Materials and methodology

Phantom model (CDRAD phantom)

The CDRAD type 2.0 phantom (Artinis Medical Systems, Zetten, Netherlands) was used for the low contrast-details objects. The CDRAD phantom is 26.5 x 26.5 cm in size with 1 cm thickness of Plexiglas plate. It contains 225 drilled holes of varying depths (0.3–8.0 mm) and different diameters (0.3–8.0 mm), so that the CDRAD phantom has circular discs

with varying contrast levels and diameter sizes (Figure 2.9 in Chapter 2). The 225 circular details are arranged in 15 columns and 15 rows. Each row has 15 holes of exactly the same diameter but different contrast levels due to the gradually varying depths of the holes. Each column has 15 holes with exactly the same contrast level but different diameters. The first three rows contain only one detail per square (Figure 2.9 in Chapter 2) while the remaining 12 rows contain two identical details per square with the same hole depth and diameter. One detail is located in the centre of the square and the second detail is located in a randomly chosen corner (Pascoal et al. 2005; Uffmann et al. 2004).

The CDRAD phantom was inserted between 10 cm thickness of Perspex sheets, with 5 cm thickness of Perspex above and 5 cm underneath the phantom (Figure 3.1). The Perspex is used to simulate attenuation of the anatomical region of an additional 10 cm of soft tissue and provides a homogenous scatter source (Pascoal et al. 2005; Uffmann et al. 2004).

Figure 3.1 CDRAD phantom is inserted in the middle of 10 cm thickness of Perspex.

Detector types

Digital radiographs of the CDRAD phantom were obtained using three systems: CR, IDR and DDR. The specifications of these systems are provided in Table 3.1. The table also shows detective quantum efficiency (DQE) of each system. Quality assurance tests of the performance of x-ray units—including half value layer (HVL), linearity and reciprocity, and accuracy and reproducibility—were undertaken and all units passed all tests (Appendix 2).

Table 3.1 Specification of digital radiographic systems

System type CR IDR DDR

Product name AGFA/CR 75.0 /IP

CDMD 4.1

Carestream DRX- 1C

Shimadzu RADspeed Safire

Tube Trex TM65 Varian A-192 Shimadzu

Focal spot Large (1.2mm) Large (1.2 mm) Large (1.2mm)

Detector material The phosphor

(BaFBrx I1-x) CsI scintillator

Amorphous selenium 1000 µm

Pixel size 150 µm/pixel

(6 pixels/mm) 139 µm 150 µm

Detector size/type 350 x 430 mm

IP code 38 350 x 430 mm FPD 432 x 432 mm FPD

Anti-scatter grid Bucky table 8:1

103/inch Bucky table 8:1 115/inch Bucky table 10:1 100/inch Resolution Standard: 3.4lp/mm High: 5.0 lp/mm 3.6 lp/mm 3.3lp/mm DQE DQE(1lp/mm) DQE(2lp/mm) 20% to 30% 18% 9% 60% to 80% 50% 35% 40% 55% 40%

QA tests Pass Pass Pass

Image acquisition

The CDRAD phantom and 10 cm Perspex sheets were imaged at various values of tube voltage (80, 90, 100 and 110 kVp) and tube current levels (1, 2, 4 and 8 mAs). The eight mAs setting was only used with 80 kVp (Table 3.2). The size of the collimation area was

fixed. FDD was maintained at a fixed distance of 100 cm. The Bucky grid table was used for all images. Three images of the CDRAD phantom at each exposure setting were acquired from each system (CR, IDR and DDR). The soft copy images were coded and saved on CD-ROMs as image files in DICOM format (Table 3.3).

Table 3.2 Exposure values of CDRAD phantom images of each system Thickness 10 cm, FFD 100cm with bucky, large focal spot

kVp 80 90 100 110

mAs 1 2 4 8 1 2 4 1 2 4 1 2 4

Table 3.3 The codes of different exposure factors images from different systems

Image exposure factors Image codes

kVp mAs CR system IDR system DDR system

80 1 CR-80/1 IDR-80/1 DDR-80/1 80 2 CR-80/2 IDR-80/2 DDR-80/2 80 4 CR-80/4 IDR-80/4 DDR-80/4 80 8 CR-80/8 IDR-80/8 DDR-80/8 90 1 CR-90/1 IDR-90/1 DDR-90/1 90 2 CR-90/2 IDR-90/2 DDR-90/2 90 4 CR-90/4 IDR-90/4 DDR-90/4 100 1 CR-100/1 IDR-100/1 DDR-100/1 100 2 CR-100/2 IDR-100/2 DDR-100/2 100 4 CR-100/4 IDR-100/4 DDR-100/4 110 1 CR-110/1 IDR-110/1 DDR-110/1 110 2 CR-110/2 IDR-110/2 DDR-110/2 110 4 CR-110/4 IDR-110/4 DDR-110/4

Image scoring

The CDRAD analyser software, version 2.1.9 (Artinis Medical Systems, Zetten, Netherlands), was used to score the images. The CDRAD analyser is dedicated software developed specifically for CDRAD phantom images and designed to provide quantitative analysis of image quality. At each of the 255 matrix locations, the software determines if a difference between the object and background exists. The Welch Satterthwaite test (Student t-tests with Welch correction) is applied in order to determine whether a certain LCD combination was detected or not (Pascoal et al. 2005). An a priori difference of means (APD) is also applied to allow a valid comparison of automated scores obtained from images stored with different bit-depth (Pascoal et al. 2005). The CDRAD analyser was used to calculate the IQFinv values using Equation 2.1 in Chapter 2 (Pascoal et al. 2005; Thijssen

et al. 1989). All image sets—each consisting of three images with identical exposure factors—were evaluated by the CDRAD analyser software.

Statistical analysis

Gaussian distributed was used to test the distribution normality of the scores on each variable. The Gaussian distribution, which is also called normal distribution, is a function that tests the probability of whether the scores on each variables real fall between any two real limits. The dependent scores of IQFinv values appear to be reasonably and normally

distributed. A two-way between-groups analysis of variance (ANOVA) using the Statistical Package for the Social Sciences (SPSS) software was conducted in data analysis. The two- way ANOVA is statistics test used to examine the influence of different categorical independent variables on one dependent variable. The two-way ANOVA is used when there

is more than one independent variable and multiple observations for each independent variable. The two-way ANOVA can determine the main effect of contributions of each independent factor and also can identify if there is a significant interaction effect between different independent factors on one dependent factor. Student t-tests, at an Alpha value of 0.05 is conducted as a part of the two-way ANOVA calculations to determine significance. So that, the two-way ANOVA test was used to determine the impact of the exposure factors including kVp and mAs and the effects of the different radiography system on the values of IQFinv. This test also used to explore if there is a significant interaction effect between

these factors (Pallant 2013). Tukey Honest Significant Difference (HSD) test was also conducted. Tukey HSD is a post-hoc test as it is performed after an analysis of variance, the two-way ANOVA test. The Tukey HSD test was used to determine which groups in the sample differ. Even though the two-way ANOVA can indicate whether groups in the sample differ, it cannot determine which groups differ. While the two-way ANOVA was used to determine if there is significant difference among the groups, the Tukey HSD test was used to determine groups in differ significantly (Pallant 2013).