比较两种人工智能方法检索记录植入式器械的临床报告以识别MRI安全风险
本研究在793张床位的学术医院回顾性评估两种人工智能方法(专家衍生法与本体衍生自然语言处理法)从影像报告、急诊记录和其他临床报告中检索植入式器械的敏感度、特异度和准确率,以465份随机抽样报告的人工审核为参考标准。两种方法的准确率、敏感度和特异度相似,检索到的含高危MRI安全风险植入式器械的报告比例为1.47%至1.88%。
要点
- 专家衍生法与本体衍生NLP法检索植入式器械的准确率分别为0.83和0.91,敏感度分别为0.88和0.96,特异度分别为0.82和0.92,差异均未达统计学显著。
- 影像报告、急诊记录和其他临床报告中检索到含高危MRI安全风险植入式器械的比例范围为1.47%至1.88%。
- 本体衍生NLP法将术语映射至Systematized Nomenclature of Medicine-Clinical Terms,专家衍生法使用开源软件和 curated 术语列表查询。
核心结论
- 专家驱动的NLP与本体驱动的NLP在识别携带对MRI构成高安全风险的植入式器械的患者方面具有相似的准确率。
研究设计与样本
- 经机构审查委员会批准的回顾性研究,在793张床位的学术医院进行,查询三种临床数据类型:25,000份影像报告、174,769份急诊记录和41,085份其他临床报告(如心脏科、手术室、医师记录、影像报告、病理报告、患者信件),以随机抽样465份报告的人工审核为参考标准,使用t检验比较两种方法的敏感度、特异度和准确率。
诊断标准与关键数值
- 专家衍生法与本体衍生法的准确率分别为0.83和0.91(P=0.080),敏感度分别为0.88和0.96(P=0.178),特异度分别为0.82和0.92(P=0.110);影像报告、急诊记录和其他临床报告中检索到含高危MRI安全风险植入式器械的比例范围为1.47%至1.88%。
临床应用与阅片要点
- 作者认为,专家驱动的NLP和本体驱动的NLP等人工智能方法在识别携带对MRI构成高安全风险的植入式器械的患者方面具有相似的准确率。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
OBJECTIVE: Assess sensitivity, specificity, and accuracy of two approaches to identify patients with implantable devices that pose safety risks for MRI-an expert-derived approach and an ontology-derived natural language processing (NLP). Determine the proportion of clinical data that identify these implantable devices.
METHODS: This Institutional Review Board-approved retrospective study was performed at a 793-bed academic hospital. The expert-derived approach used an open-source software with a list of curated terms to query for implantable devices posing high safety risk ("MRI-Red") in patients undergoing MRI. The ontology-derived approach used an NLP system with terms mapped to Systematized Nomenclature of Medicine-Clinical Terms. Queries were performed in three clinical data types-25,000 radiology reports, 174,769 emergency department (ED) notes, and 41,085 other clinical reports (eg, cardiology, operating room, physician notes, radiology reports, pathology reports, patient letters). Sensitivity, specificity, and accuracy of both methods against manual review of a randomly sampled 465 reports were assessed and tested for significant differences between expert-derived and ontology-derived approaches using t test.
RESULTS: Accuracy, sensitivity, and specificity of expert-versus ontology-derived approaches were similar (0.83 versus 0.91, P = .080; 0.88 versus 0.96, P = .178; 0.82 versus 0.92, P = .110). The proportion of radiology reports, ED notes, and other clinical reports retrieved containing implantable devices with high safety risks for MRI ranged from 1.47% to 1.88%.
DISCUSSION: Artificial intelligence approaches such as expert-driven NLP and ontology-driven NLP have similar accuracy in identifying patients with implantable devices that pose high safety risks for MRI.
引用
- Valtchinov VI, Lacson R, Wang A, et al. Comparing Artificial Intelligence Approaches to Retrieve Clinical Reports Documenting Implantable Devices Posing MRI Safety Risks. J Am Coll Radiol. 2019. PMID: 31415740. doi:10.1016/j.jacr.2019.07.018
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。