Development and Evaluation of Natural Language Processing Methods for Extracting Key Melanoma Pathology Concepts

Abstract

This study presents the development and evaluation of an annotation schema and rule-based natural language processing (NLP) system for extracting key melanoma pathology concepts from surgical pathology reports. Achieving high precision and recall, our system addresses melanoma’s complex staging criteria for use in downstream staging and cohort recruitment operational needs.