GE SenographePristinaTechnical 2D 2019

NHS Breast Screening Programme
Equipment Report
Technical evaluation of GE Senographe
Pristina digital mammography system in
2D mode
December 2019
Technical evaluation GE Senographe Pristina digital mammography system in 2D mode
About Public Health England

Public Health England exists to protect and improve the nation’s health and wellbeing,
and reduce health inequalities. We do this through world-leading science, knowledge and
intelligence, advocacy, partnerships and the delivery of specialist public health services.
We are an executive agency of the Department of Health and Social Care, and a distinct
delivery organisation with operational autonomy. We provide government, local government,
the NHS, Parliament, industry and the public with evidence-based professional, scientific
and delivery expertise and support.
Public Health England, Wellington House, 133-155 Waterloo Road, London SE1 8UG
Tel: 020 7654 8000 www.gov.uk/phe
Twitter: @PHE_uk Facebook: www.facebook.com/PublicHealthEngland
About PHE screening

Screening identifies apparently healthy people who may be at increased risk of a
disease or condition, enabling earlier treatment or informed decisions. National
population screening programmes are implemented in the NHS on the advice of the UK
National Screening Committee (UK NSC), which makes independent, evidence-based
recommendations to ministers in the 4 UK countries. PHE advises the government and
the NHS so England has safe, high quality screening programmes that reflect the best
available evidence and the UK NSC recommendations. PHE also develops standards
and provides specific services that help the local NHS implement and run screening
services consistently across the country.
www.gov.uk/phe/screening Twitter: @PHE_Screening Blog: phescreening.blog.gov.uk

For queries relating to this document, please contact: phe.screeninghelpdesk@nhs.net
© Crown copyright 2019

You may re-use this information (excluding logos) free of charge in any format or
medium, under the terms of the Open Government Licence v3.0. To view this licence,
visit OGL. Where we have identified any third party copyright information you will need
to obtain permission from the copyright holders concerned.
Published December 2019

PHE publications PHE supports the UN
gateway number: Sustainable Development Goals
2
Contents
About Public Health England
About PHE Screening
Executive summary 4
1. Introduction 5
1.1 Testing procedures and performance standards for digital mammography 5
1.2 Objectives 5
2. Methods 5
2.1 System tested 5
2.2 Output and HVL 6
2.3 Detector response 7
2.4 Dose measurement 7
2.5 Contrast-to-noise ratio 8
2.6 AEC performance for local dense areas 10
2.7 Noise analysis 11
2.8 Image quality measurements 12
2.9 Physical measurements of the detector performance 13
2.10 Other tests 14
3. Results 14
3.1 Output and HVL 14
3.2 Detector response 14
3.3 AEC performance 16
3.4 Noise measurements 21
3.5 Image quality measurements 23
3.6 Comparison with other systems 25
3.7 Detector performance 29
3.9 Other tests 31
4. Discussion 34
5. Conclusions 35
References 36
3
Executive summary
The purpose of the evaluation was to determine whether the GE Senographe Pristina
meets the main standards in the NHS Breast Screening Programme (NHSBSP) and
European protocols, and to provide performance data for comparison against other
systems.
The mean glandular dose (MGD) was found to be well below the remedial level for all
automatic exposure control (AEC) modes. For the 53mm equivalent standard breast,
the MGD was 1.19mGy in Standard mode, compared with the remedial level of 2.5mGy.
The image quality, as measured by threshold gold thickness, is at the achievable level
for the Standard mode.
The GE Senographe Pristina, operating in 2D mode, meets the requirements of the

NHSBSP standards for digital mammography systems.
4
1. Introduction
1.1 Testing procedures and performance standards for digital mammography
This report is one of a series evaluating commercially available direct digital radiography
(DR) systems for mammography on behalf of the NHS Breast Screening Programme
(NHSBSP). The testing methods and standards applied are mainly derived from
NHSBSP Equipment Report 06041 which is referred to in this document as ‘the
NHSBSP protocol’. The standards for image quality and dose are the same as those
provided in the European protocol,2,3 but the latter has been followed where it provides
a more detailed standard, for example, for the automatic exposure control (AEC)
system.
Some additional tests were carried out according to the UK recommendations for testing
mammography X-ray equipment as described in IPEM Report 89.4
1.2 Objectives
The aims of the evaluation were:
• to determine whether the GE Senographe Pristina digital mammography system,

operating in 2D mode, meets the main standards in the NHSBSP and European
protocols
• to provide performance data for comparison against other systems
2. Methods
2.1 System tested
The tests were conducted at the GE factory in Buc, Paris, on a GE Senographe Pristina
system as described in Table 1. The Pristina is shown in Figure 1.
5
Table 1. System description

