Download Compensators for dose and scatter management in cone

Compensators for dose and scatter management
in cone-beam computed tomography
S. A. Graham
Ontario Cancer Institute, Princess Margaret Hospital, Toronto, Ontario, M5G 2M9, Canada
and Department of Medical Biophysics, University of Toronto, Ontario, M5G 2M9, Canada
D. J. Moseley
Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, M5G 2M9, Canada
and Department of Radiation Oncology, University of Toronto, Toronto, Ontario, M5G 2M9, Canada
J. H. Siewerdsen
Ontario Cancer Institute, Princess Margaret Hospital, Toronto, Ontario, M5G 2M9, Canada
and Department of Medical Biophysics, and Department of Radiation Oncology, University of Toronto,
Toronto, Ontario, M5G 2M9, Canada
D. A. Jaffray
Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, M5G 2M9, Canada,
Department of Radiation Oncology, and Department of Medical Biophysics,
University of Toronto, Toronto, Ontario, M5G 2M9, Canada;
and Ontario Cancer Institute, Princess Margaret Hospital, Toronto, Ontario, M5G 2M9, Canada
共Received 22 May 2006; revised 19 April 2007; accepted for publication 20 April 2007;
published 11 June 2007兲
The ability of compensators 共e.g., bow-tie filters兲 designed for kV cone-beam computed tomography 共CT兲 to reduce both scatter reaching the detector and dose to the patient is investigated.
Scattered x rays reaching the detector are widely recognized as one of the most significant challenges to cone-beam CT imaging performance. With cone-beam CT gaining popularity as a method
of guiding treatments in radiation therapy, any methods that have the potential to reduce the dose to
patients and/or improve image quality should be investigated. Simple compensators with a design
that could realistically be implemented on a cone-beam CT imaging system have been constructed
to determine the magnitude of reduction of scatter and/or dose for various cone-beam CT imaging
conditions. Depending on the situation, the compensators were shown to reduce x-ray scatter at the
detector and dose to the patient by more than a factor of 2. Further optimization of the compensators is a possibility to achieve greater reductions in both scatter and dose. © 2007 American
Association of Physicists in Medicine. 关DOI: 10.1118/1.2740466兴
Key words: x-ray scatter, dose, compensator, bow-tie filter, cone-beam CT
I. INTRODUCTION
Standard practice in external beam radiation therapy relies
on imaging techniques that lack the soft-tissue detectability
necessary for precisely and accurately locating soft-tissue
targets. Treatment setup relies on a combination of skin
marks and/or megavoltage radiographs 共portal images兲 that
can guide treatment based largely on the location of bony
anatomy. Image guidance is an important development for
more precise treatment delivery, with numerous techniques
developed to accomplish visualization of soft-tissue structures of interest with the patient in treatment position.1 A
recently developed technology for soft-tissue visualization
and image-guided radiation therapy is kV cone-beam CT
共CBCT兲.2
CBCT systems employ conventional x-ray tubes and
x-ray detectors such as image intensifier systems3,4 or flatpanel detectors.2,5 In this work, recently developed flat-panel
detectors 共FPDs兲 are used to generate two-dimensional radiographic projection data that can be processed to form a volumetric reconstruction with sub-millimeter spatial resolution.
FPDs provide digital images read out at frame rates up to 30
2691
Med. Phys. 34 „7…, July 2007
frames per second, providing sufficient numbers of projections for reconstruction with a single rotation of the sourcedetector pair. The use of CBCT for image guidance is gaining widespread popularity with a number of manufacturers
offering CBCT imaging platforms integrated with linear accelerators. This allows radiography, fluoroscopy, and CBCT
images to be acquired in the treatment position. Although the
large area of FPDs is advantageous for CBCT acquisitions of
clinically relevant fields of view, the FPDs currently employed in CBCT systems suffer from limited dynamic range.
To achieve reasonable signal levels through the midline of
many patients it is necessary to deliver a high exposure per
projection. It is often the case that the techniques used overwhelm the signal range of the detector at the periphery of the
patient, leading to a loss of information in projections and
artifacts in reconstruction due to the truncation of anatomy.
Another issue arising in CBCT imaging is the large quantity of scattered radiation generated in the patient and reaching the detector, to be hereafter referred to simply as scatter,
due to the large projection field sizes 共⬃25⫻ 25 cm2兲 used in
patient imaging.6 In some imaging geometries, the scatter
0094-2405/2007/34„7…/2691/13/$23.00
© 2007 Am. Assoc. Phys. Med.
2691
2692
Graham et al.: Compensators for cone-beam CT
fluence exceeds the primary fluence at the detector. With
such a large fraction of the detected fluence arising from
scatter, there can be significant artifacts in the resulting reconstructions. The presence of scatter reduces contrast and
also contributes additional x-ray quantum noise and induces
localized artifacts such as streaking and reduced attenuation
estimation at the center of the object 共known as a cupping
artifact兲.6 Scattering within the patient also contributes additional dose to the patient, which does not necessarily contribute to corresponding improvements in image quality.
The harmful effects of scatter in cone-beam CT imaging
and the rapid deployment of these systems make this an important area of research. The development of robust correction schemes7–11 for removal of scatter in projections is an
active area of research. In addition to postprocessing algorithms applied to data to reduce scatter effects it is also possible to use more mechanical methods to reduce scatter. This
can be, for example, in the form of larger air gaps between
the patient and detector,12–15 or by implementing scatter rejecting grids.16–18 Each of these approaches has limitations.
The use of a grid is known to reduce the primary fluence at
the detector, despite having delivered the dose to the patient,
whereas the postprocessing methods are limited to artifact
reduction and they cannot address the additional x-ray quantum noise induced by the scatter fluence.
While the scatter correction and rejection methods mentioned above may be capable of reducing the effects of scatter in the reconstructed volumes, it would be beneficial to use
scatter rejection techniques, which also reduce the dose to
the patient. The use of a smaller longitudinal field-of-view
共FOVz兲 by reducing the collimator opening during imaging
would reduce scattered radiation, but would also limit the
total volume that could be reconstructed.
