WorksPhysics and Astronomy
Quantitative Imaging of Scandium Isotopes with a Long Axial Field-of-View PET/CT
Abstract
Objective. Sc and Sc are positron-emitting isotopes of scandium that have attracted growing interest for theranostic applications in conjunction with the therapeutic isotope Sc. Both isotopes are non-pure positron emitters, and the associated prompt-gamma emission together with the high positron energy introduce complexities in achieving precise quantitative imaging. Approach. We evaluated the quantitative imaging performance of Sc and a Sc mixture on a long axial field-of-view (LAFOV) PET/CT using a NEMA IEC body phantom. We assessed contrast recovery coefficients, image noise and bias for different numbers of iterations and for absolute versus relative scatter correction, and implemented an isotope-specific point-spread-function (PSF) correction based on FLUKA-simulated positron ranges. Main results. The quantitative imaging performance of both isotopes is comparable to F, and the vendor's prompt-gamma correction effectively suppresses the additional random coincidences and lung error caused by prompt-gamma emission. The isotope-specific PSF correction further improves contrast recovery. While the bias obtained for Sc meets the EARL acceptance criterion for non-standard radionuclides, the bias for the Sc mixture exceeds it, which we attribute to the mixture-specific cross-calibration of the dose calibrators rather than to the PET imaging. Significance. Our results support the use of LAFOV PET/CT for quantitative imaging of scandium isotopes and provide practical reconstruction recommendations for clinical protocols.
Introduction
The success story of theranostic applications of - and -labeled prostate-specific membrane antigen (PSMA) ligands and DOTA-conjugated peptides [1, 2] has fostered significant efforts to extend the theranostic concept to novel radionuclide pairs that offer a close chemical match between the diagnostic and therapeutic radionuclides. Scandium isotopes have emerged as ideal theranostic radionuclides with a chemically more coherent alternative to - and -labeled tracers [3, 4, 5, 6, 7]. In addition, the production of scandium isotopes can build upon the existing infrastructure of medical cyclotrons [8, 9, 10, 11] and there are even generator systems being developed for [12, 7]. In the context of diagnostics and treatment planning with PSMA, it is highly desirable to have positron emission tomography (PET) radionuclides with a longer half-life than the standard and in order to keep good count statistics while imaging multiple or late time points of small lesions [13].
Both PET radioisotopes of scandium, i.e. Sc and Sc, are non-pure positron emitters [14]. This means that the decay radiation contains additional prompt photons. While such prompt photons (p) might be useful for special applications [15, 16, 17], they can lead to a higher random coincidence rate and, thereby, deteriorate the PET image [14, 18, 19, 20]. Moreover, the positrons from scandium decay have a relatively large energy and the p can lead to a higher radiation dose burden for the patient [21] compared to F. Exploiting the aforementioned favorable properties of scandium therefore depends on the development and validation of reliable correction methods for the p contribution.
The isotopes Sc and Sc have already been imaged with clinical PET/CT systems [22, 23, 6], demonstrating their potential as positron emitters for diagnostic nuclear medicine. Although, first-in-human studies were reported in the literature [24, 25, 26] some time ago, scandium radioisotopes are not yet available in clinical routine. Building on these earlier works, the aim of the present study is two-fold. Firstly, we aim to extend the evaluation of scandium isotopes to long axial field-of-view (LAFOV) PET/CT [27, 28, 29, 30, 20] and to short/low-activity acquisitions. Secondly, we aim to investigate the imaging properties of Sc in more detail.
Methods
Enriched calcium targets were irradiated at the IP2 irradiation station of the Paul Scherrer Institute (PSI, Villigen-PSI, Switzerland) and chemically processed to produce pure Sc as well as a mixture of Sc and Sc [31, 8]. Despite the enrichment of Ca in the target, it is not feasible to produce isotopically pure Sc via the Ca(p,2n) nuclear reaction [32]. About 240 MBq of Sc and 334 MBq of Sc were sent from PSI to the Department of Nuclear Medicine at Inselspital (Bern, Switzerland) where a NEMA IEC body phantom (Data Spectrum Corporation, Durham, NC, USA) was filled according to the EARL standard 2 protocol [33, 34, 35]. The data were acquired with a Biograph Vision Quadra (Siemens Medical Solutions USA Inc, Knoxville, TN, USA) LAFOV PET/CT scanner. All image reconstructions were performed offline with the prototype software e7-tools (Siemens Medical Solutions USA Inc, Knoxville, TN, USA).
The dose calibrator at Inselspital (VDC-405/VIK-202, Comecer, Italy) was simply cross-calibrated with the dose calibrator from PSI [36]. No calibration factor from the vendor nor from the Swiss metrology institute was provided. For the case of Sc this is particularly challenging as the definition of Bq from the activity measurement is somewhat ambiguous when compared to the Bq mL that are measured in the PET/CT scan. We therefore applied the following procedure. At PSI a 0.999 MBq aliquot, as measured in PSI's dose calibrator, of the Sc solution was measured in an HPGe spectrometer. At reference time, the HPGe measurement identified 0.310 MBq Sc, 0.010 MBq Sc and 0.864 MBq Sc, which exceeds the total activity measured by the dose calibrator by 18.5%. In order to determine the syringe activity, i.e. the activity in the stock solution for the phantom filling, a scaling factor between PSI's and Insel's dose calibrators was computed and the (decay-corrected) isotope inventory from the HPGe measurement was rescaled accordingly (the half-life of Sc is h [37] and h for Sc [38]). After removing the Sc activity from the inventory (there is hardly any positron emission), the sum of the Sc and Sc activities at scan time can be compared to the activity concentration in the PET images. The mixture-specific positron fraction was taken into account. The two phantoms were prepared with a sphere-to-background ratio of about 10 : 1. Tab. 1 summarizes the activity concentrations at scan time. At scan time, the Sc mixture contained 72.2% Sc and 27.8% Sc, respectively.
