Digital Radiography - Chapter 11 17-Feb-20

Adjuncts to Radiology – Chapter 12

Digital Radiography – Chapter 11

Adjuncts to Radiology – Chapter 12

a copy of this lecture may be found at:

[Link]
l

Take Away: Five Things You should be able

to Explain after the DR/Adjuncts Lecture

 The various types of detectors used in digital imaging

(e.g., scintillators, photoconductors, etc.)
 The differences between the various technologies used
for digital radiography (e.g., CR, indirect and direct DR)
 Benefits of each type (e.g., resolution, dose efficiency)
 Why digital image correction and processing are
necessary or useful and how they are executed
 The various types of adjuncts to radiology (e.g., DSA or
dual-energy imaging), what issue they are trying to
resolve, mechanism exploited and end result

BME HCMUT 1
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Q: Why has radiology moved towards

the use of digital detectors?

 Insert audience participation here

Q: What are some benefits of

using film-screen technology?

 up until quite recently, ubiquitous

 simple, in use for around 100 years

 film is the triple threat – detector, viewer and storage

medium – all in one

 low equipment investment costs ($50K/room rather than

$500K/room)

 don’t need a multi-million $ PACS to view

BME HCMUT 2
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Q: What are some limiting factors of the almost

100 year old analog film-screen technology?

 limited dynamic range (only about two orders of

magnitude rather than four or five) – may require more
exposure repeats
 film can only be viewed in one location at a time (unless
a copy film produced)
 can’t visualize the image immediately – patient has to
wait while film processed and reviewed – radiographic
room throughput and utilization suffer
 film processor problems and film processor QA
monitoring time consuming

Q: What image information might we lose with the

use of digital detectors?

 loss of spatial resolution (limiting resolution about 2.5

lp/mm for 17”x17” general-purpose digital detector,
screen-film about 5-6 lp/mm)

 quantization error (depends on how many bits are used

to represent each pixel, but not as great an issue as the
loss of spatial resolution)

BME HCMUT 3
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Review: Digital Representation of Data

 Storage of Positive Integers

 In general, n bits have 2n possible permutations and can
represent integers from 0 to 2n-1 (the range usually denoted with
square brackets):
 n bits represents 2n values with range [0, 2n-1]
 8 bits represents 28 = 256 values with range [0, 255]
 10 bits represents 210 = 1024 values with range [0, 1023]
 12 bits represents 212 = 4096 values with range [0, 4095]
 14 bits represents 214 = 16384 values with range [0, 16383]
 16 bits represents 216 = 65,536 values with range [0, 65535]

Review: Conversion of Analog Data to Digital Form

c. f. Bushberg, et al., The Essential Physics

8
of Medical Imaging, 2nd ed., p. 69.

BME HCMUT 4
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Review: Digital Storage of Images

y↑
 Usually stored as a 2D array
(matrix) of data, I(x,y): I(1,1),
I(2,1), … I(n,m-1), I(n,m)
 Each minute region of the
image is called a pixel (picture
element) represented by one
value (e.g., digital value, gray
level or Hounsfield unit)
 Typical matrices:
 CT: 512x512x12 bits/pixel
 CR: 1760x2140x10 bits/pixel
 DR: 2048x2560x16 bits/pixel

x→

c.f. Huang, HK. Elements 9

of Digital Radiology, p. 8.

Review: Effect of Resolution and Bits per Pixel

10242, 642, 322, 162 matrices 8, 3, 2, 1 bits/pixel

c.f.: Bushberg, et al., The Essential Physics c.f.: Bushberg, et al., The Essential Physics
10
of Medical Imaging, 2nd ed., p. 82. of Medical Imaging, 2nd ed., p. 84.

10

BME HCMUT 5
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Q: What are some of the characteristics of ideal

materials that could be used in digital detectors?

 photoelectric interactions →
high Z, matched to photon
spectrum exiting patient

 adequate thickness to absorb

a large number of x-rays, but
not so thick as to adversely
impact spatial resolution

 low amount of x-ray energy

required to produce a light
photon or electron (signal)

c. f. Bushberg, et al., The Essential Physics

11
of Medical Imaging, 2nd ed., pp. 152 & 154.