Manufacturer GE Healthcare
Model Senographe Pristina
System serial number 000011171069167144
Target material Molybdenum, rhodium
Added filtration 30µm molybdenum, 30 µm silver
Detector type Caesium iodide with amorphous silicon
Detector serial number J125020
Pixel size 100µm
Detector size 240mm x 286mm
Pixel array 2294 x 1914, 2850 x 2394
Typical image sizes 9MB (small field size), 13MB (large field
size)
Pixel value offset 0
Source to detector distance 660mm
Source to table distance 637mm
Pre-exposure Thickness < 38mm : 26kV Mo/Mo 2mAs
Thickness 38-65mm: 34kV Rh/Ag, 2mAs
Thickness > 65mm: 34kV Rh/Ag, 4mAs
AEC modes Standard, Dose-, Standard+, Implant
Software version 1.13 (latest version is 4.2.39C)
Four AEC modes are available for use with the Pristina, as listed in table 1. The AEC is
referred to by GE as Automatic Optimisation of Parameters (AOP).
Exposure factors 26kV Mo/Mo are used for small breasts, for exposures at up to 35mm
radiological thickness, which is defined as the equivalent thickness of polymethyl
methacrylate (PMMA) plus 2mm or 5%. For thicker breasts the factors used are 34kV
Rh/Ag. The mAs is selected as appropriate for the most dense part of the breast. 29kV
Mo/Mo is used for exposures of smaller thicknesses when the magnification table is in
use.
2.2 Output and HVL
An ion chamber was used to measure the output and half-value-layer (HVL), as
described in the NHSBSP protocol, at intervals of 3kV. Tube voltage was measured with
a RMI 232 kV meter, which was only calibrated for Mo/Mo exposures.
6
Figure 1. The GE Senographe Pristina
2.3 Detector response
The detector response was measured as described in the NHSBSP protocol, except
that 2mm thick aluminium was used at the tubehead, instead of PMMA. The grid was
removed and an ion chamber was positioned above the detector cover, 40mm from the
chest wall edge (CWE). The incident air kerma was measured for a range of manually
set mAs values at 34kV Rh/Ag. The readings were corrected to the surface of the
detector using the inverse square law. No correction was made for attenuation by the
detector cover. A 10mm x 10 mm region of interest (ROI) was positioned on the midline,
40mm from the CWE of each image. The average pixel value and the standard
deviation of pixel values within the ROI were measured. The relationship between
average pixel values and the air kerma incident at the detector was determined.
2.4 Dose measurement
Doses were measured using the AEC in the different modes to expose different
thicknesses of PMMA. Each PMMA block had an area of 180mm x 240mm. Spacers
were used to adjust the paddle height to be equal to the equivalent breast thickness, as
shown in Table 3. The exposure factors were noted and mean glandular doses (MGDs)
were calculated for equivalent breast thicknesses. The value of s used in the calculation
of MGD for the Rh/Ag target filter combination was 1.087.
7
An aluminium square, 10mm x 10mm and 0.2mm thick, was used with the PMMA
blocks during these exposures, so that the images produced could be used for the
calculation of the contrast-to-noise ratio (CNR), described in Section 2.5. The aluminium
square was placed between two 10mm thick slabs of 180mm x 240mm PMMA, on the
midline, with its centre 60mm from the CWE. Additional layers of PMMA were placed on
top to vary the total thickness.
2.5 Contrast-to-noise ratio
Unprocessed images acquired during the dose measurement were analysed to obtain
the CNRs. Thirty six small square ROIs (approximately 2.5mm x 2.5mm) were used to
determine the average signal and the standard deviation in the signal within the image
of the aluminium square (4 ROIs) and the surrounding background (32 ROIs), as shown
in Figure 2. Small ROIs are used to minimise distortions due to the heel effect and other
causes of non-uniformity.5 The CNR was calculated for each image, as defined in the
NHSBSP and European protocols.
Figure 2. Location and size of ROI used to determine the CNR
To apply the standards in the European protocol, it is necessary to relate the image
quality measured using the CDMAM (Section 2.8) for an equivalent breast thickness of
60mm, to that for other breast thicknesses. The European protocol2 gives the
relationship between threshold contrast and CNR measurements, enabling the
calculation of a target CNR value for a particular level of image quality. This can be
compared to CNR measurements made at other breast thicknesses. Contrast for a
particular gold thickness is calculated using Equation 1, and target CNR is calculated
using Equation 2.
Contrast = 1 − e-μt (1)
where µ is the effective attenuation coefficient for gold, and t is the gold thickness.
CNRmeasured × TCmeasured
CNR target = (2)
TCtarget
where CNRmeasured is the CNR for a 60mm equivalent breast, TCmeasured is the threshold
contrast calculated using the threshold gold thickness for a 0.1mm diameter detail,
8
(measured using the CDMAM at the same dose as used for CNRmeasured), and TCtarget
is the calculated threshold contrast corresponding to the threshold gold thickness
required to meet either the minimum acceptable or achievable level of image quality as
defined in the UK standard.
The threshold gold thickness for the 0.1mm detail diameter is used here because it is
generally regarded as the most critical of the detail diameters for which performance
standards are set.
The effective attenuation coefficient for gold used in Equation 1 depends on the beam
quality used for the exposure, and was selected from a table of values summarised in
Table 2. These values were calculated with 3mm PMMA representing the compression
paddle, using spectra from Boone et al.6 and attenuation coefficients for materials in the
test objects (aluminium, gold, PMMA) from Berger et al.7
The European protocol also defines a limiting value for CNR, which is calculated as a
percentage of the threshold contrast for minimum acceptable image quality for each
thickness. This limiting value varies with thickness, as shown in Table 3.
Table 2. Effective attenuation coefficients for gold contrast details in the CDMAM
kV Target/filter Effective attenuation coefficient
(μm-1)
34 Rh/Ag 0.110
Table 3. Limiting values for relative CNR

Thickness Equivalent Limiting values for
of PMMA breast thickness relative CNR (%) in
(mm) (mm) European protocol
20 21 > 115
30 32 > 110
40 45 > 105
45 53 > 103
50 60 > 100
60 75 > 95
70 90 > 90
The target CNR values for minimum acceptable and achievable levels of image quality
and European limiting values for CNR were calculated. These were compared with the
measured CNR results for all breast thicknesses.
9
2.6 AEC performance for local dense areas
This test is described in the supplement to the fourth edition of the European protocol.3
To simulate local dense areas, images of a 30mm thick block of PMMA of size 180mm x
240mm, were acquired under AEC. Extra PMMA between 2 and 20mm thick and of size
20mm x 40mm was added to provide extra attenuation. The compression plate
remained in position at a height of 40mm, as shown in Figure 3. The simulated dense
area was positioned 50mm from the CWE of the table.
In the simulated local dense area the mean pixel value and standard deviation for a
10mm x 10mm ROI were measured and the signal-to-noise ratios (SNRs) were
calculated.
Measurements were carried out using each of the AEC modes.
Guidance for this test suggests that the SNR for each image should be within 20% of
the mean SNR.
Top view
Spacers (10mm thick)
AEC sensor area
Extra attenuation (20mm x 40mm)
Extra attenuation
Side view
Compression paddle
Spacers (10mm thick)
40mm
30mm
Bucky
Figure 3. Setup to measure AEC performance for local dense areas
10
2.7 Noise analysis
The images acquired in the measurements of detector response, using 34kV Rh/Ag,
were used to analyse the image noise. Small ROIs with an area of approximately
2.5mm x 2.5mm were placed on the midline, 60mm from the CWE. The average
standard deviations of the pixel values in these ROIs for each image were used to
investigate the relationship between the dose to the detector and the image noise. It
was assumed that this noise comprises three components: electronic noise, structural
noise, and quantum noise. The relationship between them is shown in Equation 3.
σp =� ke2 + kq2p + ks2p2 (3)
where σp is the standard deviation in pixel values within an ROI with a uniform exposure
and a mean pixel value p, and ke, kq, and ks are the coefficients determining the amount
of electronic, quantum, and structural noise in a pixel with a value p. This method of
analysis has been described previously.8 For simplicity, the noise is generally presented
here as relative noise defined as in Equation 4.
σp
Relative noise = (4)
p
The variation in relative noise with mean pixel value was evaluated and fitted using
Equation 3, and non-linear regression used to determine the best fit for the constants
and their asymptotic confidence limits (using Graphpad Prism version 6.05 for Windows,
Graphpad software, San Diego, California, USA, www.graphpad.com). This established
whether the experimental measurements of the noise fitted this equation, and the
relative proportions of the different noise components. The relationship between noise
and pixel values has been found empirically to be approximated by a simple power
relationship as shown in Equation 5.
σp
= ktp-n (5)
p
where kt is a constant. If the noise were purely quantum noise the value of n would be
0.5. However the presence of electronic and structural noise means that n can be
slightly higher or lower than 0.5.
The variance in pixel values within a ROI is defined as the standard deviation squared.
The total variance was plotted against incident air kerma at the detector and fitted using
Equation 3. Non-linear regression was used to determine the best fit for the constants
and their asymptotic confidence limits, using the Graphpad Prism software.
Using the calculated constants, the structural, electronic, and quantum components of
the variance were estimated, assuming that each component was independently related
11
to incident air kerma. The percentage of the total variance represented by each
component was then calculated and plotted against incident air kerma at the detector.
2.8 Image quality measurements
Contrast detail measurements were made using a CDMAM phantom (serial number
1022, version 3.4, UMC St. Radboud, Nijmegen University, Netherlands). The phantom
was positioned with a 20mm thickness of PMMA above and below, to give a total
attenuation approximately equivalent to 50mm of PMMA or 60mm thickness of typical
breast tissue. The exposure factors were chosen to be close to those selected by the
AEC, in Standard mode, when imaging a 50mm thickness of PMMA. This procedure
was repeated to obtain a representative sample of 16 images at this dose level. Further
sets of 16 images of the test phantom were then obtained at other dose levels by
manually selecting higher and lower mAs values with the same beam quality.
The CDMAM images were read and analysed automatically using Version 1.6 of
CDCOM9,10 and Version 2.1.0 of CDMAM Analysis (available on request from
www.nccpm.org). The threshold gold thickness for a typical human observer was
predicted using Equation 6.
TCpredicted = rTCauto (6)
where TCpredicted is the predicted threshold contrast for a typical observer, TCauto is the
threshold contrast measured using an automated procedure with CDMAM images. r is
the average ratio between human and automatic threshold contrast determined
experimentally with the values shown in Table 4.
Table 4. Values of r used to predict threshold contrast

