Initial investigation of novel trajectories to improve chest wall imaging in a dedicated breast computed tomography system


Journal Article

In current dedicated breast computed tomography (mammotomography) systems, comfortable patient positioning on a stationary bed restricts the practicable range of source-detector trajectories, thus compromising the system's ability to adequately image the patient's anterior chest wall. This study examines the effect on detecting small, low-contrast lesion-like-spheres using limited angle x-ray source-detector trajectories and trajectories that intentionally raise the tomographic imaging system mid-acquisition. These modified acquisition paths may increase chest wall visualization, simplify the design of the imaging system and increase patient comfort by allowing the design of an improved patient bed. Thin walled balloons of various volumes filled with iodine act as surrogate high contrast lesions to initially investigate the effect of these novel trajectories. Then, stacks of 5mm acrylic spheres regularly spaced in concentric circles are placed in water to simulate a low contrast environment in a uniform scatter medium. 360° azimuthal scans are acquired at various bed heights with contiguous projections subsequently removed to create limited angle acquisitions from 240-360°. Projections from the different bed heights are interwoven to form trajectories that mimic discontinuously raising the imaging system mid-acquisition. The resulting iteratively reconstructed volumes are evaluated with an observer study. Initial images suggest that using limited angles and raising the system is possible while increasing the observer's ability to visualize objects near the chest wall. Based on the results of this study, an improved patient bed to facilitate chest wall imaging will be designed, and the feasibility of vertical system motion to increase imaged breast volume explored. © 2009 SPIE.

Full Text

Duke Authors

Cited Authors

  • Crotty, DJ; McKinley, RL; Madhav, P; Cutler, SJ; Tornai, MP

Published Date

  • June 15, 2009

Published In

Volume / Issue

  • 7258 /

International Standard Serial Number (ISSN)

  • 1605-7422

Digital Object Identifier (DOI)

  • 10.1117/12.812500

Citation Source

  • Scopus