数据挖掘在医学影像储存与传输系统结构化报告中的应用

Application of data mining technology in the picture archiving and communication system of structured report

目的:应用数据挖掘技术的理论和方法,挖掘基于医学影像存储与传输系统的结构化报告归档数据库诊断信息资料,扩充结构化报告的影像诊断教学知识库和报告模板库。方法:结构化报告系统基于医学影像存储与传输系统的体系结构,报告的存储与传输符合医学数字影像与通讯标准。应用数据挖掘技术,实现对结构化报告文本数据资料的提取、分类和编码。结果:挖掘结构化报告归档数据库中诊断报告的数据资料,形成带有12个节点的疾病知识库决策树,在知识库中存储。按诊断报告格式,将疾病对应的条件属性值转换为报告模板,形成结构化报告的诊断报告模板库,使影像诊断医生能够依据可能的诊断结果,利用操作菜单,从中选择那些符合观察结果的影像特征值。结论:数据挖掘技术在结构化报告系统中的应用有利于提高影像诊断报告质量和效率。

AIM： Theory and method of data mining technology are applied to mine diagnosis information data in structured report （SR） based on picture archiving and communication system （PACS） and extend assistant diagnosis knowledge base and report model base of SR. METHODS： SR system was based on PACS and accorded with digital imaging and communications in medicine （DICOM） standard. Extraction, classification and encoding of SR text data were performed with data mining technology. RESULTS： To mine date in archiving database of SR and form the decision tree in disease knowledge base with 12 nodal points, which stored in knowledge base. On the basis of diagnosis report form, condition attribute value relating to diseases were changed into report model and formed into diagnosis report model base in SR. Imaging doctor could chosen imaging characteristic value meeting observed result according to possible diagnosis result with operation menu. CONCLUSION： Data mining technology in SR is in favor of improving the quality and efficiency of imaging diagnosis report.