Physics testing of image detectors
Parameters to test
• Spatial resolution
• Contrast resolution
• Uniformity/geometric distortion
Features and Weaknesses of
Phantoms for CR/DR System Testing
• Dose response/signal dynamic range
• Noise
Added filtration
Experiments/testing methods
• Direct measurements
Donald Peck, PhD
Henry Ford Hospital
• Phantoms/image tools
• Qualitative
• Quantitative
detector
Phantom/image Tool
2
Phantom Types
Attenuation test tools
Attenuation
Metals
• Aluminum, Copper, etc.
High Contrast/Spatial Resolution
Plastics/Composite Materials
• Lucite, Tissue Equivalent, etc.
Low Contrast/Contrast Sensitivity
Water/Equivalent
Anthropomorphic
3
4
Detectors - Energy Sensitivity
Attenuator positioning
Digital detectors can be significantly more sensitive to scatter
radiation as compared to traditional phosphor screens
Modifying beam quality
• Position attenuators far from
detector to minimize scatter
contribution in measurement
• Scatter needs to considered at setup when testing systems
1
Absorption Efficiency
Pelvic Phantom
Primary Radiation
Storage
Phosphor
0.8
0.6
0.6
0.4
0.4
Rare Earth
Screen/Film
Phosphor
0.2
0.2
Pelvic Phantom
Scatter Radiation
20
40
60
Energy (keV)
80
100
Simulating patient attenuation
•
Position close to detector in
same location as patient
attenuator
detector
attenuator attenuator
0
0
0
•
0.8
Relative Intensity
Se
Detector
1
Incorrect
Positioning
Correct Positioning
120
5
Attenuator Construction
6
Attenuation test tools
Easy to use
Attenuator “purity” may not
be acceptable for the
measurement
Placement of attenuator needs to be considered based
on the test
• Measurement of
mammography HVL requires
attenuators that are at
least 99.9% Aluminum
Purity or Uniformity of material may not be adequate for
some tests
Tissue equivalent materials
may not be uniform
7
8
High Contrast/Spatial Resolution Test tools
Line pair patterns
Line pair patterns
Mesh patterns
Edge phantoms
9
Aliasing and Moiré Effect
10
Line pair patterns
Line pair pattern
Detector Pitch
11
12
Mesh Patterns
Moiré Effect
13
Mesh Patterns
14
MTF/DQE Measurement
IEC 62220-01 (2003)
• Method for determining Detective
Quantum Efficiency (DQE) of
digital imaging systems
• Defines specifications for a test
device required to make these
measurements
15
16
MTF/DQE Measurement Issues
MTF/DQE Measurement Issues
Fluorescent radiation
• Only issue at high kVp
• Important if comparing to other
MTF/DQE measurements
Determination of edge response
Edge Response Function
Intensity
Requires Pre-processed image values
that are “linear” with exposure
Fluorescent
Radiation
• Need to bin pixel data along edge
Position
• Phantom positioning critical for
consistent results
Noise Power Spectrum (NPS)
determination
• Variations in method used may
produce different results
LSF
ESF
• Variations in method used may produce
different results
• Need to remove effects of trends
associate with heel effect, etc.
• Important to standardize if
comparing to other MTF/DQE
measurements
• Important to standardize if comparing
to other MTF/DQE measurements
Distance
10000
Image Value
Smoothing/fitting of edge response
curves to allow utilization of Fourier
Analysis
Line Spread Function
Edge Spread Function
12000
8000
6000
4000
2000
0
0
2
4
6
8
10 12 14 16 18 20 22 24 26 28 30 32
Image Position
Distance
17
MTF Measurements
High Contrast/Spatial Resolution Test tools
Quantitative results
Line pair patterns
Good indication of changes
•
Subtleties in the measurement and analysis of data can make comparisons
between measurements by different tests inaccurate
•
Need to understand artifacts in results due to aliasing, etc.
