Method and apparatus for multi-parameter data analysis

• US 10,451,536 B2
• Filed: 04/23/2010
• Issued: 10/22/2019
• Est. Priority Date: 04/24/2009
• Status: Active Grant

• Current Assignee: Global Life Sciences Solutions Operations UK Ltd. (Danaher Corporation)
• Original Assignee: GE Healthcare UK Limited (Danaher Corporation)
• Inventors: Thomas, Nicholas
• Primary Examiner(s): Zeman, Mary K
• Application Number:
  US13/265,439
• Publication Number:
  US 20120035859A1
• Time in Patent Office: 3,469 Days
• CPC Class Codes: G01N 15/1429 Signal processing
• CPC Class Codes: G01N 15/1433 using image recognition
• CPC Class Codes: G01N 2015/1477 Multiparameters
• CPC Class Codes: G01N 2015/1497 Particle shape

Abstract

In one aspect, the present invention relates to a method 200 for identifying one or more phenotypes from a multi-parameter data set. The method 200 comprises measuring 202 correlation between pairs of parameters within the multi-parameter data set, modifying 204 correlated parameter values within a predetermined multi-parameter data analysis set to form an analysis parameter set, and analyzing 206 the multi-parameter data set using the analysis parameter set to identify one or more phenotypes from the multi-parameter data set. Various embodiments of the present invention may, for example, be used in an automated high-content screening (HCS) apparatus 100 for biological cellular analysis.

Claims

1. A method for automated high-content screening (HCS) of a plurality of samples comprising cells and subjected to sample treatment conditions in an assay, wherein the plurality of samples comprise both samples subjected to a same treatment condition and samples subjected to different sample treatment conditions, the method comprising:

processing data from each of two or more images of the plurality of samples to obtain a multi-parameter data set for each image;

for samples subjected to the same sample treatment condition, obtaining inter-sample measurements based on distributions of different parameters in the same sample under the same sample treatment condition;

for samples subjected to different sample treatment conditions, obtaining intra-sample measurements based on distributions of same parameter under the different sample treatment conditions;

determining, based on both inter-sample and intra-sample measurements, correlated parameters in the respective multi-parameter data sets, wherein parameters from the multi-parameter data sets are graphically visualized in a phenotype map generated based on the inter-sample and intra-sample measurements and compared to determine the correlated parameters, and wherein correlation is to be determined using a Kolmogorov-Smirnov (KS) distance measurement analysis of parameters from the multi-parameter data sets;

creating an analysis parameter set by removing at least one of the correlated parameters from the multi-parameter data sets to form the analysis parameter set;

applying the analysis parameter set to identify one or more cellular phenotypes of the samples from the multi-parameter data sets; and

providing an identification, in the plurality of samples, of one or more cells that undergo changes in cellular phenotypes.

2. The method of claim 1, further comprising identifying one or more phenotypes from a plurality of multi-parameter data sets and comparing respective phenotypes from the multi-parameter data sets to identify variations therebetween.

3. The method of claim 1, further comprising forming one or more respective analysis parameter sets for a plurality of multi-parameter data sets, and comparing the analysis parameter sets to determine whether or not a correlation relationship is maintained between the multi-parameter data sets.

4. The method of claim 1, wherein correlation is determined by the KS distance measurement analysis made between the phenotypes in a predetermined multi-parameter data analysis set.

5. The method of claim 4, wherein the KS distance measurement analysis includes a comparison of the frequency distribution of data within two data sets in the predetermined multi-parameter data analysis set.

6. The method of claim 1, wherein modifying correlated parameter values within the predetermined multi-parameter data analysis set comprises removing one of a pair of parameters whose correlation exceeds a predetermined threshold value from the analysis parameter set.

7. A tangible non-transient computer program storage product comprising machine instructions operable to configure a data processing apparatus to:

process data from each of two or more images of a plurality of samples to obtain a multi-parameter data set for each image;

for samples subjected to a same sample treatment condition, obtain inter-sample measurements based on distributions of different parameters in the same sample under the same sample treatment condition;

for samples subjected to different sample treatment conditions, obtain intra-sample measurements based on distributions of same parameter under the different sample treatment conditions;

determine, based on both inter-sample and intra-sample measurements, correlated parameters in the respective multi-parameter data sets, wherein parameters from the multi-parameter data sets are graphically visualized in a phenotype map generated based on the inter-sample and intra-sample measurements and compared to determine the correlated parameters, and wherein correlation is to be determined using a Kolmogorov-Smirnov (KS) distance measurement analysis of parameters from the multi-parameter data sets;

create an analysis parameter set by removing at least one of the correlated parameters from the multi-parameter data sets to form the analysis parameter set;

apply the analysis parameter set to identify one or more cellular phenotypes of the samples from the multi-parameter data sets; and

provide an identification, in the plurality of samples, of one or more cells that undergo changes in cellular phenotypes.

