Bring Biomedical Ontologies to Personalized Healthcare: A Smart Inquiry Framework

摘要

High-quality inquiry (i.e. history taking) is the foundation of accurate diagnosis. Traditionally, inquiry occurs between doctors and patients, requiring lots of time and patience, but often receiving a lot of complaints. This paper proposes a smart inquiry framework, aiming for interacting with patient readily by mining personalized questions from biomedical big data. In other words, the patient can tell his feelings at any time and place, then the system will ask you simple questions, just like an experienced and very kind doctor. What you should do is answering Yes or No. Finally, you will get a preliminary suggestion. There are two key technologies in this paper: 1) training a probability graph from the medical big data, as the whole framework's knowledge base, 2) based on the recommendation algorithm, designing an automatic question answering system, to guide the patients speaking out other more symptoms from the initial feeling. The new proposed framework can serve everyone with low cost and high efficiency, helping them to dynamically sort out and record the state of the body. This will play a fundamental role in the field of personalized healthcare.