The implementation of compensating filters19 共a subset of
which is known as bow-tie filters兲 in CBCT offers an alternative method of addressing the issue of scatter. The scattered radiation in CBCT has been shown in a qualitative
fashion to be reduced when using compensators.20,21 These
compensating filters can also provide a method of both reducing scattered dose within the patient and relaxing the dynamic range requirements for the FPD.
Compensators have existed in some form since it was
recognized that the detectability in film radiographs would
be improved if a filter were used to deliver a more uniform
fluence at the film.22 Filters could be designed to account for
more subtle changes in patient anatomy to further increase
image quality.23 Furthermore, the use of appropriate filters in
radiography has been shown to reduce patient dose.24 Compensation schemes were implemented early on in computed
tomography scanning, with the EMI scanner using a water
bag compensator.25,26 CT scanners later used metal “dodgers” made of low atomic number material to closely mimic
water-based compensation.27
The purpose of compensation in CT is both to accommodate the dynamic range of the detectors and preferentially
harden the x-ray beam.19–21,27 Patient dose is also reduced
when introducing a compensating filter.21,28–30 Despite scatMedical Physics, Vol. 34, No. 7, July 2007
2692
FIG. 1. Schematic of the bench-top cone-beam CT system used for investigations into the dose and scatter reduction capabilities of compensating
filters. Parameters that were varied in order to alter the scatter and dose
included the compensator, z collimation, phantom size, and the air gap between the phantom and detector. The geometry of the system is also summarized in Table I.
ter not playing as large a role in conventional CT imaging as
in CBCT imaging, scatter induced artifacts have been reported as being reduced by use of a bow-tie filter.31
For compensator use in multi-row CT or CBCT it is necessary to manufacture 2D shaped filters32 to modulate the
x-ray fluence across the imager. Filters of this type can be
expected to provide similar benefits as bow-tie filters provide
in conventional CT, though they may potentially play a
larger role in scatter and dose reduction when applied to
CBCT imaging.
Although qualitatively it is recognized that compensators
offer benefits to CBCT, it is necessary to quantify the magnitude of effect obtained under realistic conditions. The investigations are performed assuming a cone-beam CT imaging geometry consistent with current radiotherapy imageguidance systems. These investigations will provide
quantitative evidence of the value of using compensators for
the purpose of improving cone-beam CT image quality, reducing the magnitude of x-ray scatter at the detector, and
reducing patient dose.
II. MATERIALS AND METHODS
A. Experimental cone-beam CT bench
Investigations of compensating filters for CBCT were performed on a bench-top CBCT system 共Fig. 1兲. The x-ray
source was a Rad-94 rotating anode x-ray tube 共Varian Medical Systems兲 in a Varian Sapphire housing with a maximum
potential of 150 kVp, 14° tungsten rhenium molybdenum
graphite target, and 0.4 and 0.8 mm focal spot sizes. The
geometry of the system was configured to have a source-toisocenter distance 共dSA兲 of approximately 100 cm, and a
source-to-detector distance 共dSD兲 of approximately 155 cm,
which is consistent with the geometry of clinical CBCT sys-
2693
Graham et al.: Compensators for cone-beam CT
2693
TABLE I. Nominal imaging geometry and acquisition/reconstruction parameters.
Imaging system geometry
Focal spot to isocenter distance 共dSA兲
Nominal focal spot to detector distance 共dSD兲
Nominal isocenter to detector distance 共dAD兲
Focal spot to compensator distance 共dcomp兲
Axial field of view 共FOVx兲
Maximum longitudinal field of view 共FOVz兲
Minimum longitudinal field of view 共FOVz兲
100
155
55
9.7
25.6
19.2
2
cm
cm
cm
cm
cm
cm
cm
Image acquisition
Tube potential
Tube current
Exposure time
Head phantom
Body phantom
20.5 cm diameter watercylinder
Frame rate
Number of exposures
Nominal beam filtration
10 ms
20 ms
4 ms
1 s/projection
320
4 mm Al, 0.13 mm Cu
Image processing and reconstruction
Defect pixel removal
Reconstruction filter
Maximum reconstruction volume
Voxel dimension
3 ⫻ 3 median filter
Hamming
25.6⫻ 25.6⫻ 19.2 cm3
0.05⫻ 0.05⫻ 0.05 cm3
120 kVp
100 mA
tems used for image guidance during radiation therapy. Object rotation during CBCT acquisition was performed with a
direct-drive rotation stage 共Dynaserv DM1060B兲. The detector used was a 14 bit indirect-detection flat-panel detector
共PaxScan 4030A, Varian Medical Systems兲 with a 600 ␮m
thick CsI:Tl scintillating layer. The pixel array was a 2048
⫻ 1536 共40⫻ 30 cm2兲 amorphous silicon flat panel matrix
with a pixel pitch of 194 ␮m. This detector provides a
25.6⫻ 19.2 cm2 maximum FOV at isocenter when using a
centered detector. Table I summarizes the relevant parameters of the bench-top CBCT system.
B. Compensating filters
There are numerous possible designs for compensators
implemented on CBCT imaging systems. Compensators
could potentially be optimized based on a variety of imaging
tasks while taking account of the tradeoffs between image
quality and dose to the patient. Patient size, scatter, dose, and
exposure dependent detective quantum efficiency 共DQE兲 of
the detector are examples of quantities that could be considered in the design of modulating filters for cone-beam CT.
Simple representative compensators were used in this work
to examine the dose and scatter effects of modulating the
input fluence. Although compensators could be manufactured
to have variable modulation across the two-dimensional
cone-beam projection images, the compensators used in this
study have constant modulation in the z direction.
The two bow-tie filters constructed for this work were
built based on the objective of achieving uniform fluence
through a cylindrical head phantom with a diameter of 16 cm
Medical Physics, Vol. 34, No. 7, July 2007
FIG. 2. Thickness profile designed for the 8:1 and 4:1 modulation factor
compensators, shown with machining tolerances for the 8:1 filter. Measurements of the compensator thickness performed after the completion of machining, shown here only for the 8:1 compensator, indicate that the 4:1
compensator corresponded well with the designed thickness of 0.1 mm,
while the 8:1 compensator was measured to be 0.14± 0.03 mm at the center.