11

Q: What types of materials are used

as detectors for digital imaging?

 gas detectors (Xe):

X-ray→e-→ADC
 photoconductors (Se, CdTe, HgI2, PbI2):
X-ray→e-→TFT→ADC
 Scintillators/phosphors (CsI, Gd2O2S):
X-ray→e-→Visible Light (VL)→e-→TFT→ADC
 photostimulable phosphors (BaFBr):
X-ray→e-→F+/F→laser→e-→VL→PMT→ADC

12

12

BME HCMUT 6
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Detectors in Digital Imaging (1)

 Solid-state materials
 Electrons arranged in bands with conduction band usually empty
 Solid-state detectors
 Photoconductor – charge collected and measured directly
 Scintillator(phosphor) – some deposited energy converted to visible light
 Photostimulable phosphors – energy stored in electron traps

c.f. Yaffe MJ and Rowlands JA.

13
Phys. Med. Biol. 42 (1997), p 10.

13

Detectors in Digital Imaging (2)

c.f. Yaffe MJ and Rowlands JA.

14
Phys. Med. Biol. 42 (1997), p 9.

14

BME HCMUT 7
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Periodic Table of the Elements

15
c.f. [Link]

15

Q: What major technologies have or are being

used to supplant analog screen/film radiography?

 Computed Radiography (CR)

 Digital Radiography (DR)

X-ray→e-→F+/F→laser→e-→VL→PMT

c.f. Bushberg, et al. The Essential Physics c.f. [Link]

16
of Medical Imaging, 2nd ed., p. 294. /Shimadzu_Safire_Dynamic_DR.html

16

BME HCMUT 8
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Q: What features of CR have made it useful as a transition

between analog film/screen and digital radiography?

 uses the same sized cassettes as in film/screen

radiography, so little impact to the then current
film/screen radiography suite equipment (i.e., Bucky tray
and phototimers)
 initial implementations of CR used film printers to make
hardcopy for review (keep old lightboxes and film
alternator equipment – didn’t have to spend $ on PACS)
Radiographic Suite CR Cassette CR Reader Film Alternator

c.f. [Link]
17
time_stamp/images/CR_casette_2.jpg c.f. [Link]

17

Computed Radiography (CR)

 Photostimulable phosphor (PSP)

 Barium fluorohalide: 85%
BaFBr:Eu + 15% BaFI:Eu
 e- from Eu2+ liberated through
absorption of x-rays
 Liberated e- fall from the
conduction band into ‘trapping
sites’ near F-centers
 Via low energy laser light (700
nm) stimulation the e- are re-
promoted into the conduction
band where some recombine with
the Eu3+ ions and emit a blue-
green (450-550 nm) visible light
(VL)

c.f. Bushberg, et al. The Essential Physics

18
of Medical Imaging, 2nd ed., p. 295.

18

BME HCMUT 9
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Computed Radiography (CR) System (1)

 Imaging plate (IP) made of PSP is

exposed identically to SF
radiography in Bucky
 IP in CR cassette taken to CR
reader where the IP is separated
from cassette
 IP is transferred across a stage
with stepping motors and scanned
by a laser beam (~700 nm) swept
across the IP by a rotating
polygonal mirror
 Light emitted from the IP is
collected by a fiber-optic bundle
and funneled into a photomultiplier
tube (PMT)
 PMT converts VL into e- current

c.f. Bushberg, et al. The Essential Physics

19
of Medical Imaging, 2nd ed., p. 294.

19

Computed Radiography (CR) System (2)

 Electronic signal output from PMT

input to an ADC
 Digital output from ADC stored
 Raster swept out by rotating
polygonal mirror and stage
stepping motors produces I(t) into
PMT which eventually translates
into the stored DV(x,y):
PMT→ADC→RAM
 IP exposed to bright light to erase
any remaining trapped e- (~50%)
 IP mechanically reinserted into
cassette ready for use
 200mm and 100mm pixel size -
(14”x17”: 1780x2160 and Summary: X-ray→e-→F+/F→laser→e-
3560x4320, respectively) →VL→PMT→ADC→RAM

c.f. Bushberg, et al. The Essential Physics

20
of Medical Imaging, 2nd ed., p. 294.

