文献速递颈动脉与弓上血管

利用自然语言处理建立颈动脉狭窄疾病队列以开展比较效果研究

Establishing a carotid artery stenosis disease cohort for comparative effectiveness research using natural language processing.
PubMed 文献 · J Vasc Surg · 2021-06-25

本研究在大型 integrated 医疗系统中识别2008至2012年接受颈动脉影像检查的成年患者并随访至2017年,开发自然语言处理(NLP)算法依据超声和NASCET指南对狭窄进行分类。共识别95,896项索引检查(76,276项超声和19,620项CTA/MRA),算法排除正常/非血流动力学显著狭窄的PPV为超声99%、CTA/MRA 96.5%,识别非正常动脉的PPV/敏感度分别为CTA/MRA 76.9%/93.1%和超声74.7%/94%。最终构建包含15,522例患者的侧别特异性疾病队列,可减少人工审查并支持比较效果研究。

要点

  • NLP算法可高精度排除正常/非血流动力学显著狭窄,超声PPV达99%。
  • 最终队列包含12,828条中度、5,283条重度及1,895条闭塞动脉,支持比较效果研究。

核心结论

  • NLP可准确排除正常/非血流动力学显著狭窄,以中等精确度识别病变,大幅减少人工审查并高效构建颈动脉狭窄疾病队列。

研究设计与样本

  • 回顾性队列研究,纳入2008至2012年在一大型integrated医疗系统内接受颈动脉影像检查的成年患者,随访至2017年。
  • 共识别95,896项索引检查(76,276项超声和19,620项CTA/MRA),来自94,822例患者,其中1,059例同日接受多项检查。
  • 开发NLP算法依据超声(Society of Radiologists in Ultrasound)和轴位影像(NASCET)指南对狭窄进行侧别特异性分类。
  • 通过简单随机样本计算阳性预测值(PPV)和敏感度以验证算法,并对所有非正常检查进行人工复核以确认患病率和组建队列。

主要发现

  • 排除正常/非血流动力学显著狭窄的PPV为超声99%、CTA/MRA 96.5%。
  • 识别非正常动脉且侧别正确的PPV/敏感度:CTA/MRA为76.9%(95%CI 74.1%-79.5%)/93.1%(95%CI 91.1%-94.8%),超声为74.7%(95%CI 69.3%-79.5%)/94%(95%CI 90.2%-96.7%)。
  • 共15,522例患者被识别为存在病变颈动脉,其中2,674例为双侧等同病变。
  • 侧别特异性队列包含12,828条中度狭窄、5,283条重度狭窄、1,895条闭塞动脉及326条狭窄程度未知的病变动脉。
  • 随访期间30.1%的患者接受了61,107项额外检查。

关键数值与推荐

  • 超声报告评估内颈动脉收缩期和舒张期流速及流速比,并与侧别匹配以提高准确性。
  • NLP算法依据Society of Radiologists in Ultrasound(超声)和NASCET(轴位影像)指南对狭窄进行分类。
  • 正常/非血流动力学显著狭窄的排除PPV:超声99%,CTA/MRA 96.5%。
  • 非正常动脉识别的PPV/敏感度:CTA/MRA 76.9%/93.1%,超声74.7%/94%。

临床怎么用

  • 可利用NLP自动排除正常/非血流动力学显著狭窄的影像报告,减少人工审查负担。
  • 所构建的侧别特异性疾病队列可用于颈动脉狭窄治疗的比较效果研究。

局限性

  • 原文未报告

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)。

原文

OBJECTIVE: Investigation of asymptomatic carotid stenosis treatment is hindered by the lack of a contemporary population-based disease cohort. We describe the use of natural language processing (NLP) to identify stenosis in patients undergoing carotid imaging. METHODS: Adult patients with carotid imaging between 2008 and 2012 in a large integrated health care system were identified and followed through 2017. An NLP process was developed to characterize carotid stenosis according to the Society of Radiologists in Ultrasound (for ultrasounds) and North American Symptomatic Carotid Endarterectomy Trial (NASCET) (for axial imaging) guidelines. The resulting algorithm assessed text descriptors to categorize normal/non-hemodynamically significant stenosis, moderate or severe stenosis as well as occlusion in both carotid ultrasound (US) and axial imaging (computed tomography and magnetic resonance angiography [CTA/MRA]). For US reports, internal carotid artery systolic and diastolic velocities and velocity ratios were assessed and matched for laterality to supplement accuracy. To validate the NLP algorithm, positive predictive value (PPV or precision) and sensitivity (recall) were calculated from simple random samples from the population of all imaging studies. Lastly, all non-normal studies were manually reviewed for confirmation for prevalence estimates and disease cohort assembly. RESULTS: A total of 95,896 qualifying index studies (76,276 US and 19,620 CTA/MRA) were identified among 94,822 patients including 1059 patients who underwent multiple studies on the same day. For studies of normal/non-hemodynamically significant stenosis arteries, the NLP algorithm showed excellent performance with a PPV of 99% for US and 96.5% for CTA/MRA. PPV/sensitivity to identify a non-normal artery with correct laterality in the CTA/MRA and US samples were 76.9% (95% confidence interval [CI], 74.1%-79.5%)/93.1% (95% CI, 91.1%-94.8%) and 74.7% (95% CI, 69.3%-79.5%)/94% (95% CI, 90.2%-96.7%), respectively. Regarding cohort assembly, 15,522 patients were identified with diseased carotid artery, including 2674 exhibiting equal bilateral disease. This resulted in a laterality-specific cohort with 12,828 moderate, 5283 severe, and 1895 occluded arteries and 326 diseased arteries with unknown stenosis. During follow-up, 30.1% of these patients underwent 61,107 additional studies. CONCLUSIONS: Use of NLP to detect carotid stenosis or occlusion can result in accurate exclusion of normal/non-hemodynamically significant stenosis disease states with more moderate precision with lesion identification, which can substantially reduce the need for manual review. The resulting cohort allows for efficient research and holds promise for similar reporting in other vascular diseases.

来源:J Vasc Surg,查看原文。版权归原作者所有。

引用

  1. Chang RW, Tucker LY, Rothenberg KA, et al. Establishing a carotid artery stenosis disease cohort for comparative effectiveness research using natural language processing. J Vasc Surg. 2021. PMID: 34182027. doi:10.1016/j.jvs.2021.05.054

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。