Watson-Aided Non-Linear Problem-Oriented Clinical Visit Preparation on Tablet Computer
P. Ruchikachorn, J. J. Liang, M. Devarakonda, and K. Mueller, Watson-Aided Non-Linear Problem-Oriented Clinical Visit Preparation on Tablet Computer, Visualizing Electronic Health Record Data (EHRVis), IEEE VIS Workshop, Paris, France, 2014.
📊 Slides
🖼 Poster
📄 Preprint
Tablet computers have been growing in popularity and finding their way to workspace. Their ubiquity is attributable to their ease of use and portability, also applicable to hospital work environment. We have designed an Electronic Medical Record (EMR) application for clinical visit preparation task on tablet computers. There are many design challenges due to the limited form factor for both input and output. According to our interviews with physicians, our solution utilizes IBM Watson’s machine learning ability to analyze unstructured clinical notes and create a relation graph of medical entities. With the medical knowledge graph, the system enables easy relation retrieval to show only relevant information as needed in simple and easy-to-understand displays. We informally evaluated our system by walking through sample visit preparation steps from a physician.