Physics of
Computed Radiography
Computed Radiography (CR)
...is the generic term applie d to an ima ging
system comprised of:
Overview
Acceptance Testing
Quality Control
Photostimulable Storage Phosphor
to a cqu ire the x-ray pr ojection image
CR Reader
J. Anthony Seibert, Ph .D.
University o f Califo rnia, Davis
Medical Ce nter , Sacramento
to extract th e el ectronic late nt image
Digital electronics
AAP M 1999 An nual Meeti ng, Nashvil le
to convert the signals to digital form
CR Detector
Computed Radiography “reader”
• Photostimulable Storage Phosphor (PSP)
• BaFBr compound, Eu activated
Information panel
Cas sette Holder
Phosphor Plate
Plate stacker
RIS interface
CR Image acquisition
1. X-ray Exposure
Patient
X-ray
system
5.
unexposed
2.
PSP
detector Image
Reader
CR reader interface
Download patient demographic data; select image processing algorithms
3.
4.
Image
Scaling
Image
Record
Computed
Radiograph
exposed
re-usable
phosphor
plate
1
ID terminal: select anatomy-specific exam
Bar-code reader: identify exposed cassette
Film laser printer
CR Networking
CR
Reader
• DICOM
– Digital Imaging COmmunications in Medicine
– Provides open architecture solutions for modality
interfaces, storage/retrieval, and print functions
Dicom
CR - QC Workstation
• Technologist QC Workstation
• Modality Worklist Input
• Processed image output
Soft-cop y review
CR vendors
CR Trends
• Fuji ........
(GE, Siemens, Philips, others )
• Lower system costs
• Agfa ........
(Toshiba)
• Smaller footprint
• Kodak
• High throughput sy stems
• Konica
• Low throughput systems “Table-top” units
• Lumisys
• Integrated QC workstations for images
• Others
• DICOM output
2
Digital x-ray detector
Conventional film/screen detector
1. A cqu isition, Display, Ar chiving
1. A cqu isition
Transmitted x-rays
through patient
2. Display
Transmitted x-rays
through patient
Film processing:
Digital
processing
light to optical density
Gray Scale
encoded on
film
Intensifying
Screens
x-rays → light
Film
An al og to Di gi tal
C on ve rsio n
3. A rchi ving
Stimulation and Emission Spectra
Electrons trapped in proportion to x-rays absorbed
3.0 eV
e-
F/ F +
τ =Eu
Eu3+ / Eu 2+
Incident x-rays
e-
e-
Laser
stimulation
2.0 eV
8.3 eV
Valence band
F centers proportional to
incident x-ray intensity
Relative intensity
PSL
τ =recombination
Di od e
68 0 nm
0.0
80 0
He Ne
63 3 nm
70 0
1.5
60 0
1.75
2
50 0
2.5
λ (n m)
40 0
30 0
3
4 Ene rgy (eV)
CR: Latent Image Readout
Referen ce
detector
Light guide
Emi ssion
0.5
Photostimulated Luminescence
Incident Laser Beam
BaFBr: Eu2+
Stimu latio n
1.0
Conduction band
τtunneling
=
X-ray converter
x-rays → electrons
Charge
collection
device
PSP Latent Image Formation
ph on on
D ig ital to An al og
C on ve rsio n
Digital Pixel
Matrix
PSL
Signal
PMT
f-θ
le ns
Lase r
Source
Polyg onal
Mirror
Cyli ndrica l mi rror
Li ght cha nnel ing gu ide
Output Signal
ADC
Exposed
Imaging
Plate
Light
Scattering
Photostimulated
Luminescence
Protective Layer
Laser beam:
Sca n direction
To i mage
proce sso r
Phosphor Layer
Las er Light Spr ead
