Chapter 09 X-ray detectors

X-ray detectors
CHAPTER 09
1 Si photodiodes
2 Si photodiode arrays
2-1
2-2
2-3
2-4
Structure
Features
Applications
New approaches
3 CCD area image sensors
3-1
3-2
3-3
3-4
Direct CCD area image sensors
CCD area image sensors with scintillator
How to use
Applications
4 CMOS area image sensors
4-1 Features and structure
4-2 How to use
4-3 Applications
5 Flat panel sensors
5-1
5-2
5-3
5-4
5-5
5-6
Features
Structure
Operating principle
Characteristics
How to use
Applications
1
X-ray detectors
X-rays were first discovered by Dr. W. Roentgen in Germany in 1895 and have currently been utilized in a wide range of
fields including physics, industry, and medical diagnosis. Detectors for X-ray applications span a broad range including a-Si
detectors, single crystal detectors, and compound detectors. There are many kinds of detectors made especially of Si single
crystals. For X-ray detectors, Hamamatsu offers Si photodiodes, Si APDs, CCD area image sensors, and CMOS area image
sensors, flat panel sensors, etc. Applications of our X-ray detectors include dental X-ray imaging and X-ray CT (computer
tomography) in medical equipment fields, as well as non-destructive inspection of luggage, foods, and industrial products;
physics experiments; and the like.
In the low energy X-ray region called the soft X-ray region from a few hundred eV to about 20 keV, direct detectors such as Si
PIN photodiodes, Si APDs, and CCD area image sensors are utilized. These detectors provide high detection efficiency and
high energy resolution, and so are used in X-ray analysis, X-ray astronomical observation, physics experiments, etc.
The hard X-ray region with energy higher than soft X-rays is utilized in industrial and medical equipment because of high
penetration efficiency through objects. Scintillator detectors are widely used in these applications. These detectors use
scintillators to convert X-rays into light and detect this light to detect X-rays indirectly. Especially in the medical field, the
digital X-ray method, which uses X-ray detectors with large photosensitive area, is becoming mainstream, replacing the
conventional film-based method. In non-destructive inspection, dual energy imaging, which allows image capturing with
deep tones by simultaneously detecting high- and low-energy X-rays, is becoming popular.
Example of detectable photon energy and spectral response range
Wavelength [nm] =
Si photodiode
for X-ray
(with scintillator)
1240
Photon energy [eV]
NMOS/CMOS
image sensor
Si photodiode for X-ray
(without scintillator)
Back-thinned CCD
Back-thinned CCD (windowless type)
X-ray imaging CCD/CMOS
image sensor (with FOS)
Front-illuminated CCD (windowless type)
Frontil uminated CCD
Flat panel sensor
1 MeV
100 keV
10 keV
1 keV
100 eV
10 eV
1 eV
0.1 eV
Photon
energy
Wavelength
0.01 nm
0.1 nm
1 nm
10 nm
100 nm
1 μm
10 μm
KMPDC0178EB
Hamamatsu X-ray detectors
Type
Si photodiode
Si photodiode array
CCD area image sensor
CMOS area image sensor
Image sensor
Flat panel sensor
Photodiode array with
amplifier
2
Features
combined with CsI(Tl) or ceramic scintillator are available.
•Products
•Back-illuminated CSP photodiodes that can be tiled (two-dimensional array) are available.
long, narrow image sensor can be configured by arranging multiple arrays in a row.
•ASupports
dual energy imaging
•
of FOS to FFT-CCD (CCD with scintillator)
•Coupling
•Front-illuminated CCD for direct X-ray detection are available.
•Coupling of FOS to CMOS image sensor
large-area two-dimensional imaging
•For
•Captures distortion-free, high-detail digital images in real time
Allows configuring a long, narrow image sensor by use of multiple arrays
•(See
chapter 5, “Image Sensors.”)
1.
[Figure 1-2] Examples of Si photodiodes combined
with scintillator
Si photodiodes
(a) Front-illuminated Si photodiode
When used for X-ray detection, Si photodiodes are typically
used with scintillators to form detectors for scintillator
coupling. Hamamatsu offers two types of Si photodiodes
for X-ray detection: Si photodiodes with scintillators and
Si photodiodes without scintillators (which assume that
users will bond the appropriate scintillators). In either
case, Si photodiodes have a spectral response matching the
emission band of scintillators.
In the case of Si photodiodes with scintillators, CsI(Tl)
scintillators or GOS ceramic scintillators are coupled with
the Si photodiodes. The area around the scintillator is
coated with a reflector to prevent the light emitted from
the scintillator from escaping outside the photosensitive
area [Figure 1-1].
Scintillator
X-rays
Protective resin
PN junction
Wire
Base
Disadvantage:
Patterns and wires may be damaged when the scintillator is mounted.
(b) Back-illuminated Si photodiode
X-rays
Scintillator
Bump
PN junction
[Figure 1-1] Si photodiode with scintillator
Advantages:
There is no wiring so mounting the scintillator is easy.
Multiple photodiodes can be tiled closely together.
KPDC0037EA
Scintillator
Si direct photodiodes
Reflector
Photosensitive area
Ceramic and the like
Epoxy resin
Si photodiode
KSPDC0003ED
Back-illuminated Si photodiodes have the PN junction on
the side opposite to (on the backside of ) the light incident
surface [Figure 1-2]. The photodiode surface bonded to the
scintillator is flat and does not have wires. This prevents
the photodiode from damage when the user attaches the
scintillator. In addition, the detector can be made small
because there is no area for wires as in a front-illuminated
type. Furthermore, multiple photodiodes can be arranged
with little dead space, so they can be used as a large-area
X-ray detector.
Because X-rays have no electric charge, they do not directly
create electron-hole pairs in a silicon crystal. However, the
interaction of silicon atoms with X-rays causes the release
from ground state of electrons whose energy equals that
lost by irradiated X-rays. The Coulomb interaction of these
electrons causes electron-hole pairs to be generated, and
these pairs are captured to detect X-rays. The probability
that X-rays will interact with silicon atoms is therefore a
critical factor when detecting X-rays directly.
Si direct photodiodes can effectively detect X-rays at
energy levels of 50 keV or less. Detection of X-rays less
than 50 keV is dominated by the photoelectric effect that
converts the X-ray energy into electron energy, so all
energy of X-ray particles can then be detected by capturing
the generated electrons with the Si photodiode.
Detection of X-rays and gamma-rays from 50 keV up
to 5 MeV is dominated by the Compton scattering, and
part of the X-ray and gamma-ray energy is transformed
into electron energy. In this case, the probability that the
attenuated X-rays and gamma-rays will further interact
with silicon (by photoelectric effect and Compton
scattering) also affects the detection probability, making
the phenomenon more complicated.
Figure 1-3 shows the probabilities (dotted lines) of
photoelectric effect and Compton scattering that may
occur in a silicon substrate that is 200 µm thick, and
the total interaction probabilities (solid lines) of silicon
substrates that are 200 µm, 300 µm, and 500 µm thick.