8. A system for automated high-content screening (HCS) of a plurality of samples comprising cells and subjected to sample treatment conditions in an assay, wherein the plurality of samples comprise both samples subjected to a same treatment condition and samples subjected to different sample treatment conditions, the system comprising:

an imager for obtaining two or more images of the plurality of samples, each image represented by a respective multi-parameter data set; and

an image processor that is operable to;

process data from each of the two or more images to obtain respective ones of the multi-parameter data sets;

for samples subjected to the same sample treatment condition, obtain inter-sample measurements based on distributions of different parameters in the same sample under the same sample treatment condition;

for samples subjected to different sample treatment conditions, obtain intra-sample measurements based on distributions of same parameter under the different sample treatment conditions;

determine, based on both inter-sample and intra-sample measurements, correlated parameters in the respective multi-parameter data sets, wherein parameters from the multi-parameter data sets are graphically visualized in a phenotype map generated based on the inter-sample and intra-sample measurements and compared to determine the correlated parameters, and wherein correlation is to be determined using a Kolmogorov-Smirnov (KS) distance measurement analysis of parameters from the multi-parameter data sets;

create an analysis parameter set by removing at least one of the correlated parameters from the multi-parameter data sets to form the analysis parameter set;

apply the analysis parameter set to identify one or more cellular phenotypes of the samples from the multi-parameter data sets; and

provide an identification, in the plurality of samples, of one or more cells that undergo changes in cellular phenotypes.

9. The system of claim 8, wherein the image processor is further operable to identify one or more phenotypes from a plurality of multi-parameter data sets and compare respective phenotypes from the multi-parameter data sets to identify variations therebetween.

10. The system of claim 8, wherein the image processor is further operable to form two or more respective analysis parameter sets for a plurality of multi-parameter data sets, and compare the analysis parameter sets to determine whether or not a correlation relationship is maintained between the multi-parameter data sets.

11. The system of claim 8, wherein the image processor is further operable to determine correlation by using the KS distance measurement analysis made between the parameters in a predetermined multi-parameter data analysis set.

12. The system of claim 11, wherein the KS distance measurement analysis includes a comparison of the frequency distribution of data within two data sets in the predetermined multi-parameter data analysis set.

13. The system of claim 8, wherein the image processor is further operable to modify correlated parameter values within a predetermined multi-parameter data analysis set by removing one of a pair of parameters whose correlation exceeds a predetermined threshold value from the analysis parameter set.

14. The system of claim 8, wherein the two or more images comprise cellular images and the imager comprises a cellular imager.

15. The system of claim 14, wherein each multi-parameter data set includes one or more parameters of:

nuclear area;

length;

diameter;

perimeter;

ellipticity; and

weighted moment of inertia (WMOI), form, and distribution.

16. The system of claim 14, wherein the intra-sample measurements are used in weighting and prioritizing data produced from the inter-sample measurements to produce a single phenotype score measure summarizing a cumulative difference between two cellular populations across a range of measured parameters.

17. The system of claim 8, wherein the image processor is further configured to produce one or more phenotype maps graphically representing phenotypic signatures based on intra-sample measurements and inter-sample measurements.

18. The system of claim 17, wherein the inter-sample measurements and the intra-sample measurements comprise population comparator measures.

19. The system of claim 18, wherein the population comparator measures comprise Kolmogorov-Smirnov (KS) distance measurements.

20. The system of claim 8, wherein the intra-sample measurements are utilized to measure changes in the inter-sample measurements to provide phenotype characterization data.

21. The system of claim 8, wherein the sample is subjected to a drug treatment.

22. The system of claim 8, wherein the sample comprises a cellular population and the two or more images comprises a microscope image of the cellular population.

23. The system of claim 8, wherein identifying cells that undergo changes in cellular phenotypes comprises detecting changes in interrelationships of correlated parameters based on inter sample measurement.

24. The system of claim 8, the image processor is further operable to generate, based on a combination of inter-sample and intra-sample measurements, phenotype maps graphically representing phenotypic signatures.