The scans were performed using the Sc and Sc settings of Quadra. The preset half-lives and positron fractions on the Quadra are 3.891 h and 0.881 for Sc and 3.97 h and 0.9427 for Sc, respectively. We adapted the decay factors in order to correct for the most recent value of 4.042 0.007 h half-life for Sc reported in Ref. [38]. While the short scan time makes the difference in decay factor negligible, the cross-calibration of the dose calibrators is significantly affected. Tab. 1 also reports the positron fraction of the Sc mixture at scan time. Although the difference to the pure positron fraction is marginal, we corrected the positron fraction in the images.
Reconstructions of the NEMA phantom data were performed with the Siemens’ proprietary ordered-subset expectation-maximization (OSEM) algorithm, including time-of-flight (TOF), attenuation and scatter correction, as well as point-spread-function (PSF) recovery. The list-mode data were restricted to a 5 min acquisition. The matrix size was and we did not apply a post-reconstruction filter (all-pass), despite being part of the clinical protocols using Quadra [27, 28]. The random sinogram was estimated through a delayed time window. We compared different numbers of iterations and scatter-correction models (absolute vs. relative [39]); five subsets were used throughout since this is the clinical standard for Quadra. It is important to note that for absolute scatter correction, no correction is computed. In the case of relative scatter, a sinogram is estimated by fitting a linear combination of a scaled uniform background sinogram and simulated TOF single scatter sinograms to the tail region of the emission sinogram [40, 14, 39].
As shown in the left panel of Figure 1, the positron energy of and is higher than that of , though still lower than that of . In order to correct for the increased positron range in water (see right panel of Figure 1), we computed the positron range using the FLUKA particle transport toolkit [41, 42, 43]. The annihilation locations were scored using a modified MGDRAW routine and the resulting kernel densities were applied in image space using a Richardson-Lucy deconvolution (same methodology as in Ref. [44]). We refrained from implementing a custom positron range correction in the OSEM projectors [45, 46, 47]. However, by deconvolving only the isotope-specific part of the PSF, we preserved the geometric/scanner-specific PSF correction [48]. In the following, we will refer to the images with an isotope-specific PSF as ‘corrected PSF’ images.
Figure 1. Positron energy spectrum (left) and positron range in water (right) of scandium isotopes in comparison with and .
We evaluated the image properties of scandium radioisotopes according to the updated EARL standard 2 protocol [33, 34, 35] for . In particular, we followed Ref. [35] and considered contrast recovery coefficients (CRC) rather than the previously-used standardized uptake value (SUV) recovery coefficients (RC). On the PET image, the maximum pixel values in each sphere , of the NEMA phantom were identified and a background-corrected 50% isocontour volume-of-interest (VOI) drawn therefrom to compute the mean activity concentration in each sphere. The peak activity concentration was computed by placing a spherical VOI with mm diameter such that it yields the highest mean activity concentration. The contrast recovery coefficients for each sphere are therefore defined as
where and are the activity concentrations in the spheres and the background compartment of the phantom reported in Table 1. The background activity concentration is determined through the mean activity in three quadratic regions of interest (ROIs) in three distinct slices in the phantom’s background compartment, i.e. .
| Phantom/isotope | [kBq mL] | [kBq mL] | Positron fraction |
| Mixture Sc | 30.18 | 3.03 | 0.881 |
| Mixture Sc | 11.61 | 1.17 | 0.943 |
| Mixture Sc | 41.79 | 4.20 | 0.898 |
| Pure Sc | 40.45 | 4.05 | 0.943 |
Table 1. Details of the phantom preparation at scan time. For the Sc mixture we list the two isotopes separately.
The bias measures the quantitative accuracy [35]. While a should be achieved for and , a deviation of up to 10% is accepted for other radionuclides [35]. The coefficient of variation (COV), i.e. the inverse of the signal-to-noise ratio, is defined through the standard deviation in each background ROI.
COV is compliant with the EARL limit for image noise. The residual lung error is defined as the mean activity concentration in a cylindrical VOI (120 mm length, 30 mm diameter) placed at the center of the phantom’s lung insert divided by . Also, for the lung error 15% is considered acceptable.
Results
For illustrative purposes, Figure 2 shows the two central slices of the and NEMA phantom images. These images were reconstructed with three iterations, relative scatter correction, standard PSF correction and without post-reconstruction filter.
Figure 2. Example images of the NEMA phantom for (left) and (right).
The bias for the two phantom scans is and for , respectively.
Figure 3 shows the three recovery curves , and for and in comparison with the EARL standard 2 for . The different recovery curves show relative vs. absolute scatter correction as well as the standard vs. positron range-corrected PSF. Table 2 and Table 3 summarize the corresponding numerical values of the recoveries as well as the COV and lung error.