As can be seen from the figure, photodiodes created with
a thicker Si substrate provide higher detection probability.
With a 500 µm thick Si substrate, the detection probability
is nearly 100% at 10 keV, but falls to just a few percent at
3
100 keV. The approximate range of electrons inside a Si
direct photodiode is 1 µm at 10 keV and 60 µm at 100 keV.
Si photodiode arrays
2.
[Figure 1-3] Detection probabilities of Si direct photodiodes
(Theoretical value)
Interaction probability
Photoelectric effect
(200 μm)
Compton scattering
(200 μm)
Total interaction
probability
Si su
b
thick strate
n
500 ess
μm
300
μm
200
μm
X-ray energy (keV)
KSPDB0018EA
Baggage inspection equipment for examining the shapes
and materials of items in baggage are used in airports
and other facilities. Recently, high-accuracy CT baggage
inspection equipment are being developed. Hamamatsu
Si photodiode arrays with scintillators are widely used
in these types of baggage inspection equipment. X-rays
directed at baggage pass through objects and are
converted into light by a scintillator. Then, the converted
light is detected by the Si photodiode array. Hamamatsu
Si photodiode arrays for baggage inspection feature low
noise and consistent sensitivity and other characteristics
between individual elements. The photodiode chips are
mounted with high accuracy allowing highly accurate
detection. Moreover, their sensitivity range matches
the emission wavelength of scintillators making them
suitable for baggage inspection.
[Figure 2-1] Imaging example of baggage inspection
equipment
Structure
2-1
Many of the Hamamatsu Si photodiode arrays for baggage
inspection equipment employ back-illuminated structure.
Since back-illuminated Si photodiode arrays do not have
patterns or wires on the surface that scintillators are
bonded to, damage to patterns and wires when mounting
scintillators can be avoided. Figure 1-2 shows cross sections
for when a front-illuminated photodiode is combined with
a scintillator and for when a back-illuminated photodiode
[Table 2-1] Scintillator comparison table
Parameter
Condition
Peak emission wavelength
GOS ceramic
Unit
560
512
nm
10
7
-
2.2
µs
X-ray absorption coefficient
100 keV
Refractive index
At peak emission wavelength
1.74
1
3
100 ms after X-ray turn off
0.3
0.01
%
4.51
7.34
g/cm3
Transparent
Light yellow-green
-
±10
±5
%
Decay constant
Afterglow
Density
Color tone
Sensitivity variation
4
CsI(Tl)
is combined with a scintillator. Examples of scintillators
include CsI(Tl) and GOS ceramic. GOS ceramic features
small variations in light emission and high reliability. We
do not recommend CWO scintillators since they contain
cadmium which falls under environmental management
substances.
[Figure 2-2] Spectral response of Si photodiode arrays
and emission spectrum of scintillators
(a) S11212-121
(Typ. Ta=25 °C)
100
100
S11212-121
80
CsI(TI)
scintillator
light emission
characteristics
60
60
40
40
20
20
0
0
200
400
600
800
1000
Quantum efficiency (%)
Relative light output (%)
80
0
1200
•
Robustness
Through the adoption of a back-illuminated structure,
the photodiode array’s output terminals are connected to
the circuit board electrodes using bumps without wires.
Robustness is achieved by running the circuit wiring inside
the board.
•
High reliability
Since back-illuminated Si photodiode arrays do not have
patterns or wires on the surface that scintillators are mounted
on, scintillators can be mounted to the photodiode arrays
without damaging the patterns and wires. High reliability is
achieved since there are no wires, which could break due to
temperature changes or be adversely affected in other ways.
•
Superior sensitivity uniformity
In back-illuminated Si photodiode arrays (S11212/S11299
series), nonuniformity in sensitivity between elements are
minimized, and the sensitivity variations at the sensor’s end
elements are suppressed. The sensitivity uniformity has been
greatly improved as compared with the previous product
(S5668 series) and enables high-quality X-ray images to be
obtained.
Wavelength (nm)
KSPDB0330EB
[Figure 2-3] Sensitivity uniformity
(b) S11212-321
(Typ. Ta=25 °C)
110
(Typ. Ta=25 °C)
100
100
Back-illuminated Si photodiode array
S11212/S11299 series
80
GOS ceramic
scintillator light
emission
characteristics
60
60
40
40
20
20
Quantum efficiency (%)
Relative light output (%)
Relative sensitivity (%)
S11212-321
80
100
Previous product S5668 series
90
80
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0
0
200
400
600
800
1000
0
1200
Element
KPDB0023EB
Wavelength (nm)
KSPDB0331EB
2-2
•
Features
Low cost
•
Allows tiling
Back-illuminated Si photodiodes do not have space for
wires as shown in Figure 1-2 (b), so multiple photodiodes
can be tiled close together.
[Figure 2-4] Tiling example
The adoption of a back-illuminated structure simplifies
scintillator mounting and other processes, and this leads
to shorter manufacturing process.
Moreover, back-illuminated Si photodiode arrays use bumps
for their electrodes. Bumps are used in the manufacturing
process of LCD monitors and the like and are suitable for
high-volume production. The use of bumps has cut cost
when compared with our previous products.
5
2-3
[Figure 2-6] Multiple arrangement example (S11212-121)
Applications
Dual energy imaging
In normal X-ray non-destructive inspection, the X-ray
transmitted through an object is detected by a single type
of sensor, and the shape, density, and other characteristics
of the object is made into an image using shading. In
comparison, in dual energy imaging, high-energy image
and low-energy image are captured simultaneously
by two types of sensors, and the images are combined
through arithmetic processing. This enables images that
show detailed information about hard and soft objects
to be obtained. Dual energy imaging is used in a wide
range of fields such as security where specific chemicals,
explosives, and other dangerous objects are detected and
in the field of grain, fruit, meat, and other inspections.
Hamamatsu back-illuminated Si photodiode arrays
S11212/S11299 series support dual energy imaging. It is
structured so that two types of Si photodiode arrays with
scintillators can be combined to create top and bottom
layers in order to simultaneously detect high-energy and
low-energy X-rays. Moreover, its construction allows
multiple arrays to arranged in close proximity to form a
line sensor. This makes measurement of long and narrow
objects possible.
2-4
New approaches
Hamamatsu is currently developing a special ASIC that
can be combined with the proven Si photodiode array
for X-ray CT/baggage inspection. Hamamatsu ASICs are
compact and operate on low power. They can be made
into custom order products.
We are developing an X-ray CT module that combines
two ASICs and a 32 × 16 (512) element back-illuminated Si
photodiode array. Four of these modules arranged side by
side can be used in 128 slice X-ray CT scanners. Hamamatsu
modules with ASICs feature high X-ray durability. We can
also provide modules with heatsinks or GOS scintillators.
