Image analysis of biological objects

• US 20060280352A1
• Filed: 06/09/2006
• Published: 12/14/2006
• Est. Priority Date: 06/10/2005
• Status: Active Grant

• Current Assignee: Cleveland Clinic Foundation
• Original Assignee: Cleveland Clinic Foundation
• Inventors: Muschler, George F., Bryan, Jason Allen, Powell, Kimerly A.
• Granted Patent: US 7,796,815 B2
• Time in Patent Office: Days
• US Class Current: 382/133
• CPC Class Codes: G06T 2207/10056 Microscopic image
• CPC Class Codes: G06T 2207/20041 Distance transform
• 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 7/11 Region-based segmentation
• CPC Class Codes: G06V 20/695 Preprocessing, e.g. image segmentation

Abstract

Systems and methods are described for performing image analysis. A computer-implemented method for analyzing images may include quantitatively analyzing image data to identify image objects relative to a background portion of the image according to predefined object criteria, the image data including a plurality of image objects that represent objects in a sample distributed across a substrate. The identified image objects are further clustered into groups or colonies of the identified image objects according to predetermined clustering criteria.

Claims

1. A computer-implemented method for analyzing an image, comprising:

quantitatively analyzing image data to identify image objects relative to a background portion of the image according to predefined object criteria, the image data including a plurality of image objects that represent objects in a sample; and

clustering the identified image objects into groups or colonies of the identified image objects according to predefined clustering criteria.

2. The computer-implemented method of claim 1, wherein the quantitatively analyzing further comprises segmenting image objects relative to the background portion of the image and to provide a segmented image.

3. The computer-implemented method of claim 2, wherein the segmenting further comprises applying a global threshold to segment image objects relative to the background portion of the image and to provide a segmented image.

4. The computer-implemented method of claim 2, further comprising removing objects from the segmented image according to object removal criteria to provide a cleaned and segmented image, wherein the identified image objects are clustered according to the cleaned and segmented image.

5. The computer-implemented method of claim 2, wherein the predefined object criteria comprises at least one user-adjustable threshold set according to the type of objects being clustered into the groups or colonies.

6. The computer-implemented method of claim 1, further comprising removing objects from the segmented image according to object removal criteria to provide a cleaned and segmented image, wherein the predefined object criteria further comprises a global intensity threshold and at least one of a size of the identified object, a predetermined intensity level, and a local density of other objects surrounding a given object in the segmented image.

7. The computer-implemented method of claim 1, wherein the quantitatively analyzing further comprises removing image objects from the image corresponding to at least one selected type of object according to predefined object criteria to provide a cleaned and segmented image, the identified image objects being clustered according to the cleaned and segmented image.

8. The computer-implemented method of claim 5, wherein the cleaned and segmented image includes image objects representing cell nuclei.

9. The computer-implemented method of claim 1, wherein the image data further comprises at least first and second image sets acquired for substantially a same region of interest of the sample, the quantitatively analyzing and the clustering being performed on cleaned and segmented image sets for at least one of the first and second image sets, the method further comprising cross-correlating the first and second image sets to substantially align the respective first and second image sets and to provide an aligned image that includes groups or colonies of the image objects identified within the aligned image.

10. The computer-implemented method of claim 9, wherein at least one of the first and second image sets corresponds to image data acquired from the sample representing the presence of a predetermined marker that identifies at least one known characteristic, the method further comprising quantitatively correlating at least two of the first image set, the second image set and the aligned image set to identify at least one characteristic for the groups or colonies of the identified image objects.

11. The computer-implemented method of claim 10, wherein the identified image objects comprise cells, and the at least one characteristic further comprises at least one of a morphologic characteristic of at least one of the cells and the groups or colonies, immunochemical characteristic of the cells, histochemical characteristic of the cells, and characteristics revealed by in situ hybridization of the cells.

12. The computer-implemented method of claim 10, wherein the at least one characteristic further comprises a characteristic of telomeres located within identified groups or colonies of cells.

13. The computer-implemented method of claim 10, wherein the identified image objects comprise cells and the at least one characteristic further comprises at least one of a chemical feature or a morphologic feature in a matrix of materials adjacent the cells.

