| 摘要: |
| ［摘要］ 目的 探讨基于人工智能（AI）差分进化算法的患者数据实时质量控制智能监控平台在10个生化免疫项目质量控制中的应用价值。方法 基于患者数据的实时质量控制（PBRTQC）智能监控平台自动收集2023年1月至12月广西壮族自治区人民医院检验科10个生化免疫项目的检验结果，采用差分进化算法优化指数加权移动均值法（EWMA）程序的计算数量、步长、截断限和控制限，使用2024年1月至5月数据进行模型验证，并于2024年6月至10月开展真实环境运行评价。运用相关质控规则，评价PBRTQC智能监控平台在这10个项目检测质量控制中的应用价值。结果 通过PBRTQC智能监控平台选出最优程序相关参数。在这10个项目检测中，当出现由于试剂剩余量过低、试剂开瓶时间过长和搅拌棒携带污染引起检测系统性能变化时，智能监控平台能正确识别并及时发出报警提示。结论 PBRTQC智能监控平台可实时监控10个生化免疫项目检测质量风险，并精准识别系统误差。 |
| 关键词: 生化免疫项目 人工智能 基于患者数据的实时质量控制 指数加权移动均值法 |
| DOI：10.3969/j.issn.1674-3806.2026.07.03 |
| 分类号:R 446 |
| 基金项目:中央引导地方科技发展资金项目（编号：桂科ZY24212050）；广西壮族自治区卫生健康委自筹经费科研课题（编号：Z-A20240120） |
|
| Application value of an AI-based differential evolution algorithm-driven patient-based real-time quality control intelligent monitoring platform in quality management of clinical chemistry and immunology assays |
|
Luo Changliang1, Su Hangjiu1, Huang Xiuli1, Yang Wenhui1, Huang Xiong2, Xiao Yu1, Liang Li1, Chen Jin1, Yuan Yulin1,2*, Ning Leping1
|
|
1.Department of Clinical Laboratory, the People′s Hospital of Guangxi Zhuang Autonomous Region, Nanning 530021, China; 2.Guangxi Clinical Center for Laboratory Medicine, Nanning 530021, China
|
| Abstract: |
| ［Abstract］ Objective To explore the application value of an artificial intelligence(AI)-based differential evolution algorithm-driven patient-based real-time quality control(PBRTQC) intelligent monitoring platform in quality management of 10 clinical chemistry and immunology assays. Methods Based on the PBRTQC intelligent monitoring platform, the test results of 10 clinical chemistry and immunology assays were automatically collected from the Department of Clinical Laboratory of the People′s Hospital of Guangxi Zhuang Autonomous Region from January 2023 to December 2023. The differential evolution algorithm was adopted to optimize the computational parameters of the exponentially weighted moving average(EWMA) procedure, including batch size, step length, truncated bounds and control limits. The resultant model was validated using data collected from January 2024 to May 2024, and its performance was subsequently evaluated under real-world operating conditions from June 2024 to October 2024. The application value of the PBRTQC intelligent monitoring platform in the quality control of the 10 assays was evaluated by using relevant quality control rules. Results The optimal procedure-related parameters were selected through the PBRTQC intelligent monitoring platform. In the testing of the 10 assays, the PBRTQC intelligent monitoring platform consistently identified and promptly flagged systematic performance disturbances arising from insufficient reagent volumes, prolonged reagent exposure after the reagent bottles were opened, or carryover contamination from the stirring rods, demonstrating reliable error detection with timely alerting function. Conclusion The PBRTQC intelligent monitoring platform can monitor the quality risk of detection of 10 clinical chemistry and immunology assays in real time and accurately identify systematic errors. |
Key words: clinical chemistry and immunology assays artificial intelligence(AI) patient-based real-time quality control(PBRTQC) exponentially weighted moving average(EWMA)
*共同通信作者 |