[Figure 2-5] Example of combining back-illuminated
Si photodiode arrays
[S11212-421 (top) and S11299-121 (bottom)]
[Figure 2-7] Dual energy imaging
X-ray source
Baggage under
inspection
Low-energy
X-rays
High-energy
X-rays
I Cross section
Low-energy X-rays
Phosphor screen
Conveyor
High-energy X-rays
Si photodiode array for
low-energy detection
Photosensitive
area
Scintillator
Si photodiode array
for low-energy detection
Photosensitive area
Si photodiode array for
high-energy detection
The top and bottom photodiode arrays detect low-energy X-rays and highenergy X-rays, respectively, to capture two types of images simultaneously.
Si photodiode array
for high-energy detection
KPDC0038EA
6
[Figure 2-8] Modules with ASICs
(a) With heatsink
(b) For 128 slice scanners (with GOS scintillators and heatsinks)
3.
CCD area image sensors
3-1
Direct CCD area image sensors
Windowless CCDs (front-illuminated type) are used for
directly detecting X-rays from 0.5 keV to 10 keV. These
CCDs cannot be used to detect X-rays whose energy is
lower than 0.5 keV since an absorption layer exists on the
CCD surface. A direct CCD (back-thinned type) must be
used to detect X-rays whose energy is lower than 0.5 keV.
To achieve high quantum efficiency in the energy region
higher than 10 keV, a direct CCD with a thick depletion
layer must be used.
Direct CCDs are capable of both X-ray imaging and
spectrophotometry. X-rays can also be detected in
photon-counting mode (method for counting individual
photons one by one). Direct CCDs are used in fields such
as X-ray astronomy, plasma analysis, and crystal analysis.
Principle of X-ray direct detection
Photons at an energy higher than a specified level generate
electron-hole pairs when they enter a CCD. If the photon
energy is small as in the case of visible light, only one
electron-hole pair is generated by one photon. In the
vacuum-UV-ray and soft-X-ray regions where photon
energy is greater than 5 eV, multiple electron-hole pairs are
generated by one photon. The average energy required for
silicon to produce one electron-hole pair is approx. 3.6 eV.
So an incident photon at 5.9 keV (Kα of manganese), for
example, generates 1620 electron-hole pairs in the CCD.
The number of electrons generated by direct X-ray detection
is proportional to the energy of the incident photons.
Characteristics
Figure 3-1 shows the result when X-rays (Mn-Kα/Kβ) emitted
from a Fe-55 radiation source are detected by a CCD.
Spectrum resolution is usually evaluated by using the FWHM
(full width at half maximum). The Fano limit (theoretical
limit of energy resolution) of Si detectors for Fe-55 is 109 eV.
Major factors that degrade energy resolution are CCD charge
transfer efficiency and CCD noise including dark current.
When a CCD is sufficiently cooled down and is operated at
a charge transfer inefficiency of 1 × 10-5 or less, the energy
resolution is determined by the readout noise. To improve
energy resolution, the CCD readout noise has to be less
than 5 e- rms. The energy resolution of optimally adjusted
Hamamatsu CCDs is below 140 eV for Fe-55.
There are two modes for evaluating the CCD quantum
efficiency in the X-ray region. One is the photon-counting
mode, and the other is the flux mode that integrates all
photons. The quantum efficiency in the visible region is
usually evaluated in the flux mode [Figure 3-2].
7
[Figure 3-1] CCD energy resolution in X-ray (Mn-Kα/Kβ)
detection (typical example)
2500
•
Mn-Kα
(5.9 keV)
2000
Number of counts
Features
High sensitivity and low noise are achieved by use of FFT
(full frame transfer) type CCD, which is widely used for
analysis and measurement.
1500
•
1000
500
5.5
6.0
6.5
Structure and characteristics
7.0
Energy (keV)
KMPDB0236EA
[Figure 3-2] Quantum efficiency vs. photon energy
(Typ. Td=-100 °C)
100
High-quality image type and low cost type available
The high-quality image type CCD uses a CsI(Tl) scintillator
to convert X-rays to visible light, and the low cost type CCD
uses a GOS scintillator.
Mn-Kβ
(6.5 keV)
0
5.0
Highly detailed images
•
CCD area image sensors with FOS
This CCD is coupled to an FOS which is an FOP with
scintillator. This CCD with FOS utilizes CsI(Tl) as the
scintillator to achieve high resolution.
Flux mode
Quantum efficiency (%)
[Figure 3-3] Structure of CCD with FOS
Scintillator CsI(Tl)
FOP
Photon-counting mode
Base
10
1
KMPDC0295EA
1
10
Photon energy (keV)
KMPDB0154EA
3-2
FOS
CCD chip
CCD area image sensors with
scintillator
Typically, CCD chips are damaged to some extent when
exposed to X-rays. However, this CCD with FOS has an FOP
on the CCD chip’s photosensitive area, and the FOP also
serves as an X-ray shield to suppress damage by X-rays.
Electric charges generated by X-rays incident near the
surface of the CCD may cause noise, where white spots are
seen at random positions. It degrades the image quality. This
CCD with FOS, however, maintains high-quality images
since the amount of X-rays incident on the CCD is small
due to the X-ray shielding effect of the FOP [Figure 3-4].
[Figure 3-4] X-ray transmittance in FOP
(X-ray tube voltage: 80 kV)
8
For panoramic/
cephalo imaging
Transmittance (%)
Besides visible, infrared, and ultraviolet light, a CCD can
directly detect and image X-rays below 10 keV. However, in
the X-ray region from several dozen to more than 100 keV
used for medical diagnosis and industrial non-destructive
inspection, scintillators are needed to convert the X-rays
into visible light. In this case, CsI(Tl) and GOS scintillators
are generally used, which convert X-rays into light at a
peak of around 550 nm. The CCD then detects this light
for X-ray detection.
In X-ray imaging applications requiring large-area detectors,
Hamamatsu provides front-illuminated CCD coupled to
an FOS (fiber optic plate with scintillator). We also respond
to requests for CCD coupled to an FOP (fiber optic plate)
(scintillator to be implemented by the user).
For intraoral imaging
FOP thickness (mm)
KMPDB0297EA
[Figure 3-5] CsI(Tl) absorption coefficient
[Figure 3-7] CsI(Tl) emission intensity
Light output (relative value)
Absorption coefficient (cm-1)
(70 kVp, 4.8 mm Al filter)
X-ray energy (keV)
Scintillator thickness (mg/cm2)
KMPDB0293EA
KMPDB0298EA
The resolution of a CCD with FOS is mainly determined
by the following factors:
· Pixel size
· Scintillator specifications (material, thickness)
· Gap between CCD chip and FOP (e.g., chip flatness)
Due to the CCD structure, the resolution determined by
the pixel size cannot be exceeded.
The thicker scintillator results in higher emission intensity, yet
the resolution deteriorates as the thickness increases (there is
a trade-off here between emission intensity and resolution)
[Figure 3-6, 3-7]. Since the resolution deteriorates as the gap
between the chip and FOP becomes wider, technology for
keeping this gap at a narrow width is essential. Note that the
FOP flatness is superior to the chip flatness and so poses no
problems.
