CDMAM 4.0 Contrast-Detail Phantom Manual Ver 2.1

Manual

45 Pages

File Type: PDF

File Size: 1.61 MB

File Name: cdmam_4-0_contrast-detail_phantom_ver_2-1.pdf

Preview

Page 1

Manual CDMAM 4.0

MANUAL CONTRAST DETAIL PHANTOM CDMAM 4.0 and CDMAM 4.0 Analyser Version 2.3

By M.J. Floor and R.J.M. van der Burght

Artinis Medical Systems Einsteinweg 17 6662 PW Elst The Netherlands Tel: + 31 481 350 980; Fax: + 31 842 105 702

Manual CDMAM 4.0

Contents

Introduction to the CDMAM

Description of the CDMAM

2.1

The aluminium base

2.2

The PMMA cover

2.3

The PMMA plates

2.4

Absorption properties

Directions for use of the CDMAM

3.1

Making images

3.2

Handling instructions

3.3

The declaration of conformity

Evaluation of the CDMAM image

4.1

Correction scheme

4.2

Correction examples

4.3

Presentation of the results

CDMAM Analyser

5.1

Introduction

5.2

Image readout and scoring

5.3

Analysis method

Program structure

6.1

Projects

6.2

Groups

6.3

Images

6.4

Analysis

Analysis results

7.1

Image result

Manual CDMAM 4.0

7.2

Group result

7.3

Project result

Toolbar

9.1

File

9.2

Edit

9.3

Project

9.4

View

9.5

Window

9.6

Help

System demands and installation

10.1

Requirements

10.2

Use of the USB Key

FAQ

11.1

Analysis

11.2

License key

Warranty Policy Artinis Medical Systems B.V.

Literature

Appendices

Appendix 1. SCORE FORM OF THE CDMAM PHANTOM

Appendix 2.EVALUATION FORM OF THE CDMAM PHANTOM

Appendix 3: Software License Agreement (CDMAM Analyser Software)

Manual CDMAM 4.0

In mammography it is essential that objects with very small contrast and diameter can be distinguished from the background. Therefore the quality of the technical aspects of mammography equipment should be monitored with regular time-intervals. This is usually performed by measurement of the physical parameters of the X-ray equipment and observation conditions. The main item in quality control however should be the assurance of information transfer from the tissue under examination to the radiologist. Therefore the information content of an image should be monitored. To evaluate the image quality of mammography systems, a contrast detail phantom (CDMAM 3.2) was developed (Bijkerk, Lindeijer and Thijssen, 1995). With this phantom, which has been adapted from the Burger-Rose phantom (Burger, 1950), the threshold contrast of an imaging system is determined as a function of object diameter, by the detection of pairs of low-contrast objects. In the type 3.2 phantom, the objects (gold disks) ranged in diameter from 0.1 - 3.2 mm and in thickness from 0.05 to 1.6 m, resulting in a radiation contrast range of about 1 - 25 % at standard conditions (molybdenum anode, 30 m molybdenum filtration, 28 kV). The constant improvement of image quality in mammography (high contrast film-screen combinations, digital mammography) made it necessary to visualize lower contrast, as well as a higher spatial resolution (smaller objects) at a higher contrast. The CDMAM phantom was modified to facilitate the evaluation of these systems. In the CDMAM 3.4 version, the gold disks range in diameter from 0.06 to 2.0 mm and in thickness from 0.03 to 2.0 m, resulting in a radiation contrast range of about 0.5 - 30 % at standard mammography exposure conditions (Bijkerk, Thijssen and Arnoldussen, 2000). As automatic scoring makes time consuming manual readout obviate, the CDMAM 4.0 could be developed with an increased number of gold discs. With the CDMAM 4.0 the contrast detail (CD) curve of the mammography system under evaluation is determined with a higher accuracy (Floor, Colier and Van der Burght, 2013; Strudley and Young, 2014). More gold discs are available in the relevant areas for psychometric curve fitting. Not only where there added more diameters but also the thicknesses where optimized for each

The CDMAM phantom is a result of the project: “Quality Assurance in Mammography, at the Department of Diagnostic Radiology, University Medical Centre Nijmegen, St. Radboud”. It was designed by M.A.O. Thijssen, K.R. Bijkerk and J.M. Lindeijer, in cooperation with the Medical Instrumental Department.  Molybdenum anode, 30 µm Molybdenum filtration, at 28 kV. 

Manual CDMAM 4.0

individual diameter. The number of gold discs increased from 410 to 672 with improved tolerance specifications based on highly accurate production processes. In the CDMAM 4.0 version, each diameter has its own gold disk thicknesses. The thicknesses are optimized per diameter to the most interesting values per diameter according to the European guidelines for Quality Assurance in Mammography. The gold disks diameter range is from 0.08 to 2.00 in 21 steps. The thicknesses are between 0.012 m – 0.105 m for the 2.00 mm diameter gold discs to 0.590 m – 2.800 m for the 0.08 mm diameter gold discs, all in 16 steps. The phantom has been designed that about half of the disks will be detected by an experienced observer, when state-of-the-art mammography equipment is used with normal clinical settings. The phantom design is kept as close as possible to the CDMAM 3.4 for ease of use of the phantom and the translation ratio from automatic to human observer scores. The PMMA plates delivered by Artinis for simulation of breast thickness have a thickness accuracy of ±1%. Please, acknowledge Artinis Medical System B.V. in all published and/or presented work involving the phantom.

Figure 1. The CDMAM 4.0 (Contrast detail Mammography) phantom.

