Download CT issues in PET / CT scanning

abc
CT issues in PET / CT scanning
ImPACT technology update no. 4
David Platten
ImPACT
October 2004
Introduction
Since its introduction in the 1970s, X-ray computed tomography (CT) scanning has
become a widespread three-dimensional (3D) imaging modality, capable of
producing anatomical images with sub-millimetre resolution. Positron emission
tomography (PET) also produces 3D images, but these reflect physiological function,
showing the uptake of injected positron-emitting radiopharmaceuticals within the
body. The recent introduction of hybrid PET / CT scanners combines the two
modalities, enabling both anatomical and functional images to be collected in a single
session. PET / CT scanning is generally provided as a nuclear medicine service,
where there may not be experience in X-ray CT. This leaflet outlines some of the
issues related to CT that should be considered when using a PET / CT scanner.
PET / CT scanners
Current PET / CT systems consist of a single, long-bore gantry with the PET and CT
systems adjacent to one another (Figure 1). In some systems a single housing is
placed over the two gantries, in others the gantries have separate covers, but are
positioned very close to one-another. The CT scanner component of the systems
tend to be existing models that are available as separate systems, with the same
image quality and radiation dose characteristics.
The patient usually undergoes the CT part of the examination first (Figure 1a), and
then the couch is moved further into the gantry to perform the PET scan (Figure 1b).
Figure 1: An example PET / CT system
(a)
(b)
CT
PET
CT issues in PET / CT scanning
1
www.impactscan.org
PET / CT image registration
PET images contain few anatomical
Figure 2: Software registration [14]
landmarks, and are often reviewed in
conjunction with a set of CT images to aid in
locating areas of tracer uptake. In the past
PET and CT images had to be collected on
separate scanners, with the patient having
to move from one scanner to the other,
possibly on different days. Workstations
have been available for some time that
overlay the PET and CT images onto one
another. These attempt to register the
images using anatomical markers to
account for differences in set-up position
between the two scans (Figure 2). This
technique works best for areas of the body
that are immobile, such as the brain. Organs
in other anatomical regions such as the
abdomen are mobile, with scope for them to
change position between the two scans.
This makes the image registration process
more difficult, and possibly less accurate.
The accuracy of software registration can be improved with the use of reference
markers, attached to the surface of the patient. The markers must be made of a
material that is visible on the images from both modalities.
The advent of hybrid PET / CT systems has simplified image registration – the PET
and CT data sets are collected sequentially, on the same system, without the need
for the patient to move to another scanner. This removes the image registration
problems introduced by different patient set-up positions. Once the CT scan is
complete, the patient couch is moved further into the gantry to commence the PET
scan. The two data sets can be considered to be inherently registered; just the
distance between the PET and CT positions needs to be taken into account.
Another registration issue that must be considered is the flex of the patient couch. A
normal CT couch is shown in Figure 3. As the couch is moved into the gantry, more
of the patient’s weight is taken by the part of the couch that is unsupported by the
base. This results in a flex of the couch as it is moved into the gantry. For accurate
image registration it is important that the degree of couch flex does not change as the
patient is moved from the CT to PET acquisition positions. This would cause
registration problems in PET / CT, because when the patient is moved from the CT to
the PET acquisition positions the couch flex will increase, resulting in a lower patient
position for the PET scan compared with the CT part. This difference in vertical
position will vary depending on the weight of the patient.
CT issues in PET / CT scanning
2
www.impactscan.org
Figure 3: A normal CT patient couch
scan plane
Base
Special couch designs can be used to reduce this flexing problem. Figure 4 shows a
re-designed couch for use in PET / CT, where the couch is fixed to a moving
pedestal. The whole pedestal / couch assembly moves into and out of the gantry.
When a patient lays on the couch it will flex a certain amount, but the degree of flex
remains constant regardless of how far the couch is moved into the gantry. This
ensures that the vertical position of the patient is the same for the CT and PET
acquisitions.
