# US 8,787,633 B2 · System and method of organism identification

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 8,787,633 B2
- Document ID: US8787633B2
- Status: Granted
- Kind: B2
- Application number: 12/008,476
- Filed: 2008-01-11
- Published: 2008-12-18
- Granted: 2014-07-22
- Earliest claimed: 2007-01-16
- Assignees: Purdue Research Foundation
- Applicants: J. Paul Robinson; Bartlomiej Rajwa; Bulent Bayraktar; Arun K. Bhunia; E. Daniel Hirleman; Euiwon Bae
- Inventors: J. Paul Robinson; Bartlomiej Rajwa; Bulent Bayraktar; Arun K. Bhunia; E. Daniel Hirleman; Euiwon Bae
- Web page: https://patentgrep.com/patent/US8787633B2

## CPC

- G06V 20/695: {Preprocessing, e.g. image segmentation}
- G06T 7/0012: {Biomedical image inspection}
- G06V 20/698: {Matching; Classification}
- G06T 2207/30024: Cell structures in vitro; Tissue sections in vitro
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A system and method for identifying organisms by analysis of scattergrams of colonies is disclosed. cattergrams are obtained by culturing samples and illuminating the resultant colonies by a laser. The forward scattered light is imaged and subject to a feature extraction process. The feature vector may include Zernike or Chebyshev moments and may also include Harelick texture features. Feature vectors may be used to train a classification process using either supervised or unsupervised machine learning techniques. The classification process may be used to associate a colony phenotype with the genotype of the sample.

## Claims

### Claim 1

A system for the identification of microorganisms comprising:
a coherent light source;
a holder adapted to receive a substrate having a surface adapted for growth of microorganisms,
a scanner adapted to scan a region of the substrate to identify a centroid position of a colony of microorganisms;
an image capture device disposed to receive forward scattered light from the colony; and
a first computer configured to extract a feature vector from the image,
wherein an orientation of the coherent light source and the colony is configured to be adjusted so that a centroid of a coherent light beam substantially coincides with the centroid of the position of the colony.

### Claim 2 (depends on 1)

The system of claim 1, wherein a second computer is configured to compare the feature vector with an identification data set so as to determine whether the feature vector is identifiable with a specific organism contained in the identification data set.

### Claim 3 (depends on 2)

The system of claim 2, wherein the identified organism type is displayed.

### Claim 4 (depends on 2)

The system of claim 2, wherein the organism type is classified as one of pathogenic or non-pathogenic.

### Claim 5 (depends on 2)

The system of claim 2, wherein the feature vector is associated with a phenotypic characteristic of the colony and the identification data set associates the phenotype with a genotype.

### Claim 6 (depends on 2)

The system of claim 2, wherein the first computer and the second computer are the same computer.

### Claim 7 (depends on 2)

The system of claim 2, wherein the first computer and the second computer are connectable over a communications network.

### Claim 8 (depends on 7)

The system of claim 7, wherein the communications network is, at least in part, a wireless circuit.

### Claim 9 (depends on 1)

The system of claim 1, wherein the feature vector is a Zernike moment.

### Claim 10 (depends on 1)

The system of claim 1, wherein the feature vector is a Chebyshev moment.

### Claim 11 (depends on 1)

The system of claim 1, wherein the feature vector is a gray-level co-occurrence matrix (GLCM).

### Claim 12 (depends on 1)

The system of claim 1 wherein the feature vector is at least two of a Zernike moment, a Chebyshev moment, or a grey-level co-occurrence matrix (GLCM).

### Claim 13

A method of identifying microorganisms, the method comprising:
culturing a sample of material on a substrate to form a colony of microorganisms;
scanning a region of the substrate to determine a centroid position of the colony;
adjusting a relative orientation of a coherent light beam and the colony such that a centroid of the coherent light beam substantially coincides with the centroid position of the colony;
obtaining image data of the colony illuminated by the coherent light beam;
analyzing the image data; and
categorizing the analyzed image data to identify a type of the colony.

### Claim 14 (depends on 13)

The method of claim 13, wherein the image data is recorded on a photographic film or by an array of photosensitive electronic elements.

### Claim 15 (depends on 13)

The method of claim 13, wherein the adjusting the relative orientation is performed by a two-axis positioner.

### Claim 16 (depends on 13)

The method of claim 13, wherein analyzing the image data further comprises:
extracting a feature vector from the image data.

### Claim 17 (depends on 16)

The method of claim 16, wherein the extracting a feature vector further comprises:
characterizing the image data using a function having an invariant magnitude in at least one dimension.

