Department of Medical Imaging, The Ottawa Hospital, The University of Ottawa, 1053 Carling Ave, Ottawa, ON K1Y 4E9, Canada.
Joint Department of Medical Imaging, University Health Network, Mount Sinai Hospital, Women's College Hospital, University of Toronto, Toronto, ON, Canada.
AJR Am J Roentgenol. 2019 Sep;213(3):619-625. doi: 10.2214/AJR.18.20990. Epub 2019 May 23.
The objective of our study was to evaluate iodine concentration and attenuation change in Hounsfield unit (ΔHU) thresholds to diagnose enhancement in renal masses at rapid-kilovoltage-switching dual-energy CT (DECT). We evaluated 30 consecutive histologically confirmed solid renal masses (including nine papillary renal cell carcinomas [RCCs]) and 27 benign cysts (17 simple and 10 hemorrhagic or proteinaceous cysts) with DECT December 2016 and May 2018. A blinded radiologist measured iodine concentration (in milligrams per milliliter) and ΔHU (attenuation on enhanced CT - attenuation on unenhanced CT) using 70-keV corticomedullary (CM) phase virtual monochromatic and 120-kVp nephrographic (NG) phase images. The accuracies of previously described enhancement thresholds were compared by ROC curve analysis. An iodine concentration of ≥ 2.0 mg/mL and an iodine concentration of ≥ 1.2 mg/mL achieved sensitivity, specificity, and the area under the ROC curve (AUC) of 73.3%, 100.0%, and 0.87 and 86.7%, 100.0%, and 0.93, respectively. On 70-keV CM phase images, ΔHU ≥ 20 HU and ΔHU ≥ 15 HU yielded sensitivity, specificity, and AUC of 80.0%, 100.0%, and 0.90 and 90.0%, 100.0%, and 0.95, respectively. The numbers of incorrectly classified papillary RCCs were as follows: iodine concentration of ≥ 2.0 mg/mL, 77.8% (7/9; range, 0.7-1.6 mg/mL); iodine concentration of ≥ 1.2 mg/mL, 44.4% (4/9; range, 0.7-0.9 mg/mL); ΔHU ≥ 20 HU on 70-keV CM phase images, 66.7% (6/9; range, 4-17 HU); and ΔHU ≥ 15 HU on 70-keV DECT images, 33.3% (3/9; 4-12 HU). No cyst pseudoenhancement occurred on DECT. For 120-kVp NG phase DECT, ΔHU ≥ 20 HU and ΔHU ≥ 15 HU yielded sensitivity, specificity, and AUC of 93.3%, 96.3%, and 0.95 and 100.0%, 88.9%, and 0.94, respectively. With ΔHU ≥ 20 HU, 22.2% (2/9) (range, 15-18 HU) of papillary RCCs were misclassified and there was one pseudoenhancing cyst. With ΔHU ≥ 15 HU, no papillary RCCs were misclassified but 11.1% (3/27) of cysts showed pseudoenhancement. Only an iodine concentration of ≥ 2.0 mg/mL showed significantly lower accuracy than other measures ( = 0.031-0.045). DECT applied in the CM phase performed best using an iodine concentration of ≥ 1.2 mg/mL or a 70-keV ΔHU ≥ 15 HU; these parameters improved sensitivity for the detection of enhancement in renal masses without instances of cyst pseudoenhancement.
我们的研究目的是评估碘浓度和亨氏单位(ΔHU)衰减变化阈值,以诊断快速千伏切换双能 CT(DECT)肾肿块增强。我们在 2016 年 12 月和 2018 年 5 月用 DECT 评估了 30 个连续的经组织学证实的实体性肾肿块(包括 9 个乳头状肾细胞癌[RCC])和 27 个良性囊肿(17 个单纯囊肿和 10 个出血或蛋白性囊肿)。一位盲法放射科医生使用 70keV 皮质髓质(CM)相虚拟单能和 120kVp 肾图(NG)相图像测量碘浓度(每毫升毫克数)和 ΔHU(增强 CT 上的衰减 - 未增强 CT 上的衰减)。通过 ROC 曲线分析比较了以前描述的增强阈值的准确性。碘浓度≥2.0mg/mL 和碘浓度≥1.2mg/mL 的灵敏度、特异性和 ROC 曲线下面积(AUC)分别为 73.3%、100.0%和 0.87,86.7%、100.0%和 0.93。在 70keV CM 相图像上,ΔHU≥20HU 和 ΔHU≥15HU 的灵敏度、特异性和 AUC 分别为 80.0%、100.0%和 0.90,90.0%、100.0%和 0.95。误分类的乳头状 RCC 数量如下:碘浓度≥2.0mg/mL,77.8%(9/9;范围,0.7-1.6mg/mL);碘浓度≥1.2mg/mL,44.4%(9/9;范围,0.7-0.9mg/mL);70keV CM 相上的 ΔHU≥20HU,66.7%(9/9;范围,4-17HU);70keV DECT 上的 ΔHU≥15HU,33.3%(9/9;范围,4-12HU)。DECT 上没有囊肿假性增强。对于 120kVp NG 相 DECT,ΔHU≥20HU 和 ΔHU≥15HU 的灵敏度、特异性和 AUC 分别为 93.3%、96.3%和 0.95,100.0%、88.9%和 0.94。ΔHU≥20HU 时,22.2%(9/9)(范围,15-18HU)的乳头状 RCC 被误分类,有一个假性增强的囊肿。ΔHU≥15HU 时,没有乳头状 RCC 被误分类,但 11.1%(3/27)的囊肿出现假性增强。只有碘浓度≥2.0mg/mL 的准确性明显低于其他指标(=0.031-0.045)。CM 相应用 DECT 时,以碘浓度≥1.2mg/mL 或 70keV ΔHU≥15HU 效果最佳;这些参数提高了检测肾肿块增强的灵敏度,而没有囊肿假性增强的情况。