CR Imaging Mythbusters

8/14/2019 CR Imaging Mythbusters

http://slidepdf.com/reader/full/cr-imaging-mythbusters 1/38

Myth vs. Realityin Digital ImageQualityand SystemPerformanceSteve MangoWW Technical Manager

Carestream Health Inc.!DT Solutions



".#

The $%&#'&.() Question*

What do users +ant,

-r

What do users think they +ant,

-r

What do users need,

What is the real requirement??



".

Is it/.

X Lp/mm resolution?

Nth duplex wire pair?

X Micron laser spot size?

X Micron pixel size?

20% MT o! X Lp/mm?

" ma#ic super duper $ &' (ma#in# )late?



".0

"*

None o! those parameters +speci!ications,- .themsel es- can #uarantee a s stem will achie e theradiographic sensitivity required !or a particular

examination tas1



".&

&ontrast 3N' MT

&N'

(ma#e 4ualit )erception

12ray Scatter

Scanner 3lare

12ray 45"osure

Scanner 6aser 45"

Imaging Plate7eg. Structure noise8

9ac:ground !oise

6aser 45"osureS"ot ;itter

6uminescent decay

4lectronic <and+idth

6aser S"ot Si=e

Imaging Plate

Scanner !oise

12ray 4nergy 7:>"8

P< Screen

Scanner?Screen gain

12ray 4nergy 7:>"8P< Screen



Image Quality Perce"tion

".%

&ontrast 3N' MT

&N'

Tas1 .ased ima#e perception



Contrast and !oise 3actors

".'

&ontrast 3N' MT

&N'


12ray energy 7:>8

12ray scatter

P< screen

scanner flare

12ray e5"osure

Scanner?screen gain

Scanner laser e5"osure

Imaging "late7eg. Structure noise8

9ac:ground noise

Scanner noise



Contrast and !oise 3actors

".)

&ontrast 3N' MT

&N'


12ray energy 7:>8

12ray scatter

P< screen

scanner flare

12ray e5"osure

Scanner?screen gain

Scanner laser e5"osure

Imaging "late7eg. Structure noise8

9ac:ground noise

Scanner noisecan <e o"timi=ed <y

techni@ueA



Modulation Transfer 3unction

It is customary to characteri=e s"atial resolution of systems using the MT3. TheMT3 re"resents the ma5imum a<ility of a system to transfer su<Bect contrast to thefinal image as a function of s"atial fre@uency.

".(



45am"le System MT3 Com"onents

".

C # 0 &. C

.C

C

EP Screen F 6aser 45"osureG C. 5

System MT3 7 F !y@uist G #.(0C cy?mm 8

9eam 7d<GC µ m8

Screen 7hGC.08

Pream" 72 d<F #H!y@uist8 3ilter 72d<

F CH!y@uist8 Decay 7 of τ c

G&.C8 System

M T 3

S"atial 3re@uencyS"atial 3re@uency 7cycles?mm8

MT3



The CR ReaderJs Role in System MT3

".

&ontrast 3N' MT

&N'


6aser e5"osure

S"ot Bitter

6uminescent decay

4lectronic <and+idth6aser s"ot si=e

Imaging "late

12ray energy 7:>8

P< screen

can <e o"timi=ed <y <etterdesign choices or <y

o"timi=ation of scan "arameters



Image Quality Perce"tion

". #

&ontrast 3N' MT

&N'


6aser e5"osure

S"ot Bitter

6uminescent decay

4lectronic <and+idth6aser s"ot si=e

Imaging "late

12ray energy 7:>8

P< screen

can <e o"timi=ed <y design or o"timi=ation of thic:ness "hos"hor grain

si=e use of a<sor<er dyes etc.



".

Imaging Plate MT3

KThe "resam"ling MT3 descri<es the system res"onse u" to <ut not includingthe stage of sam"ling. The figure <elo+ sho+s the "resam"ling MT3 curves for<oth imaging "lates.

KThe im"roved shar"ness of the I!DLSTR41 HR imaging "late comes from

smaller "hos"hor "articles and a thinner "hos"hor layer.

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6

Spacial Frequency (cycles/mm)

Sharpness(MTF%)

HR imaging plates

GP imaging plates

" 0



ey Contri<utors to CR Image Quality

6etJs ta:e a closer loo: at the effects of*

K 6aser "o+er

K S"ot si=e

K Pi5el si=e

K 45"osing energy

K 45"osure dose

". 0

" &



". &

4ffect of laser "o+er

$i#her laser power penetrates deeper into thephosphor la er- de#radin# MT

Hi Po+er 6o+ Po+er

" %



Imaging Plate MT3 as a function of 6aserPo+er

". %

S"atial 3re@uency 7cycles?mm8

MT3

" '



6aser S"ot Si=e N What is it,

". '

O laser does not ma:ea "erfect s"ot +ith ashar"ly defined <oundary.

Instead it has a gaussiandistri<ution +ith highintensity in the center and

falling off gradually a+ayfrom the center.

3o- how is it measured or characterized?

0

500

1000

1500

2000

2500

3000

3500

4000

" )



. )

Common S"ot Si=e Criteria

The t+o most common +ays of Characteri=ing s"ot si=e are*

5 ull Width $al! Max6

and

7/e 2

But there isNo standard!

