Simultaneous PET/MRI with 13C magnetic resonance spectroscopic imaging (hyperPET): phantom-based evaluation of PET quantification
© Hansen et al. 2016
Received: 23 June 2015
Accepted: 13 April 2016
Published: 22 April 2016
Integrated PET/MRI with hyperpolarized 13C magnetic resonance spectroscopic imaging (13C-MRSI) offers simultaneous, dual-modality metabolic imaging. A prerequisite for the use of simultaneous imaging is the absence of interference between the two modalities. This has been documented for a clinical whole-body system using simultaneous 1H-MRI and PET but never for 13C-MRSI and PET. Here, the feasibility of simultaneous PET and 13C-MRSI as well as hyperpolarized 13C-MRSI in an integrated whole-body PET/MRI hybrid scanner is evaluated using phantom experiments.
Combined PET and 13C-MRSI phantoms including a NEMA [18F]-FDG phantom, 13C-acetate and 13C-urea sources, and hyperpolarized 13C-pyruvate were imaged repeatedly with PET and/or 13C-MRSI. Measurements evaluated for interference effects included PET activity values in the largest sphere and a background region; total number of PET trues; and 13C-MRSI signal-to-noise ratio (SNR) for urea and acetate phantoms. Differences between measurement conditions were evaluated using t tests.
PET and 13C-MRSI data acquisition could be performed simultaneously without any discernible artifacts. The average difference in PET activity between acquisitions with and without simultaneous 13C-MRSI was 0.83 (largest sphere) and −0.76 % (background). The average difference in net trues was −0.01 %. The average difference in 13C-MRSI SNR between acquisitions with and without simultaneous PET ranged from −2.28 to 1.21 % for all phantoms and measurement conditions. No differences were significant. The system was capable of 13C-MRSI of hyperpolarized 13C-pyruvate.
Simultaneous PET and 13C-MRSI in an integrated whole-body PET/MRI hybrid scanner is feasible. Phantom experiments showed that possible interference effects introduced by acquiring data from the two modalities simultaneously are small and non-significant. Further experiments can now investigate the benefits of simultaneous PET and hyperpolarized 13C-MRI in vivo studies.
Integrated PET/MRI offers the combination of functional imaging of positron emission tomography (PET) and high soft-tissue contrast of anatomic magnetic resonance imaging (MRI) in a simultaneous acquisition. Clinical MRI is for the major part based on 1H nuclei, which are predominantly present in water. 1H-MRI has also functional capabilities, for instance to measure perfusion, characterize tissue cellularity through diffusion-weighted MRI, and study metabolites with magnetic resonance spectroscopy (MRS). The use of MRS imaging (MRSI) is, however, in general challenged by the low sensitivity of in vivo MRI.
Hyperpolarized MRI offers a signal increase by many orders of magnitude. Recently, the feasibility of using a 13C-labeled pyruvate substrate, hyperpolarized through the dynamic nuclear polarization method , for imaging of tumor metabolism in man was demonstrated . Hence metabolic imaging based on 13C-MRSI has come within the reach of routine clinical use.
PET is an established clinical method for metabolic imaging. For example, in oncology, the tracer 18F-2-fluoro-2-deoxy-d-glucose (FDG) gives a measure of glucose metabolism. Since both 13C-MRSI and FDG-PET are metabolic imaging techniques, comparison studies are highly relevant  and have been pursued using sequential 13C-MRSI and PET [4, 5]. In the setting of integrated PET/MRI, simultaneous hyperpolarized 13C-MRSI and PET could become a reality. Such a setup could be very important for generation of new knowledge in physiology and pathophysiology as well as demonstrating potential clinical benefits of 13C-MRSI.
A prerequisite for the use of simultaneous PET and MRI is the absence of interference between the two systems for reviews; see, e.g., [6–8]. Absence of interference has been documented for a clinical whole-body system using simultaneous 1H-MRI and PET  but never for 13C-MRS(I) and PET. Radio frequency (RF) electromagnetic radiation, which is both transmitted and picked up by MRI coils, is a well-known potential source of interference both from MRI to PET and vice versa . PET systems based on solid-state photodetectors have a major part of the electronics placed in close proximity to the MRI RF system. To avoid cross-talk, PET electronics is shielded by a conducting material [10–16]. The shielding material can induce eddy currents and magnetic field inhomogeneities, and the thickness of the material must therefore be minimized ; e.g., for a human brain PET insert, a 10-μm copper shield was used . The magnetic resonance frequency is approximately four times lower for 13C than that for 1H due to the lower gyromagnetic ratio. Shielding efficiency depends on the square root of the frequency of the electromagnetic radiation and can be characterized by a penetration depth. Therefore, shielding optimized for 1H-MRI may not be similarly efficient for 13C-MRSI. Apart from shielding effects, also the magnetic field gradients employed for spatial encoding can potentially introduce noise in the PET measurement system . Due to the lower gyromagnetic ratio, the amplitude of the imaging gradients may be increased. The lower 13C resonance frequency and as well the corresponding need for higher imaging gradients imply that performance measurements for simultaneous 1H-MR(S)I and PET cannot be directly translated to simultaneous 13C-MRSI and PET.