Diameter of Average ratio of human
gold disc (mm) to automatically
measured threshold
contrast (r)
0.08 1.40
0.10 1.50
0.13 1.60
0.16 1.68
0.20 1.75
0.25 1.82
0.31 1.88
0.40 1.94
0.50 1.98
0.63 2.01
0.80 2.06
1.00 2.11
12
The predicted threshold gold thickness for each detail diameter in the range 0.1mm to
1.0mm was fitted with a curve for each dose level, using the relationship shown in
Equation 7.
Threshold gold thickness = a + bx-1 + cx-2 + dx-3 (7)
where x is the detail diameter, and a, b, c and d are coefficients adjusted to obtain a
least squares fit.
The confidence limits for the predicted threshold gold thicknesses have been previously
determined by a sampling method using a large set of images. The threshold contrasts
quoted in the tables of results are derived from the fitted curves, as this has been found
to improve accuracy.
The expected relationship between threshold contrast and dose is shown in Equation 8.
Threshold contrast = λD-n (8)
where D is the MGD for a 60mm thick standard breast (equivalent to the test phantom
configuration used for the image quality measurement), and λ is a constant to be fitted.
It is assumed that a similar equation applies when using threshold gold thickness
instead of contrast. This equation was plotted with the experimental data for detail
diameters of 0.1 and 0.25mm. The value of n resulting in the best fit to the experimental
data was determined, and the doses required for target CNR values were calculated for
data relating to these detail diameters.
2.9 Physical measurements of the detector performance
The presampled modulation transfer function (MTF), normalised noise power spectrum
(NNPS) and the detective quantum efficiency (DQE) of the system were measured. The
methods used were as close as possible to those described by the International
Electrotechnical Commission (IEC).11 The radiation quality used for the measurements
was adjusted by placing a uniform 2mm thick aluminium filter at the tube housing. The
beam quality used was 26kV Mo/Mo. The test device to measure the MTF comprised a
120mm x 60mm rectangle of stainless steel with polished straight edges, of thickness
0.8mm. The grid was removed by sliding the complete bucky out and then the MTF test
device was placed on the detector entrance cover. The test device was positioned to
measure the MTF in two directions, first almost perpendicular to the CWE and then
almost parallel to it. A 10th order polynomial fit was applied to the results.
To measure the noise power spectrum the test device was removed and exposures
made for a range of incident air kerma at the surface of the table. The DQE is presented
13
as the average of measurements in the directions perpendicular and parallel to the

CWE.
2.10 Other tests
Other tests were carried out to cover the range that would normally form part of a
commissioning survey on new equipment. These included tests prescribed in IPEM
Report 894 for mammographic X-ray sets, as well as those in the UK NHSBSP protocol
for digital mammographic systems
3. Results
3.1 Output and HVL
The output and HVL measurements are shown in Table 5.
Table 5. Output and HVL

kV Target/filter Output (µGy/mAs HVL (mm Al) Focus
at 1m)
26 Mo/Mo 26.7 0.344 Large
29 Mo/Mo 38.3 0.375 Small
34 Rh/Ag 45.2 0.539 Large
The tube voltage measurements are shown in Table 6. All were within 0.3kV of
indicated values and are within the IPEM Report 894 remedial level of ±1kV.
Table 6. kV measurements made with Mo/Mo target/filter combination
kV set kV measured
26 25.8
28 27.9
30 30.2
32 32.3
3.2 Detector response
The detector response is shown in Figure 4.
14
6000
A v e ra g e p ix e l v a lu e
4000
y = 9 .8 x
2000
0
0 200 400 600
In c id e n t a ir k e r m a a t d e te c to r ( µ G y )
Figure 4. Detector response acquired at 34kV Rh/Ag anode/filter combination with 2mm
Al at the tube port
15
3.3 AEC performance
3.3.1 Dose
The MGDs for breasts simulated with PMMA, exposed using the 4 different AEC modes, are
shown in Figure 5 and Tables 7 to 10. The MGDs include the pre-pulse exposure, which is not
included in the stated mAs values.
8
Standard
Dose-
Standard+
6 Implant
MGD (mGy)
Remedial dose level
0
0 20 40 60 80 100
Equivalent breast thickness (mm)
Figure 5 MGD for different thicknesses of simulated breasts in the 4 AEC modes. (Error
bars indicate 95% confidence limits.)
Table 7. MGD for simulated breasts, AEC Standard mode

PMMA Equivalent kV Target/ mAs MGD Remedial Displ- Displayed
thick- breast filter (mGy) dose ayed % higher
ness thickness level dose than
(mm) (mm) (mGy) (mGy) measured
20 21 26 Mo/Mo 16.9 0.49 1.0 0.51 5
30 32 26 Mo/Mo 38.8 0.80 1.5 0.84 5
40 45 34 Rh/Ag 23.8 1.13 2.0 1.16 3
45 53 34 Rh/Ag 27.7 1.19 2.5 1.25 5
50 60 34 Rh/Ag 33.4 1.33 3.0 1.39 4
60 75 34 Rh/Ag 48.6 1.75 4.5 1.85 6
70 90 34 Rh/Ag 77.2 2.35 6.5 2.55 8
16
Table 8. MGD for simulated breasts, AEC Dose- mode

