Differentiation of High-Attenuation Renal Cyst and RCC with CT Texture Analysis on Unenhanced CT

Abstract

Objective: The goal of this research is to evaluate the efficiency of computed tomography texture analysis in differentiating renal cell carcinoma from a high-attenuation renal cyst on non-contrast computed tomography. Methods: Forty-nine non-contrast abdominal computed tomography examinations, 27 patients with high-attenuation renal cyst and 22 patients with renal cell carcinoma were evaluated retrospectively. Region of interest was drawn to cover the entire lesion in the sections. Gray-level intensity (Hounsfield Unit value), entropy, standard deviation, uniformity, kurtosis, skewness, size% lower, size % mean, size% upper, values were obtained by texture analysis. The findings of both groups were compared statistically. Results: Mean and median gray-level intensity values and entropy values were significantly higher in renal cell carcinoma than in high-attenuation renal cyst (p<0.001). There was no significant difference in other parameters. When receiver operator characteristics analysis was performed for the mean value, the area under the curve value was found to be 0,754. When the threshold value was selected as 34.5708, 72.7% sensitivity and 66.7% specificity was found. Conclusion: Texture analysis may be useful in differentiating renal cell carcinoma from high-attenuation renal cysts on non-contrast computed tomography.

Keywords

• Texture analysis
• High-attenuation renal cyst
• Renal cell carcinoma

References

1. 1.Ng CS, Wood CG, Silverman PM, et-al. Renal cellcarcinoma: diagnosis, staging, and surveillance. AJRAm J Roentgenol. 2008; 191: 1220-32.
2. 2.Ferlay J, et al. GLOBOCAN 2012 v1. 0, Cancerincidence and mortality worldwide: IARCCancerBase No. 11. 2013. [accessed on Aug 4,2016];International Agency for Research on CancerWeb site. 2016 Available online:http://globocan.iarc.fr.
3. 3.Bakır Ş, Özekinci Selver. The PrognosticSignificance of the KI-67 and CD-44 in Renal CellCarcinomas. Dicle Med J. 2005; 32: 123-30.
4. 4.Federle MP, Jeffrey RB, Woodward PJ, et al.Diagnostic Imaging: Abdomen, Published byAmirsys®. Lippincott Williams & Wilkins. (2009)ISBN:1931884714
5. 5. Ward RD, Tanaka H, Campbell SC, Remer EM. 2017AUA renal mass and localized renal cancerguidelines: imaging implications. RadioGraphics.2018; 38: 2021–33.
6. 6.O’Connor SD, Pickhardt PJ, Kim DH, Oliva MR,Silverman SG. Incidental finding of renal masses atunenhanced CT: prevalence and analysis of featuresfor guiding management. AJR. 2011; 197: 139–45.
7. 7.Moreno CC, Hemingway J, Johnson AC, et al.Changing abdominal imaging utilization patterns:perspectives from medicare beneficiaries over twodecades. J Am Coll Radiol. 2016; 13: 894–903.
8. 8.McHugh K, Stringer DA, Hebert D, et-al. Simplerenal cysts in children: diagnosis and follow-up with US. Radiology. 1991; 178: 383-5.

9. 9.Jonisch AI, Rubinowitz AN, Mutalik PG, Israel GM.Can High-Attenuation Renal Cysts Be Differentiatedfrom Renal Cell Carcinoma at Unenhanced CT?Radiology. 2007; 243: 445-50.
10. 10.Coleman BG, Arger PH, Mintz MC, Pollack HM,Banner MP. Hyperdense renal masses: a computedtomographic dilemma. AJR Am J Roentgenol. 1984;143: 291–4.
11. 11.Herts BR, Silverman SG, Hindman NM, et al.Management of the incidental renal mass on CT: awhite paper of the ACR Incidental FindingsCommittee. J Am Coll Radiol. 2018; 15: 264–73.
12. 12.Catalano O, Nunziata A, Sandomenico F, Siani A.Acute flank pain: comparison of unenhanced helicalCT and ultrasonography in detecting causes otherthan ureterolithiasis. Emerg Radiol. 2002; 9: 146–54.
13. 13.Castellano G, Bonilha L, Li LM, Cendes F. Textureanalysis of medical images. Clin Radiol. 2004; 59:1061-9.
14. 14.McGahan JP, Sidhar K, Fananapazir G, et al. Renalcell carcinoma attenuation values on unenhancedCT: importance of multiple, small regionof-interestmeasurements. Abdom Radiol (NY). 2017; 42:2325–33.
15. 15.Schieda N, Thornhill RE, Al-Subhi M, et al.Diagnosis of sarcomatoid renal cell carcinoma withCT: evaluation by qualitative imaging features andtexture analysis. AJR. 2015; 204: 1013–23.
16. 16.Yu H, Scalera J, Khalid M, et al. Texture analysisas a radiomic marker for differentiating renaltumors. Abdom Radiol (NY). 2017; 42: 2470–8.
17. 17.Raman SP, Chen Y, Schroeder JL, Huang P,Fishman EK. CT texture analysis of renal masses:pilot study using random forest classification forprediction of pathology. Acad Radiol. 2014; 21:1587–96.18.Nathan Y Kim, Meghan G Lubner, Jered TNystrom, et al. Utility of CT Texture Analysis inDifferentiating Low-Attenuation Renal CellCarcinoma From Cysts: A Bi-InstitutionalRetrospective Study. AJR Am J Roentgenol. 2019;213: 1259-66.
18. 19.Hodgdon T, McInnes MDF, Schieda N, et al. Canquantitative CT texture analysis be used todifferentiate fat-poor renal angiomyolipoma fromrenal cell carcinoma on unenhanced CT images?Radiology. 2015; 276: 787–96.
19. 20.Varghese BA, Chen F, Hwang DH, et al.Differentiation of predominantly solid enhancinglipid poor renal cell masses by use of contrast-enhanced CT: evaluating the role of texture in tumorsubtyping. AJR. 2018; 211: 288-96.
20. 21.Kim JY, Kim JK, Kim N, Cho K-S. CT histogramanalysis: differentiation of angiomyolipoma withoutvisible fat from renal cell carcinoma at CT imaging.Radiology. 2008; 246: 472–9.
21. 22.Takahashi N, Takeuchi M, Sasaguri K, et al. CTnegative attenuation pixel distribution and textureanalysis for detection of fat in smallangiomyolipoma on unenhanced CT. Abdom Radiol(NY). 2016; 41: 1142–51.
22. 23.Lubner MG, Stabo N, Abel EJ, Del Rio AM,Pickhardt PJ. CT textural analysis of large primaryrenal cell carcinomas: pretreatment tumorheterogeneity correlates with histologic findingsand clinical outcomes. AJR. 2016; 207: 96–105.