Figure 4: A re-designed CT couch for PET / CT
Registration accuracy
Part of the assessment of a PET / CT system should include testing of the accuracy
of image registration between the PET and
CT image data sets. A test object is required Figure 5: Schematic diagram of a
that contains a known 3D distribution of
PET / CT registration phantom
objects, visible in both the CT and PET image
data sets. The two sets of phantom images
can then be registered using the system
software, and the resulting fused image
checked for accuracy.
Figure 5 shows an example of a PET / CT
image registration phantom. The grey box
contains two 68Ge / 68Ga line sources which
are visible to both PET and CT. The source
needles are orientated such that the image
registration can be checked in all three
dimensions.
CT issues in PET / CT scanning
3
www.impactscan.org
CT attenuation correction of PET images
The absorption and scatter of 511 keV PET photons in the patient’s tissue leads to a
reduced count-rate at the detectors. This results in areas of the PET image showing
activity levels that are below the true value. Attenuation maps of the patient can be
used to correct for this effect. Stand-alone PET systems usually achieve this using
radioactive line sources which measure the attenuation through the patient volume
[11]. The acquisition of this data requires many minutes, and considerably adds to
the total PET examination time. The standard deviation of values within these
attenuation maps tends to be high, which can impact upon the accuracy of the
corrected PET images.
CT numbertissue =
µtissue − µwater
× 1000
µwater
Equation 1: CT number
In a PET / CT system, the CT images can be used to generate attenuation correction
maps to apply to PET images. CT image pixel values are given in Hounsfield Units
(HU), which are proportional to the attenuation coefficient (µ) of the tissue being
represented (Equation 1).
Compared with maps acquired from radioactive line sources, CT image maps are
quicker to obtain and have very low noise, reducing the total scan time by up to 30 –
40 % [13] and the likelihood of patient movement. However, some issues specific to
CT must be considered.
Energy dependence of attenuation maps
Figure 6: Mapping CT number to PET
attenuation coefficient
0.18
0.16
0.14
µ PET (cm )
0.12
-1
CT data is gathered using X-rays with
peak energy of around 120 keV, and an
effective mean energy of the order of 70
keV. Attenuation coefficients are energy
dependent, so attenuation maps
derived from CT need to be adjusted
before they can be applied to 511 keV
PET images. Several methods have
been suggested to achieve this [1, 2].
One such method is to provide a
conversion from CT number to PET
attenuation using a bi-linear scale, as
shown in Figure 6.
0.1
0.08
0.06
0.04
0.02
0
-1000
-500
0
500
1000
1500
CT number (HU)
CT image artefacts
In general, image artefacts can be defined as any area where there are systematic
discrepancies between an object and the image of that object. Artefacts can be
thought of as structured image noise. The presence of artefacts in a CT image data
set will result in errors in the generated attenuation map. In turn, applying this
incorrect attenuation map to the PET images will result in an apparent increase or
decrease in activity levels in some areas of the images. It is important, therefore, to
ensure that CT artefacts are minimised when they are used for attenuation maps.
CT issues in PET / CT scanning
4
www.impactscan.org
In addition, PET images are often used for quantification, such as calculating
standard uptake values (SUVs). These calculations rely on the images being a true
representation of radiopharmaceutical distribution within the patient. Any image
artefacts will introduce errors into the calculated values.
CT images are subject to many types of image artefact, some of which are discussed
below.
Metal artefacts
Some materials cannot be correctly represented in a CT image due to their high
attenuation coefficients, including metal items such as dental fillings and joint
replacements. The corresponding Hounsfield Units of these materials are off the CT
number scale of many scanners, which can lead to streaking artefacts through the
images. These streaks will cause errors in any attenuation map that is derived for use
in a corresponding PET data set [3]. Some CT scanners feature metal artefact
reduction algorithms that attempt to minimise these effects. A visual comparison of
attenuation-corrected against non-attenuation-corrected PET images will help to
identify artefacts [4].
