Transcript Digital - Department of Radiology

Fundamentals
of Digital
Radiology
George David
Medical College of Georgia
So what is “Digital”?
What we mean by Digital
Filmless Department
 Digital Radiographs
 PACS
 Picture Archival & Communication
Systems
 Reading from Monitors
What we really mean by Digital
No more
File
Room!!!
Digital Image Formation
 Place mesh over image
Digital Image Formation
 Assign each square
(pixel) a number based
on density
 Numbers form the
digital image
194
73
22
Digital Image Formation
 The finer the mesh, the better the digital rendering
What is this?
12 X 9 Matrix
Same object, smaller squares
24 X 18 Matrix
Same object, smaller squares
48 X 36 Matrix
Same object, smaller squares
96 X 72 Matrix
Same object, smaller squares
192 X 144 Matrix
Numbers / Gray Shades
 Each number of a digital image corresponds to a
gray shade for one picture element or pixel
So what is a digital image?
 Image stored as 2D array of #’s representing
some image attribute such as
 optical density
 x-ray attenuation
 echo intensity
 magnetization
125
25
311
111
182
222
176
199
192
85
69
133
149
112
77
103
118
139
154
125
120
145
301
256
223
287
256
225
178
322
325
299
353
333
300
Computer Storage
125
25
311
111
182
222
176
199
192
85
69
133
149
112
77
103
118
139
154
125
120
145
301
256
223
287
256
225
178
322
325
299
353
333
300
125, 25, 311, 111, 182, 222,
176, 199, 192, 85, 69,
133, 149, 112, 77, 103,
118, 139, 154, 125, 120,
145, 301, 256, 223, 287,
256, 225, 178, 322, 325,
299, 353, 333, 300
Digital Copies
125, 25, 311, 111, 182, 222,
176, 199, 192, 85, 69,
133, 149, 112, 77, 103,
118, 139, 154, 125, 120,
145, 301, 256, 223, 287,
256, 225, 178, 322, 325,
299, 353, 333, 300
=
125, 25, 311, 111, 182, 222,
176, 199, 192, 85, 69,
133, 149, 112, 77, 103,
118, 139, 154, 125, 120,
145, 301, 256, 223, 287,
256, 225, 178, 322, 325,
299, 353, 333, 300
If you’ve got the same numbers ...
Digital Copies
then you have an identical copy
=
Digital Copies
 Digital copies are identical
 All digital images are originals
Image Matrix
 Doubling the matrix dimension quadruples
the # pixels
111
118
125
25
311
111
199
192
85
69
77
103
118
139
145
301
256
223
87
155
2 X 2 Matrix
4 pixels
4 X 4 Matrix
16 pixels
Image Matrix
Doubling the matrix dimension
quadruples # pixels
Matrix
# Pixels
512 X 512 => 262,144
1024 X1024 => 1,048,576
2048 X2048 => 4,194,304
 A 10242 matrix compared to
a 5122 matrix quadruples
 disk storage requirements
 image transmission time
 digital image manipulation
Matrix Size & Resolution
More pixels = better spatial resolution
The Bit
 Fundamental unit of
computer storage
 Only 2 allowable values
 0
 1
 Computers do all operations
with 0’s & 1’s
BUT
Computers group bits
together
Special Binary Digit Grouping
Terms
 Nibble
 4 binary bits (0101)
 Byte
 8 binary bits (1000 1011)
 Word
 16 binary bits (1100 0100 1100 0101)
 Double Word
 32 binary bits
(1110 0100 0000 1011 0101 0101 1110 0101)
Abbreviations Review
 Bit (binary digit)
 Smallest binary unit; has value 0 or 1 only
 Byte
 8 bits
 Kilobyte
 210 or 1024 bytes
 sometimes rounded to 1000 bytes
 Megabyte
 213 or 1,048,576 bytes or 1024 kilobytes
 sometimes rounded to 1,000,000 bytes or 1,000
kilobytes
# of unique values which can be
represented by 1 bit
2 unique combinations / values
1
2
# of unique values which can be
represented by 2 bits
1
2
4 unique combinations / values
3
4
# of unique values which can be
represented by 3 bits
1
5
2
6
3
7
4
8
8 unique combinations / values
Digital Image Bit Depth
 the number of computer bits (1’s or 0’s)
available to store each pixel value
Bits
1
2
3
.
.
.
8
Values
# Values
0, 1
21=2
00, 01, 10, 11
22=4
000, 001, 010, 011, 100, 101, 110, 111
23=8
.
.
.
.
.
.
00000000, 00000001, ... 11111111 2 8 = 256
Digital Image Bit Depth
 bit depth indicates # of possible brightness
levels for a pixel
 presentation of brightness levels
 pixel values assigned brightness levels
 brightness levels can be manipulated without
affecting image data