Figure 3. Contrast recovery curves for (top row) and (bottom row) for three iterations and no post-reconstruction filter.
| Recovery | Reconstruction | COV [%] | Lung err. [%] | ||||||
| abs. std. PSF | 1.00 | 1.02 | 1.06 | 1.06 | 1.06 | 1.08 | 11.97 | 9.54 | |
| abs. corr. PSF | 1.16 | 1.12 | 1.15 | 1.14 | 1.12 | 1.14 | 12.61 | 9.46 | |
| rel. std. PSF | 1.08 | 1.10 | 1.14 | 1.15 | 1.14 | 1.16 | 12.23 | 6.05 | |
| rel. corr. PSF | 1.26 | 1.22 | 1.24 | 1.22 | 1.20 | 1.22 | 12.92 | 5.97 | |
| abs. std. PSF | 0.68 | 0.74 | 0.79 | 0.83 | 0.84 | 0.87 | 11.97 | 9.54 | |
| abs. corr. PSF | 0.79 | 0.81 | 0.85 | 0.86 | 0.88 | 0.90 | 12.61 | 9.46 | |
| rel. std. PSF | 0.73 | 0.80 | 0.85 | 0.89 | 0.91 | 0.94 | 12.23 | 6.05 | |
| rel. corr. PSF | 0.86 | 0.88 | 0.92 | 0.93 | 0.94 | 0.97 | 12.92 | 5.97 | |
| abs. std. PSF | 0.35 | 0.65 | 0.91 | 0.97 | 0.96 | 0.98 | 11.97 | 9.54 | |
| abs. corr. PSF | 0.37 | 0.69 | 0.95 | 0.99 | 0.97 | 1.00 | 12.61 | 9.46 | |
| rel. std. PSF | 0.37 | 0.70 | 0.97 | 1.04 | 1.03 | 1.06 | 12.23 | 6.05 | |
| rel. corr. PSF | 0.39 | 0.74 | 1.03 | 1.07 | 1.04 | 1.07 | 12.92 | 5.97 |
Table 2. Contrast recovery values, COV and lung error for (three iterations and no post-reconstruction filter).
| Recovery | Reconstruction | COV [%] | Lung err. [%] | ||||||
| abs. std. PSF | 0.91 | 1.03 | 1.08 | 1.05 | 1.07 | 1.06 | 10.45 | 11.88 | |
| abs. corr. PSF | 1.01 | 1.13 | 1.15 | 1.10 | 1.11 | 1.11 | 11.09 | 11.85 | |
| rel. std. PSF | 1.00 | 1.15 | 1.19 | 1.16 | 1.18 | 1.17 | 10.78 | 5.57 | |
| rel. corr. PSF | 1.12 | 1.26 | 1.28 | 1.22 | 1.23 | 1.22 | 11.48 | 5.53 | |
| abs. std. PSF | 0.61 | 0.73 | 0.80 | 0.81 | 0.84 | 0.86 | 10.45 | 11.88 | |
| abs. corr. PSF | 0.68 | 0.80 | 0.85 | 0.85 | 0.86 | 0.89 | 11.09 | 11.85 | |
| rel. std. PSF | 0.68 | 0.82 | 0.88 | 0.90 | 0.93 | 0.95 | 10.78 | 5.57 | |
| rel. corr. PSF | 0.75 | 0.90 | 0.94 | 0.94 | 0.95 | 0.97 | 11.48 | 5.53 | |
| abs. std. PSF | 0.33 | 0.62 | 0.90 | 0.95 | 0.96 | 0.99 | 10.45 | 11.88 | |
| abs. corr. PSF | 0.35 | 0.66 | 0.95 | 0.97 | 0.98 | 1.01 | 11.09 | 11.85 | |
| rel. std. PSF | 0.35 | 0.68 | 1.00 | 1.05 | 1.06 | 1.09 | 10.78 | 5.57 | |
| rel. corr. PSF | 0.37 | 0.73 | 1.05 | 1.07 | 1.08 | 1.11 | 11.48 | 5.53 |
Table 3. Contrast recovery values, COV and lung error for (three iterations and no post-reconstruction filter).
In Figure 4 and Figure 5 we present the convergence of the reconstruction algorithm with increasing number of iterations. The COV is plotted against for each sphere of the NEMA phantom. The accepted limit of 15% for the COV is marked with the dashed line while the colored band depicts the EARL standard 2 limits. Related to the convergence of the OSEM reconstruction algorithm, Figure 6 shows the lung error and the COV as a function of the number of iterations. As for the recovery curves, the number of subsets is fixed to five and there is no post-reconstruction filtering.
Figure 4. as a function of COV for for an increasing number of iterations and different scatter corrections. EARL limits are shown as the dashed line and the colored band.
Figure 5. as a function of COV for for an increasing number of iterations and different scatter corrections. EARL limits are shown as the dashed line and the colored band.
Figure 6. Lung error and COV as a function of the number of iterations with the corrected PSF and no post-reconstruction filter. The red dashed line indicates the EARL limits.
Finally, Figure 7 shows the normalized trues sinograms for and at 0° view angle in the case of relative scatter correction. The total correction sinogram, i.e. scatter and the pγ contribution, is plotted alongside the pγ correction illustrating the difference in (relative) pγ contribution in the tail regions of the two sinograms.
Figure 7. Trues sinogram together with total correction and pγ correction sinograms at view angle.
