Systems and methods for processing electronic images for auto-labeling for computational pathology

A system for processing electronic medical images, the system comprising:

• at least one memory storing instructions;
• and
• at least one processor configured to execute the instructions to perform operations comprising:
• receiving, by a first machine learning system, a plurality of electronic medical images, each of the plurality of electronic medical images being associated with one or more subcategories;
• filtering the plurality of electronic medical images to determine electronic medical images having only a single subcategory present in a particular electronic medical image, the electronic medical images having only the single subcategory being designated as pure category images;
• training one or more additional machine learning systems, each of the one or more additional machine learning systems being previously trained on a separate pure category;
• generating, by each of the one or more additional machine learning systems, a segmentation mask;
• and
• training a second machine learning system using the segmentation masks generated by the one or more additional machine learning systems, wherein the second machine learning system is trained to identify the one or more subcategories.

A non-transitory computer-readable medium storing instructions that, when executed by a processor, perform operations for processing electronic medical images, the operations comprising:

• receiving, by a first machine learning system, a plurality of electronic medical images, each of the plurality of electronic medical images being associated with one or more subcategories;
• filtering the plurality of electronic medical images to determine electronic medical images having only a single subcategory present in a particular electronic medical image, the electronic medical images having only the single subcategory being designated as pure category images;
• training one or more additional machine learning systems, each of the one or more additional machine learning systems being previously trained on a separate pure category;
• generating, by each of the one or more additional machine learning systems, a segmentation mask;
• and
• training a second machine learning system using the segmentation masks generated by the one or more additional machine learning systems, wherein the second machine learning system is trained to identify the one or more subcategories.