20 21 26 Mo/Mo 11.8 0.35 1.0 0.37 4
30 32 26 Mo/Mo 27.3 0.57 1.5 0.60 5
40 45 34 Rh/Ag 17.8 0.86 2.0 0.89 3
45 53 34 Rh/Ag 21.1 0.93 2.5 0.97 5
50 60 34 Rh/Ag 25.5 1.04 3.0 1.07 3
60 75 34 Rh/Ag 37.2 1.37 4.5 1.44 5
70 90 34 Rh/Ag 63.5 1.96 6.5 2.12 8
Table 9. MGD for simulated breasts, AEC Standard+ mode

20 21 26 Mo/Mo 16.9 0.49 1.0 0.51 5
30 32 26 Mo/Mo 42.1 0.86 1.5 0.90 4
40 45 34 Rh/Ag 30.8 1.43 2.0 1.45 1
45 53 34 Rh/Ag 40.3 1.70 2.5 1.78 5
50 60 34 Rh/Ag 53.5 2.09 3.0 2.17 4
60 75 34 Rh/Ag 78.5 2.74 4.5 2.90 6
70 90 34 Rh/Ag 100.3 3.02 6.5 3.27 8
Table 10. MGD for simulated breasts, AEC Implant mode

20 21 26 Mo/Mo 22.7 0.58 1.0 0.63 -12
30 32 26 Mo/Mo 46.7 0.91 1.5 0.95 1
40 45 34 Rh/Ag 31.8 1.39 2.0 1.48 -2
45 53 34 Rh/Ag 42.5 1.70 2.5 1.78 -1
50 60 34 Rh/Ag 48.8 1.84 3.0 1.91 -1
60 75 34 Rh/Ag 75.2 2.50 4.5 2.63 -1
70 90 34 Rh/Ag 108.3 3.14 6.5 3.26 2
17
3.3.2 Contrast-to-noise ratio
The results of the CNR measurements are shown in Figure 6 and Tables 11 to 12. The
following calculated values are also shown:
• CNR to meet the minimum acceptable image quality (IQ) standard at the 60mm
breast thickness
• CNR to meet the achievable image quality standard at the 60mm breast thickness
• CNRs at each thickness to meet the limiting value in the European protocol
S ta n d a rd
40 D ose-
S ta n d a rd +
Im p la n t
C N R fo r 0 .2 m m A l
C N R a t m in im u m IQ
30
C N R a t a c h ie v a b le IQ
C N R to m e e t E u ro p e a n lim itin g v a lu e
20
10
0
0 20 40 60 80 100
E q u iv a le n t b r e a s t th ic k n e s s ( m m )
Figure 6. CNR measured at the 4 AEC modes. (Error bars indicate 95% confidence
limits.)
Table 11. CNR measurements, AEC Standard mode

PMMA Equivalent Measured CNR for CNR for European
thickness breast thick- CNR minimum achievable limiting
(mm) ness (mm) acceptable IQ IQ CNR value
20 21 20.4 10.0 14.9 11.5
30 32 20.1 10.0 14.9 11.0
40 45 16.7 10.0 14.9 10.5
45 53 15.4 10.0 14.9 10.3
50 60 14.7 10.0 14.9 10.0
60 75 12.5 10.0 14.9 9.5
70 90 12.0 10.0 14.9 9.0
18
Table 12. CNR measurements, AEC dose modes compared

