Fluoroscopy
Robert Metzger, Ph.D.
Real-Time Imaging




Fluoroscopy is an imaging procedure that allows real-time x-ray viewing of
the patient with high temporal resolution
Use TV technology, which provides 30 frames per second imaging
Allows acquisition of a real-time digital sequence of images (digital video),
that can be played back as a movie loop
Cine cameras offer up to 120 frame per second acquisition rates using 35mm cine film. Digital cine also available
Fluoroscope Imaging Chain
c.f. Bushberg, et al.
The Essential Physics
of Medical Imaging, 2nd
ed., p. 232.
The Image Intensifier

There are 4 principal components
of an II:





(a) a vacuum bottle to keep the air
out
(b) an input layer that converts the
x-ray signal to electrons
(c) electronic lenses that focus the
electrons, and
(d) an output phosphor that
converts the accelerated electrons
into visible light
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 233.
The Image Intensifier
CCD TV
CAMERA
MIRROR
ADC
LENS
OUTPUT
PHOSPHOR
APERTURE
FOCUSING
ELECTRODES
DISPLAY
ELECTRONS
INPUT
PHOSPHOR
...CsI
PHOTO-CATHODE
LAYER
X-RAYS
The Input Screen

The input screen of the II consists of
4 different layers:


(a) vacuum window, a 1 mm
aluminum window that is part of the
vacuum bottle
 keeps the air out of the II, and its
curvature is designed to
withstand the force of the air
pressing against it
 a vacuum is necessary in all
devices in which electrons are
accelerated across open space
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 233.
The Input Screen

The input screen of the II consists of
4 different layers:


(b) support layer, which is strong
enough to support the input
phosphor and photocathode layers,
but thin enough to allow most x-rays
to pass through it
 0.5 mm of aluminum, is the first
component in the electronic lens
system, and its curvature is
designed for accurate electronic
focusing
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 233.
The Input Screen

The input screen of the II consists of
4 different layers:


(c) input phosphor, whose function is
to absorb the x-rays and convert
their energy into visible light
 cesium iodide (CsI) is used
 long, needle-like crystals which
function as light pipes,
channeling the visible light
toward the photochathode with
minimal lateral spreading
 400 mm tall, 5 mm in diameter
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 233.
The Input Screen

The input screen of the II consists of
4 different layers:


(d) photocathode is a thin layer of
antimony and alkali metals that
emits electrons when struck by
visible light
 10 to 20% conversion efficiency


23 to 35 cm diameter input image
(FOV)
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 233.
Input Phosphor Energy Conversion
Aluminum Support
Photocathode
CsI Needles
Figure courtesy from Jonathan Tucker, Brooke Army
Medical Center, SA, TX
Input Phosphor Energy Conversion
60 keV X-Ray
Aluminum Support
Photocathode
Figure courtesy from Jonathan Tucker, Brooke Army
Medical Center, SA, TX
CsI Needles
Input Phosphor Energy Conversion
Aluminum Support
3,000 light photons
 = ~ 420 nm
Photocathode
Figure courtesy from Jonathan Tucker, Brooke Army
Medical Center, SA, TX
CsI Needles
Input Phosphor Energy Conversion
Aluminum Support
Photocathode
~ 400 electrons
CsI Needles
To
Anode
Figure courtesy from Jonathan Tucker,
Brooke Army Medical Center, SA, TX
Electron Optics



Electrons are accelerated by an
electric field
Energy of each electron is
substantially increased and this
gives rise to electron gain
Focusing is achieved using an
electronic lens, which requires the
input screen to be a curved
surface, and this results in
unavoidable pincushion distortion
of the image

c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 235.
Electron Optics



The G1, G2, G3 electrodes along
with the input screen and the anode
near the output phosphor comprise
the five-component electronic lens
system of the II
The electrons under the influence of
the 25K to 35K V electric field, are
accelerated and arrive at the anode
with high velocity and considerable
kinetic energy
After penetrating the very thin
anode, the energetic electrons strike
the output phosphor
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 235.
The Output Phosphor




The output phosphor is made of zinc
cadmium sulfide
Anode is very thin coating of
aluminum on the vacuum side of the
output phosphor, which is electrically
conductive to carry away the
electrons once they deposit their
energy in the phosphor
Each electron causes the emission of
approximately 1000 light photons
from the output phosphor
2.5 cm diameter output phosphor
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 235.
The Output Phosphor

The reduction in image diameter
leads to amplification (analogy:
magnifying glass and sunlight)


