Sonographic texture characterization of salivary gland tumors by fractal analyses

Ultrasound Med Biol. 2005 Oct;31(10):1297-304. doi: 10.1016/j.ultrasmedbio.2005.05.012.

Abstract

The purpose of this study was to quantify the sonographic images of salivary gland tumors. The study population consisted of 21 pleomorphic adenomas, seven Warthin's tumors and 15 malignant tumors. We selected the region-of-interest within the lesion. We made a contour map and evaluated the complexity of the contour lines using the 2-D box-counting (BC) method. We also evaluated the 2-D distribution of the pixel values by the 3-D BC method and the fractional brownian movement (F BM) model. The fractal dimension by the 2-D BC method of pleomorphic adenoma was significantly larger than that of malignant tumors. The FD by the 3-D BC method of pleomorphic adenoma was significantly larger than both that of Warthin's tumors and that of malignant tumors. The Hurst index obtained by the F BM model of pleomorphic adenoma was significantly lower than that of the malignant tumor.

MeSH terms

  • Adenolymphoma / diagnostic imaging
  • Adenoma / diagnostic imaging*
  • Adenoma, Pleomorphic / diagnostic imaging
  • Carcinoma / diagnostic imaging*
  • Diagnosis, Differential
  • Fractals*
  • Humans
  • Image Interpretation, Computer-Assisted*
  • Imaging, Three-Dimensional
  • Neoplasms, Multiple Primary / diagnostic imaging*
  • Parotid Neoplasms / diagnostic imaging
  • Salivary Gland Neoplasms / diagnostic imaging*
  • Salivary Glands / diagnostic imaging
  • Submandibular Gland Neoplasms / diagnostic imaging
  • Ultrasonography