Praveen Dassanayake1,2, Lumeng Cui3, Elizabeth Finger2,4, Andrea Soddu5,6, Bjoern Jakoby7, Keith St. Lawrence1,2, Gerald Moran8, and Udunna Anazodo1,2
1Department of Medical Biophysics, University of Western Ontario, London, ON, Canada, 2Lawson Health Research Institute, London, ON, Canada, 3Division of Biomedical Engineering, University of Saskatchewan, Saskatoon, SK, Canada, 4Department of Clinical Neurological Sciences, University of Western Ontario, London, ON, Canada, 5Brain and Mind Institute, University of Western Ontario, London, ON, Canada, 6Department of Physics and Astronomy, University of Western Ontario, London, ON, Canada, 7Department of Physics, University of Surrey, Guildford, United Kingdom, 8Siemens Healthineers, Oakville, ON, Canada
Synopsis
Absolute quantification of tracer uptake in positron emission
tomography (PET) requires the knowledge of an arterial input function (AIF) which involves
invasive arterial blood sampling. Alternatively, input functions can be
extracted by identifying feeding arteries from PET images. In this study we
validated a software that uses magnetic resonance images to identify the
feeding arteries in PET images to generate image derived input function (IDIF)
for absolute quantification of PET. The ratio of area under curve between IDIFs
and AIFs revealed that this tool can generate accurate IDIFs for non-invasive
PET quantification.
Introduction
Positron emission tomography (PET) offers an
approach for probing molecular and neurochemical changes in the body with
superior sensitivity. Traditionally, PET tracer distribution in the body is
quantified non-invasively using standardized uptake values (SUVs). Unlike absolute
quantification of tracer uptake, SUVs are semi-quantitative and require a valid
reference region to reduce variability which is not available for some tracers2.
These have necessitated the move from semi-quantitative SUVs to fully
quantification of tracer uptake. However, full quantification requires
measuring the arterial input function (AIF) using invasive serial arterial
blood sampling3. An alternative non-invasive approach would be to
generate image-derived input functions (IDIFs) by defining the feeding arteries
to the tissue from dynamic PET images. The main challenge of using PET only
approach for IDIF is its low spatial resolution, mixing arterial tissue voxels
with surrounding tissue voxels leading to partial volume
errors (PVE)3-5. The combination of PET with magnetic resonance
imaging (MRI) offers the possibility to measure the IDIFs noninvasively, using
MRI anatomical images to help define the feeding arteries 3,5,6. Here
we introduce, an automated parametric Patlak mapping using PET/MRI input
function (CALIPER) tool, developed to extract IDIF by combining PET and MRI and
quantify irreversible tracers (i.e.18F-fluorodeoxyglucose (FDG)) based on the Patlak
graphical approach7. Our objective is to validate this tool against AIFs
determined from serial arterial blood sampling with the future goal of quantifying
reversible PET tracers without the requirement of serial blood sampling. Method
First, we validated our PET/MRI IDIF
approach on a porcine model to evaluate the implemented PVE correction method2
and see how well our IDIF approach can resolve IF from relatively small vessels
in the brain. Then, evaluated the performance of the PET/MRI IDIF approach
healthy human participants. And finally, applied CALIPER to quantify FDG in a
cohort of frontotemporal dementia (FTD) patients, with an effort to improve diagnostic
classification and characterization of FTD. PET and MRI data for all subjects
were acquired on a Biograph mMR (Siemens Healthineers, DE) for 60 minutes
immediately after injection of FDG (5 MBq/kg) at fasting plasma glucose levels
no more than 2-7 mmol/L.Validation
The IDIF were acquired from animal and
human models using CALIPER by extracting the median PET activity from carotid
vessel masks at each timeframe. The vessel mask was defined by a region-growing
segmentation algorithm to identify both carotid arteries visible on either
magnetization prepared rapid gradient echo (MPRAGE) MRI or 3D time of flight
(TOF) MR angiography images registered to PET (figure 1C and D). The IDIF was
corrected for PVE and spill-in contamination by using simulated correction
factors2,4. The IDIFs from 7 pigs were compared against AIFs
acquired from serial blood samples, while the IDIF acquired from 11 healthy
human participants were compared against a population-based AIF (PBAIF)
generated by normalizing the AIFs of 10 healthy controls by the injected dose
and body weight3. Individual PBAIF were generated by calibrating the
PBAIF by 2 late venous samples (45- and 60-min post injection)8,9.