Photon starvation
Streak artefacts in CT images can be due to photon starvation. This occurs when
insufficient X-ray photons pass through wide parts of the patient, leading to noisy
projections. When the projections are reconstructed the noise is magnified, resulting
in streaks. This is a particular problem for areas such as the shoulders and hips.
Some scanners use adaptive filtration of the projections to reduce this effect. Where
areas of a projection have low signal they are smoothed, reducing the noise. An
extension of this is multi-dimensional adaptive filtration, where further steps are taken
to reduce noise levels in certain projections [5].
Patient movement
Patient movement during CT scanning results in image artefact, which appear as
streaks or shaded discontinuities across an image. Voluntary motion, such as the
movement of the chest during inspiration and expiration, and involuntary movements
from the heart and peristalsis, can cause these artefacts.
A CT scan is usually short enough for patients to hold their breath, removing the
possibility of breathing artefact. However, PET scans require minutes to acquire, and
so are carried out with the patient shallow breathing. The resulting PET image data
contains some chest motion blurring. This is in contrast to the much more rapidly
acquired CT data, where the chest wall is imaged at full inspiration. Using a breathhold CT scan with a breathing PET scan can cause mis-registration of hot-spots near
the abdomen / chest boundary. This effect can be minimised by allowing the patient
to shallow breathe during the CT scan [6], or by scanning with a normal expiration
breath-hold [7].
CT issues in PET / CT scanning
5
www.impactscan.org
CT contrast agents
Iodine-based CT contrast agents can cause attenuation mismatches [8, 9]. At CT
energies, they have a high contrast with the surrounding tissue, but at 511 keV, the
attenuation coefficient of iodine is similar to that of water. This can lead to apparent
elevated areas of activity in the attenuation corrected PET images. In most cases the
artefacts can be identified by careful observation of the CT and PET images.
Radiation dose
In most PET / CT scanning situations the CT part of the scan does not need to be of
diagnostic quality, as the CT images are just being used to generate attenuation
correction maps. Where this is the case there is a lower CT image quality
requirement, enabling the use of lower CT exposure factors, with a corresponding
drop in patient radiation dose. It is common for the CT scanner to be run at reduced
tube currents, about half the normal diagnostic exposure, of around 70–80 mA.
However, there is potential for artefacts to be introduced as a result, such as the
photon starvation effects mentioned earlier.
Patient radiation doses from CT scans depend on the scan protocol and the
anatomical region being scanned. Using diagnostic exposure factors, effective doses
for head scans are generally in the range 1 to 3 mSv. Abdominal scans have a wider
range, 5 to 20 mSv, depending on the extent of the scan and selected exposure
factors. If the CT scanner is run at a lower mA, in the 70–80 mA range, the
magnitude of these doses will half.
Radiation dose to the patient from the PET component of the scan is around 10 mSv.
For whole-body PET / CT scanning, the CT scan constitutes a significant additional
radiation dose.
ImPACT’s CT Dosimetry spreadsheet is widely used to estimate the radiation doses
to patients from CT. It is available free of charge from our website,
www.impactscan.org/ctdosimetry.htm [10]. The spreadsheet requires the National
Radiological Protection Board’s SR250 Monte Carlo dataset to run [12].
Summary
PET / CT is now well-established clinically, with hundreds of systems installed
worldwide. The use of CT data for attenuation correction of PET images reduces the
total examination time when compared with a stand-alone PET scanner by as much
as 30 – 40 %. However, careful consideration must be given to the CT scan protocols
used in order to minimise the effect of artefacts on the PET images.
Co-registration of CT and PET images on a PET / CT scanner is a simple process
when compared to using anatomical or fiducial markers with separately acquired
scans. The fused images allow functional abnormalities to be accurately located
within the patient anatomy.
Where image artefacts are likely, e.g. where implants are present, care should be
taken to correctly interpret the attenuation-corrected PET images.