Manual CDMAM 4.0

The CDMAM phantom, shown in Figure 1, consists of an aluminium base with gold-disks of various thickness and diameter, which is attached to a PMMA cover. It is recommended to use the phantom with PMMA plates to simulate the breast thickness. Usually the CDMAM phantom and 4 PMMA plates with a thickness of 10 mm are used to simulate a breast thickness of 5 cm. The dimensions of phantom and PMMA plates match the standard mammography size (180 x 240 mm). It is strongly recommended to use plates with a thickness precision of ± 1% to minimize attenuation differences, please contact your local distributor or Artinis Medical System for delivery information of such PMMA plates. For use with the automatic readout software CDCOM only use PMMA plates without any markers.

The aluminium base consists of Al 1050 (99.5% pure aluminium) and has a thickness of 0.5 mm. The base has been polished and was anodized black. Gold-disks (99.99 % pure gold) of variable thickness and diameter have been attached to the base by means of a vaporization technique. The disks are arranged in a matrix of 16 rows and 21 columns (Table 2). Within a column the disk-diameter is constant, with an exponential increasing thickness. The thickness steps were chosen largest in the high contrast part of the phantom, due to the nature of contrast detail curves of X-ray systems (Rose, 1973). During vaporization of the gold disks, the aluminium plate is covered with a mask with lasered holes of the desired diameter. Due to small irregularities of the mask and specific deviations related to the vaporization process, the disks can deviate slightly from the intended diameter, and might not be exactly circular. Also the thickness of the disks may deviate from the indicated values, mainly due to small variations in the gold vaporization profile. The disc diameters are logarithmic linear increasing from 0.08 mm to 2.0 mm. The diameter tolerance is ± 0.005 um (Mean ± 2SD).

Manual CDMAM 4.0

The phantom contains 16 optimized thicknesses per diameter. For the 2.0 mm diameter the gold disc thickness varies between 0.012 m and 0.105 m. For the 0.08 mm diameter the gold discs thickness varies between 0.590 m and 2.800 m (Table 2). The thicknesses are optimized to be used in conjunction with the European Guidelines and for enlarging the number of discs which can be used for the fit of the psychometric curve. The thickness tolerance is ± 3% (mean ± 2 SD). The phantom is divided into four blocks for production purposes and inspection and to create space needed for reference discs. These reference discs are placed on the sides and between the blocks and used by CDCOM readout software to accurately localize the position of the gold discs. Each square contains two identical disks, one in the centre and one in a randomly chosen corner, to allow verification of the detection of each object. Easily memorisable patterns have been avoided for human image readout.

The cover consists of a 5 mm thick PMMA plate. In the cover a cavity with a depth of 2 mm has been milled, to accommodate the aluminium base with the gold-disks. The assembly ( 3 mm PMMA + 0.5 mm aluminium) has a PMMA equivalent thickness of 10 mm, under standard mammography-exposure conditions. Between the PMMA cover and the aluminium base a silkscreen printed transparent with X-ray contrasting paint has been placed, showing a grid, the disk thickness and disk diameter. Consequently the X-ray image will show a number of squares arranged in 21 columns and 16 rows, with the disk-diameter shown above most columns. The grid is not rotated for a more efficient use of the phantom area and to minimize effect of the heel effect on the CD Curve.

The CDMAM-phantom is delivered with 4 PMMA plates, for the simulation of different breast-thicknesses. The plates have a thickness of 10 mm each and are otherwise of the same dimensions as the phantom.

Manual CDMAM 4.0

Under standard mammography-exposure conditions (Mo-anode, 30 m Mo-filtration, 28 kV), the phantom has a PMMA-equivalent thickness of 10 mm. These conditions will result in an HVL at the image plane of about 0.65 mm Al, when four PMMA plates are added. This indicates an effective energy of about 19 keV, according to Table 1. The linear attenuation coefficient of gold g = 175 mm-1 at this energy (Hubbell, 1982).

Energy [keV]

/ [cm2/g]

HVL [mm Al]

25.82

0.099

7.836

0.327

3.392

0.765

1.115

2.300

Table 1. The energy of monochromatic radiation with the mass attenuation coefficient / and the HVL of aluminium. The radiation contrast Cr of the gold disks can be calculated with the formula: Cr  lblg  1 e  gxg lb

where Ig and Ib represent the X-ray intensities with and without a gold disk and xg the thickness of a gold disk. In table 3 the radiation contrast of each gold disk is given for standard mammography exposure conditions.