"Effective " read out di amete r
Base Support
PMT
Plat e translation:
Sub-scan direction
3
Phosphor Plate Cycle
PSP
Base support
x-ray exposure
Sub -scan Direction
plate exposure:
create latent image
Plate translation
reuse
laser beam scan
Typic al resolution:
plate readout:
extract latent image
35 x 43 cm -- 2.5 lp/mm
24 x 30 cm -- 3.3 lp/mm
18 x 24 cm -- 5.0 lp/mm
light erasure
plate erasure:
remove residuals
Scan Direction
Laser beam deflection
Characteristic Curve:
CR Image Manipulation
response of screen/film and CR
• Contrast enhancement
– Anatomy specific grayscale manipulation
• Spatial frequency enhancement
Finding the Image Location
• Image recognition phase
10 ,0 00
4
Film Optical D ensity
– Find the pertinent image information
– Scale the data to appropriate range
Film -screen
(40 0 spe ed)
3
CR pla te
1,00 0
Ove rexposed
2
10 0
Corre ctly expo se d
10
1
Und erexpose d
0
0.01
0.1
1
10
10 0
1
Exp osure , mR
20 00 0
20 00
20 0
20
2
Sen sitivi ty (S)
Rel ative intensi ty of PSL
• Image pre-processing
Collimation area(s) determined
Collimation
Border
– collimation (Agfa)
– EDR (Fuji)
– segmentation (Kodak)
• Finding collimation borders and edges
4
Processing the Image
Histogram analysis
• Contrast enhancement
• Frequenc y distribution of pixel values within a
defined area in the image
– MUSIC A (Agfa)
– Gradation (Fuji)
– Tonescaling (Kodak)
• Shape is anatomy specific
• Define dynamic range (histogram analysis)
• Sets minimum and maximum “useful” pixel values
• Transform to anatomy specific contrast
Data conversion
Histogram Distribution
Open
area
Frequency
Pixel value
Useful signal
Input to outpu t dig ita l number
1,000
Re la ti ve PSL
Anatomy
Output dig ital number
Collimated
area
Grayscale transformation
Exposure into digital number
102
101
100
10-1
10-1 100 101 102 103 0
511
1023
Exp osure input
Raw Digital Outp ut
800
600
400
200
0
200
600 1,000
Raw Input d igital numbe r
Histogram
mi n
The shape is dependent on radiographic study,
positioning and technique
max
1. Find the
signal
2. Scale to
range
3. Create film
look-alike
Histogram: pediatric image
to 9 368
to 8 323
Frequency
800
600
400
200
0
0
200
400
600
800
100 0
Useful image range for anatomy
Di gital valu e
Pre-processed no
window/level
Contrast enhanced
5
Data conversion for overexposure
Data conversion for wide latitude
Exposure into digital number
Exposure into digital number
Incre ase gr adi ent
1 02
1 02
Rel ative PSL
Re la ti ve PSL
Re du ce gai n
1 01
1 00
1 0- 1
Exposure
in put
10 -1 10 0 10 1 10 2 10 3
0
51 1
10 23
Ra w Di gital Ou tp ut
overexposure
min
max
(sca led a nd l og am pli fi ed)
1 01
1 00
1 0- 1
Exposure
in put
10 -1 10 0 10 1 10 2 10 3
high kVp
(w ide latitude)
80 kVp, 18 mAs
400 speed screen - film
10 23
(sca led a nd l og am pli fi ed)
min
max
Computed Radiography
Overexposed
Screen- Film
80 kVp, 18 mAs
51 1
Ra w Di gital Ou tp ut
Screen-F ilm
Underexposed
0
Underexposed
Overexposed
CR
80 kVp, 18 mAs
L=4, wide latitude
6
CR: Contrast Enhancement
Look-up-table transformation
Output digital number
1,000
M
L
E