This discrepancy in thickness at the center was corrected for in the measurements shown in this work. The width of the compensators continues as a flat
function out to ±62.5 mm, but the axes of this figure have been adjusted to
emphasize the shape of the machined section of the copper plates.
measured at 120 kVp. The shape of the filters 共Fig. 2兲 was
then smoothed using an unweighted sliding average to remove discontinuous first derivatives that are difficult to machine and that could induce artifacts under conditions of imperfect geometric calibration. The compensators were
manufactured from 99.9% pure copper plates, as this material made it possible to machine to the desired shapes and
would allow thin compensators that could be placed between
the x-ray tube and collimator assembly at approximately
10 cm from the focal spot. At this position it is expected that
very few x rays scattered from the compensators will reach
the detector because of the large 共145 cm兲 air gap. The bowtie filters were manufactured from 125⫻ 110 mm2 sheets of
Cu approximately 2.4 and 1.6 mm in thickness. The plates
were machined to the desired shape down to a copper thickness of approximately 0.1 mm at the center 共central axis of
the beam兲. The compensators were mounted on 2.5 mm thick
acrylic sheets for mechanical support. The chosen thicknesses and profiles provide compensators with modulation
factors of approximately 8:1 and 4:1, where the modulation
factor was defined as the ratio between the measured detector
signal at the center of a projection, through the thinnest area
of the compensators, to that at the periphery of a projection
共in plateau of compensator profile兲 where the beam is attenuated by 0.1, 1.6, or 2.4 mm of copper for the 1:1, 4:1, and 8:1
compensators, respectively. The modulation factor was defined only in terms of the measured detector signal in flood
field images where the compensators are in place but no
object is present in the beam. The added filtration used on the
CBCT system is commonly 4 mm of aluminum and
0.13 mm of copper 共i.e., no compensator, referred to here as
the 1:1 compensator兲. When placing the 4:1 or 8:1 compen-
2694
Graham et al.: Compensators for cone-beam CT
2694
FIG. 3. 共a兲 SPR and 共b兲 the corresponding SRF value at the center of
the projection in a 16 cm diameter
acrylic phantom 共head兲 and 共c兲 SPR
and 共d兲 SRF for a 32 cm diameter
acrylic phantom 共body兲 for varying
longitudinal FOV and the three compensators employed in this study. The
value of dAD was held at 55 cm for all
scans 共giving a dAG of 39 cm for the
body phantom and 47 cm for the head
phantom兲. Measurements are also
shown for the beam-blocker method
for the 1:1 共open circle兲 and 8:1 共open
square兲 compensators.
sator in the beam, the 0.13 mm copper filter was removed to
obtain a primary fluence at the center of the beam that is
nearly equivalent in all cases.
C. Scatter-to-primary ratio measurements
The SPR at the center of the detector was measured for
the 8:1, 4:1, and 1:1 compensators. The SPR is defined as
SPR =
S
,
P
共1兲
where S is the energy integrated signal of the scattered radiation measured as the average of a 100⫻ 20 pixel2 area on
the FPD, and P is the signal due to the primary radiation.
The SPR was measured at the location of the central axis in
the projection images for both an acrylic 16 cm cylindrical
head phantom and the 16 cm head phantom placed inside of
an acrylic annulus with an outside diameter of 32 cm.
Acrylic has a Compton scattering contribution to the mass
attenuation coefficient of approximately 2% to 3% higher
than that of tissue 共water兲 in the relevant energy range 共for
example, acrylic has values of 0.173 cm2 g−1 at 60 keV and
0.158 cm2 g−1 at 100 keV, while water has values of 0.177
and 0.163 cm2 g−1 at the same energies兲,33 and we believe
this is sufficiently close to demonstrate the effect the compensators would have on scatter in patient imaging. The projection images used in the measurement of the SPR were an
averaged set of 30 in-air exposures 共flood field images兲 with
dark fields subtracted. Projections were acquired at 120 kVp
Medical Physics, Vol. 34, No. 7, July 2007
with 1 mAs/projection for the head phantom and
2 mAs/projection for the body phantom.
The determination of the value of SPR in the images was
accomplished by varying the longitudinal field of view
共FOVz兲 from approximately 18.5 cm at isocenter down to the
minimum field of view achievable with the collimator assembly, which is approximately 2 cm 共±0.05 cm兲. The axial
field of view 共FOVx兲 was fixed at the maximum detectable
共25.6 cm兲. The total signal in the projections was assumed to
be the sum of the primary and scattered radiation signals. By
plotting the total signal measured in images against FOVz, a
curve could be found which was employed to extrapolate the
total signal to a FOVz of zero. For the 16 cm phantom the
data were fit to a square root function, while the 32 cm data
were found to be more linear with field size. At the point of
zero FOVz the signal was attributed completely to the primary fluence and was assumed to be constant in all of the
projections. Subtracting the primary signal estimate in all
projections permitted the determination of the scatter from
the phantom at the center of the projection images. Additional measurements of the SPR were performed with a variable air gap 共dAG兲 between the phantom and detector, while
maintaining the FOVz at 18.5 cm to compare the effects of
the modulating filters with alternative imaging geometries.
A secondary check of some of the SPR measurements was
also performed using a beam-blocker method.34 This was
done in order to compare the SPR values obtained by varying the longitudinal FOV with a second method of determin-
2695
Graham et al.: Compensators for cone-beam CT
2695
FIG. 4. Variation in SPR for different
spacing between the phantoms and
flat-panel detector. The 共a兲 SPR and
共b兲 SRF results are shown for the head
phantom along with the SPR and SRF
shown respectively in 共c兲 and 共d兲 for
the body phantom.
ing SPR. The lead blocker had approximately an 8 mm width
and 1 cm thickness. Measurements were made at a single
FOVz of 18 cm using the 16 and 32 cm phantoms and the
1:1 and 8:1 compensators. A correction accounting for shadowing of part of the phantom by the blocker was accomplished by performing measurements with blockers of varying widths 共5 mm up to 14 mm兲 and extrapolating to a
blocker size of zero.
In order to quantify the benefits offered by the compensating filters, the scatter reduction factor 共SRF兲 was defined
as the ratio 共for a given phantom, air gap, and FOVz兲 of the
SPR measured with a compensating filter divided by the SPR
corresponding to a flat filter 共i.e., 1:1 case兲. The SRF shows
the magnitude of scatter reduction offered by the compensators as well as the trends in scatter reduction when altering
FOVz and air gap. An SRF of one would indicate no change
in scatter magnitude, less than one would indicate scatter
removal, and greater than one would signify elevated scatter.