Disc Diameter (mm) 2

1.7

1.4

1.2

0.88

0.77

0.66

0.57

0.5

0.42

0.35

0.3

0.25

0.21

0.18

0.15

0.13

0.1

0.09

0.08

0.103

0.105

0.106

0.129

0.147

0.168

0.191

0.208

0.240

0.264

0.341

0.446

0.549

0.675

0.849

1.195

1.396

2.434

2.633

2.800

0.094

0.096

0.098

0.097

0.109

0.128

0.148

0.170

0.192

0.210

0.239

0.308

0.401

0.493

0.601

0.763

1.105

1.310

2.228

2.411

2.601

0.087

0.089

0.090

0.099

0.109

0.130

0.149

0.171

0.193

0.208

0.269

0.351

0.448

0.554

0.681

0.964

1.205

2.019

2.179

2.380

0.078

0.079

0.091

0.100

0.110

0.130

0.150

0.171

0.191

0.242

0.312

0.401

0.493

0.606

0.865

1.111

1.854

2.000

2.180

0.069

0.071

0.080

0.091

0.100

0.110

0.130

0.150

0.169

0.209

0.272

0.350

0.442

0.554

0.777

0.970

1.680

1.830

1.958

0.056

0.057

0.071

0.080

0.092

0.100

0.110

0.130

0.148

0.192

0.243

0.310

0.396

0.490

0.690

0.860

1.560

1.655

1.787

0.048

0.049

0.057

0.071

0.080

0.091

0.100

0.109

0.129

0.170

0.210

0.270

0.344

0.440

0.610

0.770

1.440

1.531

1.623

0.042

0.043

0.049

0.057

0.071

0.080

0.091

0.100

0.108

0.147

0.190

0.240

0.304

0.390

0.550

0.678

1.330

1.415

1.508

0.037

0.038

0.043

0.049

0.057

0.070

0.080

0.090

0.098

0.127

0.167

0.206

0.264

0.338

0.485

0.598

1.210

1.304

1.386

0.032

0.033

0.038

0.043

0.049

0.057

0.070

0.079

0.089

0.107

0.144

0.186

0.234

0.297

0.429

0.538

1.100

1.184

1.273

0.027

0.028

0.033

0.038

0.043

0.049

0.056

0.070

0.077

0.096

0.125

0.163

0.200

0.256

0.379

0.467

0.947

1.081

1.165

0.023

0.024

0.028

0.033

0.038

0.042

0.048

0.056

0.069

0.088

0.104

0.142

0.181

0.227

0.327

0.416

0.833

0.928

1.047

0.020

0.021

0.024

0.028

0.032

0.037

0.042

0.048

0.055

0.076

0.093

0.121

0.158

0.194

0.285

0.366

0.731

0.810

0.890

0.017

0.018

0.021

0.024

0.027

0.032

0.037

0.041

0.047

0.067

0.085

0.101

0.136

0.173

0.245

0.313

0.640

0.716

0.782

0.015

0.018

0.021

0.024

0.027

0.032

0.036

0.040

0.053

0.073

0.090

0.116

0.150

0.216

0.273

0.557

0.617

0.684

0.012

0.015

0.018

0.021

0.023

0.027

0.031

0.035

0.045

0.064

0.081

0.096

0.129

0.165

0.231

0.493

0.535

0.590

Table 2. Thickness [μm] of the gold disks within the phantom for each row and diameter(Ø) [mm].

Disc Diameter (mm) 2

1.7

1.4

1.2

0.88

0.77

0.66

0.57

0.5

0.42

0.35

0.3

0.25

0.21

0.18

0.15

0.13

0.1

0.09

0.08

1.79%

1.82%

1.84%

2.23%

2.54%

2.90%

3.29%

3.57%

4.11%

4.51%

5.79%

7.51%

9.16%

11.14%

13.81%

18.87%

21.67%

34.69%

36.92%

38.74%

1.63%

1.67%

1.70%

1.68%

1.89%

2.22%

2.56%

2.93%

3.30%

3.61%

4.10%

5.25%

6.78%

8.27%

9.98%

12.50%

17.58%

20.49%

32.29%

34.42%

36.57%

1.51%

1.55%

1.56%

1.72%

1.89%

2.25%

2.57%

2.95%

3.32%

3.57%

4.60%

5.96%

7.54%

9.24%

11.23%

15.52%

19.01%

29.77%

31.70%

34.06%

1.36%

1.37%

1.58%

1.73%

1.91%

2.25%

2.59%

2.95%

3.29%

4.15%

5.31%

6.78%

8.27%

10.06%

14.05%

17.67%

27.71%

29.53%

31.72%

1.20%

1.23%

1.39%

1.58%

1.73%

1.91%

2.25%

2.59%

2.91%

3.59%

4.65%

5.94%

7.44%

9.24%

12.71%

15.61%

25.47%

27.40%

29.01%

0.98%

0.99%

1.23%

1.39%

1.60%

1.73%

1.91%

2.25%

2.56%

3.30%

4.16%

5.28%

6.70%

8.22%

11.37%

13.97%

23.89%

25.15%

26.85%

0.84%

0.85%

0.99%

1.23%

1.39%

1.58%

1.73%

1.89%

2.23%

2.93%

3.61%

4.62%

5.84%

7.41%

10.12%

12.61%

22.28%

23.50%

24.73%

0.73%

0.75%

0.85%

0.99%

1.23%

1.39%

1.58%

1.73%

1.87%

2.54%

3.27%

4.11%

5.18%

6.60%

9.18%

11.19%

20.76%

21.93%

23.19%

0.65%

0.66%

0.75%

0.85%

0.99%

1.22%

1.39%

1.56%

1.70%

2.20%

2.88%

3.54%

4.51%

5.74%

8.14%

9.94%

19.08%

20.40%

21.54%

0.56%

0.58%

0.66%

0.75%

0.85%

0.99%

1.22%

1.37%

1.55%

1.86%

2.49%

3.20%

4.01%

5.06%

7.23%

8.99%

17.65%

18.71%

19.97%

0.47%

0.49%

0.58%

0.66%

0.75%

0.85%

0.98%

1.22%

1.34%

1.67%

2.16%

2.81%

3.44%

4.38%

6.42%

7.85%

15.27%

17.24%

18.44%

0.40%

0.42%

0.49%

0.58%

0.66%

0.73%

0.84%

0.98%

1.20%

1.53%

1.80%

2.45%

3.12%

3.89%

5.56%

7.02%

13.56%

14.99%

16.74%

0.35%

0.37%

0.42%

0.49%

0.56%

0.65%

0.73%

0.84%

0.96%

1.32%

1.61%

2.10%

2.73%

3.34%

4.87%

6.20%

12.01%

13.22%

14.42%

0.30%

0.31%

0.37%

0.42%

0.47%

0.56%

0.65%

0.71%

0.82%

1.17%

1.48%

1.75%

2.35%

2.98%

4.20%

5.33%

10.60%

11.78%

12.79%

0.26%

0.31%

0.37%

0.42%

0.47%

0.56%

0.63%

0.70%

0.92%

1.27%

1.56%

2.01%

2.59%

3.71%

4.67%

9.29%

10.23%

11.28%

0.21%

0.26%

0.31%

0.37%

0.40%

0.47%

0.54%

0.61%

0.78%

1.11%

1.41%

1.67%

2.23%

2.85%

3.96%

8.27%

8.94%

9.81%

Table 3. The radiation contrast Cr, calculated with formula 1, for standard mammography exposure conditions

Manual CDMAM 4.0

To make an X-ray image, the CDMAM phantom should be positioned on the bucky with the thinnest disk-thicknesses at the thorax side, in combination with one or more PMMA plates (Figure 2). For use with the automatic readout software CDCOM only use PMMA plates without any markers. X-Ray apparatus

Figure 2. Reference configuration of the CDMAM phantom on the bucky of a mammography unit.

