AUTHOR=Lassen Martin L. , Muzik Otto , Beyer Thomas , Hacker Marcus , Ladefoged Claes Nøhr , Cal-González Jacobo , Wadsak Wolfgang , Rausch Ivo , Langer Oliver , Bauer Martin 

TITLE=Reproducibility of Quantitative Brain Imaging Using a PET-Only and a Combined PET/MR System

JOURNAL=Frontiers in Neuroscience

VOLUME=11

YEAR=2017

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2017.00396

DOI=10.3389/fnins.2017.00396

ISSN=1662-453X

ABSTRACT=<p>The purpose of this study was to test the feasibility of migrating a quantitative brain imaging protocol from a positron emission tomography (PET)-only system to an integrated PET/MR system. Potential differences in both absolute radiotracer concentration as well as in the derived kinetic parameters as a function of PET system choice have been investigated. Five healthy volunteers underwent dynamic (R)-[<sup>11</sup>C]verapamil imaging on the same day using a GE-Advance (PET-only) and a Siemens Biograph mMR system (PET/MR). PET-emission data were reconstructed using a transmission-based attenuation correction (AC) map (PET-only), whereas a standard MR-DIXON as well as a low-dose CT AC map was applied to PET/MR emission data. Kinetic modeling based on arterial blood sampling was performed using a 1-tissue-2-rate constant compartment model, yielding kinetic parameters (K<sub>1</sub> and k<sub>2</sub>) and distribution volume (V<sub><italic>T</italic></sub>). Differences for parametric values obtained in the PET-only and the PET/MR systems were analyzed using a 2-way Analysis of Variance (ANOVA). Comparison of DIXON-based AC (PET/MR) with emission data derived from the PET-only system revealed average inter-system differences of −33 ± 14% (<italic>p</italic> &lt; 0.05) for the K<sub>1</sub> parameter and −19 ± 9% (<italic>p</italic> &lt; 0.05) for k<sub>2</sub>. Using a CT-based AC for PET/MR resulted in slightly lower systematic differences of −16 ± 18% for K<sub>1</sub> and −9 ± 10% for k<sub>2</sub>. The average differences in V<sub><italic>T</italic></sub> were −18 ± 10% (<italic>p</italic> &lt; 0.05) for DIXON- and −8 ± 13% for CT-based AC. Significant systematic differences were observed for kinetic parameters derived from emission data obtained from PET/MR and PET-only imaging due to different standard AC methods employed. Therefore, a transfer of imaging protocols from PET-only to PET/MR systems is not straightforward without application of proper correction methods.</p><p><bold>Clinical Trial Registration:</bold><ext-link ext-link-type="uri" xlink:href="http://www.clinicaltrialsregister.eu" xmlns:xlink="http://www.w3.org/1999/xlink">www.clinicaltrialsregister.eu</ext-link>, identifier 2013-001724-19</p>