基于人工智能病案质控系统质控指标构建及效果评价

Construction and Effect Evaluation on Quality Control Indicators of AI-Based Electronic Medical Record System

研究旨在探讨人工智能电子病历质控系统质量控制指标的构建,着重研究指标的筛选和自然语言结构化处理等。在基于各级主管部门颁布的政策文件、标准规范形成质控指标的初步构建基础上,将这些粗略且多样性的质控指标进行处理,形成更加客观的质量控制指标,包括定量维度和定性维度指标。在此基础上通过对电子病历中的诊断和医嘱等关键信息进行数据清洗、量化归类和自然语言结构化处理,实现病案质控指标的自动分析、错误检测和质量评估。最后通过建立中文病案分词引擎,提升病案的分词效果。基于该策略下构建人工智能的电子病历质量控制系统,能较为准确有效识别电子病历中的潜在质量问题,在病案质控工作中能起到显著的辅助效果。为基于人工智能电子病历质量控制产品的设计提供参考。

This paper aims to explore the construction of quality control indicators for artificial intelligence-based electronic medical record(EMR)quality control systems,specifically focusing on indicator screening and natural language structuring.Building upon the initial construction of quality control indicators based on policy documents and standards issued by competent authorities at all levels,these rough and diverse quality control indicators are processed to form more objective quality control indicators,including quantitative and qualitative dimensions index.On this basis,key information in EMRs,such as diagnoses and prescriptions,undergoes data cleansing,quantification,and natural language structuring,enabling automatic analysis,error detection,and quality assessment of EMR quality control indicators.Additionally,a Chinese EMR word segmentation engine is developed to enhance the segmentation accuracy of medical records.An artificial intelligence electronic medical record quality control system built based on this strategy can more accurately and effectively identify potential quality problems in electronic medical records and can play a significant assisting role in medical record quality control.This research provides a reference for the design of electronic medical record quality control products based on artificial intelligence.