window
level
Bit Depth & Contrast Resolution
The more bits per pixel the more possible gray shades
and the better contrast resolution.
2 bit; 4 grade shades
8 bits; 256 grade shades
Computer Storage
 Storage = # Pixels X # Bytes/Pixel
 Example: 512 X 512 pixels;
1 Byte / Pixel
512 X 512 pixel array
# pixels = 512 X 512 = 262,144 pixels
Storage = 262,144 pixels X 1 byte / pixel = 262,144
bytes = 256 KBytes = .25 MBytes
Image Size
 Related to both matrix size & bit depth
 higher (finer) matrix requires more storage
 doubling matrix size quadruples image size
 higher bit depth requires more storage
 doubling bit depth theoretically doubles image size
 Computer may require storage in multiples of 8 bits (bytes)


10 or 12 bits stored in 16 bit slot
alters image size requirements
1 2 34 5 6 7 8
9 1 1 1
0 1 2
Image Compression
 reduction of digital image storage size by
application of algorithm
 for example, repetitive data could be represented by
data value and # repetitions rather than by repeating
value
37, 37, 37, 37,
37, 37, 37, 37,
37, 37, 37, 37,
37, 37, 37, 37,
37, 37, 37, 37
(20) 37’s
Image Compression
 Image Decompression
 calculating original digital image
from previously compressed data
 Compression Ratio
original image size
-------------------------------compressed image size
 ratio depends upon


data to be compressed
algorithm
Compression Types
 Reversible Compression
 Image decompresses to original pixel values
 Low compression ratios only
 Non-reversable Compression
 Decompressed image’s pixel values not necessarily
identical to original
 much higher compression ratios possible
 variation from original image may or may not be
visible or clinically significant
Non-Reversable Compression
 variation from original image generally increases
with increasing compression ratio
 but a higher compression ratio means less
storage requirements
 variation less noticeable for dynamic (moving)
images than for still images such as radiographs
Computed Radiography (CR)
 Re-usable metal imaging plates replace film
& cassette
 Uses conventional bucky & x-ray equipment
CR Exposure & Readout
CR Readout
Another View: CR Operation
Computer Radiography (CR)
 plate is
photostimulable
phosphor
 radiation traps
electrons in high
energy states
 higher states
form latent image
-
X-Ray
Photon
Higher Energy
Elect ron
State
Photon pumps
electron to
higher energy state
- - - Lower Energy - - Elect ron
- State
- - - -
Reading Imaging Plate
 reader scans plate with
laser
 laser releases
electrons trapped in
high energy
states
 electrons fall to low
energy states
 electrons give up
energy as visible light
 light intensity is
measure of incident
radiation
Laser Beam
Higher Energy
Elect ron
St at e
-
Lower Energy
Electron State
Lower Energy
Elect ron
St at e
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Reading Imaging Plate
 Reader scans plate
with laser light
using rotating
mirror
 Film pulled
through scanner
by rollers
 Light given off by
plate measured by
PM tube &
recorded by
computer
Laser & Emitted Light are Different Colors
 Phosphor stimulated by laser light
 Intensity of emitted light indicates amount of radiation
incident on phosphor at each location
 Only color of light emitted by phosphor measured by
PMT
CR Operation
 after read-out, plate erased using a bright light
 plate can be erased virtually without limit
 Plate life defined not by erasure cycles but by
physical wear
CR Resolution
 Small cassettes have better spatial
resolution
 Smaller pixels
 More pixels / mm
CR Throughput
 Generally slower than film
processing
 CR reader must finish reading
one plate before starting to
read the next
 Film processors can run films
back to back
CR Latitude
 Much greater latitude
than screen/film
 Plate responds to many
decades of input
exposure
 under / overexposures
unlikely
 Computer scale inputs
exposure to viewable
densities
 Unlike film, receptor separate
from viewer
Film Screen vs. CR Latitude
CR Latitude:
.01 – 100 mR
100
Digital Radiography (DR)
 Digital bucky
 Incorporated
into x-ray
equipment
Digital Radiography (DR)
 Receptor provides direct digital output
 No processor / reader required
 Images available in < 15 seconds
 Much less work for technologist
Direct vs. Indirect
TFT = THIN-FILM TRANSISTOR ARRAY
Digital Radiography (DR)
 Potentially lower patient dose than CR
 High latitude as for CR
 Digital bucky fragile
 First DR portables coming
to market
Raw Data Image
 Unprocessed image as read from
receptor
 CR

Intensity data from PMT’s as a result of scanning plate
with laser
 DR

Raw Data read directly from TFT array
 Not a readable diagnostic image
 Requires computer post-processing
 Specific software algorithms must be applied
to image prior to presenting it as finished
radiograph
Enhancing Raw Image (Image
Segmentation)
*
Identify collimated image border
Separate raw radiation from anatomy
Apply appropriate tone-scale to image
1.
2.
3.