A
800
Fuji System
600
Example LUTs
400
200
0
0
200
400
600
1,000
800
Input digital number
“Raw data”
Spatial Frequency Processing
“Contrast Enhanc ed”
CR: Image manipulation
• “Edge Enhancement”
Di fference :
Edge Enh anced:
Dash : l ow pass fi ltered
Ori gin al - filtered
Di fference + Origin al
Response
Soli d: orig inal resp onse
Sum
lo w
hi gh
lo w
hi gh
lo w
hi gh
Spatial frequency
“Black Bone”
CR: Dual Energy Imaging
Low Energy Image
High Energy Image
“Edge Enhanced”
CR: Dual Energy Imaging
“Tissue only” Image
“Bone only” Image
7
CR: Spatial Resolution
Image Performance Measures
• Pho sph or p late sizes: impact on resolution
• Spatial res olution
– Dependent on IP size
– Less than corresponding speed screen-film
• Contrast sensitivity
– Dependent on exposure and SNR
• Exposure
– Variable speed detector
35x43 (14x17)
24x30 (10x12)
18x24 (8x10)
0.2 mm pixels
0.14 mm pixels
0.1 mm pixels
High Contrast (Spatial) Resolution
18 x 24 cm
MTF Curves
35 x 43 cm
1.0
0.8
Pre-sampled MT F
MTF
Screen-film
Scan
0.6
Subscan
Hi res CR
0.4
Sampled MT F:
Standard CR
2K x 2K matrix
35 x 43 cm
Standard CR
0.2
0
Photon absorption fraction
10
Low Contrast Response: Leeds TO -16
X-ray Absorption Efficiency
1
2
4
6
8
Spatial F requency (lp/mm)
BaFBr, 100 mg/cm²
0.8
Gd 2O2S, 120 mg/cm 2
0.6
0.4
BaFBr, 50 mg/cm²
0.2
0
0
20
40
60
80
Energy (keV)
100
120
140
3.5 mR
70 kVp
0.5 mR
8
Image retake rates
Retake rate evaluation -- 1st half, 1992
7
Screen-Fil m
6
Wrong exam
5%
Motion
6%
CR
5
4
Positioning
46%
Reprinting
9%
3
Under exposure
10%
2
1
0
Jan
Feb
Mar
Apr
May
O verexposure
12%
Jun
Total # repeats = 1043
from Willis, RSNA 1996
Adu lt portable chest ca lcu lated exposures
Adu lt portable chest ca lcu lated exposures
First half, 1994, 4572 exams
Second half, 1994, 4661 exams
38.3%
60 0
5 3.9 %
7.8%
Target
expo sure
rang e
60 0
50 0
7 3.5 %
3.4%
Target
expo sure
rang e
50 0
40 0
<50
10 0
20 0
0
30 0
10 0
0
40 0
10 0
50 0
Q4
20 0
System speed (S #)
Low
Incident Exposure
<50
20 0
Q3
30 0
10 0
Q2
20 0
30 0
30 0
Q1
40 0
40 0
#exams
#exams
23.1%
50 0
Percent error
Repeated
Examinations with CR
Other
12%
System speed (S #)
High
Low
Incident Exposure
High
Guidelines for QC based on Exposure,
typical adult exam
Apri l 1 - 17, 1996
Adul t Po rtabl e Ch est
Sys tem s peed
160
140
120
100
80
Grid tec hnique
without a grid
60
40
0
60
0
00
-6
50 9 9
054
9
40
04
35 4 9
03
30 7 4
032
4
25
02
20 7 4
02
15 2 4
017
4
10
012
4
50
-7
4
20
0
>1
Number of examinations
“Exposure Creep”
180
Sensitivi ty number
Indication
• >1000
<0.2 mR
• Underexposed: repeat
• 600 - 1000
0.3-0.2 mR
• Underexposed: QC exception
• 300 - 600
1.0-0.3 mR
• Underexposed: QC review
• 150 - 300
1.3-1.0 mR
• Acceptable range
• 75 -150
1.3-2.7 mR
• Overexposed: QC review
• 50 - 74
4.0-2.7 mR
• Overexposed: QC exception
• <50
>4.0 mR
• Overexposed: repeat
9
New issues for the Medical Physicist:
Digital Projection Imaging
Radiation Dose for CR
• Variable Speed Detector
• Differences between screen-film and PSP detectors
• Optimal dos e 2X higher than 400 speed screen/film
