Quantification of heterogeneity to classify benign parotid tumors: a feasibility study on most frequent histotypes

Future Oncol. 2020 Apr;16(12):763-778. doi: 10.2217/fon-2019-0736. Epub 2020 Apr 6.

Abstract

Aim: To differentiate Warthin tumors (WTs) and pleomorphic adenomas (PAs) measuring heterogeneity of intravoxel incoherent motion (IVIM) and dynamic-contrast enhanced-magnetic resonance imaging biomarkers. Methods: Volumes of interest were traced on 18 WT and 18 PA in 25 patients. For each IVIM and dynamic-contrast enhanced biomarker, histogram parameters were calculated and then compared using the Wilcoxon-signed-rank test. Receiver operating characteristic curves and multivariate analysis were employed to identify the parameters and their pairs with the best accuracy. Results: Most of the biomarkers exhibited significant difference (p < 0.05) between PA and WT for histogram parameters. Time to peak median and skewness, and D* median and entropy showed the highest area under the curve. No meaningful improvement of accuracy was obtained using two features. Conclusion: IVIM and dynamic-contrast enhanced histogram descriptors may help in the classification of WT and PA.

Keywords: DCE-MRI; DW-MRI; Warthin tumor; histogram; intravoxel incoherent motion; magnetic resonance; parotid tumors; pleomorphic adenoma; volume of interest.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Algorithms
  • Biological Variation, Population
  • Feasibility Studies
  • Female
  • Histocytochemistry
  • Humans
  • Image Processing, Computer-Assisted
  • Magnetic Resonance Imaging / methods
  • Male
  • Middle Aged
  • Models, Theoretical
  • Neoplasm Grading
  • Neoplasm Staging
  • Parotid Neoplasms / diagnosis*
  • ROC Curve
  • Retrospective Studies
  • Young Adult