Advertisement

European Radiology

, Volume 30, Issue 2, pp 1254–1263 | Cite as

Radiomics of small renal masses on multiphasic CT: accuracy of machine learning–based classification models for the differentiation of renal cell carcinoma and angiomyolipoma without visible fat

  • Ruimeng Yang
  • Jialiang Wu
  • Lei Sun
  • Shengsheng Lai
  • Yikai Xu
  • Xilong Liu
  • Ying Ma
  • Xin ZhenEmail author
Imaging Informatics and Artificial Intelligence

Abstract

Objective

To investigate the discriminative capabilities of different machine learning–based classification models on the differentiation of small (< 4 cm) renal angiomyolipoma without visible fat (AMLwvf) and renal cell carcinoma (RCC).

Methods

This study retrospectively collected 163 patients with pathologically proven small renal mass, including 118 RCC and 45 AMLwvf patients. Target region of interest (ROI) delineation, followed by texture feature extraction, was performed on a representative slice with the largest lesion area on each phase of the four-phase CT images. Fifteen concatenations of the four-phasic features were fed into 224 classification models (built with 8 classifiers and 28 feature selection methods), classification performances of the 3360 resultant discriminative models were compared, and the top-ranked features were analyzed.

Results

Image features extracted from the unenhanced phase (UP) CT image demonstrated dominant classification performances over features from other three phases. The two discriminative models “SVM + t_score” and “SVM + relief” achieved the highest classification AUC of 0.90. The 10 top-ranked features from UP included 1 shape feature, 5 first-order statistics features, and 4 texture features, where the shape feature and the first-order statistics features showed superior discriminative capabilities in differentiating RCC vs. AMLwvf through the t-SNE visualization.

Conclusion

Image features extracted from UP are sufficient to generate accurate differentiation between AMLwvf and RCC using machine learning–based classification model.

Key Points

• Radiomics extracted from unenhanced CT are sufficient to accurately differentiate angiomyolipoma without visible fat and renal cell carcinoma using machine learning–based classification model.

• The highest discriminative models achieved an AUC of 0.90 and were based on the analysis of unenhanced CT, alone or in association with images obtained at the nephrographic phase.

• Features related to shape and to histogram analysis (first-order statistics) showed superior discrimination compared with gray-level distribution of the image (second-order statistics, commonly called texture features).

Keywords

Angiomyolipoma Renal cell carcinoma Machine learning Classification 

Abbreviations

ACC

Accuracy

AMLwvf

Angiomyolipoma without visible fat

AUC

Area under the ROC curve

CECT

Contrast-enhanced CT

CMP

Corticomedullary phase

EP

Excretory phase

GLCM

Gray-level co-occurrence matrix

GLDM

Gray-level dependence matrix

GLRLM

Gray-level run length matrix

GLSZM

Gray-level size zone matrix

ICC

Interobserver correlation coefficient

NGTDM

Neighboring gray tone difference matrix

NP

Nephrographic phase

PACS

Picture archiving and communication system

RCC

Renal cell carcinoma

ROC

Receiver operating characteristic

ROI

Region of interest

SMOTE

Synthetic minority oversampling technique

t-SNE

T-distributed stochastic neighbor embedding method

UP

Unenhanced phase

Notes

Acknowledgments

We gratefully acknowledge all the members of Department of Radiology, Guangzhou First People’s Hospital, for continuous assistance. In particular, we would like to thank Dr. Zaiyi Liu for his advice during the project.

Funding

This study has received funding from the National Natural Science Foundation of China (81874216 and 81728016), the National Key Research and Development Program of China (2017YFC0112900), the Natural Science Foundation of Guangdong Province, P.R. China (2018A030313282), the Fundamental Research Funds for the Central Universities, SCUT (2018MS23), and the Guangzhou Science and Technology Project, P.R. China (201607010038).

Compliance with ethical standards

Guarantor

The scientific guarantor of this publication is Dr. Xin Zhen.

Conflict of interest

The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.

Statistics and biometry

No complex statistical methods were necessary for this paper.

Informed consent

Written informed consent was waived by the Institutional Review Board.

Ethical approval

Institutional Review Board approval was obtained.

