|Year : 2018 | Volume
| Issue : 9 | Page : 394-399
Gemstone spectral imaging dual-energy computed tomography for differentiation of renal cell carcinoma and minimal-fat renal angiomyolipoma
Yamin Wan1, Hua Guo1, Lijuan Ji2, Zhizhen Li3, Jianbo Gao1
1 Department of Radiology, First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, P.R. China
2 Department of Computed Tomography, Jiaozuo People's Hospital, Jiaozuo 454150, Henan, P.R. China
3 Department of Endocrinology, First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, P.R. China
|Date of Web Publication||29-Jun-2018|
Department of Endocrinology, First Affiliated Hospital of Zhengzhou University, No. 1 Jianshe Dong Road, Zhengzhou 450052, Henan
Source of Support: None, Conflict of Interest: None
Purpose: To investigate the values of gemstone spectral imaging (GSI)-dual-energy computed tomography (DECT) in differentiation of renal cell carcinoma (RCC) and minimal-fat renal angiomyolipoma (MF-RAML).
Patients and Methods: Twenty-one patients with ischemic RCC and 19 patients with MF-RAML were enrolled in this study. GSI was performed on them, and the spectrum signs were analyzed.
Results: I(H2O), H2O(I), I(fat), and fat(I) concentrations, normalized I concentration, and effective atomic number of corticomedullary phase and parenchymal phase in enhanced GSI-DECT in ischemic RCC group were all significantly lower than those in MF-RAML group (P < 0.05). CT value and absolute slope rate of spectral attenuation curve in two phases in ischemic RCC group were also significantly lower than those in MF-RAML group (P < 0.05).
Conclusion: GSI-DECT has provided a new idea and method for differential diagnosis of ischemic RCC and MF-RAML, with high-clinical values.
Keywords: Differentiation, gemstone spectral imaging, minimal-fat renal angiomyolipoma, renal cell carcinoma
|How to cite this article:|
Wan Y, Guo H, Ji L, Li Z, Gao J. Gemstone spectral imaging dual-energy computed tomography for differentiation of renal cell carcinoma and minimal-fat renal angiomyolipoma. J Can Res Ther 2018;14, Suppl S2:394-9
|How to cite this URL:|
Wan Y, Guo H, Ji L, Li Z, Gao J. Gemstone spectral imaging dual-energy computed tomography for differentiation of renal cell carcinoma and minimal-fat renal angiomyolipoma. J Can Res Ther [serial online] 2018 [cited 2020 Feb 21];14:394-9. Available from: http://www.cancerjournal.net/text.asp?2018/14/9/394/172714
| > Introduction|| |
There are 60,000 persons newly diagnosed as renal cell carcinoma (RCC) in the USA in 2011, and approximately 13,000 cases were expected to die of this disease. RCC is the most common but benign renal tumor, representing 10–14% of resected solid renal tumors.,, Angiomyolipoma (AML) is the most common benign tumor of kidney. A well-circumscribed renal mass with intratumoral fat on computed tomography (CT) scan is similar to the diagnostic finding of renal AML (RAML). However, the diagnosis will be challenged when the tumor contains only a minimal amount of fat. AML without visible fat accounts for approximately 5% of all AMLs. Dual-energy CT (DECT), using dual tube voltages with either two consecutive scans or dual X-ray source-dual detector assembly has existed for a number of years, providing additional information for material separation with imaging.,,,, This scanning method is also useful in several clinical applications, including preoperative detection of insulinoma, differentiation of hypervascular hepatic lesions, and diagnosis of pulmonary embolism.,,, Therefore, the purpose of the present study was to investigate the use of gemstone spectral imaging (GSI)-DECT in the differentiation of ischemic RCC from minimal-fat RAML (MF-RAML).
| > Patients and Methods|| |
Twenty-one cases of ischemic RCC and 19 cases of MF-RAML in First Affiliated Hospital of Zhengzhou University were enrolled in this study from May 2011 to January 2013. All cases were confirmed pathologically by biopsy or surgery. This study was conducted with approval from the Ethics Committee of the First Affiliated Hospital of Zhengzhou University. Written informed consent was obtained from all patients.
