Real-time whole slide pathology image cell counting
Techniques are provided for determining a cell count within a whole slide pathology image. The image is segmented using a global threshold value to define a tissue area. A plurality of patches comprising the tissue area are selected. Stain intensity vectors are determined within the plurality of patches to generate a stain intensity image. The stain intensity image is iteratively segmented to generate a cell mask using a local threshold value that is and gradually reduced after each iteration. A chamfer distance transform is applied to the cell mask to generate a distance map. Cell seeds are determined on the distance map. Cell segments are determined using a watershed transformation, and a whole cell count is calculated for the plurality of patches based on the cell segments. A client device may be configured for real-time cell counting based on the whole cell count.
1. An object-counting computing apparatus comprising:
at least one computer-readable memory storing object-counting software instructions;
at least one processor coupled with the at least one computer-readable memory and that operates as an object-counting engine upon execution of the object-counting software instructions to:
obtain, in the at least one computer-readable memory, a whole image comprising images of objects;
segment the whole image into image areas based on a global threshold;
select a plurality of patches associated with the image areas;
generate an intensity image based on an optical density of each patch of the plurality of patches;
generate at least one object mask with at least one object seed by segmenting the intensity image based on an iteratively reducing threshold;
generate a distance map based on at least one transformation of the at least one object mask;
distinguish object segments within the plurality of patches based on at least one transformation of the distance map and at least one object seed; and
configure a device to display an object count for at least one patch in the plurality of patches based on counting the object segments.
2. The apparatus of claim 1 , wherein the global threshold is based on at least one of an Otsu filter and a Yen filter.
3. The apparatus of claim 2 , wherein the global threshold is based on a minimum of the Otsu filter and the Yen filter.
4. The apparatus of claim 1 , wherein the iteratively reducing threshold is related to a known object size.
5. The apparatus of claim 4 , wherein the known object size is between about 5 microns and about 20 microns.
6. The apparatus of claim 1 , wherein the optical density comprises an RGB density.
7. The apparatus of claim 1 , wherein the plurality of patches is associated with at least one selected region of interest.
8. The apparatus of claim 7 , wherein the object count relates to the at least one selected region of interest.
9. The apparatus of claim 7 , wherein the at least one selected region of interest is based on a set of points that form a closed curve in the whole image.
10. The apparatus of claim 9 , wherein an object is determined based on a point that forms a closed curve.
11. The apparatus of claim 1 , wherein the device is configured to display the object count in real-time.
12. The apparatus of claim 1 , wherein the object count comprises a whole object count for the whole image.
13. The apparatus of claim 1 , wherein the whole image comprises a whole-slide image.
14. The apparatus of claim 1 , wherein at least some the patches of the plurality of patches are selected randomly.
15. The apparatus of claim 1 , wherein the distance map indicates a distance to a nearest background point.
16. The apparatus of claim 1 , wherein the at least one object seed is determined by gaussian smoothing the distance map.
17. The apparatus of claim 1 , wherein the at least one object seed relates to a maximum height marker in the distance map.
18. The apparatus of claim 1 , where in the object count comprises an updated object count based a change in a region of interest.
19. A non-transitory, computer-readable medium storing instructions which, when executed by at least one processor and memory, cause the at least one processor to perform operations comprising one or more steps to:
obtain, in at least one computer-readable memory, a whole image comprising images of objects;
segment the whole image into image areas based on a global threshold;
select a plurality of patches associated with the image areas;
generate an intensity image based on an optical density of each patch of the plurality of patches;
generate at least one object mask with at least one object seed by segmenting the intensity image based on an iteratively reducing threshold;
generate a distance map based on at least one transformation of the at least one object mask;
distinguish object segments within the plurality of patches based on at least one transformation of the distance map and at least one object seed; and
configure a device to display an object count for at least one patch in the plurality of patches based on counting the cell segments.
20. A computerized method comprising:
obtaining, in at least one computer-readable memory, a whole image comprising images of objects;
segmenting the whole image into image areas based on a global threshold;
selecting a plurality of patches associated with the image areas;
generating an intensity image based on an optical density of each patch of the plurality of patches;
generating at least one object mask with at least one object seed by segmenting the intensity image based on an iteratively reducing threshold;
generating a distance map based on at least one transformation of the at least one object mask;
distinguishing object segments within the plurality of patches based on at least one transformation of the distance map and at least one object seed; and
configuring a device to display an object count for at least one patch in the plurality of patches based on counting the object segments.