Manual CDMAM 4.0

Store and use the phantom and its belongings at room temperature (15o-25o) and at normal humidity protected for fluid and moisture, dust, etc. preferable in the delivered case. Handle all products with care. PMMA scratches easily which might give that the phantom or plates is useless for evaluation. Clean the materials with non-aggressive general cleaner. The phantom cannot be used in MRI systems or in the neighbourhood of other magnetic materials.

Send back to manufacturer

The Declaration of conformity to the specifications states that the CDMAM 4.0 phantom is in compliance to the norms and product description. By heavy wearing out, misuse or mistreatment of the phantom, the phantom might become unusable or unreliable. In case one of the following symptoms is noticed, please contact your local distributor or Artinis Medical Systems.  Large scratches and other damage to the body  Large scratches and other damage to one of the PMMA Attenuation plates  Dents, large scratches or other damage into the aluminium plate  The aluminium plate gets loose from the body  The grid paint is dissolved by moisture which is visible by high contrast particles at the Xray images and little white speckles between the foil and the body

Manual CDMAM 4.0

CDMAM images can be evaluated at the classical way with images and human observer or in a digital environment by using automatic CDMAM Analyser scoring software. The Xray image of the CDMAM phantom needs to be evaluated by at least 3 experienced observers. The "Score form CDMAM phantom" (see Appendix 1) can be used for this purpose. The image should be evaluated in the area where the gold-disks are just visible, by indication of the location of the eccentric disks. At least 3 fields must be observed in each column and each row, in order to comply with the correction scheme, which is described in paragraph 4.1. The indicated positions of the eccentric disks have to be compared to the true diskpositions in the phantom, for which the "Evaluation form CDMAM phantom" can be used (see Appendix 2). To evaluate the observations certain rules have to be applied, taking into account the 4 nearest neighbours (defined by a common vertice) of the field under examination. The evaluation of a particular field must refer to the original observations for the nearest neighbours. Examples of the correction scheme are discussed in paragraph 4.2.

In the correction scheme, there are three possibilities for each observation:  T: the eccentric disk was indicated at the true position  F: the eccentric disk was indicated at a false position  N: the eccentric disk was not indicated at all The two main rules within the correction scheme are:  A True needs 2 or more correctly indicated nearest neighbours to remain a True.  A False or Not indicated disk will be considered as True when it has 3 or 4 correctly indicated nearest neighbours. Exceptions on the two main rules are:  A True which has only 2 nearest neighbours (at the edges of the phantom) needs only 1 correctly indicated nearest neighbour to remain True.  A False or Not indicated disk which has only 2 nearest neighbours will be regarded True if both nearest neighbours are correctly indicated.

Manual CDMAM 4.0

Six examples (shown in Figure 3) of the correction scheme are discussed below. Example 1: The common situation. T* remains T because of its 2 correctly indicated nearest neighbours. F* remains F because it has only 2 correctly indicated nearest neighbours. Example 2: F* is considered T because it has more than 2 correctly indicated nearest neighbours. Both T*'s however have only 1 correctly identified nearest neighbour, and thus are considered to be F's. Example 3: T* remains T because it has 1 out of 2 correctly indicated nearest neighbours. The absent corner will be considered as an F because it has only 1 correctly indicated nearest neighbour. Example 4: F* will be considered as a T because of its 2 out of 2 correctly indicated nearest neighbours. T* will be considered as an F because it has only 1 correctly indicated nearest neighbour. The absent corner will be considered as an F, because it has no correctly indicated nearest neighbours. Example 5: F* remains an F, because it has only 1 out of 2 correctly indicated nearest neighbours. Both T*'s are considered as F's because they have none respectively 1 correctly indicated nearest neighbour. Example 6: T* remains T because it has 1 out of 2 correctly indicated nearest neighbours. F* will be considered as a T because of 3 correctly indicated nearest neighbours. The absent corner will be considered as a T because of 2 correctly indicated nearest neighbours.

Manual CDMAM 4.0

T T T T N T* T T N F* T T N N N T

T T T T N T* F* T N N T* T N N N T

Figure 3.1

Figure 3.2

F T* T T N N T T N N N T N N N N

F F* T T N T* T T N N T T N N N T

Figure 3.3

Figure 3.4

F F* T* T N N N T N T* N T N N N N

T T* F* T T T T T N T T T N N T T

Figure 3.5

Figure 3.6

Figure 3. Six examples for the use of the correction scheme.

The results can be presented in a logarithmic graph, in which the disk-thickness is plotted against the disk-diameter. The curve through the threshold fields is called the contrast detail (CD) curve (Thijssen, 1993). The image quality can be expressed in a figure by calculation of the ratio of correctly identified disk-positions to the total number of squares:

𝑐𝑜𝑟𝑟𝑒𝑐𝑡 𝑜𝑏𝑠𝑒𝑟𝑣𝑎𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑖𝑜 =

𝐶𝑜𝑟𝑟𝑒𝑐𝑡 𝑜𝑏𝑠𝑒𝑟𝑣𝑎𝑡𝑖𝑜𝑛𝑠 ∙ 100% 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠

Another method to quantify image quality is called the Image Quality Figure (IQF)-method (Thijssen et al, 1989), which is defined by 16

𝐼𝑄𝐹 = ∑ 𝐶𝑖 ∙ 𝐷𝑖,𝑚𝑖𝑛 𝑖=1

where Di,min denotes the threshold diameter in contrast-column i. Summation over all contrast-columns yields the IQF. Image quality increases with an increasing number of correctly identified disk-positions. In this case the IQF will become smaller because the values of diameter and thickness of the threshold-disks are smaller.