20

BME HCMUT 10
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Computed Radiography (CR) System (3)

 IP dynamic range = 104, about

100x that of S-F (102)
 Very wide latitude → flat contrast
 Image processing required:
 Enhance contrast
 Spatial-frequency filtering
 CR’s wide latitude and image
processing capabilities produce
reasonable OD or DV for either
under or overexposed exams
 Helps in portable radiography:
where the tight exposure limits of
S-F are hard to achieve
 Underexposed → ↑ quantum
mottle and overexposed →
unnecessary patient dose

c.f. Bushberg, et al. The Essential Physics

21
of Medical Imaging, 2nd ed., p. 296.

21

CR and Underexposure

22
c.f. Radiographics Lee et al. 11 (6): 1013.

22

BME HCMUT 11
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

CR – Presented Image and Exposure Range

55 kVp 55 kVp
3.0 mA 1.5 mA

55 kVp
0.5 mA

c.f. Bushberg, et al. The Essential Physics

c.f. Radiographics Lee et al. 11 (6): 1013. of Medical Imaging, 2nd ed., p. 296.
23

23

Charged-Coupled Devices (CCD)

 Form images from visible light

 Videocams & digital cameras
 Each picture element (pixel) a
photosensitive ‘bucket’ (PE)
 After exposure, the elements
electronically readout via ‘shift-
and-read’ logic and digitized
 Light focused using lenses or
fiber-optics
 Fluoroscopy (II)
 Digital cineradiography (II)
 Digital biopsy system
(phosphor screen)
 1K and 2K CCDs used

c.f. Bushberg, et al. The Essential Physics

24
of Medical Imaging, 2nd ed., pp. 298-299.

24

BME HCMUT 12
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Q: What are the two major types of digital radiography

technology and what is the difference between them?

 Indirect DR – create visible

light photons from x-rays with
scintillator then produce
electrons with photodiodes
typically lower spatial
resolution than direct DR and
lower dose efficiency than
direct DR due to limiting the
phosphor thickness so as not
to adversely impact spatial
resolution

25

25

Q: What are the two major types of digital radiography

technology and what is the difference between them?

 Direct DR – directly create

electrons from absorbed x-
rays, typically higher spatial
resolution than indirect DR and
higher dose efficiency than
indirect DR due to electric field
lines constraining electron
lateral drift (can make
photoconductor thick - > high
x-ray absorption)

26

26

BME HCMUT 13
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Indirect Flat Panel Detectors

 Use an intensifying screen -

e.g., CsI(Tl) - to generate VL
photons from an x-ray
exposure
 EK: Cs = 36 keV, I = 33.2 keV
 Light photons absorbed by
individual array photodetectors
 Each element of the array
(pixel) consists of transistor
(readout) electronics and a
photodetector area
 The manufacture of these
arrays is similar to that used in
laptop screens: thin-film
transistors (TFT)

c.f. Bushberg, et al. The Essential Physics

27
of Medical Imaging, 2nd ed., pp. 234 & 301.

27

Thin-Film Transistors (TFT)

 After the exposure is complete

and the e- have been stored in
the photodetection area
(capacitor), rows in the TFT
are scanned, activating the
transistor gates
 Transistor source (connected
to photodetector capacitors is
shunted through the drain to
associated charge amplifiers
 Amplified signal from each
pixel then digitized and stored
 Summary: X-ray→e-→VL→e-
→TFT→ADC→RAM

c.f. Bushberg, et al. The Essential Physics

28
of Medical Imaging, 2nd ed., p. 301.

28

BME HCMUT 14
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Resolution and Fill Factor

 Dimension of detector element largely determines spatial resolution

 200mm and 100mm pixel size typical
 For dimension of ‘a’ mm - Nyquist frequency: FN = 1/(2a)
 If a = 100mm → FN = 5 cycle/mm or lp/mm
 Fill factor = (light sensitive area)/(detector element area)
 Trade-off between spatial resolution and contrast resolution

c.f. Bushberg, et al. The Essential Physics

29
of Medical Imaging, 2nd ed., p. 303.