Discussion
Our study confirms the expected quantitative imaging performance of both and to be comparable to . Although this was already shown in Refs. [22, 23] for a clinical PET/CT scanner, we established that the scandium radioisotopes meet EARL’s updated contrast recovery limits when imaged with a LAFOV PET / CT. Furthermore, we showed how a radioisotope-specific PSF correction can significantly improve the contrast recovery.
Our main concern is the rather large bias, in particular for . The 10% limit of Ref. [35] for radionuclides other than and is met for . Likely, the bias we obtained for could be improved by implementing a calibration factor from the vendor or from the national metrology institute in Insel’s dose calibrator. The bias , however, lies outside the allowed range. The total activity in the aliquot measured by the HPGe spectrometer (including ) exceeds the activity measured with the dose calibrator (both measurements at PSI) by 18.5%. This might be seen already as an indication that a mixture-specific calibration factor for the dose calibrator is necessary. Although the activity is small compared to and (grows to 2% at scan time due to the long half-life), the additional prompt photons might offset the dose calibrator measurement. Furthermore, determining the scaling factor between Insel’s and PSI’s dose calibrators based on a simple cross-calibration bears a rather big risk of increasing the bias. Ref. [23] does not specify the details of the definition of Bq for the mixture. They cross-calibrate the local dose calibrator with the reference activity from PSI’s dose calibrator. No details of positron fraction corrections or treatment of the exact mixture are given. For a clinical translation of a correct gauging of a dose calibrator by a metrology institute for the produced isotope mixture and the implementation of the mixture parameters in the scanner software would be necessary. Until then, we believe that an absolute image quantification will remain challenging for . Beyond quantification, the limited radionuclidic purity of poses a challenge for its clinical translation, since the isotopic
composition varies over time and depends on details of the production process).
The physical decay properties of would suggest superior imaging properties compared to . The positron energy is smaller (see Figure 1) and the emission is much smaller and of lower energy (main line at keV at intensity vs. keV at for ). However, we observed that there is only a marginal difference between and in the recovery curves of Figure 3. The main (visible) difference appears in the tail regions of the sinograms in Figure 7. The tails are caused mainly by false coincidences [14]. The ratio of the (almost constant) sinogram tails relative to the maximum between and was around . This is consistent with the in the mixture, indicating that the randoms are caused by the in the mixture. The branching ratio for within or above the coincidence energy window is below for and will therefore not really contribute to the random coincidences outside the source regions.
As shown also in Ref. [44], the vendor’s correction algorithm is effective on Quadra. This is clearly visible when considering the absolute vs. relative scatter correction. The absence of a correction brings closer to 1.0 compared to the case of relative scatter correction, but clearly deteriorates . Most importantly, however, the missing correction in absolute scatter correction leads to a larger lung error (see Figure 6). It should also be noted that the lung error is significantly larger for in the absence of a correction. For relative scatter, the lung error is hardly distinguishable for and , indicating an effective correction of randoms. Absolute scatter correction without correction should therefore be avoided in any clinical protocol for scandium isotopes on Quadra.
Figure 4 and Figure 5 show a very fast convergence of the OSEM reconstruction algorithm. Mostly independent of the isotope, scatter or PSF correction, there is little improvement in after three iterations, even for the mm phantom sphere. Note that clinical protocols on Quadra typically use four iterations [27]. Of course, four iterations further decrease the lung error (see Figure 6). Depending on whether one is willing to apply a post-reconstruction filter (improving COV), the loss in COV may outweigh the gain in lung error at mostly equal contrast recovery. We therefore advocate that three iterations should be used for a clinical protocol with scandium isotopes in the case of an all-pass filtering. We refrain from exploring the reconstruction parameter space with different numbers of subsets but rely on the clinical standard of five subsets [27].
While Refs. [22, 23] relied on the standard PSF correction in the vendor’s reconstruction algorithm, we apply an isotope-specific PSF correction. In Table 2 and Table 3 we note that improves up to 0.13 for and 0.08 for in the case of relative scatter correction. A similar improvement in contrast recovery was observed by Ref. [46] in the case of , which has a comparable positron range in water to that of . Note, however, that we do not implement isotope-specific projectors in the OSEM algorithm as done in Ref. [46]. This is also the reason why we do not observe a significant difference in the convergence (see Figure 4 and Figure 5). Since we apply a post-reconstruction deconvolution, is expected to increase as well as the COV. Since the ROI position for is optimized while keeping the size constant, the PSF correction has only a marginal impact. Though the clinical standard is still without nuclide-specific PSF modeling (even for [47]), we hope that vendors will implement (material-dependent) positron range corrections for standard as well as novel radionuclides.
In summary, we confirmed the findings of Refs. [22, 23] that clinical PET/CT systems can image scandium isotopes with an image quality that meets the EARL standard 2 criteria for . The vendor’s corrections applied in the image reconstruction process are effective for both mixture and pure , but could be improved with an isotope-specific PSF correction. The increased sensitivity of LAFOV scanners in conjunction with the longer half-life of and could allow for longer uptake times [13]. Despite the challenges in producing with a medical cyclotron [32], challenging radionuclidic composition and the difficulties with the bias, we would like to motivate further investment in research and development towards the clinical translation of , given its favorable imaging properties and significantly reduced radiation dose to a patient compared to [21].