There were 15 male and 25 female patients, aged 16–68 years, of whom 41 tumors were detected out. In 21 cases of ischemic RCC included 12 males and nine females (average age, 50.86 ± 14.02 years), there were nine cases of clear cell carcinoma, 7 cases of papillary RCC, and five cases of chromophobe RCC. All patients had unilateral single tumor. The maximum tumor diameter was about 72.87 ± 37.50 mm. Eight cases were located in the left kidney and 13 cases were located in the right kidney. Thirteen cases were discovered in the routine physical examination, four cases were accompanied by painless gross hematuria, three cases had waist dull pain or lower back discomfort, and one case had upper abdominal discomfort. Nineteen cases of MF-RAML (20 lesions) were detected including three males and 16 females (among who two lesions in left kidney was found in one female case), with average age of 37.15 ± 11.23 years. The maximum tumor diameter was about 43.72 ± 15.30 mm. Eleven cases were located in the left kidney and nine cases were in the right kidney. Fourteen cases were discovered in the routine physical examination, of which one case suffered from intermittent abdominal pain for more than 10 months, two cases had lower back discomfort, one case had unknown fever, and one case had painless gross hematuria.
Computed tomography examination
CT examination was performed with a high-definition CT scanner (Discovery CT750HD, GE Healthcare, Wisconsin, USA). The patients were in supine position, and the routine plain CT scan toward both kidneys was performed to determine the location of kidneys and tumors before enhancement scanning. The dual-phase enhancement scanning was conducted, with spectroscopy imaging mode as follows: 1.5 ml/kg nonionic contrast agent (ioversol injection, Optiray 320; Tyco Healthcare, Quebec, Canada) was injected by binocular high-pressure syringe through ulnar vein, and the flow rate was 3–5 ml/s. The enhancement scanning was triggered by CT value monitoring of abdominal aorta at diaphragmatic level (Smart Prep technique). The corticomedullary phase was automatically triggered when CT value of abdominal aorta was close to 200 HU, and the parenchymal phase was automatically triggered 50–60 s after corticomedullary phase (i.e., about 90 s after iohexol injection). The plain CT scan used 120 kVp and 240–600 mA automatic current, and the enhanced scan used instantaneous high-speed switching of 80 and 140 kVp, and 600 mAs maximum current, with speed of 0.6 s/rotation and pitch of 0.984:1. The scanning thickness was 5.0 mm, with spacing of 5.0 mm. The image reconstruction thickness was 0.625 mm, with single energy image distance of 0.625 mm (40–140 keV, interval of 10 keV). The spectral CT images were analyzed with the GSI viewer software 4.4 (GE Healthcare, Wisconsin, USA).
CT imaging data were blinded, read by two advanced radiologists, and the consistent results were obtained after discussion. The single energy images were loaded into the spectrum imaging browser (GSI viewer) software. The maximum tumor sections in dual-phase enhanced images were selected, respectively, and placed round or oval region of interest (ROI) within the tumor solid part. The ROI size was about ½–⅔ of tumor, avoiding the tumor calcification region, liquefaction necrosis region, tumor blood vessels, or abnormal enhanced zone. To reduce the errors, the measurement was performed in cross, sagittal, and coronal sections. The corresponding ROI data were measured and saved for mean value calculation. The size, shape, and location of ROI were kept consistent during the measurements of corticomedullary and parenchymal phase. At the same time, ROI was selected in the abdominal aorta region, and ROI data were recorded. ROI data were saved as excel formation, containing CT values (HU) of every energy level from 40 to 140 keV (interval 10 keV) and concentrations of 20 pairs of base substances. According to the purpose of this study, four pairs of base materials, namely I(H2O), H2O(I), I(fat), and fat(I), effective atomic number (EAN), CT value of every energy level from 40 to 140 keV, normalized I concentration (NIC), and spectral attenuation curve slope rate were selected for comparison analysis. The equations of NIC and spectral attenuation curve slope rate k were as the following: NIC = I concentration in lesion/I concentration in abdominal aorta on the same level; k = (HU40keV− HU100keV)/(40–100).
