L02a Radiation Contrast
L02a Radiation Contrast
L02a Radiation Contrast
X-rays
(e.g. 60 keV) (e.g. 150 keV)
Radiation Contrast, Noise and Imaging
Requirements
Lecture 02 intensity
intensity
1 2
Digital Radiology
Radiation Contrast,Level
Noise II Radiation Contrast, Noise
and Imaging Requirements and Imaging Requirements
Essential conflict: I t = I P + IS
With increasing energy, the intensity difference of a structural
detail (e.g. due to a cavity) decreases in the X-ray image
independently from the chosen detector. C = ΔI = I P1 − I P2
Grey value contrast
This has to be taken into account in each particular inspection
task! I = I0 · e -μ·w
„More is better“ is the wrong approach to adjust the tube voltage For small w:
for the perceptibility of flaws!
ΔI ≈ IP · μ · Δw
3 4
5 6
1
Classical specific contrast Two cases must be considered
Scatter Ratio k Intensity
I Penetrated object
primer (w)
The proportion of the scatter radiation is defined by the (e.g. weld with crack)
specimen I Scatter
k=
length
k= S Crs 2 = primer ( w) −
1
IP ln(1 + k ( w))
Iprimary
I Scatter DScatter
k= = w
I Prim ( DScatter + DPrim ) − DScatter IScatter
kFilm 0.1 w for contact film radiography
7 8
length for steel, X-ray and w < 60 mm (DGZfP)
Radiation Contrast
Influence of Image Noise on the Visibility of Details
Signal Signal
• scatter ratio (Base material) (Basematerial)
2
Human Observer Model: Perception Threshold (PT)
Essential Parameters in Digital Radiography
What do you see? Detector = 50 µm pixel size
}
Compensation
{
Material, Exposure time Noise = 1000
keV, Gamma source Signal = 30000
Tube current, Activity is possible 0.5T 1T 2T 4T 8T ➢ Large area flaws are better
Scatter protection Detector efficiency CNR = 2.5 visible than small ones at
Screens and filters Source-to-Detector Distance
same Contrast-to-Noise
Alateral
{
Flat bottom holes of different depth and diameter
Depends on Hardware: effective pixel size SRb image = 50 µm Calculated CNRmin = 1
Specific normalized Magnification
contrast-to-noise ratio Focal spot size, source size Alateral Cdepth PT - human perception threshold
PTconst = Alateral – area of just visible indication
SRb
Cdepth – mean contrast of just visible indication
The normalized SNRN is considered as the equivalent value to the selection of a film - noise of image (N)
system and a required minimum opt. density. 13 14
SRb – basic spatial resolution (effective pixel size)
Rose approach, 1946
Radiation Contrast, Noise Radiation Contrast, Noise
and Imaging Requirements and Imaging Requirements
Qualification of Imaging Plate Systems by the Visibility of Example: EPS test with HD CR scanner at
EPS= equivalent penetrameter sensitivity (see ASTM E 746, E 747, E 1025)
IQIs and/or EPS calculation from µeff, SNR and SRb 20 µm pixel size
EPS – Procedure is proposed for CR qualification
in the new ASTM E 2033 draft and the Draft on “Practice for the Use of …
Computed Radiology for Aerospace Casting Inspections” (USA: MAI – group)
- The EPS (equivalent penetrameter sensitivity) measurement is based on E 746
CNR 0.8
- A smooth ¾ inch (19 mm) steel plate with a set of plate holes is radiographed at 200
kV in ≥ 1 m distance Do you see the holes?
- Other IQIs are on the plate
to increase the information
on image response.
Visibility of EPS holes (E 746) Formula for Conversion of SNRN Measurements to EPS
Values and Working Range for CR (ASTM Draft E 2033)
▪ PT’
the is aboutthe
IP, whereby 2▪100 forshall
PV=0 visibility of the 2atT unexposed
be measured hole of IQIs corr. to ASTM E 1025
areas.
Do you see the holes? ▪ EPS by ASTM E 746 with 200 kV, t = 19 mm Fe plate and µeff = 0.05 mm-1
EPS vs.SNR method with 3/4" Fe
4
High resolution imaging plate (blue)
PT ' SRbimage
EPS =
3,5
3
Standard imaging plate (white) ttestplate µeff SNR
2,5
EPS (%) = PT/sqrt(SNR)
2
SNR measured with IP I and Scanner II
1/
PT SNR
EPSmax
max = 1.4
1,5 SNR measured with IP II and Scanner I
1
( 1EPS
/ SNR )== 15%
15%
0,5 PVmin = 5000 Plateau area defines the
CNR 0.8 0
working range
0 10000 20000 30000 40000 50000 60000 70000
17 18
Pixel value
3
Concept of Normalized SNR Quality of X-ray Images
Statistic data of digitized Signal-to-Noise Ratio (SNR) in a region of interest (ROI)
film image in a ROI
- Is calculated by “mean value of gray values” / “standard
deviation of gray values” penetrable wall contrast intensity SNR
thickness CNR
- Is normalized to the scanning aperture value of a
standardized microphotometer for film classification
X-ray tube voltage / energy:
corresponding to ISO 11699-1, EN 584-1, ASTM E 1815
increasing larger lower higher SNR
- This permits the comparison of sharp film higher
radiography with unsharp digital detectors
decreasing smaller higher lower SNR
SNRN – normalized SNR lower
SNRN SNR 88,6 µm / SRb
X-ray tube current:
SRb – basic spatial resolution
increasing larger equal higher higher
SRb = uT / 2 decresing smaller equal lower lower
uT = total unsharpness 19 20
- high basic spatial resolution (SRbimage), which means Detector Contrast : Signal
SRbimage should have a small value, SNR for linear detectors
CNR
- high signal-to-noise ratio (SNR), SRbdetector – Basic Spatial Resolution
→ effective Pixel size
- high relative specific contrast (µeff ), which is the effective CNR - Contrast/Noise Ratio Detectors can be qualified
attenuation coefficient (includes scatter effects). SNR - Signal/Noise Ratio
and classified by its SNR
21 and SRbdetector 22
Collimator
4
Thermal and Optical Focal Spot Real Imaging
SDD Location
SOD
thermal focal spot
(also actual focal spot)
Collimator
Real
focal spot
Object
effective
focal spot Intensity
Detector
25 26
ug
=
(SDD − SOD ) d – focal spot size
M – magnification
d SOD
SDD – source detector distance
SOD – source object distance, also f
ODD – object detector distance, also b
SDD
ug = d − 1
SOD
Parameter: parameters:
large focal spot low unsharpness increasing unsharpness
large focal spot enlarged silhouette
short ODD low magnification
SOD = ODD
large SOD
29 30
5
parameters: parameters:
large focal spot high unsharpness low unsharpness
small focal spot
large ODD large magnification large magnification
large ODD
small SOD poor resolution of details small SOD low intensity
31 32
ISO Formula
geometric unsharpness ug d ODD d/ui = 8 d/ui = 2
ug =
SOD
focal spot d or, since ui will be called also udetector
fls
ug = d − 1 d/ui = 0,5
focus – object distance
SOD
Mopt-line
SOD*
M 33
Magnification v
34
*If detector unsharpness < focal spot size
6
Effect of the Unsharpness on the Contrast Evaluation – Illumination Intensity
minimal contrast
6000
geometric unsharpness 15 cm
geometric unsharpness nearly equivalent to the
about 0.5 width of the object width of the object
Luxmeter
Distance = 100 cm
ug ug
37 38
1500 2 x Distance
30 cm =
¼ of Intensity
Luxmeter
Distance = 200 cm
39 40
41