Conclusion
In this study, we advanced the quantitative imaging of and on a clinical LAFOV PET/CT scanner. Using a NEMA phantom, we confirmed that the imaging performance of both scandium isotopes is comparable to and meets the updated EARL standard 2 contrast recovery criteria. We showed that the vendor’s correction, is essential to control the random coincidences and lung error introduced by the emission of scandium isotopes, and that an isotope-specific PSF correction further improves contrast recovery. Based on the convergence behavior of the reconstruction, we recommend three iterations with relative scatter correction (including correction) as a practical clinical protocol on Quadra. The main outstanding challenge is the bias observed for the mixture. Together with the longer half-lives of and compared to and , the increased sensitivity of LAFOV systems opens the possibility of longer uptake times and/or multi-time-point imaging, supporting the clinical translation of scandium-based theranostic pairs.
Acknowledgments
The authors thank Hasan Sari, William M. Steinberger for useful discussions and for their help with e7-tools. They also acknowledge the help of Ângelo Rafael Felgosa Cardoso, Angela Filipa Silva Mendes and Marco Viscione for the scheduling of the experiments at Inselspital.
Declarations
Funding
This research is partially supported by the grant no. 216944 of the Swiss National Science Foundation. The dangerous good transportation was financed by the Research Fund of the Swiss Society of Radiobiology and Medical Physics.
Competing interests
MT is a full-time employee of Siemens Medical Solutions USA Inc. All other authors have no conflicts of interests to report.
Ethics approval
Not applicable.
Consent to participate
Not applicable.
Data availability
The datasets generated and/or analyzed in the current study are available from the corresponding author upon reasonable request.
Authors’ contributions
LM: Conceptualization, Formal analysis, Investigation, Methodology, Software, Validation, Writing - Original Draft, Writing - Review & Editing MT: Software, Methodology, Writing - Review & Editing AEG: Software, Formal analysis, Writing - Review & Editing PVG and ANM: Investigation, Resources, Writing - Review & Editing SB: Conceptualization, Project administration, Writing - Review & Editing NPvdM: Conceptualization, Investigation, Supervision, Project administration, Writing - Review & Editing.
References
- [1]Gandhi N, Alaseem AM, Deshmukh R, Patel A, Alsaidan OA, Fareed M, Alasiri G, Patel S, and Prajapati B. Theranostics in nuclear medicine: the era of precision oncology. Medical Oncology 2025 Sep; 42(11):498. DOI: 10.1007/s12032-025-03061-0DOI
- [2]Behera D and McEwan A. Theranostics in Nuclear Medicine: Historical, Regulatory, and Evidence Context for the Practicing Nuclear Medicine Physician. PET Clinics 2026 Jul; 21(3):331–41. DOI: 10.1016/j.cpet.2026.02.006DOI
- [3]Müller C, Bunka M, Haller S, Köster U, Groehn V, Bernhardt P, van der Meulen NP, Türler A, and Schibli R. Promising prospects for ⁴⁴Sc-/⁴⁷Sc-based theragnostics: application of ⁴⁷Sc for radionuclide tumor therapy in mice. Journal of Nuclear Medicine 2014 Oct; 55(10):1658–64. DOI: 10.2967/jnumed.114.141614DOI
- [4]Müller C, Domnanich KA, Umbricht CA, and van der Meulen NP. Scandium and terbium radionuclides for radiotheranostics: current state of development towards clinical application. British Journal of Radiology 2018 Nov; 91:20180074. DOI: 10.1259/bjr.20180074DOI
- [5]Mikolajczak R, van der Meulen NP, and Lapi S. Radiometals for imaging and theranostics, current production, and future perspectives. Journal of Labelled Compounds and Radiopharmaceuticals 2019; 62:615–34. DOI: 10.1002/jlcr.3770DOI
- [6]van der Meulen NP, Strobel K, and Lima TVM. New Radionuclides and Technological Advances in SPECT and PET Scanners. Cancers 2021; 13. DOI: 10.3390/cancers13246183DOI
- [7]Malik N, Lokesha YU, Habte FG, and Daldrup-Link HE. Emerging Theranostic Radiometals (149Tb, 44Sc, 52Mn, 203Pb, 55Co)—Decay Diversity, Production Landscape, and Translational Imaging. Pharmaceuticals 2026; 19. DOI: 10.3390/ph19060889
- [8]van der Meulen NP, Hasler R, Talip Z, Grundler PV, Favaretto C, Umbricht CA, Müller C, Dellepiane G, Carzaniga TS, and Braccini S. Developments toward the Implementation of 44Sc Production at a Medical Cyclotron. Molecules 2020; 25. DOI: 10.3390/molecules25204706DOI
