ZurichOpenRepositoryand
Archive
UniversityofZurich
UniversityLibrary
Strickhofstrasse39
CH-8057Zurich
www.zora.uzh.ch
Year:2013
PET-MRimagingusingatri-modalityPET/CT-MRsystemwitha
dedicatedshuttleinclinicalroutine
Veit-Haibach,Patrick;Kuhn,FelixPierre;Wiesinger,Florian;Delso,Gaspar;vonSchulthess,Gustav
Abstract:Tri-modalityPET/CT-MRIincludesthetransferofthepatientonadedicatedshuttlefromone
systemintotheother.AdvantagesofthissystemincludeatrueCT-basedattenuationcorrection,reliable
PET-quanticationandhigherexibilityinpatientthroughputonbothsystems.Comparativestudiesof
PET/MRIversusPET/CTarereadilyaccomplishedwithoutrepeatedPETwithadierentPETscanner
atadierenttimepoint.Additionally,thereisahigherimagingexibilitybasedontheavailabilityof
threeimagingmodalities,whichcanbecombinedforthecharacterizationofthedisease.Thedownsideis
asomewhathigherradiationdoseofupto3mSvwithalowdoseCTbasedontheCT-component,longer
acquisitiontimesandpotentialmisalignmentbetweentheimagingcomponents.Overall,thetri-modality
PET/CT-MRsystemoerscomparativestudiesusingthethreedierentimagingmodalitiesinthesame
patientvirtuallyatthesametime,andmayhelptodevelopreliableattenuationalgorithmsatthesame
time.
DOI:https://doi.org/10.1007/s10334-012-0344-5
PostedattheZurichOpenRepositoryandArchive,UniversityofZurich
ZORAURL:https://doi.org/10.5167/uzh-71512
JournalArticle
PublishedVersion
Originallypublishedat:
Veit-Haibach,Patrick;Kuhn,FelixPierre; Wiesinger,Florian;Delso,Gaspar;vonSchulthess,Gustav
(2013).PET-MRimagingusingatri-modalityPET/CT-MRsystemwithadedicatedshuttleinclinical
routine.Magma,26(1):25-35.
DOI:https://doi.org/10.1007/s10334-012-0344-5
REVIEW ARTICLE
PET–MR imaging using a tri-modality PET/CT–MR system
with a dedicated shuttle in clinical routine
Patrick Veit-Haibach
•
Felix Pierre Kuhn
•
Florian Wiesinger
•
Gaspar Delso
•
Gustav von Schulthess
Received: 17 April 2012 / Revised: 5 September 2012 / Accepted: 5 September 2012 / Published online: 9 October 2012
Ó ESMRMB 2012
Abstract Tri-modality PET/CT–MRI includes the transfer
of the patient on a dedicated shuttle from one system into
the other. Advantages of this system include a true
CT-based attenuation correction, reliable PET-quantification
and higher flexibility in patient throughput on both sys-
tems. Comparative studies of PET/MRI versus PET/CT are
readily accomplished without repeated PET with a differ-
ent PET scanner at a different time point. Additionally,
there is a higher imaging flexibility based on the avail-
ability of three imaging modalities, which can be combined
for the characterization of the disease. The downside is a
somewhat higher radiation dose of up to 3 mSv with a low
dose CT based on the CT-component, longer acquisition
times and potential misalignment between the imaging
components. Overall, the tri-modality PET/CT–MR system
offers comparative studies using the three different imag-
ing modalities in the same patient virtually at the same
time, and may help to develop reliable attenuation algo-
rithms at the same time.
Keywords PET/CT PET/MRI PET/CT–MRI
Multi-modality imaging Prot ocol evaluation
Introduction of the specific concept of PET/ CT–MRI
Integrated positron emission tomography (PET)/computed
tomography (CT) has evolved into one of the major
imaging procedures in oncology imaging and partly also in
infection imaging.
However, PET/CT has several—technical as well as
diagnostic—limitations. Thus, there is increasing interest
in integrating PET with other imaging procedures like
magnetic resonance imaging (MRI) [
1–3]. Superior soft
tissue contrast of MRI compared to CT and lower radiation
exposure are the most obvious advantages. However, there
are several technical and clinical challenges that have to be
resolved before CT can be replaced with MRI in such a
multimodal system. The current approach used in our
hospital is a sequential PET/CT-MRI system composed of
two major components: a 3T MRI system (Discovery 750w
3T, GE Healthcare, Waukesha, WI, USA) and a state-
of-the-art TOF PET/CT (Time of flight, Discovery 690, GE
Healthcare, Waukesha, WI, USA). Both systems can be
connected with a dedicated shuttle system (see ‘‘Technical
considerations,’’ Fig.