To assess the performance of PET/MRI IDIF, the ratio of area under the
curve (rAUC) between IDIF and AIF/PBAIF were compared as:$$$(AUC IDIF/AUC AIF, or PBAIF)$$$Application
To assess the feasibility of CALIPER in quantifying FDG uptake, we
compared regional changes in parametric maps of the cerebral metabolic rate of
glucose consumption (CMRglc) in 10 FTD patients and 10 age-matched controls7.
Regions of interest analysis in brain regions associated with FTD were compared
between groups using a two-sample t-tests1. Results
The carotid vessel size in the porcine and human
subjects ranged from 4-5 mm and 5-7 mm respectively. The rAUC of pig validation experiments were
overestimated by 8 ± 19 %, as shown in figure 2B . This bias was observed for
both anatomical MR techniques used as a guide to generate IDIFs. For healthy
human controls, IDIFs were in good agreement with PBAIF as shown in figure 3A
and had less bias (6 ± 14 %) compared to the pig model. The rAUC for IDIFs
generated using TOF MRI as a guide was closer to the reported bias in
literature3 (figure 3B). Figure 4, A and B shows the representative
CMRglc map of a healthy control and a patient generated using CALIPER. Figure 4C
indicates regions of the brain implicated in FTD with a significant reduction
in CMRglc in FTD patients compared to healthy controls. Conclusion
Despite the relatively small vessel diameter in the porcine model,
which was close to the 4.3mm full width at half maximum of the PET/MRI10
and increased the susceptibility of the IDIF to partial volume errors, the
measured IDIFs were in good agreement with the true AIFs as shown in figure 2A.
Though TOF anatomical information due to its higher contrast-to-noise ratio is
preferred over MPRAGE for vessel mask delineation for generating the IDIFs, we
demonstrated that CALIPER could generate reliable results using both
modalities. In general, the results demonstrate the potential of CALIPER as a tool
for automated quantification of PET without the need for invasive AIFs. This
opens the door for application of non-invasive PET tracer kinetic modeling
clinically using PET and multiparametric MRI. Acknowledgements
No acknowledgement found.References
- Anazodo U et al. Using simultaneous
PET/MRI to compare the accuracy of diagnosing frontotemporal dementia by
arterial spin labelling MRI and FDG-PET. Neuroimage Clin. 2017 Oct
31;17:405-414.
- Croteau
E et al. Image-derived input function in dynamic human PET/CT: methodology
and validation with 11C-acetate and 18F-fluorothioheptadecanoic acid in muscle
and 18F-fluorodeoxyglucose in brain. Eur J Nucl Med Mol Imaging. 2010
Aug;37(8):1539-50.
- Sundar
LK et al. Towards quantitative [18F]FDG-PET/MRI of the brain: Automated
MR-driven calculation of an image-derived input function for the non-invasive
determination of cerebral glucose metabolic rates. J Cereb Blood Flow Metab.
2019 Aug;39(8):1516-1530.
- Feng ST et al. Image-derived
arterial input function in dynamic positron emission tomography-computed
tomography: a method using both positron emission tomographic and computed
tomographic images. J Comput Assist Tomogr. 2012 Nov-Dec;36(6):762-7.
- Sari
H et al. Estimation of an image derived input function with MR-defined
carotid arteries in FDG-PET human studies using a novel partial volume
correction method. J Cereb Blood Flow Metab. 2017 Apr;37(4):1398-1409.
- Ringheim
A et al. Kinetic modeling of 68Ga-PSMA-11 and validation of simplified
methods for quantification in primary prostate cancer patients. EJNMMI Res.
2020 Feb 24;10(1):12.
- Patlak CS et al. Graphical evaluation
of blood-to-brain transfer constants from multiple-time uptake data. J Cereb
Blood Flow Metab. 1983 Mar;3(1):1-7.
- Tsuchida
T et al. Noninvasive measurement of cerebral metabolic rate of glucose
using standardized input function. J Nucl Med. 1999 Sep;40(9):1441-5.
- Eberl
S et al. Evaluation of two population-based input functions for
quantitative neurological FDG PET studies. Eur J Nucl Med. 1997
Mar;24(3):299-304.
- Delso
G et al. Performance measurements of the Siemens mMR integrated
whole-body PET/MR scanner. J Nucl Med. 2011 Dec;52(12):1914-22.