Thus, the aim of the present work is to demonstrate the feasibility of simultaneous PET and 13C-MRSI in an integrated whole-body PET/MRI hybrid scanner. Phantom experiments were carried out to investigate if simultaneous 13C-MRSI and PET acquisition is possible without influencing the PET and 13C-MRSI quantification. Furthermore, we aimed to demonstrate that the system is capable of hyperpolarized 13C-MRSI.
Combined PET and MRI were performed using an integrated PET/MRI system (Siemens Biograph mMR, Erlangen, Germany).
Overview of measurements performed
1, 3, 5, 7, 9, 11, 13, 15
PET, 13C-MRSI acq. ON versus OFF
2, 4, 6, 8, 10, 12, 14, 16
13C-MRSI, PET gantry ON versus OFF; with high activity
13C-MRSI, PET gantry ON versus OFF; with low activity
1H-MRI utilized the built-in RF body coil. 13C-MRI utilized a transmit/receive 12-cm loop coil (RAPID Biomedical, Würzburg, Germany) placed on top of the NEMA phantom. Inside the loop coil were placed two 5.5-mL vials of 4.0 M [13C]urea and 4.0 M [1-13C]acetate doped with gadolinium (Dotarem®) (0.23 %, v/v). The vials were centered vertically above the set of hollow spheres. A cardboard holder above the loop coil was used to place a syringe with hyperpolarized [1-13C]pyruvate.
For later alignment with CT data, a total of six fish oil capsules were placed at various positions at the phantom.
The 13C-MRSI performance was mainly evaluated using the phantom setup shown in Fig. 2. A cylindrical phantom supplied as a part of the PET/MRI system (mMR water, 9.5 L volume) was filled with demineralized water added an aqueous solution of [18F]-FDG (16.2 kBq/mL). The 12-cm loop coil was placed on top of the mMR water phantom, with the urea and acetate vials inside the loop.
PET/MR data acquisition, NEMA phantom
The NEMA phantom (Fig. 1) was placed on the patient table with the set of hollow spheres in the isocenter of the scanner. Only the hardware described above was present.
Following localizer scans, first, a calibration of the 13C MR flip angle was performed using the urea vial. A series of free induction decay measurements with increasing transmitter voltage was acquired (repetition time (TR) 10,000 ms, echo time (TE) 0.35 ms, (nominal) flip angle 90°, pulse duration 0.50 ms, bandwidth 3000 Hz, transmitter voltage 10 to 90 V in steps of 10 V). Chemical shift spectra were centered on the [13C]urea peak. A sinusoidal fit of the [13C]urea peak height versus transmitter voltage was used to determine the relation between transmitter voltage and flip angle.
To evaluate the effect of simultaneous 13C-MRSI on the PET, a total of 16 PET list mode acquisitions were performed (1 bed position). During PET acquisitions with scan numbers 1, 3, 5, 7, 9, 11, 13, and 15, simultaneous 13C-MRSI was carried out; the remaining PET acquisitions were MR silent. The PET acquisition time was 6 min for scan numbers 1–8 and 8 min for scan numbers 9–16, to (partially) compensate for the decay of the tracer. During the first 58 s of each PET acquisition, vendor-supplied Dixon-VIBE MRI intended for MR attenuation correction (MRAC) was performed.
For 13C-MRSI, a chemical shift imaging (CSI) sequence was used (TR 130 ms, TE 2.3 ms, flip angle 29°, bandwidth 4000 Hz, 1024 points in spectrum, field-of-view 200 mm, voxel size 12.5 × 12.5 × 23.0 mm3). The sequence used a 16 × 16 grid and elliptical k-space sampling. Chemical shift spectra were centered on the [13C]urea peak. The CSI sampling time was 19 s and was repeated for the remainder of the PET acquisition following the Dixon MRI.
