METHOD AND SYSTEM FOR CHARACTERIZING CELL POPULATIONS

• US 20150087240A1
• Filed: 09/22/2014
• Published: 03/26/2015
• Est. Priority Date: 09/26/2013
• Status: Active Grant

• Current Assignee: Curi Bio, Inc.
• Original Assignee: Cellogy, Inc.
• Inventors: Loewke, Kevin, Maddah, Seyyedeh Mahnaz
• Granted Patent: US 9,619,881 B2
• Time in Patent Office: Days
• US Class Current: 455/67.11
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/30024 Cell structures in vitro; Tissue sections in vitro
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/0016 involving temporal comparison
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes
• CPC Class Codes: G06V 20/698 Matching; Classification

Abstract

A method and system for characterizing a cell population including a set of cell subpopulations, the method comprising: at a processing system, receiving image data corresponding to a set of images of the cell population captured at a set of time points; at the processing system, generating an analysis based upon processing the set of images according to: a cell stage classification module configured to automatically identify a cell class of each of the set of cell subpopulations, and a cell graph representation module configured to characterize geometric and spatial features of the set of cell subpopulations; from the analysis, determining a set of parameters characterizing the cell population; and generating an indication of quality of the cell population based upon a comparison between at least one parameter of the set of parameters and a set of reference values.

Claims

1. A method for characterizing a cell population including a set of cell subpopulations, the method comprising:

at a processing system, receiving image data corresponding to a set of images of the cell population captured at a set of time points;

at the processing system, generating an analysis based upon processing the set of images according to;

a cell stage classification module configured to automatically identify a cell stage of each of the set of cell subpopulations;

from the analysis, determining a set of parameters characterizing the cell population; and

at the processing system, generating an indication of quality of the cell population based upon a comparison between at least one parameter of the set of parameters and a set of reference values.

2. The method of claim 1, wherein receiving image data comprises receiving a set of phase contrast images, for each of a set of locations spanning the cell population at each of the set of time points.

3. The method of claim 1, wherein processing the set of images according to the cell stage classification module comprises:

partitioning each image in the set of images into a set of pixel subsets,

for each pixel subset in the set of pixel subsets for each image, generating a feature vector based upon processing of a local binary pattern histogram and a pixel intensity histogram associated with each pixel subset, thereby producing a set of feature vectors corresponding to the set of pixel subsets for each image in the set of images, and

training a machine learning classifier, with a training dataset, to identify each of a set of cell stages of the cell population captured in the set of images, using the set of feature vectors for each image in the set of images.

4. The method of claim 3, wherein processing the set of images according to the cell stage classification module comprises enabling identification of at least one of:

a single-cell stage, a medium-compaction stage, a full-compaction stage, a dead cell stage, a differentiated cell stage, and a debris stage exhibited by cells of the cell population at the cell stage classification module.

5. The method of claim 1 wherein wherein generating the analysis further comprises processing the set of images according to a cell graph representation module, including:

at a cell cluster segmentation module, segmenting, from at least one image of the set of images, a set of regions corresponding to the set of cell subpopulations of the cell population, each region defined by a set of nodes and a set of edges, and

determining a set of parameters characterizing the cell population based upon an analysis of the set of nodes and the set of edges of at least one of the set of regions.

6. The method of claim 1, wherein determining the set of parameters comprises determining parameters indicative of at least one of:

degree of cell compaction, cell doubling time, sensitivity to culture media change, and colony border spikiness based upon outputs of at least one of the cell stage classification module and the cell graph representation module.

7. The method of claim 1, further comprising generating a notification based upon the indication and providing the indication to an entity associated with the cell population, at a mobile device of the entity.

8. A method for characterizing a cell population including cells of at least one of a set of cell stages, the method comprising:

at a processing system, receiving image data corresponding to a set of images of the cell population captured at a set of time points; and

at the processing system, generating an analysis based upon processing the set of images according to a cell stage classification module configured to automatically identify a cell stage of the set of cell stages, wherein processing the set of images includes;

partitioning each image in the set of images into a set of pixel subsets,

for each pixel subset in the set of pixel subsets for each image, generating a feature vector associated with each pixel subset, thereby producing a set of feature vectors corresponding to the set of pixel subsets for each image in the set of images, and

training a machine learning classifier, with a training dataset, to identify each of the set of cell stages of the cell population captured in the set of images, using the set of feature vectors for each image in the set of images, thereby characterizing the cell population.

9. The method of claim 8, wherein processing the set of images according to the cell stage classification module comprises enabling identification of at least one of:

a single-cell stage, a medium-compaction stage, a full-compaction stage, a dead cell stage, a differentiated cell stage, and a debris stage exhibited by cells of the cell population.

10. The method of claim 9, wherein generating the set of feature vectors for the set of pixel subsets comprises for each of the set of pixel subsets:

constructing a local binary pattern histogram, generating a pixel intensity histogram corresponding to the local binary pattern histogram, and generating the feature vector as a concatenation of the local binary pattern histogram and the pixel intensity histogram.

11. The method of claim 9, wherein training the machine learning classifier comprises processing the set of feature vectors using a support vector machine with a Gaussian kernel.

12. The method of claim 8, further comprising using an output of the cell stage classification module to determine a compaction-related parameter, derived from an amount of cells of the cell population exhibiting a compaction stage across a subset of the set of images.

13. The method of claim 8, further comprising using an output of the cell stage classification module to determine an expansion rate-related parameter, including:

for each image in the set of images and for each of the set of cell stages, determining an area occupied cells exhibiting the cell stage, and weighting the area by a density corresponding to the cell stage.

14. The method of claim 8, further comprising automatically identifying at least one of the set of cell stages within a second cell population, based upon application of the machine learning classifier to a second set of images of the second cell population.

15. The method of claim 8, further comprising:

at the processing system, generating an indication of quality of the cell population based upon the analysis, generating a notification based upon the indication, and providing the indication to an entity associated with the cell population, at a mobile device of the entity.

16. A method for characterizing a cell population including a set of cell subpopulaitons, the method comprising:

at a processing system, receiving image data corresponding to a set of images of the cell population captured at a set of time points;

at the processing system, generating an analysis based upon processing the set of images according to a cell graph representation module configured to characterize geometric and spatial features of the set of cell subpopulations, wherein processing includes;

at a cell cluster segmentation module, segmenting, from at least one image of the set of images, a set of regions corresponding to the set of cell subpopulations of the cell population, each region defined by a set of nodes and a set of edges, and

determining a set of parameters characterizing the cell population based upon an analysis of the set of nodes and the set of edges of at least one of the set of regions.

17. The method of claim 16, wherein processing the set of images according to the cell graph representation module comprises constructing a geometric graph having a weight assigned to each of the set of edges based upon determination of a distance calculated between nodes of the set of nodes, and wherein the weight is determined as an inverse of a geodesic distance between nodes.

18. The method of claim 16, wherein processing the set of images according to the cell graph representation module comprises constructing a spatial graph, having a weight assigned to each node of the set of nodes based upon an associated area of a region of the set of regions corresponding to the node.

19. The method of claim 16, wherein determining the set of parameters comprises determining a morphology-related parameter indicative of colony border spikiness of the cell population, based upon a quantification of weights associated with the set of edges having values below a threshold value.

20. The method of claim 19, further comprising automatically generating an indication of quality of the cell population based upon the morphology-related parameter, generating a notification based upon the indication, and providing the indication to an entity associated with the cell population, at a mobile device of the entity.