" (



. (

4ffect of laser s"ot si=e

" 2X impro ement in spot size does not ield a&orrespondin# 2X impro ement in ima#e qualit 8

microns"ot

& microns"ot

" #



.#

45am"le Eaussian 9eam MT3s

# 0 % ) C.C

C

Eaussian 9eam Mtfs

M t f

S"atial 3re@uency cy?mmQ

&%.&µ micron (). µ micron Im"rovement 3actor

" #



45am"le System MT3 Com"onents

.#

C # 0 &. C

.C

C

EP Screen F 6aser 45"osureG C. 5

System MT3 7 F !y@uist G #.(0C cy?mm 8

9eam 7d<GC µ m8

Screen 7hGC.08 Pream" 72

d<F #H!y@uist8

3ilter 72d<

F CH!y@uist8 Decay 7 of τ

cG&.C8

System

M T 3

S"atial 3re@uency

MT3

S"atial 3re@uency 7cycles?mm8

Im"roving the laser 9eam 7s"ot8 MT3

9y /

/ +ould im"rove theSystem MT3 <y

-nly

" ##



.##

4ffect of "i5el si=e

6arge "i5el*3aster scan shorterd+ell time

Small "i5el*Slo+er scanlonger d+ell time

" smaller pixel can ha e acounterproducti e e!!ect on MT 8

".#



.#

4ffect of 45"osing 4nergy

Increasing :>

3 stem 9

3 stem X

3 stem :

Resol ution

With increasin# ener# - theper!ormance capa.ilities o!

arious &' s stems .e#in to5con er#e68



The 4ffect of Dose

# # Carestream Health

Increasing DoseIncrea sing!oise

;lectronicnoise

X<'anoise

() 3tructurenoise

5Noise6 is impro ed withincreasin# dose

".#&



S!R vs Pi5el >alue 7Dose8

Normalized 3N' s (N%=3T';X )ixel (ntensit/ +&>, !or "&' 2000iand $' (ma#in# )lates

D

#D

0D

%D

)D

CDD

C#D

C0D

C%D

C)D

#DD

D &DD CDDD C&DD #DDD #&DD

(N%=3T';X )ixel >alue +&>,

N o r m a

l i z e d 3 N ' () 7

() 2

() ?

() @() A() B

".#%



O Simulation

An input targetcontaining2% contrastpatches

5% contrast bars

0 1 2 3 4 5 6 7 8 9 100

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Spatial Frequency [cy/mm]

M t f

System Mtfbased onh-value

0 1 2 3 4 5 6 7 8 9 100.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


M t f

System Mtf basedon h-value

And two differentMTFs

LetCs illustrate a couple o! the pre ious pointsD +ith a simulation in +hich e5"osure

is varied using*

".#'



The 4ffect of MT3

0 1 2 3 4 5 6 7 8 9 100

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

SpatialFrequency [cy/mm]

M t f


1 mr50 um pixelGain 10

1 mr50 um pixelGain 10

0 1 2 3 4 5 6 7 8 9 100.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

SpatialFrequency [cy/mm]

M t f


".#)



The 4ffect of Dose

0 1 2 3 4 5 6 7 8 9 100

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


M t f

SystemMtfbasedon h-value

10 mr50 umpixelGain !

0 1 2 3 4 5 6 7 8 9 100

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1


M t f

System Mtfbased on h-value

1 mr50 um pixelGain !



O 3e+ Practical45am"les

".



Ho+ Much Resolution is 4noughU,

CR Scan at ' Micron "i5el si=e

".



MT3* 9e Careful What Vou Os: 3orA

Ty"ical HR Plate) Microns 9SR

S"ecial 9lueU Plate

% Microns 9SR

". #



Ta:e a closer loo:*

HR Plate 9lueU "late

".



S"ot si=e* not a <ig factor <y itself

& Micron Micron

##& :>" & micron "i5el .& mW :H=

Eoth laser spots produced similar results 8

". 0



Pi5el Si=e* See any difference,

Eoth samplin# pitches produce similar results8

)' micron "i5el & micron "i5el

". &



Onother e5am"le* .'& U 7 (mm8 3e "late

;ither pixel size meets the required sensiti it

'& micron s"ot' micron "i5el

'& micron s"ot& micron "i5el

". %



The 9ottom 6ine*

&' System per!ormance is a !unction o! man competin# parameters Thereis no sin#le 5ma#ic .utton68

KCR system "erformance is a function of many "arameters.

K O num<er of these factors com"ete +ith each other and are traded off against each other.

On im"rovementU in one "arameter can degrade another resulting in degraded"erformance.

KThe radiogra"hic techni@ue is a critical com"onent in gaining @uality image out"ut.

K4ven a very im"ressive shar"ness 7MT38 +ill not result in a good image +ithout good C!Rand S!R.

". '



Than: Vou

Man than1s to m collea#ues at&arestream $ealth !or their contri.utions tothis wor1*

Eo. Fulpins1iCarestream Health Inc.Research X Innovation 6a<s

Gacques 'oussilheCarestream Health 3rance!DT Solutions



CR Imaging Mythbusters

Documents