摘要
针对单层关联规则在精神障碍患者药品不良反应(Adverse Drug Reaction,ADR)风险检测与预警中存在的不足,提出一种基于多层关联规则和概念层次树的ADR风险检测与预警系统框架。该系统结合精神障碍患者ADR报告实际数据,通过领域知识构建概念层次树,并采用多层关联规则挖掘方法找出ADR临床症状与不同相关因素(用药人群、用药情况等)组合之间较高概念层次的强关联规则。结果表明,与单层关联规则相比,多层关联规则能够为ADR风险检测与预警提供更好的临床用药辅助决策。
Aiming at the shortcomings of single-level association rules in the risk detection and early warning of adverse drug reactions(ADR)in patients with mental disorders,the framework of ADR risk detection and warning system for patients with mental disorders based on concept hierarchy tree and multiple-level association rule mining was proposed.The system combines the actual data of ADR reports from patients with mental disorders,constructs a conceptual hierarchy tree through domain knowledge,and uses a multiple-level association rule mining method to discover strong association rules at higher conceptual levels between the medication population,the medication situation and the clinical symptoms of ADR.The results show that compared with single-level association rules,multi-level association rules can provide better clinical medication aided decision-making for ADR risk detection and early warning.
作者
叶明全
苏洋
童九翠
Ye Mingquan;Su Yang;Tong Jiucui(School of Medical Information,Wannan Medical College,Wuhu 241000;Research Center of Health Big Data Mining and Applications,Wannan Medical College,Wuhu 241000;Yijishan Hospital,Wannan Medical College,Wuhu 241000)
出处
《池州学院学报》
2020年第3期23-26,共4页
Journal of Chizhou University
基金
国家自然科学基金项目(61672386)
教育部人文社会科学研究规划基金项目(16YJAZH071)
安徽省自然科学基金项目(1708085MF142)。
关键词
药品不良反应
ADR预警
数据挖掘
多层关联规则
概念层次树
Adverse Drug Reaction
ADR Early Warning
Data Mining
Multi-Level Association Rule
Concept Hierarchy Tree
