Ontology-Driven Weak Supervision for Clinical Entity Classification in Electronic Health Records

We will review a framework for weakly supervised entity classification using medical ontologies and expert-generated rules. Our approach, unlike hand-labeled notes, is easy to modify, while offering performance comparable to learning from manually labeled training data. We will discuss rapid adaptation of Trove to analyze the records of patients visiting Stanford Health Care for continuous symptom profiling of patients screened for SARS-CoV-2 as well as its use to perform post-market surveillance of medical implants.