• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IEEE International Conference on Bioinformatics and Biomedicine >Clinical Entities Association Rules (CLEAR): Untangling Clinical Notes in Electronic Health Records
【24h】

Clinical Entities Association Rules (CLEAR): Untangling Clinical Notes in Electronic Health Records

机译：临床实体协会规则（CLEAR）：整理电子健康记录中的临床注释

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Clinical notes make a large part of Electronic Health Records (EHR). While applying large and automated analysis on unstructured clinical notes can be much more complex than on structured data, it can potentially provide better support and information for medical decision-making. In this paper, we propose a methodology to identify clinical entities and their associations that are widely exist across multiple types of clinical notes. The focus of this research is to discover patterns of significant relationships amongst clinical entities, such as treatments, tests, and diagnoses, particularly in clinical notes within EHR. Extracting clinical entities associations from clinical notes is a crucial task for many applications in clinical and public health informatics. We utilize population-level data to discover knowledge sets which can help health providers during their assessments for patients. We incorporate Natural Language Processing, text mining, and weighted association rule mining techniques in building a pipeline to transform clinical notes into knowledge set. Our preliminary results with proposed weighted clinical transactional itemset representation show promising results in identifying strongly related clinical entities. We evaluated our results with labeled data from a healthcare professional with an accuracy of 98.3% on the relevancy task and 54.4% of rules are labeled as interesting and worth further clinical investigation. We discovered patterns of patients developing secondary problems during their length of stay at a hospital.
机译：临床笔记在电子健康记录（EHR）中占很大一部分。尽管对非结构化临床笔记进行大型自动化分析可能比对结构化数据进行复杂得多，但它可能为医疗决策提供更好的支持和信息。在本文中，我们提出了一种方法，用于识别在多种类型的临床笔记中广泛存在的临床实体及其关联。这项研究的重点是发现临床实体之间重要关系的模式，例如治疗，测试和诊断，尤其是在EHR中的临床笔记中。从临床笔记中提取临床实体关联对于临床和公共卫生信息学中的许多应用而言都是至关重要的任务。我们利用人群水平的数据来发现知识集，这些知识集可以帮助医疗服务提供者对患者进行评估。我们将自然语言处理，文本挖掘和加权关联规则挖掘技术结合在一起，以构建将临床笔记转换为知识集的管道。我们的初步结果与拟议的加权临床交易项目集表示法相结合，显示出在确定高度相关的临床实体方面的有希望的结果。我们使用来自医疗保健专业人员的标签数据评估了我们的结果，该数据在相关任务上的准确度为98.3％，而54.4％的规则被标签为有趣且值得进一步临床研究。我们发现患者在住院期间出现继发性问题的模式。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号