Method to combine brightfield and fluorescent channels for cell image segmentation and morphological analysis using images obtained from imaging flow cytometer (IFC)
View Patent ↗A classifier engine provides cell morphology identification and cell classification in computer-automated systems, methods and diagnostic tools. The classifier engine performs multispectral segmentation of thousands of cellular images acquired by a multispectral imaging flow cytometer. As a function of imaging mode, different ones of the images provide different segmentation masks for cells and subcellular parts. Using the segmentation masks, the classifier engine iteratively optimizes model fitting of different cellular parts. The resulting improved image data has increased accuracy of location of cell parts in an image and enables detection of complex cell morphologies in the image. The classifier engine provides automated ranking and selection of most discriminative shape based features for classifying cell types.
1. A system comprising:
(i) an imaging flow cytometer configured to acquire a plurality of images of a moving cell in a sample, the plurality of images being acquired across multiple different imaging modes including a side scatter image and a fluorescence image, each of the plurality of images of the moving cell being spatially well aligned with each other; and
(ii) a processor coupled in communication with the imaging flow cytometer and configured to receive the plurality of images of the moving cell and execute a classifier engine, the classifier engine configured to:
(A) select a first image from the plurality of images of the moving cell;
(B) segment the first image into subcomponents representing parts of the moving cell;
(C) segment a second image from the plurality of images of the moving cell into at least one subcomponent representing a part of the moving cell to generate a subcomponent mask, wherein the second image is of a different imaging mode than the first image;
(D) spatially correlate the subcomponent mask to the segmented first image;
(E) apply a graph cut segmentation to the second image using the subcomponent mask as a foreground object marker for the segmented first image to generate image data having increased location accuracy of the subcomponents of the moving cell; and
(F) reprocess the first image using the image data generated in (E) to identify cell morphology of the moving cell in the first image.
2. The system of claim 1 , wherein
the first image is a brightfield image and
the second image is a fluorescent image.
3. The system of claim 1 , wherein
the second image is one of a plurality of different images,
wherein the different images were obtained by different imaging modes, and
the classifier engine is configured to iteratively perform (C), (D), and (E) for each of the plurality of different images.
4. The system of claim 3 , wherein
the plurality of different images were obtained by imaging different fluorescent channels.
5. The system of claim 1 , wherein
the classifier engine is further configured to characterize the moving cell as a live cell or a dead cell.
6. The system of claim 5 , wherein the classifier engine is further configured to
(i) output the characterization of the moving cell as a live cell or a dead cell, and
(ii) if the moving cell is a live cell, output a cell type classification for the live cell based on the identified cell morphology.
7. The system of claim 1 , wherein the classifier engine is further configured to output a cell type classification for the moving cell based on the identified cell morphology.
8. The system of claim 1 , wherein
the moving cell is a sperm cell.
9. The system of claim 8 , wherein
the first image is a brightfield image of the sperm cell, and
the second image comprises a fluorescent image of the sperm cell.
10. A non-transitory computer-readable medium having instructions stored thereon that are executable by a computing device to perform operations comprising:
(A) select a first image from a plurality of images of a moving cell in a sample acquired by an imaging flow cytometer, the plurality of images being acquired across multiple different imaging modes including a side scatter image and a fluorescence image;
(B) segment the first image into subcomponents representing parts of the moving cell;
(C) segment a second image from the plurality of images of the moving cell into at least one subcomponent representing a part of the moving cell to generate a subcomponent mask, wherein the second image is of a different imaging mode than the first image;
(D) spatially correlate the subcomponent mask to the segmented first image;
(E) apply a graph cut segmentation to the second image using the subcomponent mask as a foreground object marker for the segmented first image to generate improved image data having increased location accuracy of the subcomponents; and
(F) reprocess the first image using the image data generated in (E) to identify cell morphology of the moving cell.
11. The non-transitory computer-readable medium of claim 10 , wherein the second image is one of a plurality of different images, wherein the different images were obtained by different imaging modes, and the instructions stored on the non-transitory computer-readable medium are executable by the computing device to perform to perform operations comprising iteratively performing (C), (D), and (E) for each of the plurality of different images.
12. The non-transitory computer-readable medium of claim 10 , wherein of the instructions stored on the non-transitory computer-readable medium are executable by the computing device to perform operations comprising
characterizing the moving cell as a live cell or a dead cell.
13. The non-transitory computer-readable medium of claim 12 , wherein of the instructions stored on the non-transitory computer-readable medium are executable by the by computing device to perform operations comprising
(i) outputting the characterization of the moving cell as a live cell or a dead cell, and
(ii) if the moving cell is a live cell, outputting a cell type classification for the live cell based on the identified cell morphology.
14. The non-transitory computer-readable medium of claim 10 , wherein the instructions stored on the non-transitory computer-readable medium are executable by the computing device to perform operations comprising
outputting a cell type classification for the moving cell based on the identified cell morphology.
15. The non-transitory computer-readable medium of claim 10 , wherein
the moving cell is a sperm cell.
16. The non-transitory computer-readable medium of claim 15 , wherein
the first image is a brightfield image of the sperm cell, and
the second image comprises a fluorescent image.
17. A method comprising:
(A) receiving, with a computer system, a plurality of images of a moving cell in a sample obtained by an imaging flow cytometer, the plurality of images being acquired across multiple imaging modes including a side scatter image and a fluorescence image, each of the plurality of images being spatially well aligned with each other;
(B) selecting, with the computer system, a first image from the plurality of images of the moving cell;
(C) segmenting, with the computer system, the first image into subcomponents the first image into subcomponents representing parts of the moving cell;
(D) segmenting, with the computer system, a second image from the plurality of images of the moving cell into at least one subcomponent representing a part of the moving cell to generate a subcomponent mask, wherein the second image is of a different imaging mode than the first image;
(E) spatially correlating, with the computer system, the subcomponent mask to the segmented first image;
(F) applying, with the computer system, a graph cut segmentation to the second image using the subcomponent mask as a foreground object marker for the segmented first image to generate image data having increased location accuracy of the subcomponents of the moving cell; and
(G) reprocessing, with the computer system, the first image using the image data generated in (F) to identify cell morphology of the moving cell in the first image.
18. The method of claim 17 , further comprising:
(H) outputting, with the computer system, a cell type classification for the cell based on the identified cell morphology.
19. The method of claim 17 ,
wherein the second image is one of a plurality of different images,
wherein the different images were obtained by different imaging modes, and
the method further comprises iteratively performing, with the computer system, (D), (E), and (F) for each of the plurality of different images.