软件导刊2026，Vol.25Issue(1)：32-38,7.DOI:10.11907/rjdk.241764
融合预训练语言模型的冠心病专病库建设及应用
Construction and Application of Coronary Heart Disease Specialized Knowledge Base by Integrating Pre-trained Language Models
摘要
Abstract
The efficiency and accuracy of data processing in coronary heart disease specialty databases play a crucial role in clinical research and decision-making.Therefore,building a highly efficient and reliable disease-specific database is essential to support clinical researchers in rapidly acquiring key information,optimizing treatment decisions,and improving overall patient care quality.Based on the Clinical-BERT+Bi-LSTM+CRF model and combined with data platform and Enterprise Service Bus(ESB)integration,we optimized the data processing workflow of the specialty database.Experimental results demonstrate that the data extraction time was reduced by an average factor of 36(t=115.96,P<0.01),and the accuracy of structured data increased by 6.9％(χ²=222.41,P<0.01).These findings indicate that the proposed op-timization effectively enhances the efficiency and accuracy of data processing in the coronary heart disease specialty database,providing ro-bust data support for clinical research and decision-making in coronary heart disease.关键词
冠心病/专病库建设/数据处理/预训练模型Key words
coronary heart disease/specialized disease database construction/data processing/pre-trained model分类
信息技术与安全科学引用本文复制引用
薛扬,侯旭敏..融合预训练语言模型的冠心病专病库建设及应用[J].软件导刊,2026,25(1):32-38,7.基金项目
上海市城市数字化转型专项基金项目(202301002) （202301002）
上海市卫生健康委员会智慧医疗专项研究项目(2025ZHYL011) （2025ZHYL011）