Minification gain of an II is simply the
ratio of the area of the input phosphor
to that of the output phosphor, e.g.,
9’’ input phosphor, 1’ output
phosphor, area is square of the
diameter ratio, minification gain is 81
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 235.
The Output Phosphor



The output phosphor is coated right
onto the output window
Some fraction of the light emitted by
the output phosphor is reflected at the
glass window
Light bouncing around the output
window is called veiling glare, and can
reduce image contrast
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 235.
Image Intensifier Performance
Conversion Factor
Conversion Factor =
Light out of image intensifier (cd/m2)
Exposure rate into image intensifier (mR/sec)




Defined as a measure of the gain of an image intensifier
ratio of light output to exposure rate input
100 to 200 for new image intensifier
Degrades over time, ultimately can lead to II replacement
Image Intensifier Performance
Brightness Gain




BG = minification gain x electronic gain (flux gain)
Minification gain = increase in image brightness that results from
reduction in image size from the input phosphor to output phosphor size
(di/do)2, di is input diameter which varies, do is output diameter typically 2.5
cm
For 30 cm (12”) II, minification gain = 144
Image Intensifier Performance
Brightness Gain




BG = minification gain x electronic gain (flux gain)
Electronic gain or flux gain is typically 50
The brightness gain therefore ranges from about 2,500 – 7,000
As the effective diameter of the input phosphor decreases
(magnification increases), the brightness gain decreases
Field of View/Magnification




FOV specifies the size of the input phosphor of the image intensifier
Different sizes: 23 cm (9”), 30 cm (12”), 35 cm (14”), 40 cm (16”) FOV
Magnification is accomplished electronically using electronic focusing that projects
part of the input layer onto the output phosphor
Since brightness gain decreases in mag. mode, the x-ray exposure rate is
boosted. (12/9)2 = 1.8, (12/7)2 = 2.9



c.f. Bushberg, et al. The Essential
Physics of Medical Imaging, 2nd ed.,
p. 237.
NON-MAGNIFY MODE OF I.I.
OUTPUT
IMAGE
ALL OF INPUT SURFACE
USED TO GATHER X-RAYS
MAGNIFIED MODE OF IMAGE
INTENSIFIER
LESS IMAGED
ANATOMY IS
EXPANDED
OVER THE
SAME
OUTPUT
SURFACE
AND LOOKS
MAGNIFIED
OUTPUT
IMAGE
ONLY A PORTION OF
INPUT SURFACE USED
TO GATHER X-RAYS
LARGE NON-MAG FoV
e.g., 12 INCH
SMALL, MAG FoV
e.g., 6 INCH
Magnification
Pincushion Distortion
Optical Coupling

Parallel rays of light enter the
optical chamber, are focused
by lenses, and strike the video
camera where an electronic
image in produced

A partially silvered mirror is
used to shunt the light emitted
by the image intensifier to an
accessory port
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 239.
Video Camera
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 240.
Video Camera



Analog video systems typically have 30 frames/sec operation, but they work
in an interlaced fashion to reduce flicker, the perception of the image flashing
on and off
The human eye-brain system can detect temporal fluctuations slower than
about 47 images/sec, and therefore at 30 frames/sec flicker would be
perceptible
With interlaced systems, each frame is composed of two fields and each field
is refreshed at a rate of 60 times per second, which is fast enough to avoid
perception of flicker
Lag



Lag means that each new TV image actually contains residual image
information from the last several frames
Lag is good and bad
Lag acts to smooth the quantum noise in the image, but can also cause
motion blurring
Lag
Effect of camera lag.
Angiogram of a
rapidly moving
coronary artery
shows a trailing
"ghost" due to
excessive camera
lag (the direction of
travel is from right to
left).
Video Resolution


Spatial resolution of a video in the vertical direction (top to bottom) of the
TV image is governed by the number of scan lines
By convention, 525 lines are used in N. America for TV
 490 lines usable
 In the early days of TV, a man named Kell determined that about
70% of theoretical video resolution is appreciated visually, and this
psychophysical effect is now called the Kell factor
 490 x 0.7 = 343 lines or 172 line pairs useful for resolution
 For 9” field, resolution = 172 lp/229 mm = 0.75 lp/mm
 17 cm or 7” field, resolution is 1.0 lp/mm
 12 cm or 5” field, resolution is 1.4 lp/mm
Video Resolution