29

Direct Flat Panel Detectors

 Use a layer of photoconductive

material (e.g., α-Se) atop a TFT
array
 e- released in the detector layer
from x-ray interactions used to
form the image directly
Indirect Flat Panel Detector (for comparison)
 X-ray→e-→TFT→ADC→RAM
-
 High degree of e directionality
through application of E field
 Photoconductive material can be
made thick w/o significant
degradation of spatial resolution
 Photoconductive materials
 Selenium (Z=34, EK = 12.7 keV)
Direct Flat Panel Detector
 CdTe, HgI2 and PbI2

c.f. Bushberg, et al. The Essential Physics

30
of Medical Imaging, 2nd ed., p. 304.

30

BME HCMUT 15
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Periodic Table of the Elements

31
c.f. [Link]

31

Digital versus Analog Processes & Implementation

 Although some of the previous image reception systems

were labeled ‘digital’, the initial stage of all those devices
produce an analog signal that is later digitized
 CR: x-rays→F+/F → VL→PMT→current→voltage→ADC
 CCD, direct & indirect digital detectors: stored e- → ADC
 Benefits of CR
 Same exam process and equipment as screen-film radiography
 Many exam rooms serviced by one reader, lower initial cost
 Benefits of DR
 Throughput ↑: radiographs available immediately for QC & read
 Reduced radiation dose: 2-3x compared with CR
 Greater spatial resolution possible

32

32

BME HCMUT 16
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Huda Ch6: Digital X-ray Imaging Question

 12. Photostimulable phosphors [systems] do NOT

include:
 A. Analog-to-digital converters
 B. Barium fluorohalide
 C. Light detectors (blue)
 D. Red light lasers
 E. Video cameras

33

33

Huda Ch6: Digital X-ray Imaging Question

 11. Which of the following x-ray detector materials emits visible light:

 A. Xenon
 X-ray→e-→ADC
 B. Mercuric iodide
 X-ray→e-→TFT→ADC
 C. Lead iodide
 X-ray→e-→TFT→ADC
 D. Selenium
 X-ray→e-→TFT→ADC
 E. Cesium iodide
 X-ray→e-→VL→e-→TFT→ADC

34

34

BME HCMUT 17
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Raphex 2002 Question: Digital Radiography

 D47. Concerning computed radiography (CR), which of

the following is true?
 A. Numerous, small solid-state detectors are used to
capture the x-ray exposure patterns.
 B. It has better spatial resolution than film.
 C. It is ideal for portable x-ray examinations, when
phototiming cannot be used.
 D. It is associated with high reject/repeat rates.
 E. The image capture, storage, and display are
performed by the receiver.

35

35

Huda Ch6: Digital X-ray Imaging Question

 13. Photoconductors convert x-ray energy directly into:

 A. Light
 B. Current
 C. Heat
 D. Charge
 E. RF energy

36

36

BME HCMUT 18
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Why is Image Processing Important or Necessary?

 Insert audience participation here

37

37

Why is Image Processing Important or Necessary?

 Want to do some neat things:

 Remove bone, see only contrast, mimic CT
 Increase sharpness, enhance contrast
 Digital detector elements don’t operate identically
 Some pixels go bad or have a large offset
 The wide dynamic range of digital detectors → flat image
 Must manipulate image digitally to view → window/level
 May wish to reduce image noise
 May wish to enhance the spatial frequency → MTF
 Try to automatically set gray-scale mapping of image

38

38

BME HCMUT 19
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Look Mom, no bones!

39

39

Digital Image Correction (1)

c.f. Bushberg, et al. The Essential Physics

40
of Medical Imaging, 2nd ed., p. 310.

40

BME HCMUT 20
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Digital Image Correction (2)

 Interpolation to fill in dead pixel and row/column defects

 Subtracting out average dark noise image Davg(t)(x,y)
1 9 5 6
4 5 3
4 6 2 5 6 4
7 3 8 7 4 5
6
D(t=0)(x,y): dark noise Davg(t)(x,y): dark noise avg

 Differences in detector element digital values for flat field

 Gain image: G(x,y) =G’(x,y) - Davg(t)(x,y); Gavg =(1/N) ∙   G(x,y)