[Figure 3-6] CsI(Tl) resolution
Contrast transfer function (%)
(10 Lp/mm, 40 kVp, no filter)
•
Buttable configuration
To obtain a long photosensitive area, panoramic imaging
CCDs use two chips and cephalo imaging CCDs use three
chips, with each chip being arranged in close proximity in
a buttable configuration. There is a dead space between
each chip. See Figure 3-8 for an example of an insensitive
area caused by this dead space.
3-3
How to use
There are two methods for capturing X-ray images: oneshot and TDI operation imaging.
For one-shot imaging, in the CCD pixels, charges are
constantly generated due to dark current, so those charges
must be constantly drained when no X-rays are being
input (standby state). When using TDI operation, the pixel
transfer speed has to be made to match the motion speed
of the object. (See chapter 5, “Image sensors.”)
Image correction
Scintillator thickness (mg/cm2)
KMPDB0299EA
CCDs may sometimes have pixel defects known as white
spots where the dark current is large, and black spots
where the output is low (low sensitivity). Scintillator and
FOP performance also affect the image quality of CCDs
with FOS. To achieve high image quality, we recommend
using software to compensate for the dark current and
sensitivity. See chapter 5, “Image sensors,” for information
on compensating for pixel defects, dark current, and
sensitivity.
Multiple CCD chips are combined in CCDs for panoramic/
cephalo imaging and non-destructive inspection, and there
is a dead space between each chip. Software compensation
may help suppress effects from this dead space.
9
[Figure 3-8] Example of CCD for cephalo imaging and non-destructive inspection (S8658-01)
X-ray detection side
220.0 (H) × 6.0 (V)
5.6
FOP
228.0 ± 0.3
FOP
3.0
15.24
60.96
B20
←
B14
*1
*2
Left chip
1.6
Center chip
Left chip
A1
→
A13
B1
45.72
→
30.48
Right chip
C1 →
B13
Center chip
Edge pixels
C20 ← C14
25.4
28.0 ± 0.3
A20 ← A14
60.96
Dead area 1:
250 μm min.
350 μm max.
C13
45.72
3.4
Scintillator
0.45
Enlarged view of portion *2
2.54
Center chip
±50 μm max.
FOP
7.2 ± 0.2
223.0 ± 0.5
±50 μm max.
Enlarged view of portion *1
Right chip
Edge pixels
Dead area 2:
130 μm min.
200 μm max.
KMPDA0251EA
Precautions
[Figure 3-9] Precautions when holding the sensor
(a) Hold the board by the edges with fingers.
Take the following precautions when using an X-ray CCD.
(1) Anti-static and surge measures
For measures to avoid electrostatic charge and surge voltage
on an X-ray CCD, refer to “1-3 How to use” in section 1 “CCD
area image sensors,” in chapter 5, “Image sensors.”
(2) Operating and storage environment
X-ray CCDs are not hermetically sealed, so avoid operating
or storing them in high humidity locations. Also do not
apply excessive vibrations or shock during transportation.
(b) Do not apply force to the FOS.
(3) Deterioration by X-ray irradiation
Like other X-ray detectors, X-ray CCD characteristics
deteriorate due to excessive X-ray irradiation. In some
applications, CCDs need to be replaced as a consumable
product.
(4) Handling CCD with FOS
· FOP is made from glass, so do not apply a strong force
and shock to it.
· Do not touch the scintillator section and photosensitive
area. A scratched scintillator will cause changes in
sensitivity.
Bonding wires are coated with protective resin, but do
not touch the resin as it can damage or break the wire.
· When holding the sensor, hold the board by the edges with
your fingers and make sure not to touch the exposed areas
of the leads and wires as shown in the photos [Figure 3-9
(a)]. Touching the exposed areas of the leads and wires
may damage the sensor due to static electricity.
· Never apply force to the FOS. It may damage the scintillator
[Figure 3-9 (b)].
10
(c) Do not bend the cable excessively.
(5) Handling the module with an assembled cable
· Do not apply excessive force to the sensor section. Biting
it, applying force to it, or dropping it may cause damage
or failure.
· Applying excessive force by bending or pulling on the
cable may cause breakdowns such as the cable breaking
internally, so please handle the cable with care [Figure
3-9 (c)].
3-4
[Figure 3-12] Example of cephalo imaging
Applications
Non-destructive inspection (for industry)
X-ray CCDs can be used to perform one-shot imaging
or perform imaging in TDI operation to inspect objects
moving on a conveyor and for other purposes.
[Figure 3-10] Example of printed circuit board imaging
Radiography
(1) Panoramic imaging
Panoramic X-ray imaging devices capture images by
using the X-ray source and detector unit designed to
rotate around the patient’s head. A TDI-CCD allows
capturing panoramic diagnostic images that are greater
than the photosensitive area.
[Figure 3-11] Example of panoramic imaging
(2) Cephalo imaging
Cephalo X-ray imaging devices capture images of
the head. These devices use TDI-CCDs for acquiring
diagnostic images like panoramic imaging.
11
CMOS area image sensors
4.
X-ray CMOS area image sensors are image sensors designed
for intraoral imaging and non-destructive inspection.
These image sensors make use of advantages offered by
active pixel type CMOS devices, including high integration,
sophisticated functions, and high S/N. They contain a
timing generator, vertical and horizontal shift registers,
readout amplifier, A/D converters, and LVDS [Figure 4-1].
The digital input and output make these image sensors
very easy to use. These image sensors contain a global
shutter function (integrates charges simultaneously in all
pixels) that allows acquiring one shot of an X-ray image in
synchronization with the X-ray irradiation timing.
Since these image sensors have an internal A/D converter,
analog video wiring can be kept short to reduce noise. The
internal A/D converter also simplifies the external circuit
and helps hold down the overall cost.
[Figure 4-1] Block diagram (typical example)
Timing
generator
Monitoring
photodiode amplifier
Low-speed
A/D converter
Column CDS
Image amplifier
High-speed
A/D converter
LVDS
Photosensitive area
CLK (P)
LVDS
Vertical shift register
Monitoring photodiode
Data (P)
determines that the X-ray emission has started and instructs
the CMOS area image sensor to start charge integration
and readout.
[Figure 4-2] Monitoring photodiode
Monitoring photodiode
(red section along circumference)
KMPDC0308EA
Low power consumption
X-ray CMOS area image sensors have an internal high-speed
A/D converter (14 bits) for image data, and a low-speed A/D
converter (10 bits) for monitoring photodiode data. The highspeed A/D converter which uses up much current starts
only when image data is transferred. Only the low-speed
A/D converter which consumes low power is on during
the long periods of standby for X-ray irradiation. This keeps
the average power consumption lower [Figure 4-3].