The horizontal resolution is determined by how fast the video electronics
can respond to changes in light intensity
This is influenced by the camera, the cable, the monitor but the
horizontal resolution is governed by the bandwidth of the system
The time necessary to scan each video line (525 lines at 30 frame/sec)
is 63 msec
11 msec required for horizontal retrace, 52 msec available
To achieve 172 cycles in 52 msec, the bandwidth required is 172
cycles/52 x 10-6 sec = 3.3 x 106 cycles/sec = 3.3 MHz
Higher bandwidths are required for high-line video systems
TELEVISION IMAGE
HORIZONTAL
DIRECTION
RASTER
LINE
VERTICAL
DIRECTION
TV LINES ARE
COMPOSED OF DOTS
INTERLACED SCANS
INTERLACED SCANS
TYPICAL MEASURED
RESOLUTION
[ 1023 LINE T.V. ]
FoV
9 INCH
6 INCH
4.5 INCH
T.V.
1.8-2.2 LP/mm
2.5-2.8 LP/mm
3.2-3.7 LP/mm
I.I. [CINE]
2.7-3.2 LP/mm
3.7-4.5 LP/mm
5.0-6.0 LP/mm
Summary











Fluoroscopy is a live imaging procedure
Image Intensifier main component and consists of the input phosphor,
electronic lens system and output phosphor
Input phosphor – Cesium Iodide, converts x-rays to light
Photocathode – converts light into electrons
Output phosphor – Zinc cadmium sulphide, converts electrons into light
Artifacts – pincushion distortion, veiling glare, lag
Brightness gain = minification gain x electronic (flux) gain
Several magnification modes available, typically exposure rate increases
with magnification
Video camera produces the electronic image which we see on the TV
monitor
Use interlaced scanning to avoid flicker
Horizontal (determined by bandwidth) and vertical (determined by the
number of scan lines) video resolution
Flat Panel Digital Fluoroscopy




Flat panel devices are thin film
transistor (TFT) arrays that are
rectangular in format and are used
as x-ray detectors
CsI, a scintillator is used to
convert the incident x-ray beam
into light
TFT systems have a photodiode
at each detector element which
converts light energy to an
electronic signal
Flat panel detectors would replace
the image intensifier, video
camera, and other peripheral
devices
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 242.
DIAGRAM OF GE FLAT
PANEL IMAGE
DETECTORS
FROM GE
FLAT PANEL-LIGHT SENSOR
Very
High Fill
Factor
Fill Factor= Sensitive Area
Pitch x Pitch
FROM GE
data line
Pitch
Pitch
FET
scan line
Peripheral Equipment


Photo-spot camera
 used to generate images on photographic film, 100-mm cut film or 105mm roll film
 full resolution of the II system, hardly seen nowadays
Digital photo-spot
 high resolution, slow-scan TV cameras in which the TV signal is digitized
and stored in computer memory
 Or CCD cameras with 10242 or 20482 pixel formats
 near-instantaneous viewing of the image on a video monitor
 allows the fluoroscopist to put together a number of images to
demonstrate the anatomy important to the diagnosis
 digital images can be printed on a laser imager
Peripheral Equipment


Spot-film devices
 attaches to the front of the II, and produces conventional radiographic
screen-film images
 better resolution than images produced by II
Cine-radiography cameras
 attaches to a port and can record a very rapid sequence of images on 35mm film
 used in cardiac studies, 30 frames/sec to 120 frames/sec or higher
 uses very short radiographic pulses
 digital cine are typically CCD-based cameras that produce a rapid
sequence of digital images instead of film sequence
Fluoroscopy Modes of
Operation


Continuous fluoroscopy
 continuously on x-ray beam, 0.5 – 4 mA or higher
 display at 30 frames/sec, 33 msec/frame acquisition time
 blurring present due to patient motion, acceptable
 10 R/min is the maximum legal limit
High dose rate fluoroscopy
 specially activated fluoroscopy
 20 R/min is the maximum legal limit
 audible signal required to sound
 used for obese patients
Fluoroscopy Modes of
Operation

Pulsed fluoro:
 series of short x-ray pulses, 30 pulses at ~10 msec per pulse
 exposure time is shorter, reduces blurring from patient motion
 Can be used where object motion is high, e.g., positioning catheters in
highly pulsatile vessels
 15 frames/sec, 7.5 frames/sec also available
 Variable frame pulsed fluoroscopy is instrumental in reducing dose
 Ex., initially guiding the catheter up from the femoral artery to the aortic
arch does not require high temporal resolution and 7.5 frames/sec could
potentially be used instead of 30 frames/sec
 7.5 frames/sec instead of 30 frames/sec, dose savings of (7.5/30) 25%
Frame Averaging






