Combined Diagnosis of Whole-Lesion Histogram Analysis of T1- and T2-Weighted Imaging for Differentiating Adrenal Adenoma and Pheochromocytoma: A Support Vector Machine-Based Study

Abstract

French

Purpose:

This study aimed to determine whether a combined diagnosis of whole-lesion histogram analysis of T1- and T2-weighted imaging based on support vector machine (SVM) can distinguish pheochromocytoma from adrenal adenoma.

Methods:

A pathology database was retrospectively appraised over a period of 7 years and we obtained 40 histopathologically proven adrenal adenomas and 20 pheochromocytomas with magnetic resonance images. The T1-weighted imaging (T₁WI, including both in phase and opposed phase) and T2-weighted imaging (T₂WI) images of each patients were analyzed using Mazda software. Nine parameters were selected as indicators of comparison: variance, skewness, kurtosis, mean, 1st percentile, 10th percentile, 50th percentile, 90th percentile, and 99th percentile. The parameters with differential-diagnosis significance were used to establish the combined diagnostic model of SVM.

Results:

Among the 9 parameters extracted using histogram analysis, the 1st percentile, 10th percentile, and 50th percentile of T₁WI (in phase) and the skewness of T₂WI and almost all parameters of T₁WI (opposed phase), except variance and 99th percentile, showed statistical significance between groups. Among the above parameters, the area under the curve (AUC) of 10th percentile of T₁WI (opposed phase) was the largest with the value of 0.909 (100.0% sensitivity and 80.0% specificity). After the analysis of combined diagnosis was performed, the AUC of SVM model in testing set showed the value of 0.917 (85.0% accuracy).

Conclusions:

Whole-lesion histogram analysis of T₁WI and T₂WI may help differentiate adrenal adenomas from pheochromocytomas. Furthermore, the combined diagnosis of T₁WI and T₂WI histogram based on SVM was more effective than most of individual histogram parameters.

Keywords

pheochromocytoma adrenocortical tumor magnetic resonance imaging diagnosis support vector machine

Get full access to this article

View all access options for this article.

References

Fassnacht

Arlt

Bancos

, et al. Management of adrenal incidentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2016;175(2):G1–G34.

Song

Chaudhry

Mayo-Smith

. The incidental adrenal mass on CT: prevalence of adrenal disease in 1,049 consecutive adrenal masses in patients with no known malignancy. AJR Am J Roentgenol. 2008;190(5):1163–1168.

Crona

Beuschlein

Pacak

Skogseid

. Advances in adrenal tumors 2018. Endocr Relat Cancer. 2018;25(7):R405–R420.

Northcutt

Raman

Long

, et al.

MDCT of adrenal masses: can dual-phase enhancement patterns be used to differentiate adenoma and pheochromocytoma?

AJR Am J Roentgenol. 2013;201(4):834–839.

Lenders

Duh

Eisenhofer

, et al. Pheochromocytoma and paraganglioma: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2014;99(6):1915–1942.

Young

. Clinical practice: the incidentally discovered adrenal mass. N Engl J Med. 2007;356(6):601.

Berland

Silverman

Gore

, et al. Managing incidental findings on abdominal CT: white paper of the ACR incidental findings committee. J Am Coll Radiol. 2010;7(10):754–773.

Blake

Krishnamoorthy

Boland

, et al. Low-density pheochromocytoma on CT: a mimicker of adrenal adenoma. AJR Am J Roentgenol. 2003;181(6):1663–1668.

Derlin

Busch

Wisotzki

, et al. Intraindividual comparison of ¹²³I-mIBG SPECT/MRI, ¹²³I-mIBG SPECT/CT, and MRI for the detection of adrenal pheochromocytoma in patients with elevated urine or plasma catecholamines. Clin Nucl Med. 2013;38(1):e1–e6.

10.

Nguyen

Shah

Mortazavi

, et al. Non-invasive quantification of tumour heterogeneity in water diffusivity to differentiate malignant from benign tissues of urinary bladder: a phase I study. Eur Radiol. 2017;27(5):2146–2152.

11.

Just

. Improving tumour heterogeneity MRI assessment with histograms. Br J Cancer. 2014;111(12):2205–2213.

12.

Schieda

Alrashed

Flood

Samji

Shabana

McInnes

. Comparison of quantitative MRI and CT washout analysis for differentiation of adrenal pheochromocytoma from adrenal adenoma. AJR Am J Roentgenol. 2016;206(6):1141–1148.

13.