After the 13C-MRSI and PET acquisitions, a 1H-MRI VIBE sequence (TR 4.18 ms, TE 1.6 ms, flip angle 5.0°, bandwidth 540 Hz/pixel, field-of-view 350 mm, voxel size 1.6 × 1.6 × 3.0 mm3) was obtained as a structural reference.
At the end of the PET/MRI NEMA phantom scan session, a hyperpolarization experiment was carried out. Hyperpolarized 13C-pyruvate was produced using the SpinLab system (GE Healthcare, Milwaukee, WI, USA). The sample consisted of [1-13C]pyruvic acid mixed with an electron paramagnetic agent (sample concentration 15 mM) manufactured by Syncom (Groningen, Netherlands, PN AH111501) under contract from General Electric. The dissolution media applied contained 0.1 g/L ethylenediaminetetraacetic acid disodium salt dehydrate (EDTA disodium salt, Sigma PN E4994) in water, and the neutralization media used to neutralize the hyperpolarized [1-13C]pyruvic acid sample prior to injection contained 0.72 M NaOH, 0.4 M Tris, and 0.1 g/L EDTA disodium salt in water. A 60-mL syringe with hyperpolarized 13C-pyruvate was placed in the abovementioned holder. Imaging took place immediately hereafter using the above CSI sequence with the flip angle reduced to 1°.
CT data acquisition
Following the PET/MR NEMA phantom scan session, a CT (Siemens Biograph mCT) scan of the NEMA phantom including the coil was performed. The phantom was centered as in the PET/MR. The low-dose CT (120 kVp, 3-mm slices) scan was used for CT-based attenuation correction (CTAC).
PET images of the NEMA phantom from both the PET and MR were reconstructed using vendor-supplied 3D OP-OSEM with 4 iterations, 21 subsets, and a 4-mm Gaussian filter. Voxel size was 2.08 × 2.08 × 2.03 mm3 with matrix 344.
For PET/MR CTAC, the low-dose CT image was used. First, the PET/CT patient table was manually removed using OsiriX (v. 4.1.2, Pixmeo, Geneva, Switzerland). CT images were subsequently aligned with the MR VIBE images using Minc (minctracc, McConnell Imaging Center, Montreal), whereupon the Hounsfield units were converted to linear attenuation coefficients using the bi-linear scaling of Carney et al.  as implemented in our PET/CT systems. The patient bed of the PET/MR is automatically included in the AC map by the vendor-supplied algorithm. No additional hardware elements were present during the PET/MR scan session. The first minute of the PET acquisition (during which simultaneous Dixon MRI was performed) was excluded from all reconstructions to study the possible effect of 13C-MRSI only.
PET/MR data acquisition, mMR water phantom
The mMR phantom setup (Fig. 2) was placed on the patient table with the 13C loop coil in the isocenter of the scanner. A calibration of the 13C MR flip angle was performed using the urea vial with the same procedure as above.
To evaluate the effect of the PET data acquisition system on the 13C-MRSI, 13C-MRSI was performed with the PET gantry in an “ON” or “OFF” state. With the PET gantry in the OFF state, all PET electronics inside the Faraday cage of the PET/MRI examination room are switched off. This procedure can be accessed through the service module of the system. A total of 24 13C-MRSI acquisitions were performed. During acquisitions with scan numbers 17–20 and 25–28, the PET gantry was ON; during acquisitions with scan numbers 21–24, the PET gantry was OFF. The ON-OFF-ON design was chosen as a slight temporal drift of the 13C-MRSI urea peak height was noted during preparation of the measurements. The ON-OFF-ON set of measurements was repeated 14 h later when the activity of the phantom had decayed to less than 1 % of its initial value. Please refer to Table 1 for an overview of measurements.
For 13C-MRSI, the above CSI sequence was used again (one repetition only). A 1H-MRI VIBE sequence with parameters as above was obtained as a structural reference.
Evaluation of PET quantification
We report the average PET activity value (kBq/mL) of a spherical VOI with a diameter of 20 mm, centrally placed in the largest sphere (with an inside diameter of 37 mm). Background activity was measured using a spherical VOI with a diameter of 60 mm, placed in the center of the phantom. PET activity was decay corrected to the time of the filling of the phantom. In addition, the total number of trues for each PET acquisition on the PET/MR system was obtained from the list mode files (total prompts subtracted total randoms) and corrected for decay and the differences in acquisition times.