– Lower absorption efficiency
• Testing digital systems: vendor specific details
• Indirect (CR) vs. direct (Flat-panel) detectors
– Quantum and electronic noise
• Exposure levels and SNR measurements
– Readout inefficiencies of latent image
• QC phantoms
• Anti-scatter grids needed
• Soft-copy displays and workstations
Uniformity
Recommended acceptance tests
(Ta sk Group #10 -- AAPM)
498
508
537
480
• Physical Ins pection - Inventory
• Evaluation of image process ing parameters
490
513
• Imaging Plate Uniformity and Dark Noise
• Signal Response
– Linearity and Slope
– Calibration and Beam Quality
497
505
544
487
• Laser Beam Function
CR Parameter Settings
Demographics on Fuji CR output
Fu ji CR reader system
Ex po su re m e nu c ode
R ea d out (ED R) m ode
A - Auto m a tic
S - S e m i-a utom a t ic
F - Fix e d
U CD MC RA DIO LO G Y
00
A 0200
Top
of
film
Film Imag e
D ATE
PO R TAB LE C HE ST
L 2 .0 S 2 5 0 C *1 .6 , *1 .0
G 1 .0 E # 1 .0 -- 0 .2 0 R 0 .3
A 054
# 9 5 0 3 3 10 5 4
1 1 :4 5
B ottom
2 /3
of
film
L = La tit ude (us e ful r an ge of e x pos ur es in o rde rs of m a gnitude )
S = S e ns itiv ity (v a lue inv e rs e ly re la te d to inc ide nt e x pos ur e )
C = De ns ity / c ontr a s t s e tting m od ific a tions
G = Film ga m m a c ur ve s e tt ings (c ontr a st e nha nc e m e nt pa r am et e rs )
No te : the a s s oc ia te d le tte r s in dic a te the LU T ty pe
R = Fre que nc y pr oc e s sin g (s pa tia l a nd e dge e nh an c em en t pa ra m e te r s )
A = Num be r of film s s inc e la s t re bo ot of s y s te m
2 /3 = Im a ge re duc t ion fa c tor ( 6 7 % of a c tua l s iz e in this c a s e )
R= m
I a ge re v e rs a l indic a tor
Tim e
0 3 .3 1 .1 9 95
Pa tie nt Ide ntific a tion
R
Anatomical region
Gener al ch est (LAT)
Gener al ch est (PA)
Port Chest GRID
Port Chest NO GRI D
Peds chest NICU/PICU
Finger
Wrist
For earm
Plaste r cast (ar m)
Elbo w*
Upper Ribs*
Pelvis*
Pelvis por table
Tib/F b
i
Foo t
Foo t*
Os Calcis
Foo t cast
C-spine
T- sp n
i e
Swimme rs
Lumba r spine
Breast specimen
GA
1.0
0.6
0.8
1.0
1.1
0.9
0.8
0.8
0.8
0.8
0.8
0.9
0.9
0.9
0.8
1.2
0.8
0.8
1.1
0.8
1.2
1.0
2.5
GT
B
D
F
D
D
O
O
O
O
O
O
O
O
N
O
N
O
O
F
F
J
N
D
GC
1.6
1.6
1.8
1.6
1.6
0.6
0.6
0.6
0.6
0.6
1.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
1.8
0.9
0.9
0.6
GS
-0.2
-0.5
-0.0 5
-0.1 5
-0.2
0.3
0.2
0.3
0.4
0.4
0.0
0.2
0.2
0.25
0.3
-0.0 5
0.4
0.5
0.5
-0.0 5
0.3
0.4
0.35
RN
4.0
4.0
4.0
4.0
3.0
5.0
5.0
5.0
5.0
7.0
5.0
6.0
4.0
5.0
5.0
7.0
5.0
5.0
5.0
4.0
5.0
5.0
9.0
RT
R
R
T
R
R
T
T
T
T
T
R
T
T
F
T
T
F
F
P
T
T
T
P
RE
0.2
0.2
0.2
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
0.5
0.5
0.5
0.5
1.0
0.5
0.5
0.2
0.5
1.0
1.0
10
Da te:
7/1 0/98
M ed ci al Phy si ci st: Antho ny Sei be rt, Ph.D.
Lo ca tio n:
Sy ste m Ide ntif i ca tio n:
UCDM C, ACC, 3
CRu ni t 3
UC Da vi s Me dica l C ente r
Recommended acceptance tests
CR R eader and Screens
Sig na l Resp on se: Calib r atio n a nd Beam Qu ality
(Ta sk Group #10 -- AAPM)
No te: Use mAs va u
l es to pro vid e a n ap pro xi ma te ex po su re of 1 mR to th e IP.
