Filling the gap between microscopic and automated analysis of the tumor-stroma ratio

Abstract

Determining the tumor-stroma ratio (TSR) using a conventional microscope is an easy to apply and highly reproducible method. Due to digitalization in the pathology workflow, the demand for automated analysis of the TSR method is rising. However, the process of automation is rather time consuming and needs validation before implementation in daily practice. In addition, international studies ask for exchange of digital images instead of the actual slides. This calls for an alternative digital scoring method. This brief report describes the pitfalls of analyzing the TSR using digital images and proposes essential adaptations to create a standardized and reproducible scoring protocol. By using a circular annotation to mimic the microscopic method, these pitfalls can be avoided. Scoring the TSR digitally using a circular annotation does not take much additional effort compared to the microscopic method. When a fixed size of the annotation is saved, new cases can be scored in less than two minutes. With this brief report we propose an adjusted method for scoring the TSR on digital images to fill the gap between microscopically and automated scoring of the TSR. In addition, it opens the opportunity for application in daily diagnostics.