Department of Radiology, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA.
Department of Biostatics, University of Wisconsin-Madison, Madison, WI, USA.
Eur Radiol. 2019 Oct;29(10):5682-5690. doi: 10.1007/s00330-019-06143-6. Epub 2019 Mar 22.
This study was conducted in order to evaluate whether an MRI protocol with only fluid-sensitive sequences can be used to evaluate for musculoskeletal (MSK) infection of the pelvis and limbs in children.
This retrospective study analyzed 90 contrast-enhanced (CE) MRI studies from 88 consecutive patients (52 boys and 36 girls; mean age 9 ± 4.3 years; range 2-17) that were performed for the clinical suspicion of MSK infection. Two radiologists reviewed each study twice. The initial study review included only the fluid-sensitive sequences (fluid-sensitive study); the second review, performed at least 1 month later, included all sequences of the contrast-enhanced study (CE study). At each review, anatomic sites of abnormal signal and overall suspicion for infection were recorded. Cohen's kappa and percent agreement were performed to compare agreement between readers, types of studies, and clinical diagnoses.
Interreader agreement for both types of studies had kappa values between 0.86 and 1. For the assessment of MSK infection, the fluid-sensitive study had 100% sensitivity and 61.3% specificity, with 84.8% interreader agreement; and the CE study had 100% sensitivity and 71.0% specificity, with 88.6% interreader agreement. All cases of septic arthritis (13 cases) and osteomyelitis (25 cases) were identified as possible infection or infection until proven otherwise (negative predictive value 100%) with 100% interreader agreement on fluid-sensitive sequences.
An abbreviated MRI study using only fluid-sensitive sequences has the same high degree of sensitivity as a CE study to identify MSK infection in children and could be used to exclude septic arthritis and osteomyelitis.
• MRI with only fluid-sensitive sequences can be used to evaluate for musculoskeletal infection in children.
本研究旨在评估仅使用液体敏感序列的 MRI 方案是否可用于评估儿童骨盆和四肢的肌肉骨骼（MSK）感染。
本回顾性研究分析了 88 例连续患者（52 名男孩和 36 名女孩；平均年龄 9±4.3 岁；范围 2-17 岁）的 90 例增强对比（CE）MRI 研究，这些患者因临床怀疑 MSK 感染而进行检查。两位放射科医生对每项研究进行了两次复查。初次研究回顾仅包括液体敏感序列（液体敏感研究）；第二次回顾至少在 1 个月后进行，包括 CE 研究的所有序列。在每次复查时，记录异常信号的解剖部位和整体感染可疑程度。采用 Cohen's kappa 和百分比一致性来比较读者之间、研究类型之间和临床诊断之间的一致性。
两种类型研究的读者间一致性的kappa 值在 0.86 到 1 之间。对于 MSK 感染的评估，液体敏感研究的敏感性为 100%，特异性为 61.3%，读者间一致性为 84.8%；CE 研究的敏感性为 100%，特异性为 71.0%，读者间一致性为 88.6%。所有化脓性关节炎（13 例）和骨髓炎（25 例）均被确定为可能的感染或感染（阴性预测值 100%），在液体敏感序列上，读者间的一致性为 100%。
仅使用液体敏感序列的简化 MRI 研究与 CE 研究具有相同的高度敏感性，可用于识别儿童的 MSK 感染，并可用于排除化脓性关节炎和骨髓炎。
• 仅使用液体敏感序列的 MRI 可用于评估儿童的肌肉骨骼感染。
