Research on efficient compilation and promotion of disease knowledge base

改善临床实践的信息系统需要一个疾病知识库，其中包含每种疾病可能的临床结果和实验室检查结果。然而，编制这样的知识库需要相当多的时间和金钱。因此，本研究建议将人工编制的重大疾病数据与自动生成的罕见病数据进行混合，以确保必要的质量，以便覆盖更广的范围，以期高效地编制实用的疾病知识库。此外，针对知识库的质量控制，建立了一个具有反馈机制的诊断决策支持系统，以收集使用信息，从而有效地提高疾病知识库的质量。该研究产生了具有实用质量的疾病知识库和诊断决策支持系统，以及从医学文献中自动获取疾病知识的各种知识。

Improvement of information systems for clinical practices requires a disease knowledge base that contains possible clinical findings and laboratory examination results for each disease. However, compilation of such knowledge base necessitates considerable time and money. Accordingly, this study proposed to hybridize the data for major diseases, manually compiled to guarantee necessary quality, and the data for rare diseases, automatically generated for broader coverage, in the aim of efficient compilation of practical disease knowledge base. Additionally, toward quality control of the resulting knowledge base, a diagnostic decision support system is built with a feedback mechanism to collect usage information to efficiently improve the quality of the disease knowledge base. The research yielded disease knowledge base and a diagnostic decision support system with practical quality, as well as various knowledge about automatic acquisition of disease knowledge out of medical literature.