1a, b) which is based on a transfer-
able board mounted onto a mobile shuttle system that links
with the PET/CT and MRI from one side of the patient
handling system via a floor-mounted docking station. The
patient is first positioned on the transfer board on the fixed
MRI table and subsequently, the MRI is performed. In
order to limit the overall examination time for the patient,
the MRI is performed during the uptake period of the PET
tracer prior to the PET/CT examination. Following the
acquisition of the MRI, the patient is transferred on the
board to the shuttle system and redirected to the PET/CT.
The shuttle system docks to the floor-mounted rails and the
board is fitted on top of the PET/CT patient bed. Once
the patient is moved and positioned inside the gantry, the
P. Veit-Haibach (&) F. P. Kuhn G. Delso
G. von Schulthess
Department of Medical Imaging, University Hospital,
Ra
¨
mistrasse 100, 8091 Zurich, Switzerland
e-mail: patrick.veit-haibach@usz.ch
F. Wiesinger
GE Healthcare, Garching, Germany
G. Delso
GE Healthcare, Waukesha, WI, USA
123
Magn Reson Mater Phy (2013) 26:25–35
DOI 10.1007/s10334-012-0344-5
PET/CT examination commences. The PET/CT study is
performed with CT-based AC for PET-corrections in
clinical routine. The acquired PET/CT and the MRI are
retrospectively coregistered on a comme rcially available
imaging workstation (Advantage Windows, GE Health-
care). Images can be displayed as PET/CT, PET–MRI,
CT-only or PET-only. We always coregister the PET/CT
with the MRI data. The MRI examination is typically
performed as a whole-body examination due to current
research trials comparing PET/CT and PET–MRI, or as a
partial-body contrast-enhanced examination in the area of
interest (head/neck, abdomen, pelvis, brain, see below for a
more detailed discussion).
This paper will highlight some of the major issues and/
or advantages of introducing this new hybrid technology
into clinical routine, especially the advantages and disad-
vantages of simultaneous versus sequential PET/CT–MRI
and PET/MRI, the related technical issues, as well as
several research aspects which might arise in PET/CT–MR
and PET/MR.
Advantages and disadvantages of sequential
PET/CT–MR and simultaneous PET/MR
Several issues have to be addressed when evaluating a
sequential PET/CT–MR versus the simultaneous PET/MRI
approach. In our opinion, at the present state of the
development of PET/MR technology, the advantages of a
sequential PET/CT–MRI system outweigh the disadvan-
tages. Currently there are only very few data available in
the literature which demonstrate a potential advantage of
PET/MRI over PET/CT in routine clinical oncological
applications [
1, 4–7]. So far, the majority of available
data does not find statistically significant advantages of
PET/MRI versus PET/CT concerning staging accuracy or
detection of distant metastases [
8–11]. One clear advantage
of a sequen tial system is the flexibility in a routine clinical
environment. Since both scanners can be operated sepa-
rately, clinical PET/CT and MRI can be acquired at the
same time in different patients. With the dedicated shuttle
system, the patient can be shuttled back and forth between
both systems. An overview over possible protocols in a
clinical setting is given in Fig.
2.
Another key issue is how to achieve MRI-based atten-
uation correction (AC) and how to do adequate lung
imaging in MR. Currently there are different approaches on
how to perform MRI-based AC, e.g., template-based, atlas-
based/pseudo-CT [
12]. The first results of MR-based AC of
PET-data using ultrashort echo time (UTE) sequences are
promising, but to date are very time consuming, especially
in large FOVs [
13]. Another drawback in MR imaging is
the difficulty of lesion detection in body regions presenting
with large susceptibility changes, especially in the lung.
Although MRI can be performed with gated sequences,
those typically increase the acquisition time. Initial results
with fast breath hold GRE-sequences may offer sufficient
results in several cases [
14, 15].
The development of an appropriate workflow for a PET/
CT–MRI system as well as for a simultaneous PET/MRI
remains challenging. Typically, whole-body PET/CT can
be done in 10–15 min while a diagnostic MR of even a
limited body region typically takes 25 min or more. Thus, a
possible approach to an adequate and timely workflow
might be similar to that of PET/CT, where a whole-body,
low-dose scan is followed by a regionally focused full
diagnostic CT. In PET/MRI, such a protocol would consist
of a quick wholebody overview (e.g., possibly with a
Dixon T1 GRE sequence) and a dedicated contrast-
enhanced MRI in the region of interest. However, a major
task for the future will remain the development of faster
and more robust sequences to be used in PET/CT–MRI and
combined PET/MRI.
Radiation issues have to be addressed when discussing the
sequential PET/CT–MR solution. In principle, there is cur-
rently a clear advantage concerning radiation dose in
simultaneous PET/MR systems because there is no addi-
tional CT-component. However, as long as the attenuation
correction issues are not solved, this argument is moot
[
16–18]. Furthermore, much effort is applied to dose
reduction in CT and PET/CT. Model based iterative recon-
struction algorithms (MBIR) are already available in current
commercially available scanners and upcoming advanced
algorithms will reduce the CT dose even further. Thus, there is
considerable potential that the radiation dose from PET/CT–
MRI will be reduced by several factors in the near future.