SPSS l7.0 (SPSS Inc., IL, USA) statistical software was used for analysis. Measurement data were expressed as mean ± standard deviation. Data in line with normal distribution were analyzed by independent-sample t-test. P < 0.05 was considered as statistically significant.
| > Results|| |
Comparisons of concentrations of base substance pairs between two groups
There were significant differences of concentrations of base substance pairs, I(H2O), H2O(I), I(fat), and fat(I) of corticomedullary and parenchymal phases between ischemic RCC and MF-RAML groups (P < 0.05). Above concentrations of base substance pairs in ischemic RCC group were all lower than those in MF-RAML group [Table 1].
|Table 1: Comparisons of concentrations of base substance pairs between two groups|
Click here to view
Comparisons of normalized I concentration and effective atomic number between two groups
As shown in [Table 2], NIC of corticomedullary phase in ischemic RCC and MF-RAML groups was 0.15 ± 0.07 and 0.30 ± 0.08, respectively, and that of parenchymal phase was 0.43 ± 0.13 and 0.56 ± 0.05, respectively. NIC of two phases in ischemic RCC group was significantly lower than MF-RAML group (P < 0.05). EAN in corticomedullary phase in ischemic RCC and MF-RAML groups was 8.82 ± 0.48 and 9.41 ± 0.26, respectively, and that in parenchymal phase was 8.84 ± 0.25 and 9.27 ± 0.12, respectively. EAN of two phases in ischemic RCC group was also significantly lower than MF-RAML group (P < 0.05).
|Table 2: Comparisons of normalized iodine concentration and effective atomic number of two phases between two groups|
Click here to view
Comparisons of computed tomography value and spectral attenuation curve between two groups
In corticomedullary phase, CT values in two groups reduced with increase of keV, and CT value in MF-RAML group was significantly higher than ischemic RCC group (P < 0.01) [Table 3]. In parenchymal phase, CT value in two groups reduced with the increase of keV, and that in MF-RAML group was significantly higher than ischemic RCC group (P < 0.01) [Table 4].
|Table 3: Comparison of computed tomography value of corticomedullary phase between two groups|
Click here to view
|Table 4: Comparison of computed tomography value of parenchymal phase between two groups|
Click here to view
In two phases, the spectral attenuation curve in MF-RAML group was above that in ischemic RCC group. In corticomedullary phase, the spectral attenuation curve in MF-RAML group was located on the top, with the sharpest shape, while in parenchymal phase, it was below that of corticomedullary phase. There was no obvious difference between spectral attenuation curves in two phases in ischemic RCC group [Figure 1].
|Figure 1: Comparison of spectral attenuation curves of two phases between two groups. CMP: Corticomedullary phase; PP: Parenchymal phase|
Click here to view
Slope rates of spectral attenuation curve (k) of corticomedullary phase in ischemic RCC and MF-RAML groups were −2.55 ± 1.12 and −4.06 ± 0.67, respectively, and those in parenchymal phase were −2.53 ± 0.58 and −3.62 ± 0.33, respectively. The absolute slope rates of two phases in ischemic RCC group were significantly lower than MF-RAML group (P < 0.05). This indicated that with increase of keV, the degree of CT attenuation in MF-RAML group was bigger than ischemic RCC group [Table 5].
|Table 5: Comparisons of spectral attenuation curve slope rates between two groups|
Click here to view
| > Discussion|| |
It is known that the imaging features of RAMLs depend on the relative proportions of tumor components. However, the low-fat content of AML makes the diagnosis of these tumors difficult. The distinction between RCC and AML using qualitative imaging features analysis alone might be difficult.
