乳腺癌和宫颈癌的知识图谱构建
Knowledge Graph Construction For Breast and Cervical Cancer
吴庆军 1蒲立新 2何明杰 3周滨 4张勇4
作者信息
- 1. 深圳市龙华区妇幼保健院,广东省深圳市,518110
- 2. 深圳市龙华区数据智能与认知计算重点实验室,广东省深圳市,518110;电子科技大学(深圳)高等研究院,广东省深圳市,518110;电子科技大学,四川省成都市,611731
- 3. 深圳成电金盘健康数据技术有限公司,广东省深圳市,518110
- 4. 成都成电金盘健康数据技术有限公司,四川省成都市,611731
- 折叠
摘要
本研究提出了一种基于乳腺癌和宫颈癌医学文本进行知识图谱构建的工作流程.首先根据乳腺癌和宫颈癌的临床诊疗指南、专家共识定义本体,然后通过训练的实体关系抽取模型对乳腺癌和宫颈癌电子病历中的非结构化文本数据进行知识抽取,最后构建出乳腺癌和宫颈癌知识图谱.初步实现了乳腺癌和宫颈癌知识图谱的可视化呈现及基于知识图谱的病历高效检索,为乳腺癌和宫颈癌医学文本数据的组织、存储及应用提供了新的模式支撑,也为临床和科研提供了数据支持.
Abstract
This study proposes a workflow for knowledge graph construction based on breast and cervical cancer medical texts.Firstly,the ontology is defined according to the clinical diagnosis and treatment guidelines and expert consensus of breast cancer and cervical cancer,and the unstructured text data in electronic medical records of breast cancer and cervical cancer are extracted through the trained entity-relationship extraction model,and then the knowledge graph of breast cancer and cervical cancer is constructed,which initially realizes the visual presentation of the knowledge graph of breast cancer and cervical cancer and the efficient retrieval of medical records based on the knowledge graph,and provides a new model support to organize,store and apply medical text data of breast cancer and cervical cancer,and also provides data support to clinical and scientific research.
关键词
知识图谱/乳腺癌/宫颈癌/关系抽取Key words
knowledge graph/breast cancer/cervical cancer/relation extraction引用本文复制引用
基金项目
深圳市科技基础研究计划(JCYJ20210324134405016)
四川省区域创新合作项目(2022YFQ0097)
出版年
2024
NETL
NSTL
万方数据