Department of Radiology (CCM), Charité - Universitätsmedizin Berlin, Campus Mitte, Humboldt - Universität zu Berlin, Freie Universität Berlin, Charitéplatz 1, 10117, Berlin, Germany.
Department of Rheumatology, Charité - Universitätsmedizin Berlin, Campus Mitte, Humboldt - Universität zu Berlin, Freie Universität Berlin, Berlin, Germany.
Eur Radiol. 2022 Sep;32(9):6376-6383. doi: 10.1007/s00330-022-08744-0. Epub 2022 Apr 1.
To investigate the performance of dual-energy CT (DECT)-generated iodine maps (iMap) and CT subtraction (CT-S) in the detection of synovitis, tenosynovitis, and peritendonitis/paratenonitis compared to magnetic resonance imaging (MRI) using musculoskeletal ultrasound (MSUS) as standard of reference.
This IRB-approved prospective study consecutively investigated patients with undifferentiated arthritis. All patients underwent MSUS, MRI and contrast-enhanced DECT of the hand; from the latter conventional CT-S, image-based iMap (iMap-I) and raw data-based iMap (iMap-RD) were reconstructed. CT and MRI datasets were scored for synovitis and tenosynovitis/paratenonitis applying the modified Rheumatoid Arthritis MRI Score (RAMRIS). Sensitivity, specificity, and diagnostic accuracy were calculated. Non-inferiority was tested using the one-tailed McNemar test. Correlation of sum scores was assessed using Pearson's test. Interreader reliability was assessed using Cohen's kappa.
Overall, 33 patients were included. MSUS was positive for synovitis and tenosynovitis/paratenonitis in 28 patients with a sum score of 6.91. Excellent correlation with MSUS was shown for CT-S (sum score 6.38; r = 0.91), iMap-RD (sum score 9.74; r = 0.82), MRI (sum score 12.70; r = 0.85), and iMap-I (sum score 6.94; r = 0.50). CT-S had the highest diagnostic accuracy of 83%, followed by iMap-I (78%), MRI (75%), and iMap-RD (74%). All modalities showed non-inferiority. Reader agreement was good for CT-S and MRI (κ = 0.62; 0.64) and fair for iMap-RD and iMap-I (κ = 0.31; 0.37).
CT-S and iMap allow highly standardized arthritis imaging and are suitable for clinical practice. MSUS still has the highest availability for arthritis imaging and served as gold standard for this study.
• CT subtraction, iodine map with dual-energy CT, and MRI showed non-inferiority to musculoskeletal ultrasound. • MRI was the most sensitive but least specific imaging technique compared with CT subtraction and dual-energy CT. • CT subtraction showed the best correlation with musculoskeletal ultrasound.
使用肌肉骨骼超声(MSUS)作为参考标准,研究双能 CT(DECT)生成的碘图(iMap)和 CT 减影(CT-S)在检测滑膜炎、腱鞘炎和腱周炎/腱旁膜炎方面的性能与磁共振成像(MRI)相比。
这项经 IRB 批准的前瞻性研究连续调查了未分化关节炎患者。所有患者均接受 MSUS、MRI 和手部对比增强 DECT 检查;从后者中重建常规 CT-S、基于图像的 iMap(iMap-I)和基于原始数据的 iMap(iMap-RD)。应用改良类风湿关节炎 MRI 评分(RAMRIS)对 CT 和 MRI 数据集进行滑膜炎和腱鞘炎/腱旁膜炎评分。计算灵敏度、特异性和诊断准确性。使用单侧 McNemar 检验测试非劣效性。使用 Pearson 检验评估总分的相关性。使用 Cohen's kappa 评估读者间可靠性。
总体而言,共纳入 33 名患者。MSUS 对 28 例滑膜炎和腱鞘炎/腱旁膜炎阳性患者的总分为 6.91。CT-S(总分 6.38;r = 0.91)、iMap-RD(总分 9.74;r = 0.82)、MRI(总分 12.70;r = 0.85)和 iMap-I(总分 6.94;r = 0.50)与 MSUS 具有极好的相关性。CT-S 的诊断准确性最高为 83%,其次是 iMap-I(78%)、MRI(75%)和 iMap-RD(74%)。所有模态均显示非劣效性。CT-S 和 MRI 的读者间一致性良好(κ=0.62;0.64),iMap-RD 和 iMap-I 的一致性为中等(κ=0.31;0.37)。
CT-S 和 iMap 允许对关节炎进行高度标准化成像,适用于临床实践。MSUS 仍然是关节炎成像的最高可用性,并作为本研究的金标准。
• CT 减影、双能 CT 的碘图和 MRI 与肌肉骨骼超声相比具有非劣效性。• 与 CT 减影和双能 CT 相比,MRI 是最敏感但特异性最低的影像学技术。• CT 减影与肌肉骨骼超声相关性最好。