Fluoroscopy systems provide
excellent temporal resolution
However, fluoroscopy images are
relatively noisy, and in some
applications it is beneficial to
compromise temporal resolution for
lower noise images
This can be achieved by averaging
a series of images or frames
Real-time averaging in the
computer memory for display
Can cause noticeable image lag
but noise in image is reduced as
well
Could also reduce dose in some
circumstances
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 245.
Last Frame Hold

Last-frame hold
 when the fluoroscopist takes his or her foot off the fluoroscopy pedal,
rather than seeing a blank monitor, last-frame-hold enables the last live
image to be shown continuously
 useful at training institutions
 no unnecessary radiation used on patient
Road-Mapping

Road Mapping
 software-enhanced variant of the last-frame-hold feature
 side-by-side video monitors, one shows captured image, the other live
image
 In angiography, subtracted image can be overlayed over live image to
give the angiographer a vascular “road map” right on the fluoroscopy
image
 is useful for advancing catheters through tortuous vessels
Automatic Brightness
Control
The purpose of the automatic brightness control (ABC) is to keep the
brightness of the image constant at monitor



It does this by regulating the x-ray exposure rate (control kVp, mA or
both)
Automatic brightness control triggers with changing patient size and
field modes
Automatic Brightness
Control

The top curve increases mA more
rapidly than kV as a function of
patient thickness, and preserves
subject contrast at the expense of
higher dose

The bottom curve increases kV
more rapidly than mA with
increasing patient thickness, and
results in lower dose, but lower
contrast as well
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 247.
Image Quality
Spatial Resolution





A 2D image really has 3 dimensions: height, width, and gray scale
Height and width are spatial and have units such as millimeters
The classic notion of spatial resolution is the ability of an image system to
distinctly depict two objects as they become smaller and closer together
The closer together they are, with the image still showing them as separate
objects, the better the spatial resolution
At some point, the two objects become so close that they appear as one,
and at this point, spatial resolution is lost
Image Quality
Spatial Resolution


The spatial domain simply refers to the two spatial dimensions of an image, width
(x-dimension) and length (y-dimension)
Another useful way to express the resolution of an imaging system is to make
use of the spatial frequency domain



F (line pairs/mm or cycles/mm) =1/2, where  is the size of the object (mm)
Smaller objects (small ) correspond to higher spatial frequencies and larger
objects (large ) correspond to lower spatial frequencies


So, objects that are
 0.36 mm correspond to 1.4 lp/mm
 0.19 mm corresponds to 2.7 lp/mm
 1 mm correspond to 0.5 lp/mm
Image Quality: Spatial Resolution


Spatial frequency is just another way of thinking of object size
A device used to measure the spatial resolution is the bar pattern
c.f. Bushberg, et al. The Essential Physics of
Medical Imaging, 2nd ed., p. 249.
Image Quality
Spatial Resolution




The modulation transfer function, MTF
of an image system is a very complete
description of the resolution properties
of an imaging system
The MTF illustrates the fraction (or
percentage) of an object’s contrast that
is recorded by the imaging system, as a
function of the size (i.e., spatial
frequency) of the object
The limiting spatial resolution is the size
of the smallest object that an imaging
system can resolve
The limiting resolution of modern image
intensifiers is between 4 and 5
cycles/mm
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 248.
Image Quality
Contrast Resolution



The ability to detect a low-contrast object on an image is highly related to
how much noise (quantum noise and otherwise) there is in the image
The ability to visualize low-contrast objects is the essence of contrast
resolution. Better contrast resolution implies that more subtle objects can be
routinely seen on the image
The contrast resolution of fluoroscopy is low by comparison to radiography,
because the low exposure levels produce images with relatively low signalto-noise ratio (SNR)
Image Quality
Contrast Resolution