### Claim 18 (depends on 16)

The method of claim 16, wherein the extracting a feature vector further comprises:
characterizing the image by computing a gray-level co-occurrence matrix (GLCM).

### Claim 19 (depends on 16)

The method of claim 16, wherein the extracting a feature vector further comprises:
characterizing the image using a function having an invariant magnitude in at least one dimension and by computing a gray-level co-occurrence matrix (GLCM).

### Claim 20 (depends on 16)

The method of claim 16, wherein the image data is normalized prior to the feature vector extraction.

### Claim 21 (depends on 13)

The method of claim 13, wherein categorizing the analyzed image data further comprises:
comparing a feature vector extracted from the image data with a pre-determined data set, the data set characterizing at least one known genotype.

### Claim 22 (depends on 21)

The method of claim 21, wherein the data set is trained using feature vectors extracted from a plurality of colonies of the known genotype.

### Claim 23 (depends on 22)

The method of claim 22, wherein the training is supervised.

### Claim 24 (depends on 22)

The method of claim 22, wherein the training is unsupervised.

### Claim 25 (depends on 22)

The method of claim 22, wherein the training and categorization is performed by a support vector machine (SVM).

### Claim 26 (depends on 13)

The method of claim 13, wherein the categorization is performed by a process that has been trained by feature vectors from a plurality of colonies having a known genotype.

### Claim 27 (depends on 26)

The method of claim 26, wherein the training is performed on a separate computer from a computer used for categorization.

### Claim 28

A computer program product stored or distributed on a non-transient computer readable medium, including instructions causing a computer to:
control a scanner to scan a region of the substrate to determine a centroid position of a colony of microorganisms cultured thereon;
adjust a relative orientation of a coherent light beam and the colony such that a centroid of a coherent light beam substantially coincides with the centroid position of the colony
acquire image data of the colony from an imaging device;
characterize the image data to produce a feature vector of the image data; and
categorize the image data using the feature vector and process that has been trained by using feature vectors previously obtained from colonies of a known genotype.

### Claim 29

A computer program product stored or distributed on a non-transient computer readable medium, including instructions causing a computer to:
control a scanner to scan a region of the substrate to determine a centroid position of a colony of microorganisms cultured thereon;
adjust a relative orientation of a coherent light beam and the colony such that a centroid of a coherent light beam substantially coincides with the centroid position of the colony;
acquire image data of the colony from an imaging device;
characterize the image data to produce feature vectors of the image data; and
train a characterization process using by using feature vectors obtained from colonies of a known genotype.

### Claim 30

A method of screening material samples, the method comprising:
culturing a material sample on a substrate;
scanning the substrate to determine a centroid position of a colony of the cultured material sample;
adjusting a position of a centroid of a coherent light beam so that a centroid of the coherent light beam is substantially coincident with the centroid position of the colony;
obtaining image data for the colony;
extracting a feature vector from the image data; and
categorizing the colony based on the feature vector.

### Claim 31 (depends on 30)

The computer program product of claim 30, wherein the characterization process is a support vector machine (SVM).

### Claim 32 (depends on 30)

The method of claim 30, wherein the step of categorizing comprises:
comparing the feature vector of the image data against a data base of feature vectors;
determining whether the feature vector is characteristic of at least one of feature vectors of the data base of feature vectors.

### Claim 33 (depends on 32)

The method of claim 32, wherein an indication that the colony does not match any of the feature vectors of the data base is provided.

### Claim 34 (depends on 33)

The method of claim 33, wherein the indication is provided on a computer display.

### Claim 35 (depends on 34)

The method of claim 34, where the feature vector of a selected colony is added to the data base of feature vectors.

### Claim 36 (depends on 32)

The method of claim 32, wherein the data base of feature vectors comprises feature vectors associated with known genotypes.

### Claim 37 (depends on 36)

The method of claim 36, wherein the data base of feature vectors further comprises feature vectors associated with unknown genotypes.

### Claim 38 (depends on 37)

The method of claim 37, where the feature vectors associated with unknown genotypes are associated with colonies previously selected for analysis.

### Claim 39 (depends on 30)

The method of claim 30, wherein the characterizing is performed by a support vector machine (SVM) and the SVM is trained using a feature vector data base, the feature vector data base including feature vectors obtained during a previous sample screening.

### Claim 40 (depends on 1)

The method of claim 1, wherein the image capture device is a camera.

### Claim 41 (depends on 1)

The method of claim 1, wherein the image capture device is an electronic device.

### Claim 42 (depends on 41)

The method of claim 41, wherein the electronic device is a two-dimensional charge coupled device (CCD).