D. Dose measurements
Dose was measured in the acrylic 16 cm head and 32 cm
body phantoms exposed at 120 kVp with 1 mAs/projection
for the head phantom and 2 mAs/projection for the body
phantom. The total dose was found based on an acquisition
of 320 projections across 360°. Measurements were perMedical Physics, Vol. 34, No. 7, July 2007
formed with a 0.6 cc Farmer ionization chamber 共NE 2571,
S/N 1700兲 and a Keithley 35040 electrometer 共S/N 62968兲
with a 300 V bias. The charge integrated by the electrometer
was converted to dose using an air kerma calibration factor
共Nk兲 for the specific ionization chamber/electrometer pair determined from a standard ionization chamber/electrometer
pair calibrated by the National Research Council 共Ottawa,
Ontario, Canada兲. The Nk corresponding to 120 kVp was
420 mGy/ 10−8 C. The dose was measured for variations in
compensator modulation, longitudinal field of view 共FOVz兲,
and position of the dosimeter in the phantom. The dose reduction factor 共DRF兲 was defined similarly to the SRF and
was used to quantify the dose reductions achieved with the
compensators in place.
Slight differences in filter thickness along the central axis
for the three choices of compensator were expected to play a
small role in creating the differences in measured dose. Measurements of the copper thickness after machining was complete indicated that the 4:1 compensator was 0.10± 0.03 mm
at the center, while the 8:1 compensator was 0.14± 0.03 mm
at the center. To provide the most accurate comparison of the
1:1, 4:1, and 8:1 compensators the differences caused by the
differing primary fluence along the central axis were corrected out of the measured doses. This was performed by
measuring the detector signal under narrow beam geometry
2696
Graham et al.: Compensators for cone-beam CT
2696
that of the 1:1 filter. Appropriate adjustments were made to
the doses measured with the 8:1 filter to account for this
systematic discrepancy.
E. Reconstructed images
FIG. 5. SPR measured as a function of modulation factor for the 16 cm
共head兲 and 32 cm 共body兲 phantoms.
共2 ⫻ 2 cm2 field at isocenter兲 with each of the compensators
in place and no object in the beam. Ideally, the measured
signal through each of the compensators would be identical.
The measured signal through the 1:1 filter was compared to
the signal through the center of the 4:1 and 8:1 compensators
and deviations from the 1:1 signal were used to correct the
measured doses. These measurements showed that the transmission through the center of the 4:1 compensator matched
共within 2%兲 that of the 1:1 filter. The transmission through
the 8:1 filter was found to be approximately 7% lower than
The influence of the compensators on CBCT image quality was assessed using a simple cylindrical water phantom
for examination of cupping artifacts and noise. The acrylic
wall of the phantom had an outside diameter of 20.5 cm and
a thickness of 0.5 cm. The height of the phantom was approximately 28 cm. Projection images were acquired for
FOVz values of 2 and 18.5 cm at 120 kVp and
0.4 mAs/projection with 320 projections being acquired
through 360°. The raw projection data acquired at 2048 pixels by 1536 pixels were down-sampled to 512⫻ 384, gain
and offset corrected using floods taken with the proper compensator in place, and median filtered to remove defective
pixels from the images. Volumes of 25.6⫻ 25.6⫻ 1.3 cm3
共0.05⫻ 0.05⫻ 0.05 cm3 voxels兲 were reconstructed using the
Feldkamp algorithm35 for each case 共FOVz = 2 cm, 18.5 cm;
compensation: 1:1, 4:1, 8:1兲 so that comparisons could be
made in terms of apparent cupping, noise, and any
compensator-induced artifacts. A summary of the reconstruction parameters is given in Table I.
Images of the phantom were also performed with two
liver-equivalent tissue inserts 共GAMMEX rmi, ␳w
e = 1.07兲
suspended in the phantom for the purpose of contrast-tonoise ratio 共CNR兲 measurements. The CNR was defined as
FIG. 6. Dose measured in the 16 cm
acrylic head phantom at 共a兲 the center
and 共c兲 the periphery of the phantom.
The compensators reduce the dose in
both locations, as demonstrated by the
DRF for the 共b兲 center and 共d兲 peripheral doses.
Medical Physics, Vol. 34, No. 7, July 2007
2697
Graham et al.: Compensators for cone-beam CT
2697
FIG. 7. Dose measured in the 32 cm
acrylic body phantom at 共a兲 the center
of the phantom and 共c兲 the periphery
of the phantom, shown with the DRF
for 共b兲 the center of the phantom and
共d兲 the periphery of the phantom.
the difference between the mean attenuation value in the insert and in the surrounding water 共both measured in a 100
voxel ROI兲, divided by the standard deviation of the attenuation value in the water. The images acquired for the CNR
measurements were measured at 4 mAs/projection, and an
additional 3 mm of copper filtration added to the beam. This
was necessary to reduce beam-hardening artifacts present in
images. All other parameters for the CNR measurements
were the same as those given above.
III. RESULTS
A. Scatter-to-primary ratio measurements
The SPR at the center of each phantom was found to
decrease as the longitudinal field of view was reduced, and
the modulation factor was increased 共Fig. 3兲. The SPR in the
head phantom reached a value greater than 100% for the
highest FOVz for the 1:1 modulation factor filter. With the
4:1 and 8:1 compensators the magnitude of the scatter signal
does not reach the same level as the primary signal for even
the highest field of view. The SRF showed a constant 20%
reduction in scatter for all FOVs when applying the 4:1 compensator, and a greater than 40% reduction for the 8:1 compensator. Beam-blocker measurements agreed well with the
extrapolation measurements giving SPR values of 1.02± 0.03
for the 1:1 compensator and 0.54± 0.03 for the 8:1 compensator at a FOVz of 18 cm, compared with 1.01± 0.03 and
0.60± 0.02 for the 1:1 and 8:1 compensators, respectively, at
18.5 cm using the extrapolation method.
