Novel patient optimized acquisition trajectories for dedicated breast SPECT imaging


Journal Article

Novel acquisition trajectories developed for our dedicated breast SPECT camera move 3 dimensionally within a hemispherical volume, fully contouring a patient's pendent breast to provide a high quality, high resolution 3D functional image. Each unique trajectory, created in under a minute, is tailored for each breast of each subject to obtain the highest image quality for a particular study. If a suspected lesion location is known prior to the scan, a trajectory can be created with many close and direct views of the lesion. A torso phantom with an attached 1730 mL breast phantom containing a 2.1 mL (0.8cm radius) spherical lesion was filled with clinical levels of activity: heart:liver:torso: breast: lesion concentration ratio 12:12:1:1:6. A variety of novel acquisition trajectories were employed to image the lesion. Sequentially increasing tilted parallel beam trajectories investigated signals obtained from different polar angles for imaging the breast and chest wall with contamination from the heart and liver. These studies yielded a bound on polar positions for all azimuthal locations in order to minimize background contamination. Other trajectories were created to obtain the best lesion signal. This study shows sinusoidal trajectories can recover the breast's shape and image into the chest wall best. Changing the camera's starting position or subtracting projection views can reduce cardiac and hepatic contamination in the reconstructed image. However, more than one trajectory may provide equivalent image quality. Acquisition trajectories can be created to meet specific imaging goals which consider certain patient factors, such as breast size, lesion location and cardiac and hepatic uptake. © 2008 IEEE.

Full Text

Duke Authors

Cited Authors

  • Perez, KL; Cutler, SJ; Madhav, P; Tornai, MP

Published Date

  • December 1, 2008

Published In

Start / End Page

  • 5629 - 5634

International Standard Serial Number (ISSN)

  • 1095-7863

Digital Object Identifier (DOI)

  • 10.1109/NSSMIC.2008.4774521

Citation Source

  • Scopus