Men u = TEST
IP T yp e: ST 1 4x 17
IP SN:
Ex po su re Co n ditio ns
Su bMenu = A ve 2. 0
L = 2 , E D R = se mi
Foc al sp ot
1. 2 m m
Tim e del ay
~2m n
i
S ID (c m)
140
mR -IP
S
S 1
( mR )
OD
NA
0 .9 1
0 .9 1
1 21 .0 0
1 08 .0 0
11 0. 5 9
98 .7 1
1 .4 1
1 .3 8
NA
NA
11 3. 0 4
14 .3 3
1 .4 5
0 .0 7
NA
NA
• High Contrast Resolution
SMD (cm)
1 30
k Vp De p en d en cy
k Vp
F li trati on
m As
mR -mete r
80
1 A /l 0 .5 Cu
1 A /l 0 .5 Cu
15 . 00
4 .5
1 . 06
1 . 06
11 5
1 A /l 0 .5 Cu
1. 13
1 . 14
60
0 .9 8
1 15 .0 0
Ma xi mu m Dif f eren ce :
• Noise / Low-Contrast Response
• Distortion
F i l trati o n D e pe n d en cy
k Vp
F li trati on
m As
80
n on e
0. 50
m R-me ter
0 . 96
80
80
1 A /l 0 .5 Cu
1A /l 2 .5 Cu
4. 50
60 . 00
1 . 06
0 . 99
mR -IP
S
S 1
( mR )
OD
NA
0 .8 3
1 87 .0 0
15 4. 7 9
1 .4 0
NA
0 .9 1
0 .8 5
1 08 .0 0
1 24 .0 0
98 .7 1
10 5. 8 5
1 .3 8
1 .4 0
NA
NA
56 .0 8
0 .0 2
NA
Ma xi mu m Dif f eren ce :
12 0. 00
• Erasure Thoroughness
• Anti-aliasing
180 0. 0
160 .0 0
10 0. 00
• Positioning and collimation errors
140 0. 0
120 .0 0
Re spon se
R es po ns e
8 0. 00
100 .0 0
80 .0 0
6 0. 00
4 0. 00
• Throughput
60 0. 0
40 .0 0
2 0. 00
S (1mR)
50
70
90
kV p
110
S (1mR )
20 .0 0
0. 00
13 0
0.0 0
n one
1 A /l 0. 5 Cu
Fi l trati on
1Al / 2. 5Cu
Date:
7/10/98
Medical P hysicist: A nthony S eibert, P h.D.
Location:
Syst em Identification:
UCDM C, A CC, 3
CR unit 3
UC Da vis Me dic al Cente r
C R R eader and Screens
In sp ection Resu lts Su mmary
Acceptable
1. Physica l Inspe ction - In vento ry
2. Imag ing Pla te Unif orm ity a nd Dar k No ise
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Sig nal Respo nse: L inea rity and Slo pe
Sig nal Respo nse: C ali bra tion and Be am Qu ality
Laser Be am Fun ction
Hig h-C ontr ast R esolu tion
Noi se /Low -Co ntr ast Re sp onse
Disto rtio n
Era su re Thor ou ghne ss
Ant i-A liasing
Posit ionin g an d C olli mati on E rr ors
Thr oug hput
Ye s
Ye s
Ye s
Ye s
Ye s
Ye s
Ye s
Ye s
Y es*
Ye s
Ye s
Ye s
Com men ts:
10 mAs
20 mAs
Sprea dsheet from Ehsan Samei, Ph.D ., Med ical Un iversity o f South Caroli na
Quality Control
Three levels of system performance for quality control and
system maintenance
1. Routine: Technologist level
- no radiation measurements
2. Full inspection: Physicist level
- radiation measurements; non-invasive adjustments
3. System adjustment: Vendor service level
- hardware and software maintenance
Periodic Quality Control
• Daily (technologist)
–
–
–
–
General inspection
Film processor / Laser printer
Erase imaging plates
Verify digital interfaces and network transmission
• Weekly (technologist)
– Verify CRT calibration
– Test phantom images
– System cleanliness
11
Periodic Quality Control
• Monthly (Technologist)
– Film processor maintenance (if any)
– Inspect and clean image receptors
Periodic Quality Control
• Semi-Annually / Annually (Physicist)
– Evaluate image quality
– Acceptance tests to re-establish baseline values
– Review
– Review film retake rate
– QC review for “out-of-tolerance” issues
CR: Specifications
• Phosphor plate throughput
• Spatial res olution
• Contrast resolution and dynamic range
•
•
•
•
patie nt expo sure tren ds
retake activity
QC re co rds
Service hi story
CR: Clinical Considerations
• Sensitiv ity to scatter
• Multiple images per phosphor plate?
• Patient demographic data
• RIS-HIS-DICOM interfaces / compliance
• Image quality control
• Peripheral equipment; QC phantoms
• Input to PACS
• Service issues; plate longevity; warranties
Computed Radiography Experience
• Flexibility is a double-edged sword
– reduced retakes but higher under/over exposures
– variable speed (need to tailor exposure to exam)
– more difficult to correctly use
Summary
• CR is currently the only readily available technology
for direc t digital acquisition of projection radiographs
• Experience with CR will provide a framework for
future digital detector implementation and QC
• Provides guidelines for new digital detectors
• Indicates the need for continuous training
• Filmless radiology requires a lot more than just digital
acquisition devices -- a massive investment in PACS
and knowledgeable support personnel, including
MEDICAL PHYSICIST INPUT is necessary
12