Lubner

Smith

Sandrasegaran

Sahani

Pickhardt

. CT texture analysis: definitions, applications, biologic correlates, and challenges . Radiographics. 2017;37(5):1483–1503.

14.

Guo

Qiao

, et al. Radiomics analysis on ultrasound for prediction of biologic behavior in breast invasive ductal carcinoma. Clin Breast Cancer. 2018;18(3):e335–e344.

15.

Umanodan

Fukukura

Kumagae

, et al. ADC histogram analysis for adrenal tumor histogram analysis of apparent diffusion coefficient in differentiating adrenal adenoma from pheochromocytoma. J Magn Reson Imaging. 2017;45(4):1195–1203.

16.

Shindo

Fukukura

Umanodan

, et al. Histogram analysis of apparent diffusion coefficient in differentiating pancreatic adenocarcinoma and neuroendocrine tumor. Medicine. 2016;95(4): e2574.

17.

Quan

Wen

Hao

, et al. Histogram analysis of dynamic contrast-enhanced magnetic resonance imaging for differentiating malignant from benign orbital lymphproliferative disorders. Acta Radiol. 2019;60(2):239–246.

18.

Halefoglu

Bas

Yasar

Basak

. Differentiation of adrenal adenomas from nonadenomas using CT histogram analysis method: a prospective study. Eur J Radiol. 2010;73(3):643–651.

19.

Zhang

Liang

Gao

Wang

. Models of logistic regression analysis, support vector machine, and back-propagation neural network based on serum tumor markers in colorectal cancer diagnosis. Genet Mol Res. 2016;15(2). doi:10.4238/gmr.15028643.

20.

Vapnik

Zhang

XG.

Essence of Statistical Theory. Beijing, China: Tsinghua University Press. 2000;91–97.

21.

Szczypinski

Strzelecki

Materka

Klepaczko

. MaZda—a software package for image texture analysis. Comput Methods Programs Biomed. 2009;94(1):66–76.

22.

Chang

Lin

. LIBSVM: a library for support vector machines. ACM Trans Intell Syst Technol. 2011;2(27):1–27.

23.

Han

Jiang

. Overcome support vector machine diagnosis overfitting. Cancer Inform. 2014;13(suppl 1):145–158.

24.

Wang

, et al. Grid Search Optimized SVM Method for Dish-Like Underwater Robot Attitude Prediction. Fifth International Joint Conference on Computational Sciences and Optimization (CSO), IEEE, Harbin, China, 2012.

25.

Boland

Blake

Hahn

Mayo-Smith

. Incidental adrenal lesions: principles, techniques, and algorithms for imaging characterization. Radiology. 2008;249(3):756–775.

26.

Nakajo

Fukukura

, et al. Diagnostic performances of FDG-PET/CT and diffusion-weighted imaging indices for differentiating benign pheochromocytoma from other benign adrenal tumors. Abdom Imaging. 2015;40(6):1655–1665.

27.

Hida

Nishie

Asayama

, et al. Apparent diffusion coefficient characteristics of various adrenal tumors. Magn Reson Med Sci. 2014;13(3):183–189.

28.

Donati

Mazaheri

Afaq

, et al. Prostate cancer aggressiveness: assessment with whole-lesion histogram analysis of the apparent diffusion coefficient. Radiology. 2013;271(1):130973.

29.

Jacques

Sahdev

Sandrasagara

, et al. Adrenal phaeochromocytoma: correlation of MRI appearances with histology and function. Eur Radiol. 2008;18(12):2885–2892.

30.

Halefoglu

Altun

Disli

, et al. A prospective study on the utility of diffusion-weighted and quantitative chemical-shift magnetic resonance imaging in the distinction of adrenal adenomas and metastases. J Comput Assist Tomogr. 2012;36(4):367–374.

31.

Liu

, et al. Apparent diffusion coefficient (ADC) measurements in pancreatic adenocarcinoma: a preliminary study of the effect of region of interest on ADC values and interobserver variability. J Magn Reson Imaging. 2016;43(2):407–413.

32.

Zhang

Wang

Tang

Sun

Cao

Gao

. Support vector machine model for diagnosis of lymph node metastasis in gastric cancer with multidetector computed tomography: a preliminary study. BMC Cancer. 2011;11:10.

33.

Yan

Liu

Wang

, et al. Angiomyolipoma with minimal fat: differentiation from clear cell renal cell carcinoma and papillary renal cell carcinoma by texture analysis on CT images. Acad Radiol. 2015;22(9):1115–1121.