- [9]Braccini S, Belver-Aguilar C, Carzaniga T, Dellepiane G, Haeffner P, and Scampoli P. Novel Irradiation Methods for Theranostic Radioisotope Production With Solid Targets at the Bern Medical Cyclotron. 22nd International Conference on Cyclotrons and their Applications (CYC2019). 2020 : TUA02. DOI: 10.18429/JACoW-Cyclotrons2019-TUA02
- [10]Dellepiane G, Casolaro P, Mateu I, Scampoli P, Voeten N, and Braccini S. 47Sc and 46Sc cross-section measurement for an optimized 47Sc production with an 18 MeV medical PET cyclotron. Applied Radiation and Isotopes 2022; 189:110428. DOI: 10.1016/j.apradiso.2022.110428DOI
- [11]Meier JP, Bhuiyam M, Freifelder R, Zhang HJ, Gonzalez L, Pusateri A, Tsai HM, Leoni L, Ghosh K, Markiewicz E, Henning C, Zhang Y, Weichselbaum R, Nolen J, Rotsch DA, Kao CM, Szmulewicz RZ, Chen CT, and Chitneni SK. Synthesis of DOTA-Based 43Sc Radiopharmaceuticals Using Cyclotron-Produced 43Sc as Exemplified by [43Sc]Sc-PSMA-617 for PSMA PET Imaging. Methods and Protocols 2025; 8. DOI: 10.3390/mps8030058DOI
- [12]Schmidt CE, Gajecki L, Deri MA, and Sanders VA. Current State of 44Ti/44Sc Radionuclide Generator Systems and Separation Chemistry. Current Radiopharmaceuticals 2023; 16:95–106. DOI: 10.2174/187447101666622111154424DOI
- [13]Alberts I, Prensil G, Mingels C, Bohn KP, Visione M, Sari H, Rominger A, and Afshar-Oromieh A. Feasibility of late acquisition [68Ga]Ga-PSMA-11 PET/CT using a long axial field-of-view PET/CT scanner for the diagnosis of recurrent prostate cancer-first clinical experiences. European Journal of Nuclear Medicine and Molecular Imaging 2021 Dec; 48(13):4456–62. DOI: 10.1007/s00259-021-05438-5DOI
- [14]Conti M and Ericksson L. Physics of pure and non-pure positron emitters for PET: a review and a discussion. EJNMMI Physics 2016 Dec; 3:8. DOI: 10.1186/s40658-016-0144-5DOI
- [15]Mercolli L, Steinberger WM, Grundler PV, Moiseeva A, Braccini S, Conti M, Moskal P, Rathod N, Rominger A, Sari H, Schibli R, Seifert R, Shi K, Stepień EŁ, and van der Meulen NP. First positronium lifetime imaging with scandium-44 on a long axial field-of-view PET/CT. Frontiers in Nuclear Medicine 2025; Volume 5 - 2025. DOI: 10.3389/fnume.2025.1648621
- [16]Huang B, Dai B, Lapi SE, Liles G, Karp JS, and Qi J. High-Resolution Positronium Lifetime Tomography at Clinical Activity Levels on the PennPET Explorer. Journal of Nuclear Medicine 2025; 66:1464–70. DOI: 10.2967/jnumed.125.270130DOI
- [17]Das M, Sharma S, Yitayew Beyene E, Bilewicz A, Choiński J, Chug N, Curceanu C, Czerwiński E, Valsan Eliyan K, Hajduga J, Jalali S, Kacprzak K, Kaplanoglu T, Kapłon Ł, Kasperska K, Khreptak A, Korcyl G, Kozik T, Kubat K, Kumar D, Kunimmal Venadan A, Lisowski E, Lisowski F, Medrala-Sowa J, Moyo S, Mryka W, Niedźwiecki S, Pandey P, Parzych S, Porcelli A, Rachwal B, Perez del Rio E, Rädler M, Rominger A, Shi K, Skurzok M, Stolarz A, Szumlak T, Tanty P, Tayefi Ardebili K, Tiwari S, Walczak R, Stepień EŁ, and Moskal P. First Positronium Imaging Using 44Sc With the J-PET Scanner: A Case Study on the NEMA-Image Quality Phantom. IEEE Transactions on Radiation and Plasma Medical Sciences 2026; 10:593–602. DOI: 10.1109/TRPMS.2025.3621554DOI
- [18]Rosar F, Bohnenberger H, Moon ES, Rösch F, Denig A, Vincenz-Zörner D, Hoffmann MA, Khreish F, Ezziddin S, Schreckenberger M, Buchholz HG, and Schaefer-Schuler A. Impact of prompt gamma emission of 44Sc on quantification in preclinical and clinical PET systems. Applied Radiation and Isotopes 2021; 170:109599. DOI: 10.1016/j.apradiso.2021.109599DOI
- [19]Preylowski V, Schlögl S, Schoenahl F, Jörg G, Samnick S, Buck AK, and Lassmann M. Is the image quality of I-124-PET impaired by an automatic correction of prompt gammas? PloS One 2013; 8(8):e77129. DOI: 10.1371/journal.pone.0071729DOI
- [20]Daube-Witherspoon ME, Scheuermann J, McDonald S, Lin W, Aluicio-Sarduy E, Barnhart TE, Pyles J, Engle JW, Lapi SE, and Karp JS. Quantitative Imaging of Non-Standard Radionuclides with Long Axial Field-of-View PET. 2021 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC). 2021 :1–4. DOI: 10.1109/NSS/MIC44867.2021.9875581
