Wavelet-based feature extraction applied to small-angle x-ray scattering patterns from breast tissue: a tool for differentiating between tissue types

Phys Med Biol. 2006 May 21;51(10):2465-77. doi: 10.1088/0031-9155/51/10/007. Epub 2006 Apr 26.

Abstract

This paper reports on the application of wavelet decomposition to small-angle x-ray scattering (SAXS) patterns from human breast tissue produced by a synchrotron source. The pixel intensities of SAXS patterns of normal, benign and malignant tissue types were transformed into wavelet coefficients. Statistical analysis found significant differences between the wavelet coefficients describing the patterns produced by different tissue types. These differences were then correlated with position in the image and have been linked to the supra-molecular structural changes that occur in breast tissue in the presence of disease. Specifically, results indicate that there are significant differences between healthy and diseased tissues in the wavelet coefficients that describe the peaks produced by the axial d-spacing of collagen. These differences suggest that a useful classification tool could be based upon the spectral information within the axial peaks.

Publication types

  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Artificial Intelligence*
  • Breast Neoplasms / classification*
  • Breast Neoplasms / diagnostic imaging*
  • Diagnosis, Differential
  • Female
  • Humans
  • Pattern Recognition, Automated / methods*
  • Radiographic Image Enhancement / methods*
  • Radiographic Image Interpretation, Computer-Assisted / methods*
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Software
  • X-Ray Diffraction / methods*