Advertisement

Computed tomography texture features can discriminate benign from malignant lymphadenopathy in pediatric patients: a preliminary study

  • Alexis M. CahalaneEmail author
  • Aoife Kilcoyne
  • Azadeh Tabari
  • Shaunagh McDermott
  • Michael S. Gee
Original Article

Abstract

Background

Differentiation of benign from malignant lymphadenopathy remains challenging in pediatric radiology. Textural analysis (TA) quantitates heterogeneity of tissue signal intensities and has been applied to analysis of CT images.

Objective

The purpose of this study was to establish whether CT textural analysis of enlarged lymph nodes visualized on pediatric CT can distinguish benign from malignant lymphadenopathy.

Materials and methods

We retrospectively identified enlarged lymph nodes measuring 10–20 mm on contrast-enhanced CTs of patients age 18 years and younger that had been categorized as benign or malignant based on the known diagnoses. We placed regions of interest (ROIs) over lymph nodes of interest and performed textural analysis with and without feature size filtration. We then calculated test performance characteristics for TA features, along with multivariate logistic regression modeling using Akaike Information Criterion (AIC) minimization, to determine the optimal thresholds for distinguishing benign from malignant lymphadenopathy.

Results

We identified 34 enlarged malignant nodes and 29 benign nodes from 63 patients within the 10- to 20-mm size range. Filtered image TA exhibited 82.4% sensitivity, 86.2% specificity and 84.1% accuracy for detecting malignant lymph nodes using mean and entropy parameters, whereas unfiltered TA exhibited 88.2% sensitivity, 72.4% specificity and 81.0% accuracy using mean and mean value of positive pixels parameters.

Conclusion

This preliminary study demonstrates that the use of TA features improves the utility of pediatric CT to distinguish benign from malignant lymphadenopathy. The addition of TA to pediatric CT protocols has great potential to aid the characterization of indeterminate lymph nodes. If definitive differentiation between benign and malignant lymphadenopathy is possible by TA, it has the potential to reduce the need for follow-up imaging and tissue sampling, with reduced associated radiation exposure. However future studies are needed to confirm the clinical applicability of TA in distinguishing benign from malignant lymphadenopathy.

Keywords

Children Computed tomography Lymph nodes Lymphadenopathy Textural analysis Young adults 

Notes

Acknowledgments

The authors acknowledge the MacErlaine Research Scholarship from the Academic Radiology Research Trust, St. Vincent’s Radiology Group, Dublin, Ireland, as well as the Higher Degree Bursary from the Faculty of Radiologists, RCSI, Ireland.