Done with look-up table (LUT)
This process is
specific to a
particular body
part and
projection
 Computer must establish location
Imageof Segmentation
collimated border of image
• Computer then defines
anatomic region
• Finished image produced by
tone scaling
Requires histogram analysis of
anatomic region
Histogram
 Graph showing
how much of image
is exposed at
various levels
Tone Scaling
Post-Processing
 Body part & projection-specific algorithms determine
average exposure
 Must correctly identify anatomical region
 LUT computed to display image with proper
 Density
 Contrast
Film/Screen Limited Latitude
 Film use has
little ambiguity
about proper
radiation
exposure
Should I Worry?
In CR & DR,
image density is
no longer a
reliable indicator
of exposure factor
control.
CR / DR Latitude
DANGER
Will
Robinson!!!
 Almost impossible to under or
overexpose CR / DR
 Underexposures look noisy
 Overexposures look GOOD!!!
So how do I know if exposure is optimum by
looking at my image?
Exposure Index
 Each manufacturer provides feedback to technologist on
exposure to digital receptor
 Displayed on CR reader monitor
 Displayed on workstations
Calculated Exposure Index
Affected by
 X-Ray technique selection
 Improper centering of image on cassette
 Improper selection of study or projection
 Placing two or more views on same cassette
 Can cause image to appear dark
Phototimed Phantom Image
 75 kVp
 88 mAs
 2460 EI
Let’s Approximately Double mAs
 75 kVp
 88 mAs
 2460 EI
• 75 kVp
• 160 mAs
• 2680 EI
Let’s Go Crazy
 75 kVp
 88 mAs
 2460 EI
• 75 kVp
• 640 mAs
• 3300 EI
How Low Can You Go? Cut mAs in Half!
 75 kVp
 88 mAs
 2460 EI
• 75 kVp
• 40 mAs
• 2060 EI
Let’s Go Crazy Low
 75 kVp
 8 mAs
 1380 EI
• 75 kVp
• 1 mAs
• 550 EI
CR Artifacts
 Physical damage to imaging plates
 Cracks, scuffs, scratches
 Contamination
 Dust / dirt
 Dirt in reader
 Highly sensitive to scatter radiation
DR Artifacts
 Dead detector elements
 Spatial variations in background signal & gain
 Grid interference
 Software can help correct for above
Shifting Gears:
Fluoroscopy Issues
Digital Video Sources
 DR type image receptor
 Conventional Image Intensifier with Video
Signal Digitized (“Frame Grabber”)
I
Image
T m
u a
b g
e e
Tube
X-Ray
Input
TV
Lens System
Amplfier
Analog
to
Digital
Convert
er
Digital
Memory
(Computer)
Digital Spot Film
 Frame grabber digitizes image
 Digital image saved by computer
 Radiographic Technique used
 required to control quantum noise
Last Image Hold
 Computer displays last fluoro image
before radiation shut off.
 Image noisier than for digital spot
 Image made at fluoroscopic technique / intensity
 Allows operator to review static processes
without keeping beam on
 ideal for teaching environments
 ideal for orthopedic applications such as hip pinning
 Less radiation than digital spot
Fluoro Frame Averaging
 Conventional fluoro only displays current frame
 Frame averaging allows computer to average
current with user-selectable number of previous
frames
 Averages current frame & history
Fluoro Frame Averaging
Tradeoff
 Advantage:
 Reduces quantum noise
 Disadvantage
 Because history frames are averaged with current
frame, any motion can result in lag
Other Fluoro Features
 Real-time Edge Enhancement / Image Filtering
 Option of using lower frame rates (15, 7.5, 3.75
fps rather than 30)
 computer displays last frame until next one

reduces flicker
 Lowers patient and scatter exposure

Exposure proportional to frame rate
 dynamic studies may be jumpy
Digital Subtraction
 Immediate replay of run
 Free selection of mask
 before or after bolus
 >1 frame may be averaged for mask
 Note
 subtraction adds noise
Digital Image Manipulations
 on-screen measurements
 distances
 angles
 volumes/areas
 stenosis
 image annotation
 peak opacification / roadmapping
 peak opacification displays vessels after a test injection
 allows visualization of live catheter on top to saved image
of test injection
Digital Possibilities
 Multi-modality
imaging / Image
fusion
 PET/CT
DR & Energy Subtraction
 2 images taken milliseconds apart at 2
kVp’s
 Combine / subtract images
Soft Tissue Image
Bone Image
DR Mobile Units
 See image immediately
 Wireless transmission of
images
Other Possibilities
 Tomosynthesis
 Multi-slice linear tomography from one
exposure series
 Histogram Equalization
 Use computer to provide approximately
equal density to various areas of image.
The End
?