PMMA Equivalent Measured Measured Measured Measured
thickness breast thick- CNR CNR CNR CNR
(mm) ness (mm) Standard Dose- Standard+ Implant
20 21 20.4 16.4 20.5 24.0
30 32 20.1 16.2 20.9 22.1
40 45 16.7 14.7 19.9 19.8
45 53 15.4 13.4 18.6 19.1
50 60 14.7 12.4 18.4 17.5
60 75 12.5 11.1 16.5 16.3
70 90 12.0 10.6 13.9 14.0
3.3.3 AEC performance for local dense areas
For many systems, when the AEC adjusts for locally dense areas, the SNR remains
constant with increasing thickness of extra PMMA. The results of this test are shown in
Tables 13 to16 and Figure 7. The need for a more suitable test for the Pristina, for
which the aim is to keep CNR (rather than SNR) constant, is discussed in section 4.
For Standard, Dose- and Standard+ modes, the first 2 exposures are at 26kV Mo/Mo.
For greater thicknesses the exposures are at 34kV Rh/Ag and the SNR jumps to a
higher value, and then decreases only slowly with increasing thickness of PMMA. In
Implant mode, all exposures are at 34kV Rh/Ag, and the SNR decreases slowly from
the initial value.
Most SNR values are within the suggested limit of 20% of the mean SNR. Only for the
Mo/Mo exposures. The implant mode is designed to give exposures dependent on the
breast thickness, therefore this test is not relevant.
19
Table 13. AEC performance for local dense areas, AEC Standard mode
Total kV Target/ Tube SNR % difference
attenuation filter load from average
(mm PMMA) (mAs)
32 26 Mo/Mo 45.7 98.1 -20
34 26 Mo/Mo 54.2 97.1 -21
36 34 Rh/Ag 20.7 140.8 15
38 34 Rh/Ag 22.2 139.1 14
40 34 Rh/Ag 23.7 135.2 10
42 34 Rh/Ag 24.7 128.3 5
44 34 Rh/Ag 26.6 126.5 3
46 34 Rh/Ag 27.6 123.1 0
48 34 Rh/Ag 30.0 120.0 -2
50 34 Rh/Ag 31.8 117.2 -4
Table 14. AEC performance for local dense areas, AEC Dose- mode
(mm PMMA) (mAs)
32 26 Mo/Mo 32.0 79.4 -24
34 26 Mo/Mo 37.7 80.3 -23
36 34 Rh/Ag 15.0 117.6 13
38 34 Rh/Ag 16.0 114.9 10
40 34 Rh/Ag 17.6 114.4 10
42 34 Rh/Ag 18.6 111.4 7
44 34 Rh/Ag 20.3 111.4 7
46 34 Rh/Ag 21.2 106.6 2
48 34 Rh/Ag 23.0 102.3 -2
Table 15. AEC performance for local dense areas, AEC Standard+ mode
(mm PMMA) (mAs)
32 26 Mo/Mo 49.6 102.0 -27
34 26 Mo/Mo 59.9 101.8 -27
36 34 Rh/Ag 24.4 152.5 9
38 34 Rh/Ag 26.5 153.7 10
40 34 Rh/Ag 30.1 149.1 7
42 34 Rh/Ag 32.8 150.1 7
44 34 Rh/Ag 37.0 149.2 7
46 34 Rh/Ag 40.1 148.3 6
48 34 Rh/Ag 45.4 143.5 3
50 34 Rh/Ag 49.4 149.6 7
20
Table 16. AEC performance for local dense areas, AEC Implant mode
(mm PMMA) (mAs)
32 34 Rh/Ag 25.2 174.8 27
34 34 Rh/Ag 26.5 169.4 23
36 34 Rh/Ag 26.5 158.6 15
38 34 Rh/Ag 25.2 147.6 7
40 34 Rh/Ag 25.2 136.0 -1
42 34 Rh/Ag 26.5 134.3 -2
44 34 Rh/Ag 25.2 120.8 -12
46 34 Rh/Ag 26.5 120.1 -13
48 34 Rh/Ag 25.2 109.9 -20
50 34 Rh/Ag 25.2 103.5 -25
S ta n d a rd
200 D ose-
S ta n d a rd +
Im p la n t
150
SNR
100
50
0
30 35 40 45 50 55
P M M A th ic k n e s s ( m m )
Figure 7. AEC performance (4 modes) in local dense area test
3.4 Noise measurements
The variation in noise with dose was analysed by plotting the standard deviation in pixel
values against the incident air kerma at the detector, as shown in Figure 8. The fitted
21
power curve has an index of 0.54, which is close to the expected value of 0.5 for
quantum noise sources alone.
100
S ta n d a r d d e v ia tio n in p ix e l v a lu e
0 .5 4
y = 0 .6 1 x
10
1
100 1000
In c id e n t a ir k e r m a a t s u r fa c e o f d e te c to r ( µ G y )
Figure 8. Standard deviation of pixel values versus incident air kerma at detector
0 .0 1 5
M e a s u re d n o is e
F it to d a ta
Q u a n tu m n o is e
R e la tiv e n o is e
E le c tr o n ic n o is e
0 .0 1 0
S tru c tu ra l n o is e
0 .0 0 5
0 .0 0 0
0 200 400 600
Figure 9. Relative noise and noise components
Figure 9 shows the relative noise at different incident air kerma. The estimated relative
contributions of electronic, structural, and quantum noise are shown and the quadratic
sum of these contributions fitted to the measured noise (using Equation 3).
22
Figure 10 shows the different amounts of variance due to each component; the
quantum variance is seen to predominate.
S tru c tu ra l v a r ia n c e
Q u a n tu m v a ria n c e
E le c tro n ic v a r ia n c e
100
% o f to ta l v a ria n c e
50
0
50 100 500 1000
Figure 10. Noise components as a percentage of the total variance. (Error bars indicate
95% confidence limits.)
3.5 Image quality measurements
The exposure factors used for each set of 16 CDMAM images are shown in Table 17.
Table 17. Images acquired for image quality measurement

kV Target/filter Tube load Mean glandular dose to
(mAs) equivalent breasts 60mm
thick (mGy)
34 Rh/Ag 16 0.60
34 Rh/Ag 22 0.83
34 Rh/Ag 32 1.21
34 Rh/Ag 45 1.70
34 Rh/Ag 63 2.37
The contrast detail curves (determined by automatic reading of the images) at the
different dose levels are shown in Figure 11. The threshold gold thicknesses measured
for different detail diameters at the 5 selected dose levels are shown in Table 18. The
NHSBSP minimum acceptable and achievable limits are also shown.
23
The measured threshold gold thicknesses are plotted against the MGD for an equivalent
breast for the 0.1mm and 0.25mm detail sizes in Figure 12.
M G D = 0 .6 0 m G y
10 M G D = 0 .8 3 m G y
T h r e s h o ld g o ld th ic k n e s s ( µ m )
M G D = 1 .2 1 m G y
M G D = 1 .7 0 m G y
M G D = 2 .3 7 m G y
1 A c c e p ta b le
A c h ie v a b le
0 .1
0 .1 0 0 .1 3 0 .1 6 0 .2 0 0 .2 5 0 .3 1 0 .4 0 0 .5 0 0 .6 3 0 .8 0 1 .0 0
D ia m e te r ( m m )
Figure 11. Contrast-detail curves for 5 doses at 34kV Rh/Ag. (Error bars indicate 95%
confidence limits.)
Table 18. Average threshold gold thicknesses for different detail diameters for 5 doses
using 34kV Rh/Ag, and automatically predicted data
Threshold gold thickness (μm)
Diam
Accept Achiev
-eter
-able -able MGD = MGD = MGD = MGD = MGD =
(mm)
value value 0.60mGy 0.83mGy 1.21mGy 1.70mGy 2.37mGy
0.1 1.680 1.100 2.21 ± 0.22 1.30 ± 0.13 1.16 ± 0.11 1.00 ± 0.01 0.70 ± 0.07
0.25 0.352 0.244 0.371± 0.037 0.285 ± 0.029 0.260± 0.026 0.204 ± 0.020 0.188 ± 0.019
0.5 0.150 0.103 0.160 ± 0.019 0.31± 0.02 0.099 ± 0.012 0.085 ± 0.010 0.074 ± 0.009
1.0 0.091 0.056 0.074 ± 0.015 0.059± 0.012 0.048 ± 0.010 0.039 ± 0.008 0.031 ± 0.006
24
0.1 mm detail 0.25 mm detail

0.6
Threshold gold thickness (µm)

3
Threshold gold thickness (µm)
predicted threshold predicted threshold

fit to data (y = x-n) fit to data (y = x-n)
2 0.4 minimum
minimum
achievable
achievable
1 0.2
0 0.0
0 1 2 3 4 0 1 2 3 4
Dose (mGy) Dose (mGy)
Figure 12. Threshold gold thickness at different doses. (Error bars indicate 95%
confidence limits.)
3.6 Comparison with other systems
The MGDs to reach the minimum and achievable image quality standards in the
NHSBSP protocol have been estimated from the curves shown in Figure 12. The fitted
curves are of the form y = x-n. (The error in estimating these doses depends on the
accuracy of the curve fitting procedure, and pooled data for several systems has been
used to estimate the 95% confidence limits of about 20%.) These doses are shown
against similar data for different models of digital mammography systems in Tables 19
and 20 and Figures 13 to 16. The data for these systems has been determined in the
same way as described in this report and the results published previously.12-18 The data
for film-screen represents an average value determined using a variety of film-screen
systems in use prior to their discontinuation.
25
Table 19. The MGD for a 60mm equivalent breast for different systems to reach the
minimum acceptable threshold gold thickness for 0.1mm and 0.25mm details
System MGD (mGy) for 0.1mm MGD (mGy) for 0.25mm
GE Pristina 0.58 ± 0.12 0.53 ± 0.10