Manual CDMAM 4.0

An alternative method for calculation of the IQF has been presented (Thijssen et. al., 2000), giving a higher IQFinv at better image quality. Since for CDMAM phantoms the threshold thickness is determined per diameter, we suggest to calculate the IQF Inv according to

𝐼𝑄𝐹𝑖𝑛𝑣 =

100 ∑21 𝑖=1 𝑡𝑡ℎ𝑟,𝑖 ∙ 𝐷𝑖

where tthr,i is the threshold thickness of diameter i. Di (mm)

1.7

1.4

1.2

0.88

0.77

0.66

0.57

0.5

0.42

All invisible: tthr,i

0.113

0.116

0.118

0.140

0.167

0.195

0.218

0.235

0.265

0.295

All visible: tthr,i

0.010

0.011

0.013

0.016

0.018

0.020

0.023

0.027

0.030

Di (mm)

0.35

0.3

0.25

0.21

0.18

0.15

0.13

0.1

0.09

0.08

All invisible: tthr,i

0.384

0.505

0.608

0.744

0.951

1.349

1.513

2.662

2.890

3.061

All visible: tthr,i

0.038

0.057

0.071

0.083

0.114

0.144

0.200

0.433

0.468

0.518

Table 4. tthr,i for completely invisible columns or completely visible comlumns Since the phantom contains more diameters, the IQFinv of the CDMAM 4.0 cannot be compared to the IQFinv of the CDMAM 3.4. For comparing results between phantoms we recommend to use the contrast threshold values of the common diameters of both phantoms or comparing CD Curves.

Manual CDMAM 4.0

The CDMAM Analyser software enables fast automatic readout, scoring and analysis of digital CDMAM 4.0 images. The program uses CDCOM for CDMAM 4.0 for image readout and scoring and analyses the scores according to the described analysis method in the supplement to the European Guidelines fourth edition (Perry et al., 2013). The analysis results are observer independent. As a prerequisite for reading this manual and using the program correctly the reader should also study the previous chapters on the CDMAM carefully.

For automatic readout and scoring of digital CDMAM images the software tool CDCOM for CDMAM 4.0 is used. The CDCOM software automatically identifies the gold discs on images of the CDMAM phantom and is available from the EUREF website [www.euref.org]. The program attempts to correctly locate the position of the gold discs by first identifying the grid position and reference dots and subsequently the gold discs positions. CDCOM performs the following actions:  It reads a single digital CDMAM version 4.0 image and tries to detect the gold disks in the phantom.  CDCOM determines in which corner of each cell the gold disc is most likely to be  The program checks for each cell if the correct corner was chosen  An additional detection step is done using the 3 corners where no marker is present and the centre dot More information regarding CDCOM can be found in the CDCOM Manual. The CDCOM executable can be found in the installation folder of this software (CDCOM4.exe). If EUREF releases a new version of CDCOM for CDMAM 4.0 you can download the executable from the EUREF website and replace the current by the new executable file. Please notice you have to give the new executable file the same name, so “CDCOM4.exe”.

The analysis method is according to the Supplement to the European Guidelines. All diameters are included during the analysis. The score results of the CDCOM software are

Manual CDMAM 4.0

averaged and filtered by a Gaussian filter. The probability of gold disc detection by CDCOM is between 25% and 100% by the 4-alternative forced choice method. The gold disc thicknesses having 62.5% detection probability are the threshold thicknesses and are determined by applying psychometric curve fitting by a least squares procedure over the probabilities. The psychometric curve is fitted by (Veldkamp, 2003):

𝑝(𝑑) =

0.75 + 0.25 1 + 𝑒 −𝑓(𝐶−𝐶𝑡 )

C = logarithm of signal contrast C = log(1 − e−μd) Ct = signal contrast at the threshold of 62.5% f = fitting parameter p(d) = the probability of detection p(d) of an object with thickness d The attenuation coefficient of gold  depends on the beam quality used for the exposure. These values were calculated by DR Dance and KC Young at NCCPM with 3mm PMMA representing the compression paddle, using spectra from Boone (et al., 1997) and attenuation coefficients for materials in the test objects (aluminium, gold, PMMA) from Berger (et al, 2005). In the software these threshold thicknesses are called the Automatic Threshold Gold Thicknesses. The Automatic Threshold Gold Thicknesses are converted to Predicted Human Gold Thicknesses by correction for human read out by 𝑇𝑝𝑟𝑒𝑑𝑖𝑐𝑡𝑒𝑑 = 𝑎𝑇𝑎𝑢𝑡𝑜 𝑛 Tpredicted =predicted human threshold gold thickness Tauto = Automatic Threshold Gold thickness a, n = fitted parameters with a = 1.441, n = 0.895 The Fit to Predicted Human Gold Thickness are determined by fitting a third order polynomial through the Predicted Human Gold Thicknesses resulting in the CD Curve by

𝑇𝑓𝑖𝑡 = 𝑎 +

𝑏 𝑐 𝑑 + 2+ 3 𝑥 𝑥 𝑥

Tfit = Fit to Predicted Human Gold Thickness (um) x = detail diameter (mm)

Manual CDMAM 4.0

a, b, c and d = coefficients adjusted to achieve a least squares fit, and are ≥ 0 The IQFinv number is computed according to

