News & Events

Prof. Sun’s Paper in JAMIA, A Leading Journal In Clinical Informatics

Assistant Professor Dr. Yingcheng Sun’s research paper, titled “EvidenceMap: A Three-Level Knowledge Representation for Medical Evidence Computation and Comprehension,” has been accepted by the Journal of the American Medical Informatics Association (JAMIA) as co-first author. As a top-tier journal in the clinical informatics field, JAMIA boasts an impact factor of 7.942 for 2022-2023.

In this paper, Dr. Sun and his collaborators introduce an innovative knowledge representation method for clinical evidence and develop a pipeline for extracting such knowledge. High-quality medical evidence primarily exists in the form of clinical publications, and their free-text format poses significant information overload challenges for clinical practitioners, leading to underutilization of evidence in practice. The newly proposed evidence representation, EvidenceMap, comprises three levels of abstraction: Medical Evidence Entity, Proposition, and Map, which represent the hierarchical structure of medical evidence composition. A gold standard dataset of annotated randomized controlled trial (RCT) abstracts has been provided, accompanied by a natural language processing (NLP) pipeline for converting free-text RCT evidence in PubMed into the EvidenceMap representation. This representation serves as an interoperable format for enhanced evidence retrieval and synthesis and offers an interpretable representation for efficiently understanding RCT findings.

More of Dr. Sun’s research can be found on his research webpage: