Advertisement

Calcifications

  • Ralph Highnam
  • Michael Brady
Part of the Computational Imaging and Vision book series (CIVI, volume 14)

Abstract

One of the fundamental assumptions that underpins the h int representation is that we can ignore calcifications and assume that the breast consists either of fat or of non-fat “interesting tissue”. In practice, small regions of calcification appear quite regularly in normal and abnormal mammograms, Figure 11.1 shows examples. Microcalcifications can be the only mammographic sign of non-palpable breast disease. For this reason, the computeraided detection and classification of (micro)calcifications continues to be one of the major goals of mammographic image processing. However, despite the relatively large amount of effort expended toward the goal of assisting radiologists to detect and interpret calcifications, it is only recently that algorithms appear to have achieved a level of performance where they could be used in routine clinical practice. This chapter aims to show how the h int representation can contribute significantly to this goal, since it can be used to:
  • Detect film-screen “shot” noise which can be confused with calcifications, and thus is a major source of the false positives that downgrade the performance of algorithms for detecting microcalcifications;

  • Estimate the thickness of calcifications, further contributing to the quantitative information that can be provided to the clinician;

  • Improve the detection of calcifications by removing the effects of variations in the background and imaging parameters.

Keywords

Image Noise Shot Noise Noise Detection Focal Spot Size Digitise Mammogram 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer Science+Business Media Dordrecht 1999

Authors and Affiliations

  • Ralph Highnam
    • 1
  • Michael Brady
    • 1
  1. 1.Department of Engineering ScienceOxford UniversityOxfordUK

Personalised recommendations