Diekhoff Torsten, Kiefer Tobias, Stroux Andrea, Pilhofer Irid, Juran Ralf, Mews Jürgen, Blobel Jörg, Tsuyuki Masaharu, Ackermann Beate, Hamm Bernd, Hermann Kay-Geert A
From the *Department of Radiology, Charité - Universitätsmedizin Berlin Campus Mitte, Humboldt-Universität zu Berlin, Freie Universität Berlin; †Department of Medical Informatics, Biometry and Epidemiology, Free University Berlin, Berlin, Germany; ‡Toshiba Medical Systems Europe, BV, Zoetermeer, Netherlands; §Toshiba Medical Systems Corporation, Otawara-Shi, Tochigi, Japan; and ∥Central Pharmacy, Charité - Universitätsmedizin Berlin Campus Virchow-Klinikum, Humboldt-Universität zu Berlin, Freie Universität Berlin, Berlin, Germany.
Invest Radiol. 2015 Apr;50(4):255-60. doi: 10.1097/RLI.0000000000000099.
The aim of this study was to perform phantom measurements to prove the feasibility of single-source dual-energy computed tomography (DECT) of the extremities using a volume scan mode. In addition, we, for the first time, wanted to determine which concentrations of monosodium urate (MSU) in gout and calcium pyrophosphate (CP) in pseudogout are needed to detect or distinguish these soft tissue depositions with DECT.
We created a hand-shaped plastic phantom assembled with a descending order of concentrations of MSU (6.25%-50%) and CP (1.56%-50%) with similar attenuation in conventional computed tomographic (CT) images. Dual-energy imaging was done on a standard 320-row CT scanner with acquisition of 2 volumes: one at 80 and the other at 135 kV. Using linear regression analysis, dual-energy gradients were calculated for MSU and CP. Thereafter, we selected a specific region of interest on the dual-energy graph to color-code MSU and CP on the images. Three blinded readers scored 10 scans of the randomly equipped phantom, corresponding to 60 samples, to determine the sensitivity and specificity of this technique. Receiver operating characteristics analysis was done to determine the diagnostic power.
We found a dual-energy gradient for MSU of 1.020 ± 0.006 and for CP of 0.673 ± 0.001. Assessment of the randomized phantom scans indicates reliable detection of MSU at concentrations of 12.5 % or higher and that of CP at 6.25 % or higher, corresponding to deposits with mean Hounsfield unit values of 59.8 for MSU and 101.1 for CP. The sensitivity for MSU ranged from 83.3% to 97.3% at 15/90 mA (135/80 kV) and from 86.7% to 97.3% at 100/570 mA. Specificity was 96.7% to 100% in 15/90 mA and 100% in 100/570 mA of scans. However, there was inferior sensitivity for CP owing to lower concentrations. In the receiver operating characteristics analysis, the area under the curve for MSU ranged from 0.867 to 0.947 at 15/90 mA and from 0.867 to 0.919 at 100/570 mA and that for CP from 0.659 to 0.745 and from 0.718 to 0.750, respectively.
This phantom study shows that single-source DECT allows detection and characterization of crystal deposits when present in soft tissue at relatively low concentrations. Further studies in patients have to prove its benefits in diagnostic imaging and treatment monitoring as well as its significance compared with dual-source CT systems.
本研究的目的是进行体模测量，以证明使用容积扫描模式对四肢进行单源双能计算机断层扫描（DECT）的可行性。此外，我们首次想要确定痛风中的尿酸钠（MSU）和假性痛风中的焦磷酸钙（CP）需要何种浓度，才能通过DECT检测或区分这些软组织沉积物。
我们制作了一个手形塑料体模，其中MSU（6.25%-50%）和CP（1.56%-50%）的浓度按降序排列，在传统计算机断层扫描（CT）图像中具有相似的衰减。在标准的320排CT扫描仪上进行双能成像，采集两个容积的数据：一个在80 kV，另一个在135 kV。使用线性回归分析，计算MSU和CP的双能梯度。此后，我们在双能图上选择一个特定的感兴趣区域，以便在图像上对MSU和CP进行颜色编码。三名不知情的读者对随机配置的体模进行了10次扫描，对应60个样本，以确定该技术的敏感性和特异性。进行了受试者操作特征分析以确定诊断能力。
我们发现MSU的双能梯度为1.020±0.006，CP的双能梯度为0.673±0.001。对随机体模扫描的评估表明，浓度为12.5%或更高的MSU以及6.25%或更高的CP能够可靠检测，对应的MSU沉积物平均Hounsfield单位值为59.8，CP为101.1。MSU在15/90 mA（135/80 kV）时的敏感性范围为83.3%至97.3%，在100/570 mA时为86.7%至97.3%。15/90 mA扫描的特异性为96.7%至100%，100/570 mA扫描的特异性为100%。然而，由于CP浓度较低，其敏感性较差。在受试者操作特征分析中，MSU在15/90 mA时曲线下面积范围为0.867至0.947，在100/570 mA时为0.867至0.919，CP的曲线下面积分别为0.659至0.745和0.718至0.750。
该体模研究表明，单源DECT能够检测和表征软组织中相对低浓度存在的晶体沉积物。在患者中进行的进一步研究必须证明其在诊断成像和治疗监测中的益处以及与双源CT系统相比的意义。