126 112 116 120

122 107 113 113 113 113
108 125 117 103 119 113 113 113 113
127 120 111 120 116 1
106
105 113 113 113
G’(x,y): raw gain image G(x,y): corrected gain image Gavg: average gain image

41

41

Digital Image Correction (3)

 Make corrections for each detector element (map)

 I(x,y) = Gavg ∙ [Iraw(x,y) - Davg(t)(x,y)] / G(x,y)

█ 1281
1225 226
222 1208
1139 1201
1131 233
229 1205
1136
█
?
█
? 213
209 1276
1198 230
228 228
224 1206
1132 226
224
█ █
? 1288
1208 237
235 1139
1052 1206
1131 227 1209
1126

Target phantom: X Iraw(x,y): raw image of X I(x,y): corrected image of X

 Done for DR and in a similar manner for CT (later) – and

in your digital camera!
 Not performed for CR on a pixel by pixel basis, although
there are corrections on a column basis for differences in
light conduction efficiency in the light guide to the PMT

42

42

BME HCMUT 21
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Global Processing

 Most common global image

processing: window/level
 Global processing algorithm
 I’(x,y) = c ∙ [I(x,y) – a]:
essentially y = mx + b
 Level (brightness) set by a
 Window (contrast) set by c
 I’ = [2N/ww]∙[I-{wl-(ww/2)}],
where ww = window width and
wl = window level
 Need threshold limits when
max/min [2N-1, 0] digital
values encountered
 If I’(x,y) > Tmax→I’(x,y) = Tmax
 If I’(x,y) < Tmin→I’(x,y) = Tmin

c.f. Bushberg, et al. The Essential Physics

43
of Medical Imaging, 2nd ed., pp. 92 and 311.

43

I’(x,y) = c ∙ [I(x,y) – a]

a
I(x,y) – a

44

44

BME HCMUT 22
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Window and Level Controls

 Modification of the translation table causes changes in the displayed
image brightness and contrast and is usually done through window
(contrast) and level (brightness) controls (e.g., under mouse control)
 In the example (below), the window is kept constant as the level is
increased, causing the image to become darker and darker
 The narrower the window, the greater the displayed image contrast, but
more of the image saturated in either toe or shoulder regions of LUT

darker
brighter

c.f.: Bushberg, et al., The Essential Physics

45
of Medical Imaging, 2nd ed., p. 92.

45

Window/Level Demo

46

46

BME HCMUT 23
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Image Processing Based on Convolution

 Convolution: Ch. 10 - Image

Quality and Ch. 13 - CT
 Defined mathematically as
passing a N-dimensional
convolution kernel over an N-
dimensional numeric array (e.g.,
2D image or CT transmission
profile)
 At each location (x, y, z, t, ...) in
the number array multiply the
convolution kernel values by the
associated values in the numeric
array and sum
 Place the sum into a new numeric
array at the same location

c.f. Bushberg, et al. The Essential Physics

47
of Medical Imaging, 2nd ed., p. 312.

47

Image Processing Based on Convolution

 Delta function kernel

0 0 0
0 1 0
0 0 0

 Blurring kernel (normalization)

also known as low-pass filter
1/9 1/9 1/9
1/9 1/9 1/9
1/9 1/9 1/9

 Edge sharpening kernel

-1 -1 -1
-1 9 -1
-1 -1 -1
c.f. Bushberg, et al. The Essential Physics
48
of Medical Imaging, 2nd ed., p. 313.

48

BME HCMUT 24
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Image Processing Based on Convolution

 Convolution kernels can be much larger than 3 x 3, but

usually N x M with N and M odd
 Can also perform edge sharpening by subtracting
blurred image from original → high-frequency detail
(harmonization)
 The edge sharpened image can then be added back to
the original image to make up for some blurring in the
original image: CR unsharpmasking - freq. processing
 The effects of convolution cannot in general be undone
by a ‘de-convolution’ process due to the presence of
noise, but a deconvolution kernel can be applied to
produce an approximation: 19F MRI

49

49

Unsharpmasked Spatial Frequency Processing

c.f. Bushberg, et al. The Essential Physics

50
of Medical Imaging, 2nd ed., p. 313.