Medical Physics, Vol. 34, No. 7, July 2007
For the body phantom similar effects were noticed,
though the levels of scatter were much higher. The scatter
signal equalled that of the primary signal when applying flat
filtration even at relatively small FOVz with an SPR of 100%
at a longitudinal field of view of approximately 3.7 cm. The
same SPR when using the 4:1 and 8:1 compensators could be
achieved with a FOVz of 6.5 and 9.8 cm, respectively, demonstrating the large gains that can be made through the
implementation of compensation schemes in cone-beam CT.
The reduction of scatter was more pronounced for the body
phantom than the head phantom, with SRF values of approximately 0.6 and 0.4 for the 4:1 and 8:1 compensators,
respectively. Beam-blocker measurements gave values of
4.4± 0.4 and 2.1± 0.2 for the 1:1 and 8:1 compensators.
With FOVz fixed, compensating filters reduced the SPR
regardless of the air gap 共Fig. 4兲 for the range of geometries
examined. Although the primary signal in each pixel decreased as the detector moved away from the phantom and
x-ray tube, the scatter fell off more quickly, resulting in a
decreasing SPR. This is consistent with the well-documented
role of air gaps in reducing scatter at the detector.13,15 While
the smallest air gap caused a large value of SPR when using
a flat filter, the application of the 8:1 compensator reduced
the SPR such that the SPR at the smallest air gap for the 8:1
compensator was less than that at the highest air gap for the
1:1 case for both the head and body phantoms. SRF values
remained close to the values found at the nominal distance
between the phantom and the detector 共dAD = 55 cm兲. There
2698
Graham et al.: Compensators for cone-beam CT
2698
FIG. 8. Dose as a function of the distance from the center of the phantom
in both the 共a兲 head and 共c兲 body phantom. The DRF is shown for the 共b兲
head and 共d兲 body phantoms.
was a slight decrease in SRF as the air gap was increased,
though the slope was not significantly different from zero.
Figure 5 summarizes the SPR as a function of the modulation factor. The SPR results for an 18.5 cm longitudinal
field of view at isocenter and a dSD of 155 cm at the center of
the two phantoms clearly show a substantial decrease in SPR
as the modulation factor of the compensator increases. The
effect is larger in the body phantom where the SPR is much
greater.
B. Dose measurements
The dose at the center and periphery of the 16 cm acrylic
head phantom was shown to vary with FOVz and compensation 共Fig. 6兲. At the center of the phantom the primary intensity of the beam was approximately constant, so the reduction in dose when increasing the compensator modulation
was due to the reduction in scatter. The dose at the center of
the phantom for the 8:1 compensator was approximately
75% of the dose measured at the center of the phantom when
using a flat filter. The compensators further reduced the dose
at the periphery of the phantom, since both the primary and
scatter were reduced. When using the 8:1 compensator instead of the flat filter the dose was reduced to approximately
55% of the original value at the periphery. The reduction in
dose measured at the periphery of the phantom when employing compensating filters came largely from the reduction
in primary x rays. The DRF appears constant across all
fields-of-view for both positions in the phantom, except for a
Medical Physics, Vol. 34, No. 7, July 2007
suspicious drop in the DRF when the dose was measured
with the lowest FOVz at the periphery of the phantom.
Similar trends in dose reduction were also seen when
measuring dose in the body phantom while varying the compensation scheme and FOVz 共Fig. 7兲. The dose decreased to
about 60% and 40% of the value measured with the 1:1
modulation filter at the center and periphery, respectively,
when applying the 8:1 filter. There are large reductions in
dose at the periphery of the phantom because the compensators used in this study were designed according to the fluence
through the 16 cm acrylic head phantom and could not be
placed closer to the x-ray tube to adjust the magnification of
the compensator shape. Thus the fluence pattern applied to
the body phantom reduced the primary fluence to a larger
degree than a filter that was designed to accommodate the
exact shape of this larger phantom.
Despite this large reduction in dose to the periphery of the
phantom, it was seen that, unlike the case of the head phantom, the dose at the periphery of the phantom for the 8:1
compensator was larger than the dose at the center. The dose
as a function of distance from the center of the phantom for
the different compensators was measured in the head and
body phantoms 共Fig. 8兲. In the head phantom the dose increased towards the outside of the phantom in the 1:1 case,
was nearly constant for the 4:1 case, and was reduced for the
8:1 case. In the body phantom, the dose increased when
moving away from the center of the phantom up to a point
where the dose began to drop. This drop is attributed to the
2699
Graham et al.: Compensators for cone-beam CT
2699
C. Image reconstructions
FIG. 9. For both the head and body phantoms the dose is shown to vary with
varying modulation factors. The dose shown in 共a兲 is measured at the center
of the phantoms where the primary intensity of the beam is approximately
the same for all modulations. In 共b兲 the dose is shown at a depth of 1 cm in
the two phantoms where both the scatter and primary fluences are altered by
the compensators during a CBCT scan.
edge of the phantom moving outside the field-of-view at approximately 12.8 cm from the center of the phantom. The
peak dose was found at different distances from the center of
the phantom because of the changes in dose due to scatter.
The DRF shows, as expected, that the dose reducing capabilities of the compensators increased when moving away
from the center of the phantom—this is a combination of the
reduced primary and the reduction in primary-induced scatter
fluence within the phantom.
Figure 9 shows the achievable decrease in dose at the
center and periphery of the head and body phantoms with the
compensators used in this study. Similar trends were seen in
both phantoms when varying the modulation of the compensator. The decrease in dose measured at the center of the
phantom was due only to the decreased scatter dose arising
from the modified primary fluence pattern offered by the
compensators. The decrease in dose shown at the periphery
of the phantoms was due to a combination of the decrease in
scatter, along with the decrease in primary fluence caused by
the shape of the filter.