Percent differences between PET acquisitions obtained with 13C-MRSI active (ON) or not (OFF) were calculated using the 13C-MRSI OFF condition as the reference. Differences were evaluated using a two-sample t test.
Evaluation of 13C-MRSI quantification
Peak heights of [1-13C]acetate, [13C]urea, and [1-13C]pyruvate (hyperpolarized experiment only) were quantified for every MRSI voxel using a general linear model implemented in MATLAB (MathWorks, Natick, MA, USA). The estimation was performed in the time domain. Metabolite maps were interpolated on a 128 × 128 grid for visual presentation.
The 13C-MRSI noise was derived from the background part of the 13C spectrum (from −2000 to −1000 Hz and from 1000 to 2000 Hz) in voxel (1,1) at the corner of the 13C-MRSI grid, far from the 13C phantoms, which was verified to have no signal. The noise was estimated as the standard deviation of the real part of the signal, multiplied by the square root of the number of repetitions of the MRSI. 13C-MRSI signal-to-noise ratio (SNR) was estimated for urea and acetate using the voxel with maximum signal. The signal was defined as the integrated peak area  in an interval of ±250 Hz around the peak.
Percent differences of SNR between 13C-MRSI acquisitions for the mMR water phantom obtained with PET gantry “ON” or “OFF” were calculated using the OFF condition as the reference. Differences were evaluated using a two-sample t test. Differences were evaluated both with and without high activity in the gantry (see Table 1).
Data are reported with 95 % confidence intervals (CI), and a p value of 0.05 was considered significant.
On a general note, PET and 13C-MRSI data acquisition could be performed simultaneously without any practical issues.
Summary of PET measurements
13C-MRSI ON (n = 8)
13C-MRSI OFF (n = 8)
PET activity, largest sphere
37.62 ± 0.49 kBq/mL
37.31 ± 0.38 kBq/mL
PET activity, background
6.02 ± 0.04 kBq/mL
6.06 ± 0.08 kBq/mL
208,943 ± 723
208,959 ± 747
The presented phantom data demonstrates that simultaneous PET and 13C-MRSI in a whole-body PET/MRI scanner is possible. Interference effects introduced by acquiring data from the two modalities simultaneously are small and non-significant. Hence, full advantage can be taken of the integrated PET/MRI scanner design. Furthermore, the integrated PET/MRI system was proven capable of hyperpolarized 13C-MRSI.
A simultaneous acquisition of PET and 13C-MRSI will reduce total examination time and minimize subject motion between the imaging modalities, which is likely to be beneficial for in vivo studies. Also, functional imaging can be performed without a delay between the modalities, thus ensuring that the physiologic state of the subject is unchanged. This is particularly desirable when 13C-MRSI is combined with PET tracers using a short uptake time and in cases where a pharmacological challenge or another intervention is employed.
The PET results presented here using a standard quality assurance phantom shows no effect of simultaneous 13C-MRSI (Figs. 3 and 4). The statistical analysis of our repeated measurements (Fig. 5) allows us to conclude that the maximum possible quantitative PET estimation error due to interference from the 13C-MRSI is below 2 % (95 % CI). That can be considered a sufficiently low error for clinical and research purposes.
The CSI sequence used to test interference with PET is a typical type of sequence for hyperpolarized MRSI. The RF flip angle was set to 29°, corresponding to the Ernst angle for a typical tissue longitudinal MR relaxation time (T 1) of 1000 ms, giving a maximum signal for a thermal polarization. In case of hyperpolarized MRI, most often a lower flip angle is chosen (e.g., 10° in Ref. ) to preserve the hyperpolarization as long as possible. Hence in practice, possible interference effects caused by RF radiation will be smaller than in the present experiment.
When MRI is not performed, no RF radiation is transmitted and the magnetic field is constant. Conversely, there is no fundamental change of state of the PET detector system between situations where imaging data is collected or not. A possible RF interference from the PET detector system to the 13C-MRSI would depend on the PET gantry being powered on or off. However, no effect of power state is observed (Figs. 6, 7, and 8). The maximum 13C-MRSI SNR error due to PET gantry power state is below −6 to +4 % (95 % CI) for all measurements. A tendency to temporal drift in the measurements, seemingly unrelated to the power state (see Fig. 6b), could have inflated the error estimation. The ON-OFF-ON design of the experiment will compensate for a linear trend of system instability only. Overall though, a stability of a few percent on 13C-MRSI SNR in a single 19-s acquisition as used here can be considered satisfactory.