- [21]Gomes CV, Mendes BM, Paixão L, Gnesin S, Müller C, van der Meulen NP, Strobel K, Fonseca TCF, and Lima T. Comparison of the dosimetry of scandium-43 and scandium-44 patient organ doses in relation to commonly used gallium-68 for imaging neuroendocrine tumours. EJNMMI Physics 2024 Jul; 11(1):61. DOI: 10.1186/s40658-024-00669-5DOI
- [22]Lima T, Gnesin S, Nitzsche E, Ortega P, Müller C, and van der Meulen NP. First Phantom-Based Quantitative Assessment of Scandium-44 Using a Commercial PET Device. Frontiers in Physics 2020; 8. DOI: 10.3389/fphy.2020.00241DOI
- [23]Lima TVM, Gnesin S, Strobel K, Pérez MdS, Roos JE, Müller C, and van der Meulen NP. Fifty Shades of Scandium: Comparative Study of PET Capabilities Using Sc-43 and Sc-44 with Respect to Conventional Clinical Radionuclides. Diagnostics 2021; 11. DOI: 10.3390/diagnostics11101826DOI
- [24]Singh A, Baum R, Klette I, van der Meulen NP, Müller C, Türler A, and Schibli R. Scandium-44 DOTATOC PET/CT: First in-human molecular imaging of neuroendocrine tumors and possible perspectives for Theranostics. Journal of Nuclear Medicine 2015; 56:267–7. Available from: https://jnm.snmjournals.org/content/56/supplement_3/267
- [25]Eppard E, de la Fuente A, Benešová M, Khawar A, Bundschuh RA, Gärtner FC, Kreppel B, Kopka K, Essler M, and Rösch F. Clinical Translation and First In-Human Use of [44Sc]Sc-PSMA-617 for PET Imaging of Metastasized Castrate-Resistant Prostate Cancer. Theranostics 2017; 7(18):4359–69. DOI: 10.7150/thno.20586DOI
- [26]Zhang J, Singh A, Kulkarni HR, Schuchardt C, Müller D, Wester HJ, Maina T, Rösch F, van der Meulen NP, Müller C, Mäcke H, and Baum RP. From Bench to Bedside—The Bad Berka Experience With First-in-Human Studies. Seminars in Nuclear Medicine 2019; 49. From Basic Science to Clinical Imaging:422–37. DOI: 10.1053/j.semnuclmed.2019.06.002DOI
- [27]Prenosil GA, Sari H, Fürstner M, Afshar-Oromieh A, Shi K, Rominger A, and Hentschel M. Performance Characteristics of the Biograph Vision Quadra PET/CT System with a Long Axial Field of View Using the NEMA NU 2-2018 Standard. Journal of Nuclear Medicine 2022; 63:476–84. DOI: 10.2967/jnumed.121.261972DOI
- [28]Prenosil GA, Hentschel M, Weitzel T, Sari H, Shi K, Afshar-Oromieh A, and Rominger A. EARL compliance measurements on the biograph vision Quadra PET/CT system with a long axial field of view. EJNMMI Physics 2022 Apr; 9(1):26. DOI: 10.1186/s40658-022-00455-1DOI
- [29]Spencer BA, Berg E, Schmall JP, Omidvari N, Leung EK, Abdelhafez YG, Tang S, Deng Z, Dong Y, Lv Y, Bao J, Liu W, Li H, Jones T, Badawi RD, and Cherry SR. Performance Evaluation of the uEXPLORER Total-Body PET/CT Scanner Based on NEMA NU 2-2018 with Additional Tests to Characterize PET Scanners with a Long Axial Field of View. Journal of Nuclear Medicine 2021 Jun; 62(6):861–70. DOI: 10.2967/jnumed.120.250597DOI
- [30]Zhang H, Ren C, Liu Y, Yan X, Liu M, Hao Z, Xing H, and Huo L. Performance Characteristics of a New Generation 148-cm Axial Field-of-View uMI Panorama GS PET/CT System with Extended NEMA NU 2-2018 and EARL Standards. Journal of Nuclear Medicine 2024 Dec; 65(12):1974–82. DOI: 10.2967/jnumed.124.267963DOI
- [31]Grundler PV, Eichler R, Talip Z, Schubiger PA, Schibli R, and van der Meulen NP. The Metamorphosis of Radionuclide Production and Development at Paul Scherrer Institute. CHIMIA 2020 Dec; 74:968. DOI: 10.2533/chimia.2020.968DOI
- [32]Carzaniga TS, Auger M, Braccini S, Bunka M, Ereditato A, Nesteruk KP, Scampoli P, Türler A, and van der Meulen NP. Measurement of 43Sc and 44Sc production cross-section with an 18 MeV medical PET cyclotron. Applied Radiation and Isotopes 2017; 129:96–102. DOI: 10.1016/j.apradiso.2017.08.013DOI
- [33]Boellaard R, O’Doherty MJ, Weber WA, Mottaghy FM, Lonsdale MN, Stroobants SG, Oyen WJG, Kotzerke J, Hoekstra OS, Pruim J, Marsden PK, Tatsch K, Hoekstra CJ, Visser EP, Arends B, Verzijlbergen FJ, Zijlstra JM, Comans EFI, Lammertsma AA, Paans AM, Willemsen AT, Beyer T, Bockisch A, Schaefer-Prokop C, Delbeke D, Baum RP, Chiti A, and Krause BJ. FDG PET and PET/CT: EANM procedure guidelines for tumour PET imaging: version 1.0. European Journal of Nuclear Medicine and Molecular Imaging 2010 Jan; 37(1):181–200. DOI: 10.1007/s00259-009-1297-4DOI