14. The computer-implemented method of claim 10, wherein the identified image objects comprise cells and the at least one characteristic further comprises at least one of a morphological or chemical feature of at least one of the cells and the groups or colonies that becomes identifiable by employing at least one alternative illumination or imaging condition for each of the at least first and second image sets, by employing different wavelength of illumination for each of the at least first and second image sets, by applying wave phase contrast for each of the at least first and second image sets, and by employing topographic mapping for each of the at least first and second image sets.

15. The computer-implemented method of claim 9, wherein at least one of the first and second image sets corresponds to image data acquired for the sample having been stained to express a predetermined marker, the method further comprising:

cross-correlating the cleaned and segmented image sets for each of the first and second image sets to align the respective image sets for a common region of interest;

adjusting the clustering criteria based on the cross-correlation; and

clustering the identified image objects into groups or colonies according to the adjusted clustering criteria.

16. The computer-implemented method of claim 15, wherein the quantitatively analyzing further comprises applying predetermined object removal criteria to provide a cleaned and segmented image, the clustering further comprises:

identifying groups or colonies of the identified image objects in the cleaned and segmented image; and

determining at least one morphological characteristic of image objects in the identified groups or colonies based on the cross-correlation.

17. The computer-implemented method of claim 16, wherein the clustering further comprises applying cluster removal criteria to a given identified group or colony, the cluster removal criteria varying according to the number of objects within the given identified group or colony.

18. The computer-implemented method of claim 17, wherein the cluster removal criteria comprises a first cluster removal criteria that removes a given group or cluster from the image if the given group or cluster contains less than a predetermined number of objects, the method further comprising applying at least one second cluster removal criteria to remove a given group or cluster from the image according to at least one of a measured dimension or an object density of the given group or cluster.

19. The computer-implemented method of claim 1, wherein the clustering further comprises:

separating the identified groups or colonies into different categories; and

applying the clustering criteria for each given identified group or colony, the clustering criteria varying according to which of the different categories the given group or colony belongs.

20. The computer-implemented method of claim 19, wherein the objects comprise at least one of cells or cell nuclei or cell organelles, the identified groups or colonies being separated into the different categories based on object density.

21. The computer-implemented method of claim 20, further comprising setting at least one clustering parameter as a function of the object density for identifying whether a given group or colony belongs to a first category of the different categories.

22. The computer-implemented method of claim 19, wherein the identified groups or colonies are separated into the different categories as a function of the type of objects in the sample.

23. The computer-implemented method of claim 1, wherein the clustering further comprises:

computing a distance map image for the cleaned and segmented image; and

segmenting the distance map image to cluster objects within a predetermined distance of each other into a respective group or colony of the identified image objects.

24. The computer-implemented method of claim 1, wherein the sample includes a culture of progenitor cells, which correspond to the image objects, the method further comprising characterizing at least one feature of different groups or colonies of the progenitor cells.

25. The computer-implemented method of claim 24, wherein the at least one feature further comprises at least one of a number of the colonies, a number of progenitor cells in each of the colonies, a size of cells in each of the colonies, a size of nuclei within each colony, size of each of the colonies.

26. The computer-implemented method of claim 24, wherein the at least one feature further comprises an indication of whether the progenitor cells express a given marker.

27. The computer-implemented method of claim 26, the given marker comprising at least one of a morphological marker, an immunochemical marker, and a histochemical marker, and in situ hybridization marker.

28. The computer-implemented method of claim 24, wherein the at least one feature comprises at least one of a chemical feature or a morphologic feature in a matrix of materials adjacent the cells.

29. The computer-implemented method of claim 1, wherein the image is a montaged image, the method further comprises selecting a region of interest (ROI) in the montaged image, the ROI establishing boundaries in the image of the sample for performing the quantitative analysis and the clustering.

30. The computer-implemented method of claim 29, wherein the image data is acquired by an image acquisition system for a plurality of subsamples that collectively define the montaged image.

31. The computer-implemented method of claim 1, further comprising storing parameter values that parameterize the computer-implemented method, including at least a portion of the quantitatively analyzing and the clustering, at least a portion of the parameter values being user-selectable.

32. The computer-implemented method of claim 31, further comprising storing a plurality of templates, each of the plurality of templates comprising a set of stored parameter values that parameterize at least a portion of the computer-implemented method, the stored parameter values being adjustable based on information entered via an editor.