50

BME HCMUT 25
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Median and Sigma Filtering

 Convolution of an image with a kernel where all the

values are the same, e.g. (1/NxM), essentially performs
an average over the kernel footprint
 Smoothing or noise reduction
 This can make the resulting output value susceptible to
outliers (high or low)
 Median filter: rank order values in kernel footprint and
take the median (middle) value
 Sigma filter: set sigma (s) value (e.g., 1) and throw out
all values in kernel footprint > m + s or < m – s and then
take the average and place in output image

51

51

Multiresolution/Multiscale Processing and

Adaptive Histogram Equalization (AHE)

 Some CR systems (Agfa/Fuji) make use of

multiresolution image processing (AKA unsharpmasking)
to enhance spatial resolution
 Wavelet or pyramidal processing on multiple frequency
scales
 Histogram equalization re-distributes image digital
values to uniformly span the entire digital value range
[2N-1,0] to maximize contrast
 AHE does this on a spatial sub-region basis in an image
rather than the entire image
 Fuji ‘Dynamic Range Control’ (DRC) a version of AHE
that operates on sub-regions of digital values

52

52

BME HCMUT 26
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Histogram Equalization
Properly Exposed Image Over-exposed Image

Under-exposed Image Histogram Equalized Image

c.f. [Link]
igorpro/imageprocessing/imagetransforms/
53
[Link]

53

Global and Adaptive Histogram Equalization

The following images illustrate
the differences between global
and adaptive histogram
equalization.
MR image with the corresponding gray-scale
histogram. The histogram has a peak at
minimum intensity consistent with the
relatively dark nature of the image.

Global histogram equalization and the final

gray-scale histogram. Comparing the results
with the figure above we can see that the
Adaptive histogram equalization shows better
distribution was shifted towards higher values
contrast over different parts of the image. The
while the peak at minimum intensity remains.
corresponding gray-scale histogram lacks the
mid-levels present in the global histogram
equalization as a result of setting a high contrast
c.f. [Link]
level.
igorpro/imageprocessing/imagetransforms/
54
[Link]

54

BME HCMUT 27
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Why is Image Processing Important or Necessary?

 Digital detector elements don’t operate identically (gain)

 Some pixels go bad or have a large offset (correction)
 The wide dynamic range of digital detectors → flat image
 Must manipulate image digitally to view → window/level
 May wish to reduce image noise (convol., median, σ)
 May wish to enhance the spatial frequency → MTF
 Convolution processing (unsharpmasking)
 Fourier and wavelet processing
 Try to automatically set gray-scale mapping of image
 Histogram equalization

55

55

Contrast vs. Spatial Resolution in Digital Imaging

 S-F mammography can

produce images w/ > 20 lp/mm
 According to Nyquist criterion
would require 25 mm/pixel
resulting in a 7,200 x 9,600
image (132 Mbytes/image)
 Digital systems have inferior
spatial resolution
 However, due to wide dynamic
range of digital detectors and
image processing capabilities,
digital systems have superior
contrast resolution

c.f. Bushberg, et al. The Essential Physics

56
of Medical Imaging, 2nd ed., p. 315.

56

BME HCMUT 28
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Digital Imaging Systems and DQE

 Remember the equation for

DQE(f)?
α-Si flat panel DR
 DQE(f) =

 How can we account for this?

 Both CR and the screens in
film/screens made thin
 Film higher spatial resolution
than CR
 DQE higher for α-Si systems
using CsI rather than α-Se
(mean x-ray E & Z)
 α-Si DQE falling off more
rapidly than α-Se (indirect)

c.f. Granfors and Aufrichtig. Medical

57
Physics. 27 (2000), p 1330.

57

Huda Ch6: Digital X-ray Imaging Question

 15. Which of the following does NOT involve image

processing:

 A. Background subtraction
 B. Energy subtraction
 C. Histogram equalization
 D. K-edge filtering
 E. Low-pass filtering

58

58

BME HCMUT 29
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Huda Ch6: Digital X-ray Imaging Question

 14. Processing a digital x-ray image by unsharpmask

enhancement would increase the:

 A. Bit depth per pixel

 B. Limiting spatial resolution
 C. Matrix size
 D. Patient dose
 E. Visibility of edges

59

59

Geometric (Linear) Tomography

 With the advent of CT, geometric

tomography has only limited
clinical utility where only one or a
few planes of objects with high
contrast are desired, e.g., IVP
 Desired slice through patient set
at pivot point (focal plane)
 The tomographic process blurs
out regions outside the focal
plane, but still contributes to
overall loss of contrast
 Larger tomographic angles result
in a lessening of out of plane
contributions
 High dose, comparable to CT for
many tomographic slices

c.f. Bushberg, et al. The Essential Physics

60
of Medical Imaging, 2nd ed., p. 318.