𝐼𝑄𝐹𝑖𝑛𝑣 =

100 ∑21 𝑖=1 𝑡𝑖,𝑡ℎ𝑟 ∙ 𝐷𝑖

where ti,thr denotes the threshold thickness (Fit to Predicted Human Gold Thickness) in diameter-column i. The thickness is given in µm whereas the diameter is taken in mm. The total detected (%) is determined by

𝑇𝑜𝑡𝑎𝑙 𝑑𝑒𝑡𝑒𝑐𝑡𝑒𝑑 (%) =

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑒𝑡𝑒𝑐𝑡𝑒𝑑 𝑔𝑜𝑙𝑑 𝑑𝑖𝑠𝑐𝑠 × 100% 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑔𝑜𝑙𝑑 𝑑𝑖𝑠𝑐𝑠

Manual CDMAM 4.0

The data and images in the program are organized in projects with groups (Figure 4). The project may contain several groups, each having by example other settings during imaging of the phantom, allowing easy comparison of the influence of the settings to the image quality. A group should contain multiple images made with the same settings. We recommend to use at least 8 images, but preferable 16 or more per group for reliability of the results. The readout and score result can be viewed per project, group or image. Analysed images are marked with a green sign. In case the analysis failed, the image is marked with a yellow question mark, mostly caused by wrong or incomplete DICOM header information (see chapter 6.3).

Figure 4. Program structure

A new project can be made via the menu, File, New project. Existing projects can be opened via menu, File, Open project or via the toolbar opening button. The project can be saved via menu, File, Save project (as) or via the toolbar saving button, which also changes the project name from the standard “new project” to the project file name. Latest projects can be opened via menu, File. The project has properties, which can be viewed via menu, Project, properties or via the right mouse button at the project. The first tab offers to add comments for project description and the second to use relative file paths or changing the setting if DICOM files are opened or not by addition to a project. The relative file paths save the image file paths relatively to the project file path, which is easy for copying project with images from a computer to another computer, USB or CD by example.

Manual CDMAM 4.0

A new group can be added to the project via the right mouse button at the project or via the menu, Project, add group. Groups can be removed via the right mouse button at the group, remove or via the menu, Project, remove. Group renaming or removal can be done via the right mouse button at the group, rename or remove. Each group has properties, which can be viewed via the right mouse button at the group, properties or via menu, Project, properties. The properties contain the image acquisition settings and image rotation. Image acquisition settings are used for estimation of the linear attenuation coefficient. The settings are read automatically from the DICOM headers of the images but can be set here by hand. The rotation is needed for correct detection of the image gold discs and can be clockwise, counter-clockwise or no rotation. Changing this setting, all images of the group are rotated. Images can also be rotated per image. The background is for interpretation of the pixel intensity. In most cases this can be read from the pixel intensity relationship sign in the DICOM header (“default”), but if this is missing it might be necessary to set this as “high” if the pixel intensity relationship is not reversed or as “low” if the pixel intensity is reversed.

Figure 5. Rotation and background tab for group properties

CDCOM software only analysis DICOM images containing the following tags in the DICOM header:

Manual CDMAM 4.0

 (0018,1164) Imager Pixel Spacing, if not present: (0028,0030) Pixel Spacing  (0028,0010) Rows  (0028,0011) Columns  (0028,0100) Bits Allocated  (0028,0101) Bits Stored  (0028,1041) Pixel Intensity Relationship Sign, if not present: (0028,0004) Photometric Interpretation  (7fe0,0010) Pixel Data Please note that compressed image data is not supported by CDCOM. DICOM Meta information can be checked via right mouse button at the image, properties, file properties. Images can be added to a group via the right mouse button at the group or a DICOM file, add DICOM file or via menu, Project, add Dicom file. This opens a Windows opening window, showing default only files with the “*.DCM” extension. Files with other extensions can be showed by changing the extension from DICOM Files (*.dcm) to All Files (*.*). Multiple images can be selected at once. Files can be removed via the right mouse button at the DICOM File, remove or via the menu, Project, remove. Files can also be added by dragging and dropping from Windows Explorer. Files are (default) automatic opened after addition, which shows the image in the program. In the project properties this can be changed. Maximal 8 images can be opened and then the message of Figure 6 is shown. Click “Yes” to close the images. Another quick way to close all opened images at once is via Menu, Window, Close all documents. A single file can be opened via the right mouse button at the file, open file or by double clicking at the file. Image properties can be viewed via the right mouse button at the file, properties or menu, Project, properties. This shows the DICOM header information and the selected rotation and background. Renaming and removing the file can be done via the right mouse button at the file.

Figure 6. Maximum number of files is open After opening the file the image can be rotated counter clockwise and clockwise via the toolbar by the buttons

or via the file properties. Details at the image can be viewed

Manual CDMAM 4.0

via the zoom options in the toolbar

or via the right mouse button at the image. The

image gray scale can be adapted via

or Menu, view, window level, which opens Figure

7. `Auto` sets the gray scale optimal for the whole image or for the selected area, ´Select area` offers the possibility to select an area in the image used for optimal levelling. By ´Level` and ´Window` the levelling can be changed by hand. The image can be copied or printed via the right mouse button at the image.

Figure 7. Window level to set the image gray scale

Manual CDMAM 4.0

An image can be analysed by pressing

or via the menu, Project, Analyse CDMAM4

Image or Batch Run. The first are to analyse a single file, the second to analyse a batch of images in a group. The analysis can be cancelled via menu, Project, cancel run. In the left bottom corner is shown which file is being analysed. Only not already analysed files are analysed. To reanalyse a file, the results has to be cleared first via the right mouse button at the file, clear detection result or menu, Project, clear detection result. Analysed files are green marked. In case the analysis failed, the image is marked with a yellow question mark.