Methodology

• retrospective

• experimental

• multicenter study

Supplementary material

330_2019_6384_MOESM1_ESM.docx (944 kb)
ESM 1 (DOCX 944 kb)

References

  1. 1.
    Simpfendorfer C, Herts BR, Motta-Ramirez GA et al (2009) Angiomyolipoma with minimal fat on MDCT: can counts of negative-attenuation pixels aid diagnosis? AJR Am J Roentgenol 192:438Google Scholar
  2. 2.
    Kim JY, Kim JK, Kim N, Cho KS (2008) CT histogram analysis: differentiation of angiomyolipoma without visible fat from renal cell carcinoma at CT imaging. Radiology 246:472–479CrossRefGoogle Scholar
  3. 3.
    Simpson E, Patel U (2006) Diagnosis of angiomyolipoma using computed tomography—region of interest ≤−10HU or 4 adjacent pixels ≤−10HU are recommended as the diagnostic thresholds. Clin Radiol 61:410–416CrossRefGoogle Scholar
  4. 4.
    Catalano OA, Samir AE, Sahani DV, Hahn PF (2008) Pixel distribution analysis: can it be used to distinguish clear cell carcinomas from angiomyolipomas with minimal fat? Radiology 247:738–746CrossRefGoogle Scholar
  5. 5.
    Silverman SG, Israel GM, Herts BR, Richie JP (2008) Management of the incidental renal mass. Radiology 249:16–31CrossRefGoogle Scholar
  6. 6.
    Xie P, Yang Z, Yuan Z (2016) Lipid-poor renal angiomyolipoma: differentiation from clear cell renal cell carcinoma using wash-in and washout characteristics on contrast-enhanced computed tomography. Oncol Lett 11:2327–2331CrossRefGoogle Scholar
  7. 7.
    Sung CK, Kim SH, Woo S et al (2016) Angiomyolipoma with minimal fat: differentiation of morphological and enhancement features from renal cell carcinoma at CT imaging. Acta Radiol 57:1114–1122CrossRefGoogle Scholar
  8. 8.
    Hodgdon T, McInnes MD, Schieda N, Flood TA, Lamb L, Thornhill RE (2015) Can quantitative CT texture analysis be used to differentiate fat-poor renal angiomyolipoma from renal cell carcinoma on unenhanced CT images? Radiology 276:787–796CrossRefGoogle Scholar
  9. 9.
    Yan L, Liu Z, Wang G et al (2015) Angiomyolipoma with minimal fat: differentiation from clear cell renal cell carcinoma and papillary renal cell carcinoma by texture analysis on CT images. Acad Radiol 22:1115–1121CrossRefGoogle Scholar
  10. 10.
    Lee HS, Hong H, Jung DC, Park S, Kim J (2017) Differentiation of fat-poor angiomyolipoma from clear cell renal cell carcinoma in contrast-enhanced MDCT images using quantitative feature classification. Med Phys 44:3604–3614CrossRefGoogle Scholar
  11. 11.
    Feng Z, Rong P, Cao P et al (2018) Machine learning-based quantitative texture analysis of CT images of small renal masses: differentiation of angiomyolipoma without visible fat from renal cell carcinoma. Eur Radiol 28:1625–1633CrossRefGoogle Scholar
  12. 12.
    van Griethuysen JJM, Fedorov A, Parmar C et al (2017) Computational radiomics system to decode the radiographic phenotype. Cancer Res 77:e104Google Scholar
  13. 13.
    Pyradiomics Documentation. Available via https://pyradiomics.readthedocs.io/en/latest/. Accessed 6 July 2019
  14. 14.
    Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP (2002) SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321–357CrossRefGoogle Scholar
  15. 15.
    He H, Yang B, Garcia EA, Li S (2008) ADASYN: adaptive synthetic sampling approach for imbalanced learning. In: IEEE International Joint Conference on Neural NetworksGoogle Scholar
  16. 16.
    Silverman SG, Mortele KJ, Tuncali K, Jinzaki M, Cibas ES (2007) Hyperattenuating renal masses: etiologies, pathogenesis, and imaging evaluation. Radiographics 27:1131–1143CrossRefGoogle Scholar
  17. 17.
    Woo S, Cho JY, Kim SH, Kim SY (2014) Angiomyolipoma with minimal fat and non-clear cell renal cell carcinoma: differentiation on MDCT using classification and regression tree analysis-based algorithm. Acta Radiol 55:1258–1269CrossRefGoogle Scholar
  18. 18.
    Yang CW, Shen SH, Chang YH et al (2013) Are there useful CT features to differentiate renal cell carcinoma from lipid-poor renal angiomyolipoma? AJR Am J Roentgenol 201:1017–1028CrossRefGoogle Scholar

Copyright information

© European Society of Radiology 2019

Authors and Affiliations

  1. 1.Department of Radiology, Guangzhou First People’s Hospital, School of MedicineSouth China University of TechnologyGuangzhouChina
  2. 2.Department of Radiology, Guangzhou First People’s HospitalGuangzhou Medical UniversityGuangzhouChina
  3. 3.School of Biomedical EngineeringSouthern Medical UniversityGuangzhouChina
  4. 4.Department of Medical EquipmentGuangdong Food and Drug Vocational CollegeGuangzhouChina
  5. 5.Department of Radiology, Nanfang HospitalSouthern Medical UniversityGuangzhouChina

Personalised recommendations