Abstract Archives of the RSNA, 2004
Alexander L. C. Kwan PhD, Presenter: Nothing to Disclose
Nikula Shah MS, Abstract Co-Author: Nothing to Disclose
John Michael Boone PhD, Abstract Co-Author: Nothing to Disclose
The purpose of this investigation was to evaluate the magnitude of detected x-ray scatter (characterized as the scatter to primary ratio, SPR) in geometries relevant to breast CT. The spatially dependent SPR was studied as a function of the x-ray beam energy, breast phantom diameter, and with and without a bowtie filter.
The spatially dependent SPR was measured on a prototype breast CT scanner for 50% glandular cylindrical phantoms at three diameters (10, 14 and 18 cm) with different beam energies (60, 80 and 100 kVp). The primary-plus-scatter images (P+S) were acquired by placing the phantoms at the isocenter of the scanner, while the scatter-only images (S) were obtained using a 70.0 mm thick steel plate (linear beam stop) in front of the phantom. One hundred images at each setting were acquired and averaged to produce the respective image pair (S and P+S). The position of the steel plate was located from the scatter-only image using a minima detection algorithm. The values from these locations at each of the image pairs were used to compute the SPR for each pixel across the image.
While the increase in beam energy has minimal effect on the SPR, an increase in the phantom diameter (from 10 to 18 cm) increases the SPR at the center of the phantom from 0.26 to 0.45 when the bowtie filter is present, and from 0.22 to 0.92 without the bowtie filter. The SPR remains constant across the width of the phantom for the bowtie geometries (except at the edges), whereas the non-bowtie SPR gradually decreases towards the edge of the phantom (from 0.92 to 0.12 for the 18 cm phantom).
The effect of x-ray beam energy, phantom size and the application of a bowtie filter on the SPR of a dedicated cone-beam breast CT scanner was examined. The results from the non-bowtie filter cases are consistent with the general understanding of SPR, while the use of a bowtie filter produces lower (and constant) SPR values at the center region of the phantom. Understanding the geometric scattering characteristics (and their dependencies) that exist in the cone-beam breast CT scanner will facilitate scatter reduction techniques for this emerging technology.
Kwan, A,
Shah, N,
Boone, J,
Analysis of the Scatter to Primary Ratio for a Dedicated Cone-beam Breast CT Scanner. Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL.
http://archive.rsna.org/2004/4414650.html