The present study was not designed to investigate a hypothetical overall (i.e., PET acquisition and activity independent) degradation of the 13C-MRSI due to the PET system. For example, the presence of the PET detector system could degrade the static magnetic field inhomogeneity; however, this was shown not to be the case for the PET/MRI system used here by Delso et al. . The focus of the present work was on RF effects which depend on the gyromagnetic ratio and hence potentially differ from 1H- to 13C-based imaging.
Importantly, the 13C-MRSI quality in this integrated PET/MRI setup was sufficient to perform metabolic imaging using hyperpolarization in a phantom setting (Fig. 9) and also recently for in vivo experiments [19, 20].
We noticed an unexpected non-uniform activity in the background region of the NEMA phantom (Fig. 4b). The non-uniformity could originate from suboptimal attenuation or scatter correction of the setup and deserves further investigation. However, the profile appeared to be independent of the 13C-MRSI acquisition. Further challenges specific to simultaneous PET and 13C-MRSI include characterization and correction for photon attenuation due to 13C coils. Preliminary investigations at our institution indicate that this effect can change PET activity values with up to 10 %, however, without any change in visual interpretation of images.
The present work demonstrated the feasibility of simultaneous PET and 13C-MRSI in an integrated whole-body PET/MRI hybrid scanner. Phantom experiments showed that simultaneous 13C-MRSI and FDG-PET acquisition is possible without influencing the PET quantification. Furthermore, the system is capable of hyperpolarized 13C-MRSI. Hereby, the way is paved for simultaneous PET and hyperpolarized 13C-MRI in vivo studies.
Ralf Laadebeck (Siemens Healthcare) is gratefully acknowledged for technical assistance. The financial support from the John and Birthe Meyer Foundation, the Capital Region of Denmark, the Novo Nordisk Foundation, the Lundbeck Foundation, the Innovation Fund Denmark, the Research Council for Independent Research, and the Svend Andersen Foundation is gratefully acknowledged.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Ardenkjaer-Larsen JH, Fridlund B, Gram A, Hansson G, Hansson L, Lerche MH, Servin R, Thaning M, Golman K. Increase in signal-to-noise ratio of >10,000 times in liquid-state NMR. Proc Natl Acad Sci U S A. 2003;100:10158–63.View ArticlePubMedPubMed CentralGoogle Scholar
- Nelson S, Kurhanewicz J, Vigneron D, Larson P, Harzstark A, Ferrone M, van Criekinge M, Chang J, Bok R, Park I, Reed G, Carvajal L, Small E, Munster P, Weinberg V, Ardenkjaer-Larsen J, Chen A, Hurd R, Odegardstuen L, Robb F, Tropp J, Murray J. Metabolic imaging of patients with prostate cancer using hyperpolarized [1-13C]pyruvate. Sci Transl Med. 2013;5:198ra108.PubMedPubMed CentralGoogle Scholar
- Gallagher FA, Bohndiek SE, Kettunen MI, Lewis DY, Soloviev D, Brindle KM. Hyperpolarized 13C MRI and PET: in vivo tumor biochemistry. J Nucl Med. 2011;52:1333–6.View ArticlePubMedGoogle Scholar
- Witney TH, Kettunen MI, Day SE, Hu D, Neves AA, Gallagher FA, Fulton SM, Brindle KM. A comparison between radiolabeled fluorodeoxyglucose uptake and hyperpolarized (13)C-labeled pyruvate utilization as methods for detecting tumor response to treatment. Neoplasia. 2009;11:574–82. 1 p following 582.View ArticlePubMedPubMed CentralGoogle Scholar
- Menzel MI, Farrell EV, Janich MA, Khegai O, Wiesinger F, Nekolla S, Otto AM, Haase A, Schulte RF, Schwaiger M. Multimodal assessment of in vivo metabolism with hyperpolarized [1-13C]MR spectroscopy and 18F-FDG PET imaging in hepatocellular carcinoma tumor-bearing rats. J Nucl Med. 2013;54:1113–9.View ArticlePubMedGoogle Scholar