[Figure 4-3] Block diagram
CLK (N)
Data (N)
Monitoring
photodiode
charge
Photosensitive
area
Horizontal shift register
Pixel
charge
MCLK MST
Vdd GND
KMPDC0307EB
Photosensitive area
Monitoring
photodiode
amplifier
Low-speed
A/D converter
Monitoring
photodiode data
Image
amplifier
High-speed
A/D converter
Image data
Starts only when transferring image data
↓
Achieves low power consumption
KMPDC0309EB
4-1
Features and structure
Comparing CMOS and CCD
In regions where signal levels are low, CCD area image
sensors provide better image quality. CMOS area image
sensors use CMOS technology, which makes them
superior in terms of multifunctionality and easy to use.
CMOS area image sensors also offer the advantage of a
lower total cost because peripheral circuit functions can
be built into the CMOS chip.
Internal monitoring photodiode
In the X-ray CMOS area image sensors, a monitoring
photodiode for detecting the X-ray irradiation start timing
(trigger) is mounted as a narrow strip along the entire
circumference on the outer side of the photosensitive area
[Figure 4-2]. Monitoring photodiode signals are transmitted
repetitively at specific intervals. The output is sent to an
external control circuit. When this output exceeds the
specified threshold value, the external control circuit
12
Global shutter function
Hamamatsu X-ray CMOS area image sensors employ a global
shutter function. The global shutter enables integration of all
pixels simultaneously and therefore produces high-resolution
images even when X-rays are emitted during image data
readout or in machine-vision and other applications where
images of moving objects are captured, almost without any
of their adverse effects.
APS (active pixel sensor) type
Unlike the charge transfer type image sensors exemplified
by CCDs, X-Y address type CMOS area image sensors read
out integrated pixel charges and thus have long data line
wiring. This wiring capacitance becomes a large noise
source when the transistors in each pixel switch. As such,
Hamamatsu X-ray CMOS area image sensors employ an
APS type that houses an amplifier in the pixel. Since the
integrated charge is converted into voltage for each pixel,
low-noise images can be achieved.
[Figure 4-5] Wiener spectrum vs. spatial frequency
(typical example)
Detective quantum efficiency (DQE) is one of the parameters
that define the performance of an X-ray detector. It shows
the level of the output image signal S/N (SNROUT ) with
respect to the S/N (SNRIN) of the X-ray irradiated on the
X-ray detector. Since X-ray noise is closely related to the
radiation dose, the DQE can be used as a measure of the
photon detection efficiency of the X-ray detector.
The DQE can be used as a measure for evaluating the
incident X-ray photon detection efficiency and image
quality. Higher DQE indicates higher efficiency in
obtaining high-quality image from the incident X-rays.
The DQE is given by equation (1) or (2).
DQE =
(Ta=25 °C)
10-4
Wiener spectrum (mm2)
High detective quantum efficiency
10-5
10-6
10-7
10-8
0
(SNROUT)2 .......
(1)
(SNRIN)2
10
15
20
25
Spatial frequency (cycles/mm)
KMPDB0387EA
[Figure 4-6] DQE vs. spatial frequency (typical example)
MTF2
DQE =
5
....... (2)
ϕWS
(Ta=25 °C)
1.0
MTF : modulation transfer function
ϕ : number of incident X-ray photons
WS : Wiener spectrum
0.8
[Figure 4-4] MTF vs. spatial frequency (S11684, typical example)
(Ta=25 °C)
1.0
0.8
MTF
0.6
0.4
0.6
DQE
In an ideal imaging system without noise, the DQE is
equal to 1. In a real imaging system, noise introduced
in various processes such as noise generated by the pixels
and electronic circuit increases the Wiener spectrum, and
especially in the high-frequency region where the effect is
great, the DQE decreases. Hamamatsu X-ray CMOS area
image sensors use high-emission-efficiency and highresolution CsI(Tl) for the scintillator to achieve higher
DQE, which produces high-quality image and lower X-ray
radiation dose [Figure 4-4, 4-5, 4-6].
For information on MTFs, see “1-2 Characteristics and
Resolution” in section 1 “CCD image sensors,” in chapter
5, “Image sensors.”
0.4
0.2
0
0
5
10
15
20
25
Spatial frequency (cycles/mm)
KMPDB0388EA
4-2
How to use
Since X-ray CMOS area image sensors have an internal
timing generator, it is possible to monitor X-ray emission
timing and also integrate and read out image data just
by applying a master start pulse (MST) and master clock
pulse (MCLK). Data from the monitoring photodiode and
image data are switched by an internal switch so that they
are transmitted from the same output wiring [Data(P),
Data(N)] [Figure 4-7].
For information on image correction, see chapter 3, “CCD
area image sensors.”
0.2
[Figure 4-7] Input/output wiring diagram
0
0
5
10
15
20
25
Timing
generator
High-speed
A/D converter
Spatial frequency (cycles/mm)
KMPDB0386EA
Cable
Low-speed
A/D converter
Chip
CLK(P)
CLK(N)
Data(P)
Data(N)
MCLK
MST
Vdd
GND
Base
KMPDC0310EB
13
Precautions
The precautions on using X-ray CMOS area image sensors
are the same as those for X-ray CCDs (see section 3-3,
“Precautions,” in chapter 3, “CCD area image sensors”).
4-3
Applications
Intraoral imaging
Detailed diagnostic images of two to three teeth can be
obtained by inserting a CMOS area image sensor for
intraoral imaging and non-destructive inspection into the
patient’s mouth.
For intraoral imaging, Hamamatsu provides CMOS modules
that use a relatively large area CMOS area image sensor
of 1000 (H) × 1500 (V) pixels or 1300 (H) × 1700 (V) pixels,
both with a pixel size of 20 × 20 µm, assembled with a cable
[Figure 4-8]. The scintillator uses CsI(Tl) that achieves a high
resolution of 15 to 20 line pairs/mm.
Coupling the FOS to the CMOS gives high durability against
X-ray exposure. For example, these modules can operate
up to 100000 times or more under X-ray irradiation of
approximately 250 µGy at 60 kVp. Besides this feature, the
sensor unit of these CMOS modules is thin and compact,
allowing X-ray imaging even in a narrow section.
5.
Flat panel sensors
Flat panel sensors are X-ray imaging modules that use a
large-area CMOS image sensor combined with a scintillator.
The detector (two-dimensional photodiode array), highperformance charge amplifier, and scanning circuit are
all integrated onto a large-area CMOS monolithic singlecrystal silicon chip. The A/D converters, memories, interface
circuit, and the control signal generator that controls these
components are assembled into a module. There is no need
to use an external circuit to operate the device. The flat
panel sensor can capture megapixel-class, high definition
digital images which are distortion-free in both still and
moving images. The thin profile and light weight make the
flat panel sensor easy to install into other equipment. Flat
panel sensors are now widely used in various types of X-ray
imaging systems including CT.
We also offer flat panel sensors that use advanced
a-silicon, which features large photosensitive area and
high-speed response, for the detector.
[Figure 5-1] Flat panel sensors
[Figure 4-8] CMOS module
[Figure 5-2] Imaging examples
(a) Hornet
14
(b) Fish
•
High sensitivity
•
High resolution
This structure lowers the noise level, which is about one
figure less than the current mode passive pixel type.