CT issues in PET / CT scanning
6
www.impactscan.org
The additional radiation dose burden to the patient as a result of the CT scan should
be considered. Optimisation of CT scan parameters is necessary to minimise patient
radiation dose, but care must be taken to avoid the introduction of additional image
artefacts.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Visvikis D, Costa DC, Croasdale I et al. CT-based attenuation correction in the
calculation of semi-quantitative indices of [18F]FDG uptake in PET. Eur J Nucl
Med 2003; 30: 344-53.
Burger C, Goerres G, Schoenes S et al. PET attenuation coefficients from CT
images: experimental evaluation of the transformation of CT into PET 511-keV
attenuation coefficients. Eur J Nucl Med 2002; 29: 922-7.
Kamel EM, Burger C, Buck A et al. Impact of metallic dental implants on CTbased attenuation correction in a combined PET/CT scanner. Eur Radiol 2003;
13: 724-8.
Goerres GW, Ziegler SI, Burger C et al. Artifacts at PET and PET / CT caused
by metallic hip prosthetic material. Radiology 2003; 226: 557-84.
Kachelriess M, Watzke O, Kalender WA. Generalized multi-dimensional adaptive
filtering for conventional and spiral single-slice, multi-slice, and cone-beam CT.
Med Phys 2001; 28(4): 475-90.
Beyer T, Antoch G, Müller S et al. Acquisition protocol considerations for
combined PET/CT imaging. J Nucl Med 2004; 45 (1 suppl): 25S-35S.
Goerres GW, Burger C, Kamel E et al. Respiration-induced attenuation artifact at
PET / CT: technical considerations. Radiology 2003; 226: 906-10.
Antoch G, Kuehl H, Kanja J et al. Dual-modality PET / CT scanning with
negative oral contrast agent to avoid artifacts: introduction and evaluation’,
Radiology 2004; 230: 879-85.
Dizendorf E, Hany TF, Buck A et al. Cause and magnitude of the error induced
by oral CT contrast agent in CT-based attenuation correction of PET emission
studies. J Nucl Med 2003; 44 (5): 732-8.
ImPACT CT patient dosimetry calculator, www.impactscan.org/ctdosimetry.htm
Ostertag H, Kubler WK, Doll J et al. Measured attenuation correction methods.
Eur J Nucl Med 1989; 15(11):722-6.
www.nrpb.org/publications/software/sr250.htm
Townsend DW, Carney JPJ, Yap JT et al. PET/CT today and tomorrow. J Nucl
Med 2004; 45 (1 suppl): 4S-14S.
Image courtesy of Universitätsklinikum Hamburg-Eppendorf,
www.uke.uni-hamburg.de/imi
CT issues in PET / CT scanning
7
www.impactscan.org
Bibliography
Valk PE, Bailey DL, Townsend DW, Maisey MN. Positron emission tomography.
Basic science and clinical practice. Springer 2003.
Vogel WV, Oyen WJG, Barentsz JO et al. PET/CT: panacea, redundancy, or
something in between? J Nucl Med 2004; 45 (1 suppl): 15S-24S.
Bockisch A, Beyer T, Antoch G et al. Positron emission tomography / computed
tomography – imaging protocols, artifacts, and pitfalls. Mol Imaging Biol 2004; 6 (4):
188-99.
von Schulthess GK, ‘Positron emission tomography versus positron emission
tomography / computed tomography: from “unclear” to “new-clear” medicine. Mol
Imaging Biol 2004; 6 (4): 183-7.
UK PET Special Interest Group, www-pet.umds.ac.uk/UKPET (sic)
ImPACT, St George’s Hospital, Blackshaw Road, London SW17 0QT
T: 020 8725 3366
F: 020 8725 3969
E: [email protected]
W: www.impactscan.org
ImPACT is the UK’s national CT scanner evaluation centre, providing publications, information and advice on all
aspects of CT scanning. Funded by the Medicines and Healthcare products Regulatory Agency (MHRA), it is part
of a comprehensive medical imaging device evaluation programme.
© Crown copyright 2004
CT issues in PET / CT scanning
8
www.impactscan.org