After an image is analysed, the contrast detail diagram can be viewed (Figure 8). Per contrast detail is shown if both gold discs were visible (red dot), only the center or the outer gold disc was visible (pink dot) or none was visible (no dot). If you cannot view the diagram completely, you can zoom in via the zoom buttons at the toolbar.

Contrast Detail Diagram 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 2.00 1.70 1.40 1.20 1.000.880.77 0.66 0.570.50 0.42 0.35 0.30 0.25 0.21 0.18 0.150.13 Diameter (mm)

0.100.090.08

Figure 8. Contrast detail diagram

Manual CDMAM 4.0

In case one or more images show unexpected analysis results, check  the DICOM header information  the rotation of the image(s)  background setting of the images  presence of strange objects, markings or other high contrasts not belonging to the CDMAM phantom in the image. This might disturb the grid detection by CDCOM. It is possible to help the CDCOM analysis program to detect the grid in the right position, which might solve bad detection by the latter case. Mark the top left corner of the grid via the menu, Edit, Mark Left Top (Figure 9).

Figure 9. Mark left top of the grid

The group result shows in the first table the IQFInv and Total Detected (%) per image and over all images. The second table shows the determined automatic threshold gold thickness, the Predicted Human Gold Thickness and the fitted threshold gold thickness. The figures show the contrast detail curve of the group and for each individual image (Figure 10), the contrast detail score diagram (Figure 8) and the result of the psychometric curve fitting (Figure 11). Selection of one of the image names besides the contrast detail curve will highlight the CD Curve of that image forwards. Use this to check for bad (analysed) images. Selection of another diameter besides the psychometric curve diagram will highlight the psychometric curve and scores of that diameter.

Thickness (

0.20

I19 I2 I20 I21 I22 I23 I24 I25 I26

0.10

0.05

Manual CDMAM 4.0 0.02 0.01

00 2. 70 1.

40 1. 20 1. 00 1. 8 8 0. 7 7 0. 66 0. 57 0. 0 5 0. 42 0. 35 0. 30 0. 25 0. 21 0. 18 0. 15 0. 13 0.

10 0. 9 0 0. 8 0 0.

Contrast Detail Curve

Diameter (mm)

Totals for 110mAs I27 I0 I28 I1 I29 I10 I3 I11 I30 I12 I31 I13 I4 I14 I5 I15 I6 I16 I7 I17 I8 I18 I9 I19 I2 I20 I21 I22 I23 I24 I25 I26

Contrast Detail Score Diagram 5.00 0 1 2 3

2.00 1.00

4 50.50

Thickness (µm)

row

70.20 80.10 9

0.05

11 12 13

0.02

140.01

00 2. 70 1.

40 1. 8 20 .0 1. 0.09 0 0 00 .1 1. 0 88 0. 7 13 7 0. 0. 5 .1 66 0 0. 18 57 0. 0. 0 21 5 0. 0. 25 42 0. 0. 0 .3 35 0 5 0. .3 30 0 0. 42 0. 25 0 5 0. 0. 21 57 0. 0. 6 18 .6 0. 0 7 .7 15 0 8 0. .8 13 0 0 0. 1.0

00 2.

.2 10 1 0. 9 .40 0 0. 8 1 0 0 0. 1.7

Diameter (mm)

Figure 10. Contrast detail curve of the group result Diameter (mm)

Contrast Detail Score Diagram Curve Fit 1.0 0.9

0 Ø2 Ø 1.7 Ø 1.4 Ø 1.2 Ø1 Ø 0.88 Ø 0.77 Ø 0.66 Ø 0.57 Ø 0.5 Ø 0.42 Ø 0.35 Ø 0.3 Ø 0.25 Ø 0.21 Ø 0.18 Ø 0.15 Ø 0.13 Ø 0.1 Ø 0.09 Ø 0.08

1 2

0.8 0.7

3 4

0.5

row

probability

5 0.6

0.4

8 9

0.3 10

11 0.2

0.1 13 0.0

14 00 5.

00 2.

00 1.

50 0.

20 0.

10 0.

05 0.

08 0. 9 0 0. 0 1 0.

Thickness (µm)

13 0. 15 0. 18 0. 21 0. 25 0. 30 0. 35 0. 42 0. 50 0. 7 5 0. 66 0. 77 0. 8 8 0. 0 0 1. 20 1. 40 1.

00 2.

0 02 1.7 0.

01 0.

Diameter (mm) Figure 11. Psychometric curves ( ___ ) and scores ( ⁰ ) for all diameters

Curve Fit 1.0

The project result shows a table with the IQFInv and Total detectedØ (%) 0.9 1.7 per group and a Ø2

table 0.8 with the threshold thicknesses per group according to the FitØto Predicted Human 1.2 Ø 1.4

Gold Thickness (µm). This thicknesses are also shown in the CD Curve diagram (Figure 12). Ø 0.88 Ø1

0.7

Ø 0.77 Also diagrams for the IQFInv (Figure 13) and Total detected (%) (Figure 14) are shown. Ø 0.66 0.6

Ø 0.57 Ø 0.5 Ø 0.42 Ø 0.35 Ø 0.3 Ø 0.25 Ø 0.21 Ø 0.18 Ø 0.15 Ø 0.13 Ø 0.1 Ø 0.09 Ø 0.08

probability

0.5 0.4 0.3

00 5.

00 2.

00 1.

50 0.

20 0.

10 0.

05 0.

02 0.

01 0.

Manual CDMAM 4.0

Contrast Detail Curve

110mAs 125mAs 140mAs 40mAs 45mAs 50mAs 56mAs 63mAs 71mAs 80mAs

5.00

2.00 1.00

0.50

Thickness (µm)

Contrast Detail Curve

0.20

4006h 4006n 4007h 4007n 4008h 4008n

5.00 0.10

2.00 1.00 0.05

0.02 0.20

0.01

0.10

00 2. 70 1.