GE Essential 0.49 ± 0.10 0.49 ± 0.10
Fujifilm Innovality 0.61 ± 0.12 0.49 ± 0.10
Giotto Class 0.50 ± 0.10 0.40 ± 0.08
Hologic 3Dimensions 0.40 ± 0.08 0.33 ± 0.07
Philips MicroDose L30 C120 0.67 ± 0.13 0.47 ± 0.09
Planmed Clarity 0.60 ± 0.12 0.49 ± 0.10
Siemens Revelation 0.43 ± 0.09 0.39 ± 0.08
Film-screen 1.30 ± 0.26 1.36 ± 0.27
Table 20. The MGD for a 60mm equivalent breast for different systems to reach the
achievable threshold gold thickness for 0.1mm and 0.25mm details
System MGD (mGy) for 0.1mm MGD (mGy) for 0.25mm
GE Pristina 1.23 ± 0.25 1.25 ± 0.25

GE Essential 1.13 ± 0.23 1.03 ± 0.21
Fujifilm Innovality 1.15 ± 0.23 1.02 ± 0.20
Giotto Class 1.06 ± 0.21 1.05 ± 0.21
Hologic 3Dimensions 0.78 ± 0.16 0.74 ± 0.15
Philips MicroDose L30 C120 1.34 ± 0.27 1.06 ± 0.21
Planmed Clarity 1.15 ± 0.23 1.02 ± 0.20
Siemens Revelation 0.82 ± 0.17 0.85 ± 0.17
Film-screen 3.03 ± 0.61 2.83 ± 0.57
26
remedial dose level
3
MGD (mGy)
0
ss
s
l
en
y
a
n
ia
on
rit
lit
12
tin
io
la
nt
re
va
la
at
si
is
C
C
se
sc
C
el
no
en
Pr
0
tto
Es
ev
ed
L3
lm
In
im
E
io
R
m
G
Fi
3D
ji
e
G
Fu
G
s
an
os
en
c
Pl
D
i
em
og
ro
ic
ol
Si
M
H
s
ilip
Ph
Figure 13. MGD for a 60mm equivalent breast to reach minimum acceptable image
quality standard for 0.1mm detail. (Error bars indicate 95% confidence limits.)
remedial dose level
3
MGD (mGy)
0
ss
s
l
en
y
a
n
ia
on
rit
lit
12
tin
io
la
nt
re
va
la
at
si
is
C
C
se
sc
C
el
no
en
Pr
0
tto
Es
ev
ed
L3
lm
In
im
E
io
R
m
G
Fi
3D
ji
e
G
Fu
G
s
an
os
en
c
Pl
D
i
em
og
ro
ic
ol
Si
M
H
s
ilip
Ph
Figure 14. MGD for a 60mm equivalent breast to reach achievable image quality standard
for 0.1mm detail. (Error bars indicate 95% confidence limits.)
27
remedial dose level
3
MGD (mGy)
0
ss
s
l
en
y
a
n
ia
on
rit
lit
12
tin
io
la
nt
re
va
la
at
si
is
C
C
se
sc
C
el
no
en
Pr
0
tto
Es
ev
ed
L3
lm
In
im
E
io
R
m
G
Fi
3D
ji
e
G
Fu
G
s
an
os
en
c
Pl
D
i
em
og
ro
ic
ol
Si
M
H
s
ilip
Ph
Figure 15. MGD for a 60mm equivalent breast to reach minimum acceptable image
quality standard for 0.25mm detail. (Error bars indicate 95% confidence limits.)
remedial dose level
3
MGD (mGy)
0
ss
s
l
en
y
a
n
ia
on
rit
lit
12
tin
io
la
nt
re
va
la
at
si
is
C
C
se
sc
C
el
no
en
Pr
0
tto
Es
ev
ed
L3
lm
In
im
E
io
R
m
G
Fi
3D
ji
e
G
Fu
G
s
an
os
en
c
Pl
D
i
em
og
ro
ic
ol
Si
M
H
s
ilip
Ph
Figure 16. MGD for a 60mm equivalent breast to reach achievable image quality standard
for 0.25mm detail. (Error bars indicate 95% confidence limits.)
28
3.7 Detector performance
The MTF is shown in Figure 17 for the two orthogonal directions. Figure 18 shows the
NNPS curves for a range of air kerma incident to the detector.
1 .0
M T F (v ) p e r p e n d ic u la r to tu b e a x is
M T F (u ) p a ra lle l to tu b e a x is
0 .8
0 .6
M TF
0 .4
0 .2
0 .0
0 2 4 6 8 10
-1
S p a tia l fr e q u e n c y ( m m )
Figure 17. Pre-sampling MTF
1 6 .3 µ G y N P S ( u ) 6 6 .2 µ G y N P S ( v )
1 6 .3 µ G y N P S ( v ) 1 3 4 µ G y N P S (u )
-4
10
3 2 .8 µ G y N P S ( u ) 1 3 4 µ G y N P S (v )
3 2 .8 µ G y N P S ( v ) 2 6 7 µ G y N P S (u )
6 6 .2 µ G y N P S ( u ) 2 6 7 µ G y N P S (v )
-5
10
N N P S (m m )
2
-6
10
-7
10
0 2 4 6
-1
Figure 18. NNPS curves for a range of air kerma incident to the detector
29
Figure 19 shows the DQE averaged in the two orthogonal directions for a range of
entrance air kerma. The MTF and DQE measurements were interpolated to show
values at standard frequencies in Table 21.
0 .8 1 6 .3 µ G y
3 2 .8 µ G y
6 6 .2 µ G y
0 .6
134 µG y
267 µG y
DQE
0 .4
0 .2
0 .0
0 1 2 3 4 5 6
-1
Figure 19. DQE averaged in both directions for a range of air kerma incident to the
detector
Table 21. MTF and DQE measurements at standard frequencies (DQE at 66.2µGy)
Frequency (mm-1) MTF (u) MTF (v) DQE
0.0 1.00 1.00 -
0.5 0.92 0.92 0.68
1.0 0.85 0.86 0.68
1.5 0.76 0.77 0.65
2.0 0.65 0.67 0.63
2.5 0.55 0.57 0.59
3.0 0.46 0.49 0.55
3.5 0.38 0.41 0.50
4.0 0.31 0.35 0.42
4.5 0.25 0.29 0.33
5.0 0.20 0.23 0.24
30
3.9 Other tests
The results of all the other tests that were carried out were within acceptable limits as
prescribed in the UK protocol1 and IPEM Report 89.4
3.9.1 Compression
The measured compressed breast thicknesses are compared with the displayed values
in Table 22. They were within 2mm of displayed values. This is well within the IPEM
Report 894 remedial level of > 5mm.
Table 22. Indicated compressed breast thickness

