Foreground segmentation and nucleus ranking for scoring dual ISH images

A system for segmenting and ranking nuclei in a digital image of a tissue specimen, the tissue specimen stained with at least a first and a second in situ hybridization stain having different colors, the system comprising (i) one or more processors, and (ii) a memory coupled to the processor, the memory to store computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

• segmenting the digital image for generating a foreground segmentation mask, the foreground segmentation mask being selectively indicative of image regions representing nuclei;
• identifying individual nuclei in foreground regions identified by the foreground segmentation mask whose size exceeds a nucleus size threshold value;
• for each of the identified nuclei, performing an initial dot detection routine in a first channel image derived from the digital image for identifying first dots in said identified nucleus and in a second channel image derived from the digital image for identifying second dots in said identified nucleus, each first dot being a blob whose pixel values in the first channel image correlate with an amount of first stain, each second dot being a blob whose pixel intensities in the second channel image correlate with an amount of second stain;
• for each of the identified nuclei, counting a number of first dots and a number of second dots in said identified nucleus;
• and
• ranking a plurality of the identified nuclei in descending order according to a ratio of the first and second dots in each nucleus.