Multi-detector-row CT is considered the state-of-art method for evaluating renal masses.,,,, GSI-DECT can not only provide all routine CT values, but also provide unique various quantitative indicators and analytical tools. Comprehensive application of these indicators and tools can compare and measure the diseases with different origins or different tissue components within the same tumor, based on appropriate parameters, no matter the tumor is abundant in blood supply or ischemic. GSI-DECT can integrate the statistically meaningful multi-parameters for joint diagnostic assessment, which will help to improve the diagnosis effects of tumor location, qualitation, staging, and grading. In this study, the material quantitative analysis and comprehensive analysis of GSI-DECT were applied to primarily investigate their application value in the identification of ischemic RCC and MF-RAML.
X-ray absorption coefficient of substance can be determined by that of any two base materials. Therefore, the attenuation of a substance can converted to densities of two base materials generating the same attenuation, and this is the basic principle of analysis and separation of material compositions. In medical diagnostic imaging, H2O and I are often selected as the base pair for material-decomposition image presentation, because their atomic numbers span the range of atomic numbers for materials generally found in medical imaging and approximate those of soft tissue and iodinated contrast material to from material-attenuation images intuitive to interpret. In this study, not only I-H2O, but also I-fat are used as the base material for measurement and analysis, for achieving the purpose of quantification and identification.
In this study, the concentrations of I(H2O) and I(fat) in corticomedullary and parenchymal phases in ischemic RCC group were statistically lower than MF-RAML group (P < 0.05). The reasons may be that there is no sinus-like blood vessel of typical renal clear cell carcinoma inside the ischemic RCC or the blood vessels inside the tubular or papillary structures are thin, with scarce interstitial capillaries, rich connective tissues in stroma of chromophobe cell carcinoma, and less interstitial vessels. Hence, the amount of contrast agent entering the tumors is less, and the tumor enhancement effect is not obvious, exhibiting low I intake and minimal corresponding I content. In parenchymal phase, I content in ischemic RCC reduced slightly (0.13100 μg/cc lower than corticomedullary phase). Although the average CT value in parenchymal phase of ischemic RCC in conventional enhanced CT is 1.87 HU higher than corticomedullary phase, this slight increased CT value is not sufficient to cause a change because the inherent hardening effect of X-ray in conventional CT imaging may cause a slight shift of CT value. At the same time, it indicates that I content of contrast agent can directly affect the level of I concentration, while the level of I concentration may not be entirely decided by I content of contrast agent. MF-RAML is mainly composed with smooth muscles and blood vessels. When the vascular components are relatively large, the content of the contrast agent entering into the tumor will be more in corticomedullary phase and the tumor enhancement effect will be strong with high I intake, so the I concentration is high. In parenchymal phase, the tumor enhancement is reduced, so the I intake is reduced and I concentration also decreases. When the smooth muscle components are relatively large, the content of contrast agent entering tumor will increase gradually, and the tumors will present the trend of persistent enhancement. So I concentration will gradually increase. In 20 cases of MF-RAML, obvious whole enhancement in corticomedullary phase in conventional enhanced CT image can be observed, with reduction in parenchymal phase. This is consistent with the results that the overall I intake of MF-RAML is reduced in parenchymal phase, with decreased I concentration.
In MF-RAML group of this study, although the intratumoral fat contents are not observed from CT images, there are scattered or clustered-distributed fat cells in 18 cases under cancer light microscope. The fat cells cannot be observed clearly in only two patients, of whom AML is confirmed by pathological diagnosis. Ischemic RCC rarely contains fat ingredients. Very small part of this tumor contains a small amount of fat due to tumor cell steatosis or invasion and surrounding of tumor by surrounding fat. In this study, the CT images and pathological diagnosis suggest that the ischemic RCC samples do not contain fat ingredient. The quantitative analysis of fat-based substances in two groups reveals that the fat concentration in MF-RAML is significantly higher than ischemic RCC. The explanation toward difference of H2O(I) concentration may lie in the differences of pathological tissues, cells arrangements, and water contents between two groups, which needs further studies on H2O concentration.
