of features comprises a plurality of outputs of a hidden layer of a trained object detection deep learning convolution neural network that was provided with values based on the pixels in the image as an input;

concurrently providing the one or more features to a trained support vector machine,
wherein the support vector machine was trained using images captured of other liquid crystals when exposed to a first analyte and the other liquid crystals when exposed to a second analyte; and

receiving an indication from the support vector machine indicating whether the liquid crystals have been exposed to the first analyte.

Assignees

• Wisconsin Alumni Res Found

Inventors

• Abbott Nicholas
• Cao Yankai
• Zavala Tejeda Victor
• Yu Huaizhe

Classifications

• G01N21/77 Primary

  by observing the effect on a chemical indicator · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06F18/2411

  based on the proximity to a decision surface, e.g. support vector machines · CPC title

• H04N23/60

  Control of cameras or camera modules · CPC title

• G06F18/2148

  characterised by the process organisation or structure, e.g. boosting cascade · CPC title