文献速递骨肌儿科影像MR

使用流体敏感序列检测儿童肌肉骨骼病变

Detection of pediatric musculoskeletal pathology using the fluid-sensitive sequence.
PubMed 文献 · Pediatr Radiol · 2018-09-19

该回顾性研究于2016年2月1日至10月31日纳入96例患儿(99次检查,48男48女,平均年龄11.1±4.6岁),由两名放射科医生先后独立判读流体敏感序列(分诊研究)与增强MRI。结果显示两种检查对病变检出一致性高(97%–100%),流体敏感序列可能可用于筛查需要增强MRI的儿童。

要点

  • 流体敏感序列与增强MRI对病变检出一致性达97%–100%,对最终印象一致性为93%。
  • 增强MRI的读者间一致性(κ范围0.91–1)高于流体敏感序列(κ范围0.49–1)。
  • 完整增强检查提高5次检查的诊断信心、2次检查的准确性,但未改变处理方案。

核心结论

  • 流体敏感序列与增强MRI在检测儿童肌肉骨骼病变方面具有接近完美的一致性,可能可用于筛查需要增强MRI的儿童。

研究设计与样本

  • 回顾性研究,纳入2016年2月1日至10月31日期间96例患者(48男48女,平均年龄11.1±4.6岁)的99次检查,排除综合征性畸形、近期外伤、感染性或炎症性关节病史、既往内固定物或记录不完整者;两名放射科医生间隔至少1个月先后独立判读流体敏感序列(分诊研究)和增强MRI,并以共识产生最终印象;采用Cohen's kappa和百分比一致性比较读者间和检查间的一致性。

关键影像表现/征象

  • 临床诊断包括:应激反应或过度使用31%、感染21%、占位性病变17%、正常15%、炎症性14%、炎症合并过度使用1%。
  • 增强MRI的读者间一致性(κ范围0.91–1)高于分诊研究(κ范围0.49–1)。
  • 两种检查对病变检出的一致性为97%–100%,对最终印象的一致性为93%。
  • 完整增强检查提高5次检查的诊断信心、2次检查的准确性,但未改变处理方案。

诊断标准与关键数值

  • 以增强MRI为参考,流体敏感序列检出病变的百分比一致性为97%–100%,最终印象一致性为93%。
  • 读者间一致性:增强MRI κ范围0.91–1,流体敏感序列 κ范围0.49–1。

临床应用与阅片要点

  • 流体敏感序列可能可用于筛查儿童肌肉骨骼病变,以判断是否需要增强MRI。
  • 采用流体敏感序列作为分诊研究时,需注意其读者间一致性(κ范围0.49–1)低于增强MRI(κ范围0.91–1)。

局限性

  • 原文为回顾性研究。
  • 本研究为单中心研究。
  • 样本量有限(96例患者、99次检查)。
  • 作者结论中使用「could possibly be used」提示需进一步研究。

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

BACKGROUND: Musculoskeletal complaints are common among children, and magnetic resonance (MR) is increasingly used to supplement the clinical assessment. The validation of a short triage protocol could reduce the number of unnecessary contrast-enhanced MR studies that sometimes also require the need for sedation.

OBJECTIVE: To compare the diagnostic accuracy between fluid-sensitive sequence and contrast-enhanced MR study in the detection of musculoskeletal pathology in the pelvis and the appendicular skeleton in children older than 2 years.

MATERIALS AND METHODS: We performed a retrospective review between Feb. 1, 2016, and Oct. 31, 2016, and identified 99 studies from 96 patients (48 boys and 48 girls; mean age ± standard deviation, 11.1±4.6 years) without syndromic deformity, recent trauma, a history of infectious or inflammatory arthropathy, prior instrumentation or incomplete records. Two radiologists reviewed each study twice, at least 1 month apart, first using only the fluid-sensitive sequences (triage study) and later using the contrast-enhanced study. Readers rated the presence or absence of pathology independently and generated final impressions in consensus. We used Cohen's kappa (κ) and percentage agreement to compare agreement between readers and between studies, respectively.

RESULTS: Inter-reader agreement was overall higher for the contrast-enhanced studies (κ range = 0.91-1) than for the triage studies (κ range = 0.49-1). Percentage agreement between studies was high for the detection of pathology (97-100%) and for the impressions (93%). Clinical diagnoses were stress reaction or overuse in 31%, infection in 21%, space-occupying process in 17%, normal in 15%, inflammatory in 14%, and both inflammatory and overuse in 1%. The full study increased diagnostic confidence in five studies and accuracy in two but did not alter management.

CONCLUSION: The fluid-sensitive sequence had a near-perfect percentage of agreement with the contrast-enhanced study in the detection of musculoskeletal pathology and could possibly be used to screen children who need a contrast-enhanced MR study.

来源:Pediatr Radiol,查看原文。版权归原作者所有。

引用

  1. Nguyen JC, Yi PH, Woo KM, et al. Detection of pediatric musculoskeletal pathology using the fluid-sensitive sequence. Pediatr Radiol. 2018. PMID: 30232532. doi:10.1007/s00247-018-4256-z

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。