Automating assessment of silver-enhanced in situ hybridization for evaluation of cancer properties

Abstract

To correctly assess breast cancer properties, doctors have to compare and evaluate histopathological slides stained for the presence of certain proteins; in the event that the initial evaluation is inconclusive, additional assessment is done. The assessment is necessary to accurately determine the type of cancer and select fitting treatment that increases the likelihood of the patient’s recovery. In this paper, approaches to automate different steps of this assessment are explored with the immediate goal of implementing an end-to-end algorithm pipeline capable of performing this task with minimal human input, with the potential goal of incorporating this pipeline into an existing larger slide processing software.