- [34]Kaalep A, Sera T, Rijnsdorp S, Yakub M, Talsma A, Lodge MA, and Boellaard R. Feasibility of state of the art PET/CT systems performance harmonisation. European Journal of Nuclear Medicine and Molecular Imaging 2018 Jul; 45(8):1344–61. DOI: 10.1007/s00259-018-3977-4DOI
- [35]Sunderland JJ, Boellaard R, Dickson JC, Graves SA, and Bailey DL. Standardization of PET/CT Performance Requirements for Whole-Body Quantitative Imaging: An International Proposal. Journal of Nuclear Medicine 2025 Oct; 66(10):1506–15. DOI: 10.2967/jnumed.124.269349DOI
- [36]Juget F, Durán T, Nedjadi Y, Talip Z, Grundler PV, Favaretto C, Casolaro P, Dellapiane G, Braccini S, Bailat C, and van der Meulen NP. Activity Measurement of 44Sc and Calibration of Activity Measurement Instruments on Production Sites and Clinics. Molecules 2023; 28. DOI: 10.3390/molecules28031345DOI
- [37]Kondev F, Wang M, Huang W, Naimi S, and Audi G. The NUBASE2020 evaluation of nuclear physics properties. Chinese Physics C 2021 Mar; 45:030001. DOI: 10.1088/1674-1137/abddae. Available from: https://nds.iaea.org/relnsd/nubase/nubase_min.htmlDOI
- [38]Durán MT, Juget F, Nedjadi Y, Bailat C, Grundler PV, Talip Z, van der Meulen NP, Casolaro P, Dellapiane G, and Braccini S. Half-life measurement of 44Sc and 44mSc. Applied Radiation and Isotopes 2022; 190:110507. DOI: 10.1016/j.apradiso.2022.110507DOI
- [39]Bal H, Panin VY, Schaefferkoetter J, Cabello J, and Conti M. Fully 3D Scatter Estimation in Axially Long FOV PET/CT Scanners: Residual Estimation Approach. 2021 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC). 2021 :1–4. DOI: 10.1109/NSS/MIC44867.2021.9875665DOI
- [40]Watson CC. Extension of Single Scatter Simulation to Scatter Correction of Time-of-Flight PET. IEEE Transactions on Nuclear Science 2007; 54:1679–86. DOI: 10.1109/TNS.2007.901227DOI
- [41]Battistoni G, Boehlen T, Cerutti F, Chin PW, Esposito LS, Fassò A, Ferrari A, Lechner A, Empl A, Mairani A, Mereghetti A, Ortega PG, Ranft J, Roessler S, Sala PR, Vlachoudis V, and Smirnov G. Overview of the FLUKA code. Annals of Nuclear Energy 2015 Aug; 82:10–18. DOI: 10.1016/j.anucene.2014.11.007DOI
- [42]Ahdida C, Bozzato D, Calzolari D, Cerutti F, Charitonidis N, Cimmino A, Coronetti A, D’Alessandro GL, Donadoni Servelle A, Esposito LS, Froeschl R, García Alía R, Gerbershagen A, Gilardoni S, Horváth D, Hugo G, Infantino A, Kouskoura V, Lechner A, Lefebvre B, Lerner G, Magistris M, Manousos A, Moryc G, Ogallar Ruiz F, Pozzi F, Prelipcean D, Roesler S, Rossi R, Sabaté Gilarté M, Salvat Pujol F, Schoofs P, Stránský V, Theis C, Tsinganis A, Versaci R, Vlachoudis V, Waets A, and Widorski M. New Capabilities of the FLUKA Multi-Purpose Code. Frontiers in Physics 2022 Jan; 9:788253. DOI: 10.3389/fphy.2021.788253DOI
- [43]Donadoni A, Hugo G, Theis C, and Vlachoudis V. FLAIR3 – recasting simulation experiences with the Advanced Interface for FLUKA and other Monte Carlo codes. EPJ Web of Conferences 2024; 302. Ed. by Zoia A, Diop C, and Saïkali E:11005. DOI: 10.1051/epjconf/202430211005
- [44]Mercolli L, Grundler PV, Moiseeva AN, Eggimann L, Braccini S, Köster U, and van der Meulen NP. Terbium-149 PET/CT: First quantitative imaging with a clinical long axial field-of-view scanner. EANM Innovation 2026 :100294. DOI: 10.1016/j.eanmi.2026.100294DOI
- [45]Gavrilidis P, Koole M, Annunziata S, Mottaghy FM, and Wierts R. Positron Range Corrections and Denoising Techniques for Gallium-68 PET Imaging: A Literature Review. Diagnostics 2022 Sep; 12(10). DOI: 10.3390/diagnostics12102335
- [46]Kertész H, Conti M, Panin V, Cabello J, Bharkhada D, Beyer T, Papp L, Jentzen W, Cal-Gonzalez J, Herraiz JL, López-Montes A, and Rausch I. Positron range in combination with point-spread-function correction: an evaluation of different implementations for [124I]-PET imaging. EJNMII Physics 2022 Aug; 9. DOI: 10.1186/s40658-022-00482-yDOI
- [47]Lassen ML, Kertész H, Rausch I, Panin V, Conti M, Zuehlsdorff S, Cabello J, Bharkhada D, DeKemp R, Kjaer A, Beyer T, and Hasbak P. Positron Range Correction Helps Enhance the Image Quality of Cardiac 82Rb PET/CT. Journal of Nuclear Medicine 2025 Mar; 66(3):466–72. DOI: 10.2967/jnumed.124.267855DOI
- [48]Panin VY, Kehren F, Michel C, and Casey M. Fully 3-D PET reconstruction with system matrix derived from point source measurements. IEEE Transactions on Medical Imaging 2006; 25(7):907–21. DOI: 10.1109/tmi.2006.876171DOI