Because of its low noise and high S/N features, the active
pixel type flat panel sensor acquires high definition images
from low energy X-rays.
•
High frame rate
[Figure 5-3] Internal circuits of CMOS chip
•
Wide dynamic range
•
Distortion-free images
•
Directly deposited CsI(Tl) scintillator type is available.
(a) Current mode passive pixel type
Address switch
∙∙∙
U
∙∙∙
U
Photodiode
Structure
∙∙∙
U
∙∙∙
U
∙∙∙
Figure 5-3 shows the internal circuit of a CMOS chip for
flat panel sensors. Two-dimensional X-ray image signals
converted into fluorescence by a scintillator are accumulated
as an electric charge in the junction capacitance of each
photodiode with excellent linearity. The accumulated
charges are then output one row at a time through the data
line by the vertical shift register when the address switch
turns on.
Since the flat panel sensor operates in charge integration
mode, the output video signal voltage V(t) is expressed by
equation (1).
∙∙∙
5-2
Vertical shift register
Features
Data line
5-1
Charge amp array
∙∙∙
∙∙∙
Horizontal shift register
Horizontal shift register
Video signal
Video signal
Multiport readout
Video signal
U: unti-over flow circuit
KACCC0224EB
V(t) = G × Q(t) = G × I(t) × t1 = G × I(t) × 1/Sf .....(1)
(b) Active pixel type
G : amplifier gain
Q(t): integrated charge
I(t) : photodiode photocurrent
t1 : integration time
Sf : frame rate
∙∙∙
Photodiode
Vertical shift register
When a constant X-ray radiation dose is striking an object,
the photocurrent generated in a photodiode is constant.
The output voltage can be increased by slowing the frame
rate (making the integration time longer). The frame rate
can be controlled by the external trigger mode described
later on. The saturation charge is determined by the
photodiode junction capacitance. The maximum video
output value after A/D conversion is set to the saturation
charge value.
Address switch
Amplifier
Data line
∙∙∙
Horizontal shift register
Video signal
KACCC0308EA
Amplifier circuit
There are two types of amplifier circuits for flat panel sensors:
a current mode passive pixel type and an active pixel type.
The current mode passive pixel type has an amplifier for
each column of the photodiode array, where the amplifier
is connected to each photodiode via address switch. The
amplifiers are formed on one side of the two-dimensional
photosensitive area as an amplifier array. The current mode
passive pixel type allows a high fill factor and high radiation
durability. However, the input capacitance caused by the
data line limits the reduction in amplifier thermal noise.
The active pixel type structure eliminates the foregoing
problem in the current mode passive pixel type. The
active pixel type has an amplifier for each pixel, and the
accumulated charges are converted into voltage there.
Scintillator
Flat panel sensors employ an indirect X-ray detection
method that converts X-rays into light using a scintillator
and then detects that light. By optimizing the wafer
process technology, Hamamatsu has succeeded in
developing a high-sensitivity photodiode that matches
the spectral characteristics of the scintillator.
The CsI(Tl) scintillator [Figure 5-4] used for most flat
panel sensors has needle-like crystals through which
scintillation light propagates, and flat panel sensors with
CsI(Tl) scintillator therefore have superior resolution and
emission intensity compared to flat panel sensors using
other scintillators composed of grain (particle) crystals
(such as GOS).
15
[Figure 5-4] Cross-sectional photo of CsI(Tl) scintillator
showing needle-like crystals
(b) Direct deposition type
X-rays
Directly deposited
needle-like structure CsI(Tl)
Photodiode
CMOS chip
KACCC0271EB
Tiling
[Figure 5-5] Cross-sectional photo of GOS scintillator
showing grain crystals
Some flat panel sensors have a large-format photosensitive
area achieved by using the “tiling” technique. Highly
precise tiling techniques now allow acquiring an X-ray
image which contains no gaps at the tiling edges. Though
the sensitivity of the pixels at the tiling edges decreases, it
is still high enough to be corrected by software, so seamless
images can be obtained.
[Figure 5-7] Cross-sectional diagram of tiling edges of
photosensitive area
1.9 to 2.3 mm
CsI(Tl) scintillator
Photodiode array B
CsI(Tl) scintillator
There are two scintillator-to-photodiode coupling methods.
One method uses an FSP (flipped scintillator plate), which
is a glass plate on which the scintillator is deposited, and
the scintillator side of the FSP is attached in close contact
with the photodiode. The other method is direct deposition
of scintillator onto the photodiode. The method using a
CsI(Tl) FSP has better fluorescent intensity and resolution
than medical screens utilizing GOS. The direct deposition
method further improves the resolution because it
suppresses fluorescence scattering compared to the FSP
method.
Hamamatsu provides both FSP type and direct deposition
type flat panel sensors that can be selected according to
the application.
[Figure 5-6] Cross-sectional structures of chip
(a) FSP type
Tiling position accuracy: 0 ± 60 μm
Photodiode array A
KACCC0386EB
5-3
Operating principle
Signal readout method
The following methods are generally used to read out
digital signals.
(1) Serial drive method
This method reads out video data by serially driving all
pixels, so the frame rate slows down when there are a
large number of pixels.
(2) Parallel drive method
X-rays
Glass substrate
Needle-like
structure CsI(Tl)
•
Air gap
Photodiode
CMOS chip
KACCC0270EB
16
Single port readout method
This method divides the monolithic photosensitive area
into multiple blocks, and reads out video data through a
single port by driving each block in parallel.
Figure 5-8 shows a schematic of an photosensitive area
divided into “n” blocks. Since flat panel sensors have
many pixels numbering more than one million, the serial
drive method causes the frame rate to drop. The single
port readout method, however, offers high speed and
easy processing of video data and so is used for most flat
panel sensors.
[Figure 5-9] Example of spectral response and CsI(Tl)
scintillator emission spectrum
This method reads out video data through multiple ports
to achieve even higher speed drive than the single port
readout method. Providing multiple ports for video data
readout can increase the image data transfer amount
per unit time, which is larger than that of the single port
readout method. Some flat panel sensors use this method.
[Figure 5-8] Schematic of parallel drive method
a2
a3
b1
b2
b3
n1
n2
n3
Shift register Shift register
Shift register
A/D converter A/D converter
A/D converter
FIFO memory FIFO memory
FIFO memory
100
Spectral
response
Vertical shift register
a1
(Typ. Ta=25 °C)
50
40
80
30
60
20
CsI(Tl)
scintillator
emission spectrum 40
10
20
0
400
500
600
700
Relative emission output (%)
Multiport readout method
Quantum efficiency (%)
•
0
800
Wavelength (nm)
KACCB0062EB
Linearity
Digital signal
KACCC0225EB
Flat panel sensors exhibit excellent linearity versus the
incident X-ray levels. Figure 5-10 shows the output linearity
of a flat panel sensor (14-bit output). The upper limit of the
14-bit output is 16383 gray levels.
[Figure 5-10] Output linearity (14-bit output, typical example)
Flat panel sensors support the following video output
interfaces: RS-422, LVDS, USB 2.0, and Gig-E. USB 2.0
and Gig-E support our digital camera interface DCAM.