Compliance with ethical standards

Conflicts of interest

None

References

  1. 1.
    Seber T, Caglar E, Uylar T et al (2015) Diagnostic value of diffusion-weighted magnetic resonance imaging: differentiation of benign and malignant lymph nodes in different regions of the body. Clin Imaging 39:856–862Google Scholar
  2. 2.
    van den Brekel MW, Stel HV, Castelijns JA et al (1990) Cervical lymph node metastasis: assessment of radiologic criteria. Radiology 177:379–384Google Scholar
  3. 3.
    McLoud TC, Bourgouin PM, Greenberg RW et al (1992) Bronchogenic carcinoma: analysis of staging in the mediastinum with CT by correlative lymph node mapping and sampling. Radiology 182:319–323Google Scholar
  4. 4.
    Curtin HD, Ishwaran H, Mancuso AA et al (1998) Comparison of CT and MR imaging in staging of neck metastases. Radiology 207:123–130Google Scholar
  5. 5.
    Friedmann AM (2008) Evaluation and management of lymphadenopathy in children. Pediatr Rev 29:53–60Google Scholar
  6. 6.
    Locke R, MacGregor F, Kubba H (2016) The validation of an algorithm for the management of paediatric cervical lymphadenopathy. Int J Pediatr Otorhinolaryngol 81:5–9Google Scholar
  7. 7.
    Papadopouli E, Michailidi E, Papadopoulou E et al (2009) Cervical lymphadenopathy in childhood epidemiology and management. Pediatr Hematol Oncol 26:454–460Google Scholar
  8. 8.
    Sheikhbahaei S, Ahn SJ, Young B et al (2017) Comparative effectiveness: 18F-FDG-PET/CT versus CT for post-treatment follow-up of patients with lung cancer. Nucl Med Commun 38:720–725Google Scholar
  9. 9.
    Cha J, Kim S, Wang J et al (2018) Evaluation of (18)F-FDG PET/CT parameters for detection of lymph node metastasis in cutaneous melanoma. Nucl Med Mol Imaging 52:39–45Google Scholar
  10. 10.
    Tang J, Salloum D, Carney B et al (2017) Targeted PET imaging strategy to differentiate malignant from inflamed lymph nodes in diffuse large B-cell lymphoma. Proc Natl Acad Sci U S A 114:E7441–e7449Google Scholar
  11. 11.
    Guerrera F, Renaud S, Schaeffer M et al (2017) Low accuracy of computed tomography and positron emission tomography to detect lung and lymph node metastases of colorectal cancer. Ann Thorac Surg 104:1194–1199Google Scholar
  12. 12.
    Tourassi GD (1999) Journey toward computer-aided diagnosis: role of image texture analysis. Radiology 213:317–320Google Scholar
  13. 13.
    Kassner A, Thornhill RE (2010) Texture analysis: a review of neurologic MR imaging applications. AJNR Am J Neuroradiol 31:809–816Google Scholar
  14. 14.
    Canellas R, Mehrkhani F, Patino M et al (2016) Characterization of portal vein thrombosis (neoplastic versus bland) on CT images using software-based texture analysis and thrombus density (Hounsfield units). AJR Am J Roentgenol 207:W81–w87Google Scholar
  15. 15.
    Giganti F, Antunes S, Salerno A et al (2017) Gastric cancer: texture analysis from multidetector computed tomography as a potential preoperative prognostic biomarker. Eur Radiol 27:1831–1839Google Scholar
  16. 16.
    Raman SP, Chen Y, Schroeder JL et al (2014) CT texture analysis of renal masses. pilot study using random forest classification for prediction of pathology. Acad Radiol 21:1587–1596Google Scholar
  17. 17.
    Eilaghi A, Baig S, Zhang Y et al (2017) CT texture features are associated with overall survival in pancreatic ductal adenocarcinoma — a quantitative analysis. BMC Med Imaging 17:38Google Scholar
  18. 18.
    Ng F, Ganeshan B, Kozarski R et al (2013) Assessment of primary colorectal cancer heterogeneity by using whole-tumor texture analysis: contrast-enhanced CT texture as a biomarker of 5-year survival. Radiology 266:177–184Google Scholar
  19. 19.
    Singh S, Kalra MK, Moore MA et al (2009) Dose reduction and compliance with pediatric CT protocols adapted to patient size, clinical indication, and number of prior studies. Radiology 252:200–208Google Scholar
  20. 20.
    Goh V, Ganeshan B, Nathan P et al (2011) Assessment of response to tyrosine kinase inhibitors in metastatic renal cell cancer: CT texture as a predictive biomarker. Radiology 261:165–171Google Scholar
  21. 21.
    Ganeshan B, Panayiotou E, Burnand K et al (2012) Tumour heterogeneity in non-small cell lung carcinoma assessed by CT texture analysis: a potential marker of survival. Eur Radiol 22:796–802Google Scholar
  22. 22.
    Ganeshan B, Skogen K, Pressney I et al (2012) Tumour heterogeneity in oesophageal cancer assessed by CT texture analysis: preliminary evidence of an association with tumour metabolism, stage, and survival. Clin Radiol 67:157–164Google Scholar
  23. 23.
    Ganeshan B, Abaleke S, Young RC et al (2010) Texture analysis of non-small cell lung cancer on unenhanced computed tomography: initial evidence for a relationship with tumour glucose metabolism and stage. Cancer Imaging 10:137–143Google Scholar
  24. 24.
    Weiss GJ, Ganeshan B, Miles KA et al (2014) Noninvasive image texture analysis differentiates K-ras mutation from pan-wildtype NSCLC and is prognostic. PLoS One 9:e100244Google Scholar
  25. 25.
    Hayano K, Tian F, Kambadakone AR et al (2015) Texture analysis of non-contrast-enhanced computed tomography for assessing angiogenesis and survival of soft tissue sarcoma. J Comp Assist Tomogr 39:607–612Google Scholar
  26. 26.
    Miles KA, Ganeshan B, Hayball MP (2013) CT texture analysis using the filtration-histogram method: what do the measurements mean? Cancer Imaging 13:400–406Google Scholar
  27. 27.
    Shenoy-Bhangle A, Nimkin K, Goldner D et al (2014) MRI predictors of treatment response for perianal fistulizing Crohn disease in children and young adults. Pediatr Radiol 44:23–29Google Scholar
  28. 28.
    Song HJ, Kim JO, Eun SH et al (2007) Endoscopic ultrasonograpic findings of benign mediastinal and abdominal lymphadenopathy confirmed by EUS-guided fine needle aspiration. Gut Liver 1:68–73Google Scholar
  29. 29.
    Krestan C, Herneth AM, Formanek M, Czerny C (2006) Modern imaging lymph node staging of the head and neck region. Eur J Radiol 58:360–366Google Scholar
  30. 30.
    Luciani A, Itti E, Rahmouni A et al (2006) Lymph node imaging: basic principles. Eur J Radiol 58:338–344Google Scholar
  31. 31.
    Yamazaki Y, Saitoh M, Notani K et al (2008) Assessment of cervical lymph node metastases using FDG-PET in patients with head and neck cancer. Ann Nucl Med 22:177–184Google Scholar
  32. 32.
    Bure L, Boucher LM, Blumenkrantz M et al (2017) Can magnetic resonance spectroscopy differentiate malignant and benign causes of lymphadenopathy? An in-vitro approach. PLoS One 12:e0182169Google Scholar
  33. 33.
    Goske MJ, Strauss KJ, Coombs LP et al (2013) Diagnostic reference ranges for pediatric abdominal CT. Radiology 268:208–218Google Scholar
  34. 34.
    Kim ER, Chang DK (2014) Colorectal cancer in inflammatory bowel disease: the risk, pathogenesis, prevention and diagnosis. World J Gastroenterol 20:9872–9881Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of RadiologyMassachusetts General HospitalBostonUSA

Personalised recommendations