Implement CLI for segment nuclei and generate annotations for them

DigitalSlideArchive / HistomicsTK

A Python toolkit for pathology image analysis algorithms.

https://digitalslidearchive.github.io/HistomicsTK/

Apache License 2.0

376 stars 113 forks source link

Implement CLI for segment nuclei and generate annotations for them #86

Closed cdeepakroy closed 8 years ago

cdeepakroy commented 8 years ago

@cooperlab I want to start working on this issue. Can you point me and @slee172 to the function to use for nuclei segmentation?

Also which of the two color deconvolution (standard/sparse) was working better so i could use that?

cooperlab commented 8 years ago

@cdeepakroy use the standard color deconvolution.

The function is really a sequence: normalize color -> deconvolve to hematoxylin (H) & eosin (E) -> mask -> cLoG response -> max clustering -> feature extraction.

I will write an example script if I get some free time tomorrow.

cdeepakroy commented 8 years ago

This is fixed by PRs #94 and #102.