Medical Physics, Vol. 34, No. 7, July 2007
Reconstructions were performed on projection images of
a cylindrical water phantom for a FOVz of 2 cm 共low scatter,
Fig. 10兲 and 18.5 cm 共high scatter, Fig. 11兲. In the low scatter case the 1:1 compensator gives an image with a very
uniform reconstructed attenuation value. The 4:1 image is
similar in signal uniformity to the 1:1 case, though some
reduction in the attenuation value is seen at the edges of the
phantom. Large discrepancies are seen with the 8:1 compensator image where a considerable “cap” is seen in the image
with a peak in reconstructed attenuation values at the center
of the image. The cause of the nonuniformity in the reconstruction is hypothesized to be the change in the energy spectrum of the x-ray beam when passing through the compensators. The shape of the compensators results in a spectrum
with a higher mean energy as the beam moves away from the
central axis. The harder x-ray spectrum directed at the outside of the cylinder, caused by the thicker copper on the
compensator, results in a larger percentage of the x rays passing through the cylinder resulting in a perceived reduction in
x-ray attenuation at the periphery of the cylinder. A signalto-noise 共SNR兲 analysis of the images was performed in 10
voxel by 10 voxel regions near the center and edges of reconstructed images. This analysis demonstrates that the compensators provide a more uniform pattern of noise across the
images, with SNR values at the edges of the phantom of
72± 4 for the 1:1 case, 54± 5 for the 4:1 compensator, and
41± 4 for the 8:1, and SNR values near the center of the
phantom of 33± 3, 36± 5, and 37± 5 for the 1:1, 4:1, and 8:1
compensators, respectively.
Increasing the FOVz up to 18.5 cm significantly increases
the percentage of scatter present in the system. In the image
acquired with the 1:1 compensator a severe cupping artifact
caused by the increased scatter is now seen. The reduction in
scatter afforded by the 4:1 compensator provides a substantial reduction in the cupping due to scatter. An additional
benefit is that the 4:1 image is acquired using an estimated
15% to 20% decrease in dose compared to the 1:1 case,
though the SNRs for these two images are not significantly
different 共72± 4 at the edge and 42± 3 at the center of the 1:1
image, compared to 71± 5 at the edge and 48± 4 at the center
of the 4:1 reconstruction兲. In the 8:1 case the change in the
spectrum across the beam continues to cause an artifact in
reconstruction. In fact, the capping artifact in the 8:1 image
has increased in the presence of higher amounts of scattered
x rays. This effect is hypothesized to be due to the change in
the distribution of scattered x rays relative to primary x rays
across the phantom in the presence of the 8:1 compensator.
The SNR at the center of the 8:1 image is comparable to the
1:1 and 4:1 images with a value of 42± 6. At the edge of the
8:1 reconstruction the SNR dropped to 42± 4.
With the large errors caused by beam hardening in the 8:1
images it is difficult to assess any improvements in image
quality afforded by the compensating filters. To reduce these
errors images were acquired with an additional 3 mm of copper filtration to considerably harden the x-ray beam. Although this may not be a realistic approach for everyday
2700
Graham et al.: Compensators for cone-beam CT
2700
FIG. 10. Reconstructions of the cylindrical water phantom with a FOVz of
approximately 2 cm for the 1:1, 4:1,
and 8:1 compensating filters.
imaging because of the heating of the x-ray tube, it does
provide some idea of the benefits of compensation. Figure 12
shows images of the phantom with liver inserts in place for
the 1:1 and 8:1 compensators, as well as profiles across the
images. It is evident from these profiles that the cupping
artifact due to scatter is greatly reduced by the compensator.
With these imaging conditions the degree of cupping, given
by6
tcup =
␮edge − ␮center
⫻ 100 % ,
␮edge
共2兲
was found to be 12.5% in 1:1 images and −1% in 8:1 images
共negative because of the presence of beam hardening capping
artifacts兲 acquired with a FOVz of 18.5 cm. With a 2 cm
FOVz the cupping was 3% in the 1:1 case and −0.6% in the
8:1 case. Measurements of the CNR when imaging with a
2 cm FOVz give values at the center of the phantom of
2.6± 0.2 and 2.4± 0.2 for the 1:1 and 8:1 compensators, respectively, and at the edge of the phantom of 3.1± 0.2 for the
1:1 compensator and 2.7± 0.2 for the 8:1 compensator. So,
for small longitudinal FOVs compensators do not offer any
improvements in image quality. If the FOV is increased,
there is a larger amount of scatter, and the CNR in the 1:1
case drops to 1.9± 0.2 at the center of the phantom and
2.5± 0.1 at the periphery of the phantom. The CNR for the
8:1 compensator was measured as 2.4± 0.1 at the center and
2.3± 0.2 at the periphery. The 8:1 compensator does not imMedical Physics, Vol. 34, No. 7, July 2007
prove the CNR at the edge of the phantom compared to the
1:1 compensator when imaging under high scatter conditions, but this is expected since the 8:1 compensator greatly
reduces the primary radiation at the periphery of the phantom. The improvement in CNR from the 8:1 compensator is
seen at the center of the phantom where the CNR increased
from 1.9 up to 2.4 when compared to the 1:1 filter.
IV. DISCUSSION
The influence of modulating filters on both dose and scatter in cone-beam CT was evaluated. The application of compensating filters for all choices of phantom, air gap, and field
of view resulted in decreased scatter and dose, with the compensators removing over half the scatter and reducing the
dose at the center of the phantom by nearly a factor of 2 in
some cases. One known limitation in this work is the presence of extra-focal radiation, which cannot be separated from
x rays scattered from the imaged objects. Potential future
investigations will look to Monte Carlo methods for separating these effects and verifying the trends seen for the DRF
and SRF.
As scatter is likely the largest factor in the reduction of
image quality in CBCT many approaches are being investigated to address it. This management may take the form of
algorithms applied after imaging, or as physical modifications to the imaging system that reduce the amount of scat-
2701
Graham et al.: Compensators for cone-beam CT
2701
FIG. 11. Reconstructions of the cylindrical water phantom with a FOVz of
approximately 18.5 cm for the 1:1,
4:1, and 8:1 compensating filters.
tered radiation reaching the detector. Scatter correction algorithms could rely on Monte Carlo simulation methods or
estimations of the scatter based on experimentally acquired
scatter fluences.36 Some of the physical modifications that
could be made to a cone-beam CT imaging system for the
purpose of reducing scatter include adjusting the air-gap between the patient and detector, reducing the longitudinal
field-of-view employed during imaging to only image the
clinically relevant anatomy, utilizing anti-scatter grids to reject scatter, and, as evidenced by this work, application of
compensating filters. This work gives indication of the magnitude of scatter reduction that can be accomplished with
simple compensators, which could be combined with other
methods of scatter reduction to achieve more quantitatively
accurate and artifact-free reconstructions. An added benefit
of compensators is that, unlike most of the methods for removing the effects of scatter in CBCT imaging, compensators also have the ability to modulate the primary and scatter
dose delivered during image acquisition.
