New Approaches in SPECT Breast Imaging

Abstract

The primary objective of this research is to develop new techniques in single photon emission computed tomography (SPECT) breast cancer imaging providing improved detection and characterization of early stage breast cancer. Projection data were acquired by implementing a tiltable SPECT system on a breast and torso phantom. The data were reconstructed with both an ordered subset expectation maximization (OSEM) algorithm and a filtered backprojection (FPP) algorithm with a ramp filter. Results indicated that the OSEM values for SNR and contrast were higher at all tilt angles and may offer better shape and uniform activity distribution of the breast compared to FBP images. Also, by using a tilted head SPECT system, the radius of rotation decreases and the chest wall near the breast can be adequately imaged. This results in increased resolution and sensitivity. A pinhole collimator is advantageous in imaging a breast due to its ability to get close to the breast and achieve a small radius of rotation, though lead shielding was ineffective in eliminating background activity from the myocardium and liver. Statistical analysis of differing acquisition geometries for incomplete circular orbits for breast pinhole SPECT imaging have shown greater reduction in sampling artifacts.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 2005
Accession Number
ADA443756

Entities

People

  • J. Bowsher
  • K. Bobkov
  • Kelly E. Braun
  • R. Jaszczak

Organizations

  • Duke University

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Abstracts
  • Acquisition
  • Algorithms
  • Biomedical Engineering
  • Breast Cancer
  • Circular Orbits
  • Data Acquisition
  • Detection
  • Detectors
  • Diagnostic Imaging
  • Electronic Mail
  • Geometry
  • Image Reconstruction
  • Statistical Analysis
  • Three Dimensional
  • Tomography
  • X-Ray Computed Tomography

Fields of Study

  • Medicine
  • Physics

Readers

  • Image Processing and Computer Vision.
  • Medical Imaging.

Technology Areas

  • Space
  • Space - Space Objects