One clear disadvantage of a sequential PET/CT–MR
solution is the total scanning time for the patients because,
no matter how short the MRI protocol might be, it is
always an add-on to the PET/CT-examination. Although
the overall patient stay ti me of PET/CT–MR might be the
same in case the MRI is done during uptake time, patients
might consider the ‘‘double-examination’’ as an extra
burden. On the other hand, only the sequential solution
currently offers the possibility to scan patients in a wide-
bore MRI, which is more comfortable for claustrophobic
and overweight patients. The lack of simultaneity is
another disadvantage. Simultaneity can be important in
interesting research applications such as dual neuro-acti-
vation or mapping studies or cardiac perfusion validation
studies. Since nuclear imaging is usually done in a pseudo-
steady-state tracer distribution, the meaning of simultaneity
is currently unclear in such measurements. Although this
approach (in neuro-imaging) currently does not have a
relevant application in clinical routine, this might change
with the installation of more simultaneous systems.
26 Magn Reson Mater Phy (2013) 26:25–35
123
Fig. 1 a Top row Side-loading
shuttle system in frontal view
(left) and top view (right)
consisting of a metal trolley
with counter balance weights
(60 kg) on each side (colored in
green) and two arms (colored in
blue) holding the glass fiber
board (colored in red). The
patient board can flexibly be slid
either to the right or the left of
the shuttle system, which
permits loading the patient onto
a scanner table from either side.
Bottom row 3D schematic of the
axial, front-loading patient
transporter docked to the
PET/CT table (left). When
connected, the rail-guided
transfer board (right) can easily
be slid between the shuttle and
the scanner. b Current design
concept with a 3T MRI system
and a TOF-PET/CT in two
rooms directly adjacent to each
other. The explanation of the
shuttle mechanism and
workflow is given in the
‘‘
Introduction’’
Magn Reson Mater Phy (2013) 26:25–35 27
123
A significant disadvantage of current simultaneous
PET/MRI in clinical routine is in MR lung imaging. Here, CT
is still the best modality for nodule detection, evaluation of
parenchymal texture, speed and robustness. Due to signif-
icantly longer imaging time, even with breath hold GRE-
sequences, MRI has never matched the detection rates of
CT. In a recent study from our group there is still a higher
lesion detection rate in CT based on a lesion-by-lesion
evaluation when compared to breath hold GRE Lava Flex
T1 imaging, even when using the low-dose, limited axial
resolution CT-component from routine PET/CT with tidal
breathing [
19]. However, in patient-based analysis, no
statistically significant difference was seen in this pre-
liminary study.
The current costs for simultaneous and sequential PET/
MRI systems (one-room solutions) are approximately 5
millions euros (in Europe). The costs for the tri-modality
system are comprised of the prize for a PET/CT, a MRI and
the dedicated shuttle solutions. Th us, depending on the
chosen components, we estimate that the total cost will be
between €3.5–5 million, depending on the system config-
uration. Those costs do not include (in any case) the costs
for staffing, floor space and service. The costs for the
procedure are comparable as well, of course depending on
the local circumstances (supply costs, reimbursement).
Technical considerations: patient transporter system
Multimodal imaging using separate, stand-alone PET/CT
and MR scanners is based on the critical assumption of
accurate multimodal image registration. Patient shuttle
systems have been developed to avoid patient repositioning
and minimize associated image misregistration, as well as
to improve the overall workflow. Such a patient transporter
system virtually connects the independent PET/CT and MR
scanners into a single, sequential tri-modality PET/CT–MR
imaging platform.
Conceptually, overall image misregistration can be
decomposed into hardware and patient-induced errors. The
former depends on the actual implementation of the shuttle
system and its mechanical tolerances. Because of the
rigidity of the patient shuttle, hardware registration errors
can, to a good approximation, be described via rigid
transformations. In comparison, patient-induced misregis-
tration is of non-rigid nature, is dependent on patient
Fig. 2 Overview over possible PET/CT–MRI protocols. For whole-
body (WB) indications, PET/CT with additional basic whole body
MRI might be applied (grey column). In more specific indications or
diseases, contrast-enhanced MRI confined to the area of interest might
be applied (partial body, PB) (blue column). For example, PET/CT
can be done as a whole-body examination and the following (full
diagnostic) MRI can be acquired, confined to the area of interest.
Alternatively, the MRI can be acquired first in the desired area of
interest (or whole body) during the uptake times. This concept holds
the advantage of not prolonging the total examination time since the
patient has to wait anyway (at least for most 18F tracers) for
approximately 1 h during the uptake period prior to PET. Imaging
times given in the figure are scan acquisition times and may vary
based on the medical status of the patient and ability to cooperate. The
protocols given here are examples and are not a complete overview,
as there are deviations of those protocols depending on the indication
and clinical question
28 Magn Reson Mater Phy (2013) 26:25–35
123