60

BME HCMUT 30
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Digital Tomosynthesis

 Improved version of geometric

tomography where a digital
detector saves an image at
each of several tube angles
 This allows reconstruction of
multiple planes through the
object through shifting the
various images through a
certain distance before
summing them
 Much more dose efficient, but
still suffers from out of plane
blurring effects
 Either CR or DR used

c.f. Bushberg, et al. The Essential Physics

61
of Medical Imaging, 2nd ed., p. 320.

61

Temporal Subtraction

 Digital Subtraction Angiography

(DSA) – usually 1K resolution
 Mask (background) subtracted
from images during/post contrast
injection: Δ < 1% trans. visualized
 Motion can cause misregistration
artifacts
 Digital value proportional to
contrast concentration and vessel
thickness
 Is = ln(Im) – ln(Ic) = mvessel ∙ tvessel
 Temporal subtraction works best
when time differences between
images is short
 Possible to spatially warp images
taken over a longer period of time

c.f. Bushberg, et al. The Essential Physics

62
of Medical Imaging, 2nd ed., p. 322.

62

BME HCMUT 31
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Bas-relief Effect from Subtraction Misregistration

63

63

Dual-Energy Subtraction

 Exploits differences between

the Z of bone (Zeff ≈ 13) and
soft tissue (Zeff ≈ 7.6)
 Images taken either at two
different kVp (two-shot), or
 One image (one-shot) taken
with energy separation
provided by a filter (sandwich)
 Iout = loge(Ilow) – R ∙ loge(Ihigh),
where R is altered to produce
soft-tissue predominant or
bone predominant images
 Chest DR @ UWMC/SCCA

c.f. Bushberg, et al. The Essential Physics

64
of Medical Imaging, 2nd ed., p. 324.

64

BME HCMUT 32
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Dual-Energy Subtraction

60 kVp 120 kVp

Exposure Exposure

Soft
Bone
Tissue
Computed
Computed
DE Image
DE Image
c.f. Bushberg, et al. The Essential Physics
65
of Medical Imaging, 2nd ed., p. 325.

65

Huda Ch6: Digital X-ray Imaging Question

 22. The matrix size in a DSA image is typically:

 A. 128 x 128
 B. 256 x 256
 C. 512 x 512
 D. 1024 x 1024
 E. 2048 x 2048

66

66

BME HCMUT 33
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Huda Ch6: Digital X-ray Imaging Question

 25. Changing the DSA matrix from 10242 to 20482 would

NOT increase the:

 A. Data digitization rate

 B. Data storage requirement
 C. Image processing time
 D. Pixel size
 E. Spatial resolution

67

67

Raphex 2003 Question: Digital Radiography

 D51. A flat panel digital radiographic detector has a

square 20 x 20 cm image receptor field. The full field of
the detector is coupled to a nominal 1024 x 1024 CCD
array. The relative spatial resolution (lp/mm) when going
from a 20 x 20 cm to a 10 x 10 cm field of view is:
 A. Four times better
 B. Twice as good
 C. The same
 D. Half as good
 E. One fourth as good

68

68

BME HCMUT 34
 Digital Radiography - Chapter 11 17-Feb-20
Adjuncts to Radiology – Chapter 12

Huda Ch6: Digital X-ray Imaging Question

 17. The Nyquist frequency for a 1K digital photospot

image (25 cm image intensifier diameter) is:

 A. 1 lp/mm
 B. 2 lp/mm
 C. 4 lp/mm
 D. 8 lp/mm
 E. 10 lp/mm

 FN (lp/mm) = 1/2a = 1/2(250mm/1024 lines) = 2.048 ≈ 2

69

69

BME HCMUT 35