Actual Indicated Difference
thickness (mm) thickness (mm) (mm)
20 18 2
40 38 2
70 68 2
Measurements of compression force together with the IPEM Report 894 remedial levels
are shown in Table 23.
Table 23. Compression force and thickness

Measured force (N) IPEM Report 89
remedial level (N)
Maximum motorised compression 198 < 150 or > 200
Maximum compression in any mode 198 > 300
Compression change over 30 seconds 2 > 20
3.9.2 Image retention
The image retention factor was 0.01, compared to the NHSBSP upper limit of 0.3.
3.9.3 Missed tissue at CWE
The missed tissue was measured as 5mm, which is equal to the NHSBSP limit.
3.9.4 Mesh
No discontinuities were seen in the image of the fine wire mesh.
31
3.9.5 Distortion
Measurements showed that there was no distortion in an image of small aluminium balls
spaced 50mm apart across the whole image (the tomosynthesis test tool).
3.9.6 AEC repeatability
For a series of 5 repeat images, acquired in quick succession, the maximum deviation
of mAs from the mean was 1.3%. For 6 images, acquired at intervals over several days
of testing, the maximum deviation was 1.9%. The NHSBSP remedial level is 5%.
3.9.7 Uniformity and artefacts
Uniformity measurements showed a variation in pixel values of between 2 and 7%

relative to the central area of a 24cm x 29cm image of 45mm thick PMMA. The results
are shown in Table 24; all are below the NHSBSP remedial level of 10%.
Table 24. Uniformity

Position Mean pixel value Deviation from
centre value
Centre 898
CWE left corner 883 2%
CWE right corner 884 2%
Nipple edge left corner 842 6%
Nipple edge right corner 835 7%
There was slight visible non-uniformity in the unprocessed image, as shown in Figure
20. A plot of mean pixel value along the midline, from CWE to nipple edge, is shown in
Figure 21.
Chest wall edge
Figure 20. Unprocessed image of 45mm thick PMMA.
32
1000
800
M e a n p ix e l v a lu e
600
400
200
0
0 50 100 150 200 250
D is ta n c e fr o m C W E ( m m )
Figure 21. Profile plot along midline, perpendicular to CWE
3.9.9 Cycle time
For a typical exposure of 34kV Rh/Ag at 28mAs, a subsequent exposure could be made
17 seconds after the start of the previous one.
3.9.10 Backup timer
When an AEC exposure was attempted with a steel plate blocking the X-ray beam, the
exposure terminated after a short time of less than 1s after the pre-exposure. There was
no main exposure and no image acquired. A message “AOP aborted, change
acquisition mode” was displayed.
3.9.11 Movement and safety
All movements were smooth. The minimum height of the breast support table was 65cm
and the maximum was 150cm. There were no rough edges.
A yellow radiation symbol was displayed on the console during exposures.
The safety cut-out (red button) on the console stopped operation when pressed. The
display remained on and the system could be rebooted from the console.
33
4. Discussion
The detector response was found to be linear, as expected.
MGDs measured using PMMA were well within the NHSBSP limits for all equivalent
breast thicknesses in all 4 AEC modes (Figure 5). The MGDs to a 53mm equivalent
breast thickness were 1.19mGy, 0.93mGy, 1.70mGy and 1.70mGy respectively in AEC
modes Standard, Dose-, Standard+ and Implant (Tables 7 to 10). The displayed doses
were almost all higher than the calculated MGDs, on average by approximately 5%.
This may be partly accounted for by small differences in measurement, for example the
HVLs in the DICOM headers are 0.37 for 26kV Mo/Mo and 0.57 for 34kV Rh/Ag, slightly
different from the measured values in Table 5.
CNR measurements made with plain PMMA showed a steady decrease with increasing
equivalent breast thickness (Figure 6). Target CNR values of 10.2 and 15.0, for
minimum acceptable and achievable image quality respectively, were calculated. All
CNR values exceeded the European limiting values for CNR (Tables 11 to 12). All AEC
modes exceeded the CNR target for minimum image quality from 20 to 90mm
equivalent breast thicknesses. In the Standard mode the CNR target for achievable
image quality was equalled or exceeded up to 60mm equivalent breast thickness. In
Standard+ mode this target was exceeded up to 75mm equivalent breast thickness.
Consideration should be given to using the Standard mode for breasts up to 60mm thick
and the Standard+ mode for greater thicknesses. The Dose- mode is not recommended
for routine use because of the resulting reduction in image quality.
The European guidelines include a test for whether the SNR remains approximately
constant as the thickness of added PMMA increases. The results (shown in Tables 13
to 15) for Standard, Dose- and Standard+ modes showed that a nearly constant SNR
was maintained for total thicknesses of 36mm and above. At 32 and 34mm thickness, a
lower SNR value was seen, related to the choice of 26kV Mo/Mo as exposure factors for
these thicknesses only. For Implant mode, 34kV Rh/Ag was selected by the AEC for all
thicknesses, and no step in SNR value occurred, but there was a gradual decrease with
total thickness of PMMA, as shown in Table 16.
However, the test for constant SNR is not the most appropriate test for this system.
Since the design aim is to keep CNR constant, rather than SNR, it would be better to
modify the test to include a 0.2mm thick aluminium square, and measure CNR instead.
Noise analysis showed that quantum noise is the main contribution to the noise over a
wide range of incident air kerma (Figure 10). There are minimal contributions from
electronic and structural noise.
34
Threshold gold thicknesses for a range of detail diameters are shown in Figure 11. At
an MGD of 1.26mGy (close to that selected for the equivalent thickness of PMMA in
Standard mode), the image quality was very close to the achievable level for all contrast
detail diameters.
Threshold gold thickness measurements at different dose levels for the 0.1mm and
0.25mm diameter details were used to calculate MGDs to a simulated 60mm equivalent
breast required for the minimum and achievable levels of image quality (Figure 12). This
allowed comparisons to be made between this and other systems previously tested.
The dose required for the Pristina to reach the achievable level of image quality was
comparable to that calculated for other digital mammography systems (Table 20).
The detector performance, as indicated by MTF, NNPS and DQE curves (Figures 17 to 19), was
satisfactory.
Results of other miscellaneous tests, presented in Section 3.9, were satisfactory.
5. Conclusions
The MGD is well below the remedial level for all AEC modes. In Standard mode, the
MGD to the standard breast (53mm equivalent breast) is 1.19mGy. The image quality,
as measured by threshold gold thickness, is at the achievable level for the Standard
mode.
The GE Senographe Pristina, operating in 2D mode, meets the requirements of the

