文献速递儿科影像妇产影像腹部MR

北美胎儿治疗网络(NAFTNet)各中心对先天性膈疝产前磁共振成像预后评估的变异性

Variability in antenatal prognostication of congenital diaphragmatic hernia by magnetic resonance imaging across the North American Fetal Therapy Network (NAFTNet).
PubMed 文献 · Ultrasound Obstet Gynecol · 2025-05-08

14个NAFTNet中心的14名放射科医生回顾15例左侧先天性膈疝的MRI,用Rypens与Meyers公式评估观察值与预期值总胎儿肺体积(o/e-TFLV)、预测肺体积百分比(PPLV)、肝疝出及肝疝出百分比(%LH)。o/e-TFLV与PPLV一致性可接受(中位变异系数24与26),%LH异质性显著(中位变异系数46),中心病例量不影响一致性,提示需标准化MRI评估。

要点

  • o/e-TFLV与PPLV的组间一致性可接受(中位变异系数分别为24与26),可用于胎儿肺体积评估。
  • 肝疝出百分比的组间一致性最低(中位变异系数46),提示其量化可靠性有限。
  • 图像质量差时o/e-TFLV与PPLV变异系数升高(中位分别为39与43),一致性变差。

核心结论

  • NAFTNet 14个中心的放射科医生对左侧先天性膈疝的o/e-TFLV、PPLV及肝疝出判定一致性可接受,但%LH异质性显著;中心病例量不影响一致性,提示需标准化MRI成像方案与预后评估。

研究设计与样本

  • 多中心观察性研究,14名NAFTNet放射科医生回顾性阅片15例左侧先天性膈疝的产前MRI,其中8例手术证实肝疝出;图像采集中位孕龄29.3周(范围25.0-37.6周),采集于2020至2022年。

关键影像表现/征象

  • 多数阅片者对14/15例的肝疝出判定与手术最终诊断一致。
  • 3例被多数阅片者评为图像质量差的MRI中,o/e-TFLV中位变异系数39、PPLV中位变异系数43,一致性差。

诊断标准与关键数值

  • 采用Rypens等与Meyers等公式计算观察值与预期值总胎儿肺体积(o/e-TFLV)。
  • 计算预测肺体积百分比(PPLV)及肝疝出百分比(%LH)。
  • 图像质量采用1-4分评分(1分代表质量优,4分代表质量差、无法测量)。
  • 组间一致性以变异系数(CV)评估,CV<30定义为可接受。
  • o/e-TFLV中位CV为24(四分位距19-34),PPLV中位CV为26(四分位距18-42),%LH中位CV为46(四分位距44-53)。

临床应用与阅片要点

  • 报告o/e-TFLV与PPLV时需注意图像质量差可致测量一致性下降(中位CV分别升至39与43)。
  • %LH的组间一致性低(中位CV46),量化肝疝出百分比时需谨慎。
  • 中心病例量高低不影响肺体积与肝疝出评估的一致性。

局限性

  • 原文未报告局限性。

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: To evaluate the variability in magnetic resonance imaging (MRI)-based parameters used for fetal lung volume estimation in the prediction of pulmonary hypoplasia and the degree of liver herniation in cases of antenatally diagnosed left congenital diaphragmatic hernia (CDH) across North American Fetal Therapy Network (NAFTNet) centers.

METHODS: In this study, 14 NAFTNet radiologists reviewed MRI exams of 15 cases of left CDH of variable severity, eight of which had liver herniation confirmed at surgery. Images were obtained at a median gestational age of 29.3 (range, 25.0-37.6) weeks, between 2020 and 2022. All participants were asked to rate image quality using a scale of 1-4 (where 1 represents excellent quality and 4 represents poor quality (unable to perform measurements)) and to determine the observed-to-expected total fetal lung volume (o/e-TFLV) using the formulae of Rypens et al. and Meyers et al., the percent predicted lung volume (PPLV), the presence or absence of liver herniation and the percentage of liver herniation (%LH). Fleiss' κ was used to assess inter-rater agreement for image-quality ratings. Concordance between participants was evaluated by determining a coefficient of variation (CV), with CV < 30 defined as acceptable. Additionally, the variation of individual participant's assessment of a case from the group average was also assessed. Data were also evaluated by center case volume, for which high volume was indicated by ≥ 15 CDH cases/year and low volume was indicated by < 15 CDH cases/year managed prenatally.

RESULTS: Overall, there was acceptable concordance for o/e-TFLV among reviewers using the formula of either Rypens et al. or Meyers et al. (median CV, 24 (interquartile range (IQR), 19-34)). Slightly lower but acceptable concordance was noted for PPLV (median CV, 26 (IQR, 18-42)). For the determination of liver herniation, most participants agreed with the final diagnosis at surgery in 14/15 cases; however, concordance was lowest among reviewers for the quantification of %LH (median CV, 46 (IQR, 44-53)). Among the three MRI exams rated as being of poor quality by the majority of participants, CV was higher for o/e-TFLV (median CV, 39) and PPLV (median CV, 43), indicating poor concordance among reviewers. No significant difference was noted in concordance among reviewers for the assessment of lung volume and liver herniation based on a center's CDH volume.

CONCLUSION: Noticeable variability with acceptable agreement was noted for o/e-TFLV, PPLV and determination of liver herniation between NAFTNet radiologists from 14 centers in cases of left CDH. However, significant heterogeneity was noted for %LH. Concordance among reviewers was similar, irrespective of center case volume, highlighting the need for standardization of imaging protocols and CDH prognostication by MRI. © 2025 The Author(s). Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

来源:Ultrasound Obstet Gynecol,查看原文。版权归原作者所有。

引用

  1. Abbasi N, Kajal D, Johnson A, et al. Variability in antenatal prognostication of congenital diaphragmatic hernia by magnetic resonance imaging across the North American Fetal Therapy Network (NAFTNet). Ultrasound Obstet Gynecol. 2025. PMID: 40344307. doi:10.1002/uog.29233

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。