Binning mode
Some flat panel sensor models have a binning mode function
that simultaneously reads out multiple pixel data. Up to 4 ×
4 pixels can be selected for binning though this depends on
the sensor models. Increasing the number of binning pixels
also increases the sensor frame rate. Note that the highest
resolution is obtained by single operation (1 × 1 mode)
without using binning mode.
Video output (gray level)
Video output interface
Irradiation dose (relative value)
KACCB0063EB
Dark video output
5-4
Characteristics
Spectral response
The photosensitive area of flat panel sensors consists of
a two-dimensional photodiode array. Figure 5-9 shows the
spectral response of a typical flat panel sensor and the
emission spectrum of a CsI(Tl) scintillator. To achieve high
sensitivity, the photodiode array is designed to have high
sensitivity in the vicinity of the peak emission wavelength
of CsI(Tl).
The X-ray energy range at which flat panel sensors are
sensitive differs depending on the sensor model. Refer to
their datasheets for details.
When the integration time is set longer, dark video output
slightly increases due to the photodiode dark current.
Figure 5-11 shows the relationship between dark video
output and integration time for a flat panel sensor (14-bit
output). The photodiode dark current (ID) is expressed by
equation (2).
ID = K/G [C/s] ..........(2)
K : increasing rate [gray levels/s] of dark video output versus integration time
G : conversion gain
17
Dark video output (gray level)
[Figure 5-11] Dark video output vs. integration time
(14-bit output, typical example)
( )
V2tot(rms) =
1 Ct
8
kT
gm Cf
3
Vtot(rms) =
Ct
Cf
2
β1 +
( )
Kf
Ct
Cox 2 W L Cf
2
β2 ....(3)
1
Kf
8
kT
β1 +
β2 .......(4)
gm
Cox2 W L
3
Ct = Cp + Cf + Cd ........................................................(5)
Vtot :
T
:
gm
:
β1, β2 :
Kf
:
Cox :
W
:
L
:
Integration time (s)
KACCB0064EA
A slight drift occurs in the dark video output after the power
is turned on. Figure 5-12 shows examples from measuring
this dark video output drift. In internal trigger mode (2
frames/s), the dark video output shows little change after
the power is turned on. At a slow frame rate, however, the
dark video output shifts. In applications where fluctuations
in the dark video output cause problems, determine how
often dark images should be acquired for correction to meet
the allowable drift level range.
Dark video output (gray level)
[Figure 5-12] Drift characteristics of dark video output
(14-bit output, typical example)
total noise voltage
absolute temperature [K]
transconductance of charge amplifier first-stage transistor
constants determined by charge amplifier
1/f noise constant of charge amplifier first-stage transistor
gate oxide film capacitance of charge amplifier first-stage transistor
W length of charge amplifier first-stage transistor
L length of charge amplifier first-stage transistor
The noise level of current mode passive pixel type CMOS
image sensors depends on the pixel size and the number
of pixels.
The lower limit of flat panel sensor dynamic range is
determined by noise and the upper limit by the saturation
charge. This means that the dynamic range is derived
from the ratio of saturation charge to noise.
In the active pixel type, the video line parasitic capacitance
is extremely low, so the noise is small.
Resolution
Resolution is a degree of detail to which image sensors can
reproduce an input pattern in the output. The photosensitive
area of a flat panel sensor consists of a number of regularly
arrayed photodiodes, so the input pattern is output while
being separated into pixels. Therefore as shown in Figure
5-13, when a square wave pattern of alternating black and
white lines with different intervals is input, the difference
between black and white level outputs becomes smaller as
the pulse width of the input pattern becomes narrower. In
such a case, the contrast transfer function (CTF) is given by
equation (6).
VWO - VBO
× 100 [%].......... (6)
VW - VB
CTF =
VWO:
VBO :
VW :
VB :
output
output
output
output
white level
black level
white level (when input pattern pulse width is wide)
black level (when input pattern pulse width is wide)
[Figure 5-13] Contrast transfer function characteristics
Elapsed time after power-on (min)
1 line pair
KACCB0065EA
Noise and dynamic range
Input pattern
White
Flat panel sensors were developed based on CMOS image
sensors. CMOS image sensors transfer charges accumulated
in the photodiodes to the readout circuit through the video
line.
In the current mode passive pixel type CMOS image sensors,
noise is expressed by equation (4). The video line parasitic
capacitance (Cd) is very large compared to the photodiode
junction capacitance (Cp) and charge amplifier feedback
capacitance (Cf ), so the video line parasitic capacitance
becomes a dominant source of noise.
18
Output
Black
Pixel pitch
KMPDC0070EA
The fineness of the black and white lines on the input pattern
is given by the spatial frequency of the input pattern. The
spatial frequency is the number of black and white line
pairs per unit length. In Figure 5-13, the spatial frequency
corresponds to the reciprocal of the distance from one
white edge to the next white edge in the pattern. It is usually
represented in units of line pairs/mm. The finer the input
pattern or the higher the spatial frequency, the lower the
CTF will be.
[Figure 5-14] Contrast transfer function vs. spatial
frequency [pixel size: 50 × 50 µm,
CsI(Tl) direct deposition, typical example]
X-ray irradiation damage
For example, on the C7942CA-22, if 80 kVp of X-rays are
irradiated over 4 hours a day (1 × 1 mode, frame rate: 2
frames/s), the detector life is 152 days.
Contrast transfer function (%)
5-5
How to use
Connection method
Spatial frequency (line pairs/mm)
KACCB0193EA
The resolution and sensitivity of flat panel sensors to X-rays
depend on the scintillator thickness. Both are in a tradeoff
relation. Our flat panel sensors are designed for optimal
scintillator thickness by taking the application and pixel size
into account to deliver high resolution and high sensitivity.
Setup is simple. All that is needed is to connect the flat
panel sensor to a PC and power supply using the data cable
and power cable (some models require an external trigger
input cable). Then supplying the voltage to the flat panel
sensor will start real-time X-ray image acquisition from
the PC control. Figure 5-15 shows a connection example
of an X-ray imaging system using a flat panel sensor. Use
a monotonically increasing series power supply with a
transformer for the voltage source.
[Figure 5-15] Connection example (C10500D-03)
OS +
Image acquisition software
Video output
(14-bit digital output) Frame
grabber board
Vsync, Hsync,
Pclk
PC/AT compatible
computer (rear)
Reliability
In ordinary X-ray detectors, deterioration in performance
such as a drop in sensitivity and an increase in dark video
output occurs due to X-ray irradiation. Likewise, flat panel
sensor characteristics deteriorate due to X-ray irradiation.
For example, an FSP type flat panel sensor with an aluminum
top cover intended for non-destructive inspection is designed
for use at an X-ray energy from 20 kVp to 100 kVp, and
can be used up to an accumulated irradiation dose of one
million roentgens if used under 100 kVp X-ray energy.