- Wehrl HF, Judenhofer MS, Wiehr S, Pichler BJ. Pre-clinical PET/MR: technological advances and new perspectives in biomedical research. Eur J Nucl Med Mol Imaging. 2009;36 Suppl 1:S56–68.View ArticlePubMedGoogle Scholar
- Vandenberghe S, Marsden PK. PET-MRI: a review of challenges and solutions in the development of integrated multimodality imaging. Phys Med Biol. 2015;60:R115–54.View ArticlePubMedGoogle Scholar
- Delso G, Ziegler S. PET/MRI system design. Eur J Nucl Med Mol Imaging. 2009;36:86–92.View ArticleGoogle Scholar
- Delso G, Fürst S, Jakoby B, Ladebeck R, Ganter C, Nekolla SG, Schwaiger M, Ziegler SI. Performance measurements of the Siemens mMR integrated whole-body PET/MR scanner. J Nucl Med. 2011;52:1914–22.View ArticlePubMedGoogle Scholar
- Catana C, Wu Y, Judenhofer MS, Qi J, Pichler BJ, Cherry SR. Simultaneous acquisition of multislice PET and MR images: initial results with a MR-compatible PET scanner. J Nucl Med. 2006;47:1968–76.PubMedGoogle Scholar
- Pichler BJ, Judenhofer MS, Catana C, Walton JH, Kneilling M, Nutt RE, Siegel SB, Claussen CD, Cherry SR. Performance test of an LSO-APD detector in a 7-T MRI scanner for simultaneous PET/MRI. J Nucl Med. 2006;47:639–47.PubMedGoogle Scholar
- Judenhofer MS, Wehrl HF, Newport DF, Catana C, Siegel SB, Becker M, Thielscher A, Kneilling M, Lichy MP, Eichner M, Klingel K, Reischl G, Widmaier S, Röcken M, Nutt RE, Machulla H-J, Uludag K, Cherry SR, Claussen CD, Pichler BJ. Simultaneous PET-MRI: a new approach for functional and morphological imaging. Nat Med. 2008;14:459–65.View ArticlePubMedGoogle Scholar
- Catana C, Procissi D, Wu Y, Judenhofer MS, Qi J, Pichler BJ, Jacobs RE, Cherry SR. Simultaneous in vivo positron emission tomography and magnetic resonance imaging. Proc Natl Acad Sci U S A. 2008;105:3705–10.View ArticlePubMedPubMed CentralGoogle Scholar
- Wehrl HF, Judenhofer MS, Thielscher A, Martirosian P, Schick F, Pichler BJ. Assessment of MR compatibility of a PET insert developed for simultaneous multiparametric PET/MR imaging on an animal system operating at 7 T. Magn Reson Med. 2011;65:269–79.View ArticlePubMedPubMed CentralGoogle Scholar
- Kolb A, Wehrl HF, Hofmann M, Judenhofer MS, Eriksson L, Ladebeck R, Lichy MP, Byars L, Michel C, Schlemmer H-P, Schmand M, Claussen CD, Sossi V, Pichler BJ. Technical performance evaluation of a human brain PET/MRI system. Eur Radiol. 2012;22:1776–88.View ArticlePubMedGoogle Scholar
- Wehner J, Weissler B, Dueppenbecker PM, Gebhardt P, Goldschmidt B, Schug D, Kiessling F, Schulz V. MR-compatibility assessment of the first preclinical PET-MRI insert equipped with digital silicon photomultipliers. Phys Med Biol. 2015;60:2231–55.View ArticlePubMedGoogle Scholar
- Carney JPJ, Townsend DW, Rappoport V, Bendriem B. Method for transforming CT images for attenuation correction in PET/CT imaging. Med Phys. 2006;33:976.View ArticlePubMedGoogle Scholar
- Drost DJ, Riddle WR, Clarke GD. Proton magnetic resonance spectroscopy in the brain: report of AAPM MR task group #9. Med Phys. 2002;29:2177–97.View ArticlePubMedGoogle Scholar
- Gutte H, Hansen AE, Henriksen ST, Johannesen HH, Ardenkjaer-Larsen J, Vignaud A, Hansen AE, Børresen B, Klausen TL, Wittekind AN, Gillings N, Kristensen AT, Clemmensen A, Højgaard L, Kjær A. Simultaneous hyperpolarized (13)C-pyruvate MRI and (18)F-FDG-PET in cancer (hyperPET): feasibility of a new imaging concept using a clinical PET/MRI scanner. Am J Nucl Med Mol Imaging. 2015;5:38–45.PubMedPubMed CentralGoogle Scholar
- Gutte H, Hansen AE, Larsen M, Rahbek S, Henriksen S, Johannesen H, Ardenkjær-Larsen J, Kristensen A, Højgaard L, Kjær A. Simultaneous hyperpolarized 13C-pyruvate MRI and 18 F-FDG-PET (hyperPET) in 10 dogs with cancer. J Nucl Med. 2015;56(11):1786–92.View ArticlePubMedGoogle Scholar