Although the compensators are shaped to account for the
shape of a head phantom, and not for the body phantom, the
decrease in image quality associated with the decrease in
primary fluence at the periphery of the phantom would not
be a problem if, for example, the anatomy at the center of the
patient volume is the most clinically relevant. Similar to region of interest imaging, this is one example of how comMedical Physics, Vol. 34, No. 7, July 2007
pensators could be employed while taking into account the
various factors that they are capable of influencing.
Images acquired of a cylindrical water phantom have
shown that for compensators with considerable modulation
factors there is a possibility of introducing considerable artifacts into the images due to the spectral hardening of the
x-ray beam. Although the compensators may reduce scatterinduced artifacts, the distortion of the x-ray energy spectrum
across the FOV introduced by the compensators may be
more of a problem than the cupping from scatter. Beam hardening corrections have been investigated for CBCT
systems,37,38 though beam hardening corrections generally
correct spectral changes caused only by the patient, and not
by compensating filters that are not present in the reconstructed images. Implementation of beam hardening corrections for cone-beam CT that can account for the large spectral changes introduced by compensators with considerable
modulation factors may be necessary in order to use compensators with larger modulation factors. Further investigation into the selection of materials and compensation methods that minimize the spectral perturbations caused by the
compensator is required. Despite the fact that the compensators employed here induce artifacts, there is still great potential for the use of compensators to reduce both the scatter at
the detector and the patient dose. Compensators with lower
modulation factors than those discussed in this work are a
2702
Graham et al.: Compensators for cone-beam CT
2702
tered x rays. The challenge and next logical step of this work
is to seek the appropriate compromises that still satisfy the
clinical requirements.
V. CONCLUSIONS
In this study, two representative compensators have been
used to demonstrate some of the capabilities of modulating
filters in cone-beam CT. Large reductions in scatter and patient dose could be made, as well as increased uniformity in
reconstructed images, a decrease in cupping artifacts in reconstructed volumes, and an increase in CNR at the center
phantoms imaged under high-scatter conditions. All of these
results are expected to be similar to those that could be
achieved with compensators of different shape or material.
One complicating factor that requires more investigation is
the introduction of beam-hardening artifacts into reconstructed images. One possibility for the design of a compensator would be to optimize the compensator such that it introduced the least amount of spectral hardening artifacts.
An approach that would potentially be more beneficial to
patients would be that the attenuation profile of compensating filters could be optimized to account for numerous properties of an imaging system. Depending on the desired imaging task, various aspects of the cone-beam CT imaging
system, such as the magnitude of scatter, dose, and distribution of primary fluence, can be influenced by compensating
filters. Further investigations into the possibility of creating
optimal fluence patterns with compensators to be utilized
during CBCT imaging continue to be an active area of research.
ACKNOWLEDGMENTS
The authors wish to acknowledge the radiation physics
machine shop at Princess Margaret Hospital for construction
of the compensators, and Bronwyn Hyland for assistance in
editing the manuscript. This project was supported by the
Ontario Graduate Scholarship program, the National Institutes of Health/National Institutes of Aging 共R21/R33AG19381兲, and the National Institutes of Health/National Institute of Biomedical Imaging and Bioengineering
共8R01EB002470-04兲.
FIG. 12. Reconstructions of the cylindrical water phantom with a FOVz of
approximately 18.5 cm for the 1:1 and 8:1 compensators. Liver equivalent
soft-tissue inserts have been placed at two locations inside the phantom to
allow measurement of the contrast-to-noise ratio.
possible solution leading to a compromise between beamhardening artifacts and the removal of scattered x rays from
the system.
Cone-beam CT is used in a number of clinical systems,
including linear accelerator and C-arm based CBCT systems.
While the performance of these systems is sufficient to be
valued, image quality is still recognized as an area requiring
improvement. The results of this work demonstrate that compensators, if applied carefully, can play a role in reducing
patient dose and increasing CNR through reduction of scatMedical Physics, Vol. 34, No. 7, July 2007
1
“Management of target localization uncertainties in external-beam
therapy,” Seminars in Radiation Oncology, edited by G. S. Mageras, Vol.
15, pp. 133–216 共2005兲.
2
D. A. Jaffray, J. H. Siewerdsen, J. W. Wong, and A. A. Martinez, “Flatpanel cone-beam computed tomography for image-guided radiation
therapy,” Int. J. Radiat. Oncol., Biol., Phys. 53, 1337–1349 共2002兲.
3
R. Fahrig, A. J. Fox, S. Lownie, and D. W. Holdsworth, “Use of a C-arm
system to generate true three-dimensional computed rotational angiograms: Preliminary in vitro and in vivo results,” AJNR Am. J. Neuroradiol. 18, 1507–1514 共1997兲.
4
R. Fahrig and D. W. Holdsworth, “Three-dimensional computed tomographic reconstruction using a C-arm mounted XRII: Image-based correction of gantry motion nonidealities,” Med. Phys. 27, 30–38 共2000兲.
5
D. A. Jaffray and J. H. Siewerdsen, “Cone-beam computed tomography
with a flat-panel imager: Initial performance characterization,” Med.
Phys. 27, 1311–1323 共2000兲.
6
J. H. Siewerdsen and D. A. Jaffray, “Cone-beam computed tomography
with a flat-panel imager: Magnitude and effects of x-ray scatter,” Med.
Phys. 28, 220–231 共2001兲.
2703
Graham et al.: Compensators for cone-beam CT
7
J. M. Boone and J. A. Seibert, “An analytical model of the scattered
radiation distribution in diagnostic radiology,” Med. Phys. 15, 721–725
共1988兲.
8
J. M. Boone and J. A. Seibert, “Monte Carlo simulation of the scattered
radiation distribution in diagnostic radiology,” Med. Phys. 15, 713–720
共1988兲.
9
L. Spies, M. Ebert, B. A. Groh, B. M. Hesse, and T. Bortfeld, “Correction
of scatter in megavoltage cone-beam CT,” Phys. Med. Biol. 46, 821–833
共2001兲.