When the photosensitive area is uniformly irradiated with
X-rays, the dark current also increases almost uniformly
over the photosensitive area. The dark current might
partially increase in the photosensitive area, but this can
be eliminated by dark image correction. When the partial
increase in dark video output caused by increased dark
current has exceeded the dark image correction limit, the
flat panel sensor should be replaced as a consumable part.
The life of flat panel sensors can be extended by setting the
X-ray dose to a lower level within the detectable range and
by preventing X-rays from irradiating the flat panel sensor
except during imaging. Another effective way to extend the
detector life is to use pulsed X-rays.
Monitor
IntExtGrb
ExtTrgGrb
IntExtIO
ExtTrgIO
C10500D-03
X-ray source
Voltage source
(A.vdd, D.vdd)
Object
KACCC0642EA
Trigger mode
Flat panel sensors have two trigger modes (internal trigger
mode and external trigger mode).
In internal trigger mode, the sensor always operates at
the maximum frame rate and constantly outputs the sync
signals and video signal.
To capture images in external trigger mode, apply external
trigger pulses as shown in Figure 5-16. Vsync+, Hsync+,
and video signal are output after time Tvd elapses from
the rising edge of the external trigger pulse.
To synchronize with the pulse X-ray source, apply X-rays
during Txray.
19
[Figure 5-16] External trigger mode
Frame time (Tvc to Tmax)
50% of frame time recommended
ExtTrgLemo
ExtTrgGrb
(TTL)
Tvd
Tvc - Tvdpw
Txray
Tvl
Vsync+
(RS-422)
Hsync+, Pclk, and effective video output are the same as those for internal trigger mode.
· Tmax is defined using the reciprocal of the minimum value of Sf(ext).
· Txray = Frame time - Tvd - (Tvc - Tvdpw)
· Tvl = Frame time - (Tvc - Tvdpw)
KACCC0341ED
Defect lines
Charges accumulated in the photodiodes are transferred
to the readout circuit through the data line by turning on
the CMOS switch for each pixel using the gate line from
the shift register. An open-circuit fault occurring in the gate
line or data line will make it impossible to read out some
pixels. These continuous pixels are called the defect line.
Although defect lines are inevitable in image sensors with
a large photosensitive area, correcting them by software
based on values of the surrounding pixels makes it possible
to eventually acquire images with no defects.
Charges leaking out of a defect line might increase the output
of the pixels adjacent to the defect line. This phenomenon
can also be corrected by software.
If using a flat panel sensor with a pulsed X-ray source,
then setting the flat panel sensor to external trigger mode
will be convenient. In external trigger mode, inputting an
external trigger signal to the flat panel sensor allows reading
out the charges that have been kept accumulated in the
photodiodes up until then. The charges are in this case
continually accumulated until an external trigger signal
is input. To acquire an image in synchronization with the
pulsed X-ray source, the X-ray source must emit X-rays at
the appropriate trigger intervals.
Figure 5-17 shows a timing chart for acquiring images with
pulsed X-rays using an external trigger signal.
Here, an external trigger signal is input prior to pulsed
X-ray emission, and starts readout of charges integrated
in the photodiodes up until that time (). Readout of the
integrated charges ends after Tprdy from the rising edge
of the external trigger signal (), and the photodiodes
are reset. Refer to the datasheet for information on other
parameters.
Tprdy = Tvd + Tvc - Tvdpw ......... (7)
Tvdpw: period during Vsync+ is at low level in internal trigger mode
Pulsed X-rays are emitted in the period between  and
‘ (rising edge of the next external trigger signal) on the
timing chart. The next external trigger signal is input after
the X-ray emission. The operation of  to ‘ then repeats.
[Figure 5-17] Timing chart
Frame time = Integration time (Tvc to 10 s)
Image correction
ExtTrg(TTL)
Flat panel sensors utilizing the latest CMOS process
technology and CDS circuits can acquire images with very
high uniformity, yet they also offer an even higher level of
image quality by software correction.
Vsync+
Tvd Tvc - Tvdpw
Thd + Tdd
Data1-12
(14 bits)
Tprdy
X-ray emission #1
X-ray emission
X-ray image by X-ray emission #1
X-ray image
Precautions
Flat panel sensors deteriorate due to X-ray irradiation.
After long term use or after use under large radiation doses,
the sensor sensitivity decreases and the dark video output
increases. Coping with this deterioration requires correcting
the image by software to meet the desired detection accuracy,
as well as periodically replacing the flat panel sensor as a
consumable part.
5-6
Applications
X-ray imaging using pulsed X-ray source
In most X-ray imaging using a continuous X-ray source,
there is no need to synchronize the detector with the X-ray
source during use. However, in general, when using a pulsed
X-ray source that emits a high radiation dose in a short
time compared to continuous X-ray sources, the detector
must be synchronized with the emission timing of the X-ray
source to acquire an image.
20
KACCC0460EA
Acquiring enlarged images of small objects
Flat panel sensors can acquire an enlarged image since they
capture images with no distortion and have high resolution.
The image magnification is expressed by equation (8).
Magnification =
D1 ............
(8)
D2
D1: distance between X-ray source focal point and flat panel sensor
D2: distance between X-ray source focal point and object
If the distance between the flat panel sensor and X-ray
source is fixed, then the magnification will increase as
the object is brought closer to the X-ray source.
During enlargement, the image becomes fuzzier as the
focal spot size of the X-ray source becomes larger. This
means that using an X-ray source with a small focal spot
size will yield sharp, clear images even when enlarged.
[Figure 5-18] Image distortion by X-ray source with
different focal spot size
[Figure 5-19] Concept image of X-ray diffraction
D1
D2
Diffracted X-rays
X-ray source with
small focal spot size
Flat panel sensor
Parallel X-rays
Object
Object
Focal point
KACCC0462EA
[Figure 5-20] X-ray diffraction pattern example
Ordinary X-ray source
Flat panel sensor
KACCC0245EB
Cone beam CT
As a method for making full use of the features of flat panel
sensors with a large photosensitive area, there is a cone
beam CT that uses a cone beam X-ray source capable of
emitting X-rays over a wide area.
The cone beam X-ray source and the flat panel sensor are
installed opposite each other with the object positioned
in the center. Images of the object are then acquired
while the X-ray source and flat panel sensor are rotated
at the same speed around the object.
The two-dimensional image data acquired in this way
is then reconstructed by a computer to create threedimensional X-ray transmission images. The cone beam
CT can also acquire three-dimensional X-ray images of
large objects in a short time by using high-frame-rate flat
panel sensor with a large photosensitive area.
X-ray diffraction
Flat panel sensors are useful for analysis of X-ray Laue
diffraction method because of a large photosensitive area
and high resolution. As shown in Figure 5-19, parallel
X-rays irradiate the object, and interference fringes formed
by the X-rays diffracted by the object are detected with
the flat panel sensor. In this way high definition images
equivalent to those obtained with an imaging plate can
be obtained. The flat panel sensor is used for applications
including structural analysis of crystals and proteins.
21