CsI indirect detection
1
• Need to consider detector pitch in relation to resolution
pattern
Selenium direct detection
1
RQA5, Ver XQ/i
RQA5, Hor XQ/i
Edge Phantoms
• Objective
RQA5, Ver DR-1000
0.8
RQA5, Hor DR-1000
0.7
0.7
0.6
0.6
MTF
MTF
• Subjective
0.9
0.9
0.8
0.5
0.4
0.3
0.3
0.2
• MTF Determination
• Valid for determining if changes have occurred over time if
performed “consistently”
0.5
0.4
• Requires standardization of methods used if comparison between
systems or results from different physicists is desired
0.2
0.1
• Would benefit from development of “standardized” software
package to do the calculations
0.1
0
0
2
4
6
18
8
Spatial frequency (cycles/mm)
10
0
0
2
4
6
8
• Task Group No. 162 “Research Software for 2D Image”
10
Spatial frequency (cycles/mm)
19
20
Low Contrast/Contrast Sensitivity test tools
Threshold Contrast Detail Detection index (TCDD)
TCDD gives an indication of the
lowest contrast detectable (CT) as
a function of the detail size
Contains objects of varying
size and attenuation
given in terms of square root of the
object area (A½)
Can be quoted in terms of the
threshold detection index (HT)
Requires observers to
determine which objects are
visible
• Subjective
HT(A) = 1/[CT * A½]
High value for HT(A) indicates
good visibility
21
22
IPEM Criteria (example)
Institute of Physics and Engineering in Medicine (IPEM)
Goals:
Most results are subjective !
• Improving standards in clinical practice
• Providing advice on scientific and
engineering issues in healthcare to other
healthcare professionals, government
and the public.
Develops Reports and other publications
to achieve these goals
• Owns several journals:
• Physics in Medicine and Biology
• Physiological Measurement
• Medical Engineering and Physics
• Report 91 Recommended Standards for
the Routine Performance Testing of
Diagnostic X-Ray Imaging Systems
• Specifies the use of phantoms
throughout the testing procedures
23
24
ACR Accreditation Phantoms
ACR Radiography/Fluoroscopy
CT
Mammography
Modules included
• Chest
• General
• Fluoroscopy
Radiography/Fluoroscopy
Phantom image
• Radiography Chest/Abdomen
MRI
ACR Discontinued
Radiography/Fluoroscopy
Accreditation Program in 2005
SPECT
25
Original Equipment Manufacturer (OEM) Products
26
Test Phantom for Kodak (i.e. CareStream)
DIRECTVIEW Total Quality Tool
Automated Image Quality Control Tool
• Reproducible quantitative results
• May detect sub-visible changes in image quality
performance to initiate timely preventive maintenance
• Highly automated procedure
• Most provide data reporting in spreadsheet format
24 x 30 cm
18 x 24 cm
*Images provided by Eastman Kodak Company
35 x 43 cm
KODAK User Interface
Kodak Test Result Details
Uniformity
MTF (slow scan)
Noise
Spatial frequency
response (MTF)
Exposure linearity
Pixel size accuracy and
aspect ratio
29.5%
Phantom image
artifacts
Laser Beam Function
Residual signal erase
29
Test Limits
30
DIN 6868-58 (2001) and 6868-13 (2002)
Pre-set by OEM
Acceptance testing and constancy
checks of projection radiography
systems with digital image receptors
Basis for limit may not be justified in OEM literature
• German standard for testing of
Storage Phosphor systems using a
specially designed phantom to
measure image quality parameters
If system fails a test, Service Engineer may not be
educated how to correct problem
• Can purchase a phantom that will
meet the requirements of this
standard from several vendors
AAPM Report 93: Acceptance Testing and Quality Control
of Photostimulable Storage Phosphor Imaging Systems
Recommends using vendor/manufacturer supplied phantom
for Quality Control testing
• Since each vendor/manufacturer system would be different,
the Report could not specify exactly what to do or look for in
the results
31
32
Anthropomorphic phantoms
Shape “mimicking”
Shape “mimicking”
Anatomically Accurate
33
Shape “mimicking”
34
Anatomically Accurate
35
36
European Protocol for QC of … mammography
Addendum on Digital Mammography
Image Processing
It is acknowledged that at present it
is not possible to get unprocessed
images from some systems
Un-Processed
The image processing may introduce
artifacts on phantom images and may
be different from image processing
for mammograms due to histogram or
local texture based processing
techniques
Therefore care needs to be taken in
interpretation of these processed
images
Un-Processed
Processed
Processed
37
Image Processing
38
Phantom vs. Real Life
To optimize image processing
parameters the phantom would
need to accurately mimic “true”
anatomy
At least one reference of a
study attempting to accomplish
this with a phantom for a
specific vendor processing
method*
Impractical to do this for all
Body Parts/Views and all
vendor’s image processing
systems
*Moore CS, et al, “A method to optimize the processing algorithm of a computed radiography system for chest radiography”,
British Journal of Radiology, 80 (2007), 724–730
39
40
AAPM Report 93: Acceptance Testing and Quality Control of
Photostimulable Storage Phosphor Imaging Systems
Detroit? Phantom
Government
Phantom
Lists anthropomorphic phantoms in the recommended
equipment list
• Doesn’t specify how to use the anthropomorphic phantoms
in the Report
41
42