Keywords:
Pelvis, Digital radiography, Diagnostic procedure
Authors:
K. S. Alzyoud1, A. England2, B. Snaith3, K. Flintham, P. Hogg2; 1salford/UK, 2Manchester/UK, 3Wakefield/UK
DOI:
10.1594/ecr2018/C-0425
Methods and materials
An adult anthropomorphic pelvis phantom (transparent sectional pelvis model number RS-113T) was imaged using direct digital radiography.
Commercially available catering lard (fat) was used to simulate fat equivalent material as it has similar x-ray characteristics to human fat.
This was established by CT Hounsfield Unit analysis.
Figure.1 shows the experimental setup; the container sitting on the pelvis phantom has 1cm gradations,
this was for fat thicknesses of 1-15cm,
at 1cm intervals,
to be poured in.
The acquisition parameters were chosen according to the clinical practice used for pelvis AP radiography and recommended in many published works [7– 9].
A control image was acquired and later used as a reference image for analysis purposes.
Exposure conditions for the reference image included: no added fat,
automatic exposure control (AEC),
table Bucky and standard 100cm SID without added filtration and 80kVp.
144 experimental images were acquired,
using 1-15cm of fat stepping through 1cm thicknesses and for 70-110kVp stepping through 5 kVp intervals; other than this all exposure conditions were the same as the control/reference image.
Image quality was evaluated using physical measurements (SNR and CNR).
Four ROIs were drawn in homogeneous structures within the anthropomorphic pelvic phantom images,
in order to sample the mean and standard deviation of the pixel value image j was used.
SNR and CNR were calculated using the following equations:
CNR=(ROI1-ROI2)/σnoise
and
SNR= ROI1/σnoise where ROI1 is the mean signal from interest anatomy and ROI2 is the mean signal from noise
Image quality was also assessed using relative visual grading.
Six diagnostic radiographers with more than five years clinical experience who were blinded to image acquisition parameters assessed the images.
Effective dose (E) was calculated using PCXMC software.