Classification based on characterization analysis methods and systems
A method at a computing device for classifying elements within an input, the method including breaking the input into a plurality of patches; for each patch: creating a vector output; applying a characterization map to select a classification bin from a plurality of classification bins; and utilizing the selected classification bin to classify the vector output to create a classified output; and compiling the classified output from each patch.
1. A method at a computing device for classifying elements within image data of a histopathology slide, the method comprising:
breaking the image data into a plurality of patches, each patch of the plurality of patches corresponding to a portion of the image data;
for each patch:
creating a vector output;
applying a density map to obtain a cell density;
selecting a classification bin from a plurality of classification bins according to the cell density, wherein each of the plurality of classification bins corresponds to a different density range for the element; and
utilizing the selected classification bin to classify the vector output to create a classified output; and
compiling the classified output from each patch.
2. The method of claim 1 , further comprising applying a mask to the image data prior to creating the vector input.
3. The method of claim 1 , wherein the creating the vector output is performed using a convolutional neural network.
4. The method of claim 1 , wherein each classification bin from the plurality of classification bins contains linear and non-linear classifiers.
5. The method of claim 4 , wherein the linear and non-linear classifiers in each classification bin are optimized based on the density map.
6. The method of claim 1 , wherein the image data comprises an image of a tumor.
7. The method of claim 6 , wherein each classification bin contains linear and non-linear classifiers to identify between different types of cancer cells at a cell density level associated with the classification bin.
8. The method of claim 7 , wherein the tumor is a non-small cell lung cancer and the different types of cancer cells include adenocarcinoma and squamous cell carcinoma cells.
9. The method of claim 8 , further comprising applying a tumor mask to the image data.
10. The method of claim 7 , wherein the tumor is a breast cancer and the different types of cancer cells can be divided into two or more of: Luminal A, Luminal B, Triple-negative/basal-like, and HER2-enriched cells.
11. A computing device for classifying elements within image data of a histopathology slide, the computing device comprising a processor configured to execute instructions to:
break the image data into a plurality of patches, each patch of the plurality of patches corresponding to a portion of the image data;
for each patch:
create a vector output;
apply a density map to obtain a cell density;
select a classification bin from a plurality of classification bins according to the cell density, wherein each of the plurality of classification bins corresponds to a different density range for the element; and
utilize the selected classification bin to classify the vector output to create a classified output; and
compile the classified output from each patch.
12. The computing device of claim 11 , wherein the computing device is further configured to utilize a mask for the image data prior to creating the vector input.
13. The computing device of claim 11 , wherein the vector output is created using a convolutional neural network.
14. The computing device of claim 11 , wherein each classification bin from the plurality of classification bins contains linear and non-linear classifiers.
15. The computing device of claim 11 , wherein the linear and non-linear classifiers in each classification bin are optimized based on the density map.
16. The computing device of claim 11 , wherein the image data comprises an image of a tumor.
17. The computing device of claim 16 , wherein each classification bin contains a support vector machine to identify between different types of cancer cells at a cell density level associated with the classification bin.
18. The computing device of claim 17 , wherein the tumor is a non-small cell lung cancer and the different types of cancer cells include adenocarcinoma and squamous cell carcinoma cells.
19. The computing device of claim 18 , wherein the processor is further configured to execute instructions to apply a tumor mask to the whole slide image.
20. The computing device of claim 17 , wherein the tumor is a breast cancer and the different types of cancer cells can be divided into two or more of: Luminal A, Luminal B, Triple-negative/basal-like, and HER2-enriched cells.
21. A non-transitory computer readable medium for storing instruction code for classifying elements within an input, which, when executed by a processor of a computing device cause the computing device to:
break the image data into a plurality of patches, each patch of the plurality of patches corresponding to a portion of the image data;
for each patch:
create a vector output;
apply a density map to obtain a cell density;
select a classification bin from a plurality of classification bins according to the cell density, wherein each of the plurality of classification bins corresponds to a different density range for the element; and
utilize the selected classification bin to classify the vector output to create a classified output; and
compile the classified output from each patch.