33. The computer-implemented method of claim 31, wherein each of a plurality of images are defined by respective sets of the image data, the method further comprising concurrently displaying selected portions of each of the plurality of images according to the parameter values, such that changes in at least one of the parameter values cause corresponding changes in the computer-implemented method, the concurrently displayed selected portions of the plurality of images being modified in response to the changes in at least one of the parameter values.

34. An image analysis system, comprising:

an image acquisition system that provides image data that defines an image having a plurality of image objects that represent corresponding objects in a physical sample; and

an analysis engine that segments image objects relative to image background according to predefined object criteria to provide a segmented image, the analysis engine clusters image objects in the segmented image into groups or colonies of the image objects according to predetermined clustering criteria.

35. The system of claim 34, wherein the image acquisition system further comprises:

a stage that supports the sample;

an optical system that magnifies an image of at least a portion of the sample to a desired level; and

an image capture system that collects an image as an array of pixels, each pixel corresponding to part of the image of at least a portion of the sample, the image data representing at least a light intensity level for each of the pixels.

36. The system of claim 35, wherein the image data comprises a montaged image formed from a plurality of images acquired by the image acquisition system for different parts of the sample, the analysis engine performing background correction and selecting a region of interest (ROI) in the montaged image, the ROI establishing boundaries of the montaged image for performing the segmentation and the clustering.

37. The system of claim 34, wherein the analysis engine is programmed to segment image objects and to provide the segmented image relative to the background portion of the image according to at least one segmentation parameter that parameterizes a segmentation algorithm.

38. The system of claim 37, wherein the analysis engine is programmed to calculate a size for the segmented image objects in the segmented image and to remove objects from the segmented image according to object removal criteria to provide a cleaned and segmented image, the predefined object criteria being set according to the at least one segmentation parameter, the analysis engine clustering the identified image objects in the cleaned and segmented image.

39. The system of claim 34, the image data further comprises at least first and second image sets acquired for the sample, the analysis engine performs the segmentation and clustering on at least one of the first and second image sets to provide segmented and clustered image sets for the at least one of the first and second image sets, the analysis engine cross-correlating the first and second image sets to align spatially the respective first and second image sets to facilitate analysis of the identified image objects in the first and second image sets.

40. The system of claim 39, wherein at least one of the first and second image sets corresponds to image data acquired for the sample having been stained to express at least one predetermined marker, the analysis engine correlating the first and second image sets to identify at least one characteristic for the groups or colonies of the identified image objects based on image objects in the groups or colonies of the identified image objects determined to express the at least one predetermined marker.

41. The system of claim 39, wherein the analysis engine determines the groups or colonies of the identified image objects for the first image set, the second image set corresponding to image data acquired for the sample having been stained to express a predetermined marker, the analysis engine correlating the first image set and the second image set to identify at least one characteristic for the groups or colonies of the identified image objects according to which image objects in the second image represent objects that express the predetermined marker.

42. The system of claim 39, wherein the identified image objects comprise cells, and the at least one characteristic further comprises at least one of a morphologic characteristic of at least one of the cells and the groups or colonies, immunochemical characteristic of the cells, histochemical characteristic of the cells, and characteristics revealed by in situ hybridization of the cells.

43. The system of claim 34, wherein the analysis engine applies cluster removal criteria to each given identified group or colony, the cluster removal criteria varying according to the number of objects within the given identified group or colony.

44. The system of claim 34, wherein the analysis engine provides identifying data that specifically localizes at least one of an object and a group or colony within two regions of interest (ROI) in at least first and second images based on the at least one object and the group or colony exhibiting defined selection criteria.

45. The system of claim 44, wherein the at least one object and the group or colony at the location corresponding to the identifying data represents one or more viable cell.

46. The system of claim 34, wherein the predefined object criteria and the predefined clustering criteria correspond to stored parameters that parameterize at least a portion of an image analysis process employed by the analysis engine, the system further comprising an editor control programmed to change at least one of the parameters.

47. The system of claim 46, further comprising a display that concurrently displays a representation of each of a plurality of images defined by respective sets of the image data, the concurrently displayed representation of each of the plurality of images being modified in response to the changes in the at least one of the parameters that cause changes in the parameterization of the image analysis process employed by the analysis engine.

48. The system of claim 46, further comprising at least one template that includes a set the stored parameters, the at least one template being selectable by the editor control to parameterize at least a portion of the of the image analysis process employed by the analysis engine, the editor control also being programmed to modify at least one of the parameters of the at least one template.