In this study, to eliminate the impacts on the I concentration caused by interfering factors such as injection rate, dose of contrast agent, body mass index, and individual circulation difference, the concept of NIC is introduced. I concentration in a lesion that can be calculated by spectral-based material extraction. Ascenti et al. found that the whole-tumor iodine quantification is significantly better than standard CT enhancement measurements in distinguishing enhancing from nonenhancing renal masses. Our study shows that NIC in ischemic RCC group in corticomedullary and parenchymal phases is 0.15 ± 0.07 and 0.43 ± 0.13, respectively, and those in MF-RAML group is 0.30 ± 0.08 and 0.56 ± 0.05, respectively. NIC of two phases in ischemic RCC group is lower than MF-RAML group (P < 0.05). I concentrations of ischemic RCC in corticomedullary and parenchymal phases do not differ too much, while the NIC difference is much greater. According to the formula, considering that the blood flow in abdominal aorta is a little quicker, low I concentration of abdominal aorta in parenchymal phase causes ratio increase. Although I concentration of MF-RAML in parenchymal phase is reduced compared with corticomedullary phase, the little amount difference is not sufficient to compete the denominator in formula, namely, big difference of I concentration in abdominal aorta. Therefore, I concentration in parenchymal phase is also increased. The difference of NIC between two groups in parenchymal phase is not as significant as that in corticomedullary phase (0.01< P < 0.05), which may have a great relationship with reduction of I concentration in abdominal aorta. The results analysis shows that NIC and pure I concentration have significant difference between two groups, and pure I concentration can better reflect the blood supply and enhancement characteristics of different tumors, while NIC will be more comparable and has higher diagnostic value because its dispersion and crossover phenomenon are less than those of I concentration.
The energy attenuation curve of different substance is determined by its own chemical molecular structure, and the different chemical composition of different materials can be distinguished by difference of CT value attenuation curve, and this difference can be quantitatively evaluated by the slope rate of attenuation curve.
In this study, CT values, spectral attenuation curves in two groups exhibit different degrees of reduction with the increasing of keV. The overall shapes and trends are similar, exhibiting similar performance as most solid soft tissue tumors' curves. Observation of attenuation curves reveals that the differences of curves are mainly concentrated in the section of 40–100 keV. Therefore, the line of 40–100 keV in above region and angles of abscissa are selected for calculation of curve slope rate. From spectral attenuation curve, it can be seen intuitively that the spectral attenuation curve of MF-RAML group is always above that in ischemic RCC group. CT values of former in 40-140 keV monoenergetic images are higher than latter to various degrees, and the difference is, especially obvious in relative low-energy region. MF-RAML is significantly enhanced in corticomedullary phase and the amount of I contrast agent reaches the most. The injection of I makes the differences of mass absorption coefficient of different tumors enhanced, thus causing increased difference of CT values, especially in low-keV condition. So, the corresponding CT value at 40 keV is the highest, with the steepest shape of corresponding spectral attenuation curve. The changes of absolute CT value are the most obvious. In corticomedullary phase, the enhancement of ischemic RCC is not as obvious as MF-RAML, exhibiting mild to moderate enhancement. I content inside the tumor is relatively low, so the difference of CT value is small, and the corresponding density of overall matter is also low, resulting in small slope rate of spectral attenuation curve of ischemic RCC. The trend will be relatively flat and is located beneath MF-RAML. The enhanced difference of ischemic RCC and MF-RAML in corticomedullary phase is significant, and the corresponding density of overall matter varies greatly. Hence, the differences of corresponding attenuation curve are relatively obvious. The spectral attenuation curve in parenchymal phase in MF-RAML group locates under that in corticomedullary phase and its tendency is also relatively mild. As there is no significant change of I content in both phases in ischemic RCC group, the density of overall matter changes little, so the difference of the overall shape and trend of its spectral attenuation curve of two phases is small, with relatively close slope rate.
