Cross-frame object reconstruction for image-based cytology applications
View Patent ↗Methods, computer software, and systems for analyzing a biological specimen (e.g., a cytological specimen) on a slide are provided. Magnified image data frames of the biological specimen are taken at different locations on the slide. An object that is not entirely contained within at least one of the image data frames is identified, and complementary portions of the object respectively located in different ones of the image data frames are matched. A fully integrated object is created from the matched object portions. Attributes of the integrated object are then analyzed.
1. An automated method for analyzing a biological specimen on a microscope slide, comprising:
receiving magnified image data frames of the biological specimen taken at different locations on the slide;
using a processor to identify an object that is not entirely contained within at least one of the image data frames;
reconstructing the identified object by matching complementary portions of the identified object in different image data frames, thereby
creating a fully integrated object that corresponds to the identified object; and
analyzing attributes of the integrated object.
2. The method of claim 1 , wherein the biological specimen is a cytological specimen.
3. The method of claim 1 , further comprising identifying the integrated object as an object of interest based on the analysis.
4. The method of claim 1 , further comprising classifying the biological specimen based at least partially upon the analyzed attributes of the integrated object.
5. The method of claim 1 , wherein the integrated object is created by digitally stitching the complementary object portions together.
6. The method of claim 1 , wherein identification of an object that is not entirely contained within at least one of the image data frames comprises identifying an object that intersects two or more adjacent borders of the respective image data frames.
7. The method of claim 1 , further comprising:
identifying an object entirely contained within any one of the image data frames; and
analyzing attributes of the identified entirely contained object.
8. The method of claim 1 , further comprising locating the integrated object within a coordinate system correlated to the slide.
9. A non-transitory computer readable medium for analyzing a biological specimen on a microscope slide, containing instructions, which when executed by a computer processor, cause the computer to perform the acts of:
receiving magnified image data frames of the biological specimen taken at different locations on the slide;
identifying an object that is not entirely contained within at least one of the magnified image data frames taken of the biological specimen;
reconstructing the identified object by matching complementary portions of the identified object in different image data frames, thereby
creating a fully integrated object that corresponds to the identified object; and
analyzing attributes of the integrated object.
10. The non-transitory computer readable medium of claim 9 , wherein the biological specimen is a cytological specimen.
11. The non-transitory computer readable medium of claim 9 , wherein the instructions, when executed, cause the computer to perform the further act of identifying the integrated object as an object of interest based on the analysis.
12. The non-transitory computer readable medium of claim 9 , wherein the instructions, when executed, cause the computer to perform the further act of classifying the biological specimen based at least partially upon the analyzed attributes of the integrated object.
13. The non-transitory computer readable medium of claim 9 , wherein the integrated object is created by digitally stitching the complementary object portions together.
14. The non-transitory computer readable medium of claim 9 , wherein identification of an object that is not entirely contained within the at least one of the image data frames comprises identifying an object that intersects two or more adjacent borders of the respective image data frames.
15. The non-transitory computer readable medium of claim 9 , wherein the instructions, when executed, cause the computer to perform the further acts of:
identifying an object entirely contained within any one of the image data frames; and
analyzing attributes of the identified object.
16. The non-transitory computer readable medium of claim 9 , wherein the instructions, when executed, cause the computer to perform the further act of locating the integrated object within a coordinate system correlated to the slide.