NHSBSP standards for digital mammography systems.
35
References
1. Kulama E, Burch A, Castellano I et al. Commissioning and routine testing of full field
digital mammography systems (NHSBSP Equipment Report 0604, Version 3). Sheffield:
NHS Cancer Screening Programmes, 2009
2. van Engen R, Young KC, Bosmans H, et al. European protocol for the quality control of
the physical and technical aspects of mammography screening. In European guidelines
for quality assurance in breast cancer screening and diagnosis, Fourth Edition.
Luxembourg: European Commission, 2006
3. van Engen R, Bosmans H, Dance D et al. Digital mammography update: European

protocol for the quality control of the physical and technical aspects of mammography
screening. In European guidelines for quality assurance in breast cancer screening and
diagnosis, Fourth edition – Supplements. Luxembourg: European Commission, 2013
4. Moore AC, Dance DR, Evans DS et al. The Commissioning and Routine Testing of
Mammographic X-ray Systems. York: Institue of Physics and Engineering in Medicine,
Report 89, 2005
5. Alsager A, Young KC, Oduko JM. Impact of heel effect and ROI size on the
determination of contrast-to-noise ratio for digital mammography systems. In
Proceedings of SPIE Medical Imaging, Bellingham WA: SPIE Publications, 2008,
691341: 1-11
6. Boone JM, Fewell TR and Jennings RJ. Molybdenum, rhodium and tungsten anode
spectral models using interpolating polynomials with application to mammography
Medical Physics, 1997, 24: 1863-1974
7. Berger MJ, Hubbell JH, Seltzer SM, Chang et al. XCOM: Photon Cross Section
Database (version 1.3) http://physics.nist.gov/xcom (Gaithersburg, MD, National Institute
of Standards and Technology), 2005
8. Young KC, Oduko JM, Bosmans H, Nijs K, Martinez L. Optimal beam quality selection in
digital mammography. British Journal of Radiology, 2006, 79: 981-990
9. Young KC, Cook JH, Oduko JM. Automated and human determination of threshold
contrast for digital mammography systems. In Proceedings of the 8th International
Workshop on Digital Mammography, Berlin: Springer-Verlag, 2006, 4046: 266-272
10. Young KC, Alsager A, Oduko JM et al. Evaluation of software for reading images of the
CDMAM test object to assess digital mammography systems. In Proceedings of SPIE
Medical Imaging, Bellingham WA: SPIE Publications, 2008, 69131C: 1-11
36
11. IEC 62220-1-2, Determination of the detective quantum efficiency – Detectors used in
mammography. International Electrotechnical Commission, 2007
12. Mackenzie A, Oduko JM. Technical evaluation of the Hologic 3Dimensions digital
mammography system in 2D mode. London: Public Health England, 2019
13. Young KC, Oduko JM, Gundogdu O and Asad M. Technical evaluation of profile
automatic exposure control software on GE Essential FFDM systems (NHSBSP
Equipment Report 0903). Sheffield: NHS Cancer Screening Programmes, 2009
14. Tyler N, Mackenzie A. Technical evaluation of Siemens Revelation Digital

Mammography System in 2D mode. London: Public Health England, 2019
15. Oduko JM, Young KC. Technical evaluation of Philips MicroDose L30 with AEC software
version 8.3 (NHSBSP Equipment Report 1305). Sheffield: NHS Cancer Screening
Programmes, 2013
16. Strudley CJ, Oduko JM, Young KC. Technical evaluation of the Fujifilm AMULET
Innovality Digital Mammography System (NHSBSP Equipment Report 1601). London:
Public Health England, 2017
17. Tyler N, Young KC, Oduko JM, Mackenzie A. Technical evaluation of IMS Giotto Class
Digital Mammography System in 2D mode. London: Public Health England, 2019
18. Tyler N, Oduko JM, Strudley C, Mackenzie A. Technical evaluation of planmed Clarity
Digital Mammography System in 2D mode. London: Public Health England, 2019
37

GE SenographePristinaTechnical 2D 2019

Uploaded by

Copyright:

Available Formats

GE SenographePristinaTechnical 2D 2019

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GE SenographePristinaTechnical 2D 2019

Uploaded by

Copyright:

Available Formats

NHS Breast Screening Programme

About Public Health England

About PHE screening

www.gov.uk/phe/screening Twitter: @PHE_Screening Blog: phescreening.blog.gov.uk

© Crown copyright 2019

Published December 2019

The GE Senographe Pristina, operating in 2D mode, meets the requirements of the

The aims of the evaluation were:

• to determine whether the GE Senographe Pristina digital mammography system,

• to provide performance data for comparison against other systems

Table 1. System description

2.2 Output and HVL

Figure 1. The GE Senographe Pristina

2.3 Detector response

2.4 Dose measurement

2.5 Contrast-to-noise ratio

Figure 2. Location and size of ROI used to determine the CNR

Contrast = 1 − e-μt (1)

Table 3. Limiting values for relative CNR

2.6 AEC performance for local dense areas

Measurements were carried out using each of the AEC modes.

Spacers (10mm thick)

AEC sensor area

Extra attenuation (20mm x 40mm)

Spacers (10mm thick)

Figure 3. Setup to measure AEC performance for local dense areas

2.7 Noise analysis

σp =� ke2 + kq2p + ks2p2 (3)

2.8 Image quality measurements

TCpredicted = rTCauto (6)

Table 4. Values of r used to predict threshold contrast

Threshold gold thickness = a + bx-1 + cx-2 + dx-3 (7)

Threshold contrast = λD-n (8)

2.9 Physical measurements of the detector performance

as the average of measurements in the directions perpendicular and parallel to the

2.10 Other tests

The output and HVL measurements are shown in Table 5.

Table 5. Output and HVL

Table 6. kV measurements made with Mo/Mo target/filter combination

3.2 Detector response

The detector response is shown in Figure 4.

3.3 AEC performance

Remedial dose level

Table 7. MGD for simulated breasts, AEC Standard mode

Table 8. MGD for simulated breasts, AEC Dose- mode

Table 9. MGD for simulated breasts, AEC Standard+ mode

Table 10. MGD for simulated breasts, AEC Implant mode

3.3.2 Contrast-to-noise ratio

Table 11. CNR measurements, AEC Standard mode

Table 12. CNR measurements, AEC dose modes compared

3.3.3 AEC performance for local dense areas

Figure 7. AEC performance (4 modes) in local dense area test

3.4 Noise measurements

Figure 9. Relative noise and noise components

3.5 Image quality measurements

Table 17. Images acquired for image quality measurement

0.1 mm detail 0.25 mm detail

Threshold gold thickness (µm)

predicted threshold predicted threshold