When GSI generates the spectral attenuation curve of a substance, EAN of this substance can be obtained through calculation of spectral attenuation curve. The main component of contrast agent, i.e. I, has high-EAN, so the injection of the contrast agent will increase the difference of mass attenuation coefficient of different kinds of tumors. In this study, EAN of ischemic RCC in corticomedullary and parenchymal phases is 8.82 ± 0.48 and 8.84 ± 0.25, respectively, while that of MF-RAML is 9.41 ± 0.26 and 9.27 ± 0.12, respectively. The intergroup difference is statistically significant (P < 0.05). In corticomedullary phase, the blood supply of ischemic RCC is not rich, and the strengthening is not obvious. The contrast agent entering tumor is relatively small, leading to decrease of mass attenuation coefficient difference caused by I, so EAN of ischemic RCC in corticomedullary phase is small. However, the vessels of MF-RAML are relatively abundant, and the enhancement is obvious, with high-contrast agent within tumor. So, the mass attenuation coefficient of this MF-RAML “mixture” varies greatly, leading to high-EAN of MF-RAML in corticomedullary phase. In parenchymal phase, the blood supply of ischemic RCC does not change much, with no obvious enhancement degree, and the content of contrast agent within tumor is almost the same. So, the difference of mass attenuation coefficient of ischemic RCC “mixture” is not great, and EAN change of ischemic RCC in parenchymal phase is little. However, the enhancement level of MF-RANL is reduced, and the intratumoral contrast agent concentration is correspondingly reduced, leading to small difference of mass attenuation coefficient of MF-RAML. So, EAN will also reduce, while EAN difference between two groups still remains statistically significant (P < 0.05).
This study still has some shortcomings which are as follows:First, the number of cases collected was small. Second, no grouping toward the different subtypes of RCC is performed. Finally, as the readers assess the images in consensus, intra- or inter-observer variability data are lacked.
| > Conclusion|| |
GSI-DECT may be helpful for increasing the accuracy of differentiating RCC from AML. Further research will be necessary to evaluate the clinical value of quantitative measurements of I concentration.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: The impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011;61:212-36.
Frank I, Blute ML, Cheville JC, Lohse CM, Weaver AL, Zincke H. Solid renal tumors: An analysis of pathological features related to tumor size. J Urol 2003;170(6 Pt 1):2217-20.
Fujii Y, Komai Y, Saito K, Iimura Y, Yonese J, Kawakami S, et al.
Incidence of benign pathologic lesions at partial nephrectomy for presumed RCC renal masses: Japanese dual-center experience with 176 consecutive patients. Urology 2008;72:598-602.
Kutikov A, Fossett LK, Ramchandani P, Tomaszewski JE, Siegelman ES, Banner MP, et al.
Incidence of benign pathologic findings at partial nephrectomy for solitary renal mass presumed to be renal cell carcinoma on preoperative imaging. Urology 2006;68:737-40.
Halpenny D, Snow A, McNeill G, Torreggiani WC. The radiological diagnosis and treatment of renal angiomyolipoma-current status. Clin Radiol 2010;65:99-108.
Jinzaki M, Tanimoto A, Narimatsu Y, Ohkuma K, Kurata T, Shinmoto H, et al.
Angiomyolipoma: Imaging findings in lesions with minimal fat. Radiology 1997;205:497-502.
Pache G, Krauss B, Strohm P, Saueressig U, Blanke P, Bulla S, et al.
Dual-energy CT virtual noncalcium technique: Detecting posttraumatic bone marrow lesions – Feasibility study. Radiology 2010;256:617-24.