0.05

40 1. 20 1. 00 1. 8 8 0. 7 7 0. 66 0. 57 0. 0 5 0. 42 0. 35 0. 30 0. 25 0. 21 0. 18 0. 15 0. 13 0.

10 0. 9 0 0. 8 0 0.

Thickness (µm)

0.50

Diameter (mm) 0.02

0.01 Figure 12. CD curve of the groups showing the threshold thicknesses according to fit to predicted

00 2. 70 1.

40 1. 20 1. 00 1. 8 8 0. 7 7 0. 66 0. 57 0. 0 5 0. 42 0. 35 0. 30 0. 25 0. 21 0. 18 0. 15 0. 13 0.

10 0. 9 0 0. 8 0 0.

human gold thicknesses (µm). Selection of the group name besides the diagram highlights the IQFInv Scores Diameter (mm) corresponding CD Curve. IQFInv Scores

110

100 90

80 70

IQFInvIQFInv

50 40

30 20 10

n 07 40

h 08 40

n 08 40

As m 40

As m 45

As m 50

h 07 40

n 06 40

h 06 40

10 IQFInv results per group Figure 13. Total Detected (%)

As m 80

As m 71

As m 63

As m 56

As 0m 14

As 5m 12

As 0m 11

100

80 70

Total Detected (%) 60

70 20

30 20 10 0 Manual CDMAM 4.0

h 07 40

n 07 40

h 08 40

n 08 40

h 07 40

n 07 40

h 08 40

n 08 40

n 06 40

h 06 40

Total Detected (%) 100 90 80 70

60 50 40 30 20 10 0

n 06 40

h 06 40

Figure 14. Percentage total detected per group

Manual CDMAM 4.0

1 Open a project 2 Save the project 3 Counter clockwise rotation of an image 4 Clockwise rotation of an image 5 Zoom in to an image or analysis result 6 Zoom out an image or analysis result 7 Image window levelling 8 Analyse 9 Analyse batch 10 Toggle diagram view 11 Print results

Manual CDMAM 4.0

New Project

To start a new project

Open Project

To open an existing project

Save Project

To save the current project

Save Project as

To save the current project with another name

Save Report

To save report for presentation, etc.

Export

To export to a text file for further analysis etc.

To print the result. Set to landscape for printing complete table

Print preview

To preview of the print result

Print Setup

To set the printer settings

Recent projects

To open a recent project

Exit

To close the program

Copy

To copy the image or results to another program

Drag mode

Drag the image

Mark top left

Mark left top to help CDCOM to detect the grid

Remove top left mark

Remove the top left mark for a new image analysis

Rotate 90 degrees

To rotate the image 90 degrees (counter clockwise)

Properties

To set the project properties or group properties

Add group

To add a new group to the project

Add DICOM file

To add a DICOM file to a group

Remove

To remove a group or DICOM file

Analyse CDMAM4 Image

To analyse a single image

Batch run

To analyse a batch of images

Cancel run

To cancel the analysis

Clear Detection result

Clears the detection result of the group or image

Toolbar

To show or hide the toolbar

Manual CDMAM 4.0

Statusbar

To show or hide the statusbar

Show diagrams

To show or hide the diagram of an image

Zoom in

Zoom in to an image

Zoom out

Zoom out of an image

Window level

To set the window level of an image

Cascade

cascades the opened images and results

Tile horizontal

Orders images and results

Tile vertical

Orders images and results

Close all documents

Closes all opened images

Help Topic

Opens this manual

About Artinis CDMAM4 Analyser

Opens Figure 15

Figure 15. About Artinis CDMAM4 Analyser

Manual CDMAM 4.0

 The program runs on Microsoft Windows 7® and 8® platforms.  A CD player to install the program should be present.  It accepts DICOM images. Read the CDCOM manual for additional requirements.  One free USB port for the license key.  For a correct working of the program the USB license key should always stay connected.

Connect the delivered USB key in one of your free USB ports after the software has been installed. It can be necessary to run first a driver installation manually for the dongle. The standard drivers can be found in the folder “Utilities” at the software CD. Most recent drivers are available at www.artinis.com. If after starting the program the message “Rockey dongle not detected” (the USB key) appears either restart your computer and/or disconnect – connect the USB Key. In case you suspect malfunction of the USB key you can return it to the premises of Artinis Medical Systems, and in case the key is defect it shall be replaced by a new one, free of charge within the warranty period, and for a small fee outside the warranty period. An extra license key costs EUR 250,-. In case the key is lost we will charge you for the amount of EUR 250,-. By using a key lock device you could secure the key to the analysis computer.

Manual CDMAM 4.0

My analysis fails. Why? Check if your files contain the required DICOM header parameters (paragraph 6.3) My analysis result is not as expected.  Is the rotation set in the right direction?  Is the background set right?  Do all files show the same unexpected result or is it just one? Check the properties of this file.  Use at least 8 images per group and preferable 16 images per group for a statistic reliable result  In case one or more images show unexpected analysis results, check o

the DICOM header information

the rotation of the image(s)

presence of strange objects, markings or other high contrasts not belonging to the CDMAM phantom in the image. This might disturb the grid detection by CDCOM.

It is possible to help the CDCOM analysis program to detect the grid in the right position, which might solve bad detection by the latter case.

I cannot install my license key Use the drivers from the software CD (folder Utilties) and select the right driver for your system. Notice you might have to restart your system after installation. I lost my license key Contact Artinis for another one. A new license key is 250 Euro.

Previous Page Next Page

artinis

X-Ray Phantoms

CDMAM 4.0 Contrast-Detail Phantom Manual Ver 2.1