10
R. Ning, X. Tang, and D. Conover, “X-ray scatter correction algorithm for
cone beam CT imaging,” Med. Phys. 31, 1195–1202 共2004兲.
11
J. H. Siewerdsen, M. J. Daly, B. Bakhtiar, D. J. Moseley, S. Richard, H.
Keller, and D. A. Jaffray, “A simple, direct method for x-ray scatter
estimation and correction in digital radiography and cone-beam CT,”
Med. Phys. 33, 187–197 共2006兲.
12
J. Persliden and G. A. Carlsson, “Scatter rejection by air gaps in diagnostic radiology. Calculations using a Monte Carlo collision density method
and consideration of molecular interference in coherent scattering,” Phys.
Med. Biol. 42, 155–175 共1997兲.
13
J. H. Siewerdsen and D. A. Jaffray, “Optimization of x-ray imaging geometry 共with specific application to flat-panel cone-beam computed tomography兲,” Med. Phys. 27, 1903–1914 共2000兲.
14
J. M. Boone, K. K. Lindfors, V. N. Cooper III, and J. A. Seibert, “Scatter/
primary in mammography: Comprehensive results,” Med. Phys. 27,
2408–2416 共2000兲.
15
U. Neitzel, “Grids or air gaps for scatter reduction in digital radiography:
A model calculation,” Med. Phys. 19, 475–481 共1992兲.
16
T. M. Bernhardt, U. Rapp-Bernhardt, T. Hausmann, G. Reichel, U. W.
Krause, and W. Doehring, “Digital selenium radiography: Anti-scatter
grid for chest radiography in a clinical study,” Br. J. Radiol. 73, 963–968
共2000兲.
17
A. Malusek, M. Sanbourg, and G. A. Carlsson, “Simulation of scatter in
cone beam CT—effects on projection quality,” Proc. SPIE 5030, 740–
751 共2003兲.
18
J. H. Siewerdsen, D. J. Moseley, B. Bakhtiar, S. Richard, and D. A.
Jaffray, “The influence of antiscatter grids on soft-tissue detectability in
cone-beam computed tomography with flat-panel detectors,” Med. Phys.
31, 3506–3520 共2004兲.
19
E. Seeram, Computed Tomography: Physical Principles, Clinical Applications & Quality Control 共Saunders, Philadelphia, 1994兲.
20
R. Ning, University of Rochester, “Apparatus and method for x-ray scatter reduction and correction for fan beam CT and cone beam volume CT,”
U.S. Patent No. 6,618,466共2003兲.
21
R. Ning, X. Tang, D. Conover, and R. Yu, “Flat panel detector-based cone
beam computed tomography with a circle-plus-two-arcs data acquisition
orbit: preliminary phantom study,” Med. Phys. 30, 1694–1705 共2003兲.
22
C. D. Smith, “X-ray filter,” U.S. Patent No. 2,216,326 共1940兲.
Medical Physics, Vol. 34, No. 7, July 2007
2703
23
P. Edholm, N. B. Jacobson, and Medinova A B, “Exposure compensating
device for radiographic apparatus,” U.S. Patent No. 3,755,672 共1973兲.
24
T. Katsuda, M. Okazaki, and C. Kuroda, “Using compensating filters to
reduce radiation dose,” Radiol. Technol. 68, 18–22 共1996兲.
25
G. N. Hounsfield, EMI Limited, “Body portion support for use with penetrating radiation examination apparatus,” U.S. Patent No. 3,867,634
共1975兲.
26
R. A. Brooks and G. Di Chiro, “Beam hardening in x-ray reconstructive
tomography,” Phys. Med. Biol. 21, 390–398 共1976兲.
27
H. H. Barrett and W. Swindell, Radiological Imaging: The Theory of
Image Formation, Detection, and Processing 共Academic, New York,
1981兲.
28
R. G. Walters and R. W. Carlson, Technicare Corporation, “Computerized
tomographic scanner with shaped radiation filter,” U.S. Patent No.
4,288,695 共1981兲.
29
J. E. Tkaczyk, Y. Du, D. Walter, X. Wu, J. Li, and T. Toth, “Simulation of
CT dose and contrast-to-noise as a function of bowtie shape,” Proc. SPIE
5368, 403–410 共2004兲.
30
S. Itoh, S. Koyama, M. Ikeda, M. Ozaki, A. Sawaki, S. Iwano, and T.
Ishigaki, “Further reduction of radiation dose in helical CT for lung cancer screening using small tube current and a newly designed filter,” J.
Thorac. Imaging 16, 81–88 共2001兲.
31
G. H. Glover, “Compton scatter effects in CT reconstructions,” Med.
Phys. 9, 860–867 共1982兲.
32
J. Hsieh, GE Medical Systems Global Technology Co., LLC, “Method
and apparatus for optimizing dosage to scan subject,” U.S. Patent No.
6,647,095 共2003兲.
33
M. J. Berger, J. H. Hubbell, S. M. Seltzer, J. Chang, J. S. Coursey, R.
Sukumar, and D. S. Zucker, XCOM: Photon Cross Section Database
(version 1.3) 共National Institute of Standards and Technology, Gaithersburg, MD, 2005兲. 关Available online: http://physics.nist.gov/xcom 共13 February 2007兲兴.
34
P. C. Johns and M. Yaffe, “Scattered radiation in fan beam imaging systems,” Med. Phys. 9, 231–239 共1982兲.
35
L. A. Feldkamp, L. C. Davis, and J. W. Kress, “Practical cone-beam
algorithm,” J. Opt. Soc. Am. A 1, 612–619 共1984兲.
36
G. Jarry, S. A. Graham, D. J. Moseley, D. A. Jaffray, J. H. Siewerdsen
and F. Verhaegen, “Characterization of scattered radiation in kV CBCT
images using Monte Carlo simulations,” Med. Phys. 33, 4320–4329
共2006兲.
37
J. Hsieh, R. C. Molthen, C. A. Dawson, and R. H. Johnson, “An iterative
approach to the beam hardening correction in cone beam CT,” Med. Phys.
27, 23–29 共2000兲.
38
L. Zhang, H. Gao, S. Li, Z. Chen, and Y. Xing, “Cupping artifacts analysis and correction for a FPD-based cone-beam CT,” Proc. SPIE 6065,
282–291 共2006兲.