Contrast resolution is increased when higher exposure rates are used, but
the disadvantage is more radiation dose to the patient
Fluoroscopic systems with different dose settings allow the user flexibility
from patient to patient to adjust the compromise between contrast
resolution and patient exposure
Noise and Contrast
Comparison of x-ray
noise amplitudes in
coronary
angiograms
acquired at
fluoroscopic (2 µR
per frame) (a) and
angiographic (16 µR
per frame) (b)
exposure levels.
Noise and Contrast
16 µR per frame.
Note improved
resolution and
contrast due to the
higher exposure.
Digital Image Quality
Effect of Matrix Size.
512 x 512 matrix
Digital Image Quality
Effect of Matrix Size.
256 x 256 matrix
Digital Image Quality
Effect of Matrix Size.
128 x 128 matrix
Digital Image Quality
Effect of Matrix Size.
64 x 64 matrix
Digital Image Quality
Gray Levels at a
constant 512 x 512
matrix size.
256 Grey Levels (8
bit)
Digital Image Quality
Gray Levels at a
constant 512 x 512
matrix size.
4 Grey Levels (2
bits)
Digital Image Quality
Gray Levels at a
constant 512 x 512
matrix size.
8 Grey Levels (3
bits)
Image Quality
Temporal Resolution




Fluoroscopy has excellent temporal resolution, that is over time
Blurring in the time domain is typically called image lag
Lag implies that a fraction of the image data from one frame carries over
into the next frame
Video cameras such as the vidicon demonstrate a fair amount of lag
Image Quality
Temporal Resolution



Lag in general is undesirable, beneficial for DSA
Frame averaging improves contrast resolution at the expense of temporal
resolution
With DSA and digital cine, cameras with low-lag performance (plumbicons or
CCD cameras) are used to maintain temporal resolution
Fluoroscopy Suites

Gastrointestinal Suites
 R and F room, large table that
can be rotated from horizontal
to vertical to put the patient in
a head-down or head-up
position
 II above or under the table,
spot film device usually there


Remote Fluoroscopy Rooms
 Designed for remote operation
by the radiologist
 Tube above table, II under
table
 Reduce dose to the physician
and no lead apron needed
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 250.
Fluoroscopy Suites

Peripheral Angiography Suites
 Table floats, allows patient to be
moved from side to side and head to
toe
 C-arm or U-arm configuration
 30 to 40 cm image intensifier used
 Power injectors are normally ceilingor table-mounted


Cardiology Catheterization Suite
 Similar to angiography suite, 23 cm II
used to permit more tilt in cranial
caudal direction
 Cine cameras used, biplane rooms
common
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 250.
Fluoroscopy Suites

Biplane Angiographic Systems
 Two complete x-ray tube/II systems
used, PA and Lateral
 Simultaneous acquisition of 2 views
allows a reduction of the volume of
contrast media injected in patient


Portable Fluoroscopy- C Arms
 C-Arm devices with an x-ray tube
placed opposite from the II
 18-cm (7-inch) and 23-cm (9-inch)
and several other field sizes available
 Operating rooms and ICUs
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 250.
Radiation Dose

Patient Dose
 The maximum exposure rate permitted in the US is governed by the Code
of Federal Regulations (CFR), and is overseen by the Center for Devices
and Radiological Health (CDRH), a branch of the Food and Drug
Administration (FDA)


The maximum legal entrance exposure rate for normal fluoroscopy to the
patient is 10 R/min


For specially activated fluoroscopy, the maximum exposure rate allowable
is 20 R/min
Radiation Dose

Patient Dose


Typical entrance exposure rates for
fluoroscopic imaging are
 About 1 to 2 R/min for thin (10cm) body parts
 3 to 5 R/min for the average
patient
 8 to 10 R/min for the heavy
patient


Maximum dose at 120 kVp for most
vendors
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 252.
Dose to
Personnel

Rule of Thumb: standing 1 m from the
patient, the fluoroscopist receives from
scattered radiation (on the outside of
apron) approximately 1/1,000 of the
exposure incident upon the patient


The scatter field incident upon the
radiologist while performing a
fluoroscopic procedure is shown


A radiologist of average height, 178 cm
(5’10”) is shown overlaid on the graph
and key anatomic levels are indicated
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 253.
Dose to Personnel

The dose rate as a function of height
above the floor in the room is shown for 6
different distances D, representing the
distance between the edge of the patient
and the radiologist


80 kVp beam and 20 cm patient
thickness assumed for calculation


Roentgen-area product (RAP) or dosearea product (DAP) meters can be used
to provide real-time estimate of the
amount of radiation the patient has
received
c.f. Bushberg, et al. The Essential Physics of Medical
Imaging, 2nd ed., p. 253.
Additional Reading
Additional topics on digital fluoroscopy
and digital subtraction angiography can
be found at the RSNA Education Portal.
http://www.rsna.org/education/archive/a
apm/toc.html#fluoroscopy