Takahashi N, Vrtiska TJ, Kawashima A, Hartman RP, Primak AN, Fletcher JG, et al.
Detectability of urinary stones on virtual nonenhanced images generated at pyelographic-phase dual-energy CT. Radiology 2010;256:184-90.
Thieme SF, Johnson TR, Lee C, McWilliams J, Becker CR, Reiser MF, et al.
Dual-energy CT for the assessment of contrast material distribution in the pulmonary parenchyma. AJR Am J Roentgenol 2009;193:144-9.
Graser A, Johnson TR, Hecht EM, Becker CR, Leidecker C, Staehler M, et al.
Dual-energy CT in patients suspected of having renal masses: Can virtual nonenhanced images replace true nonenhanced images? Radiology 2009;252:433-40.
Stolzmann P, Frauenfelder T, Pfammatter T, Peter N, Scheffel H, Lachat M, et al.
Endoleaks after endovascular abdominal aortic aneurysm repair: Detection with dual-energy dual-source CT. Radiology 2008;249:682-91.
Lin XZ, Wu ZY, Tao R, Guo Y, Li JY, Zhang J, et al.
Dual energy spectral CT imaging of insulinoma-Value in preoperative diagnosis compared with conventional multi-detector CT. Eur J Radiol 2012;81:2487-94.
Lv P, Lin XZ, Chen K, Gao J. Spectral CT in patients with small HCC: Investigation of image quality and diagnostic accuracy. Eur Radiol 2012;22:2117-24.
Geyer LL, Scherr M, Körner M, Wirth S, Deak P, Reiser MF, et al.
Imaging of acute pulmonary embolism using a dual energy CT system with rapid kVp switching: Initial results. Eur J Radiol 2012;81:3711-8.
Yu Y, He N, Sun K, Lin X, Yan F, Chen K. Differentiating hepatocellular carcinoma from angiomyolipoma of the liver with CT spectral imaging: A preliminary study. Clin Radiol 2013;68:e491-7.
Ji JS, Lu CY, Wang ZF, Xu M, Song JJ. Epithelioid angiomyolipoma of the liver: CT and MRI features. Abdom Imaging 2013;38:309-14.
Ascenti G, Mileto A, Krauss B, Gaeta M, Blandino A, Scribano E, et al.
Distinguishing enhancing from nonenhancing renal masses with dual-source dual-energy CT: Iodine quantification versus standard enhancement measurements. Eur Radiol 2013;23:2288-95.
Israel GM, Silverman SG. The incidental renal mass. Radiol Clin North Am 2011;49:369-83.
Graser A, Becker CR, Staehler M, Clevert DA, Macari M, Arndt N, et al.
Single-phase dual-energy CT allows for characterization of renal masses as benign or malignant. Invest Radiol 2010;45:399-405.
Millet I, Doyon FC, Hoa D, Thuret R, Merigeaud S, Serre I, et al.
Characterization of small solid renal lesions: Can benign and malignant tumors be differentiated with CT? AJR Am J Roentgenol 2011;197:887-96.
Chandarana H, Megibow AJ, Cohen BA, Srinivasan R, Kim D, Leidecker C, et al.
Iodine quantification with dual-energy CT: Phantom study and preliminary experience with renal masses. AJR Am J Roentgenol 2011;196:W693-700.
Lee MJ, Kim S, Lee SA, Song HT, Huh YM, Kim DH, et al.
Overcoming artifacts from metallic orthopedic implants at high-field-strength MR imaging and multi-detector CT. Radiographics 2007;27:791-803.
Lv P, Lin XZ, Li J, Li W, Chen K. Differentiation of small hepatic hemangioma from small hepatocellular carcinoma: Recently introduced spectral CT method. Radiology 2011;259:720-9.
Kang MJ, Park CM, Lee CH, Goo JM, Lee HJ. Dual-energy CT: Clinical applications in various pulmonary diseases. Radiographics 2010;30:685-98.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]