Research ArticleBasic Science Investigation

Performance Characteristics of a New-Generation Digital Bismuth Germanium Oxide PET/CT System, Omni Legend 32, According to NEMA NU 2-2018 Standards

Shin Yamagishi, Kenta Miwa, Shun Kamitaki, Kouichi Anraku, Shun Sato, Tensho Yamao, Hitoshi Kubo, Noriaki Miyaji and Kazuhiro Oguchi
Journal of Nuclear Medicine December 2023, 64 (12) 1990-1997; DOI: https://doi.org/10.2967/jnumed.123.266140

Abstract

The Omni Legend 32 PET/CT system features silicon photomultiplier (SiPM)–based detectors with bismuth germanium oxide crystals and a 32-cm axial field of view (FOV). The present study aimed to determine the performance characteristics of the Omni Legend 32 PET/CT system according to National Electrical Manufacturers Association (NEMA) NU 2-2018 standards. Methods: The PET component of this system comprises 22 detector modules; each module contains 24 detector blocks with 72 bismuth germanium oxide crystals with a volume of 4.1 × 4.1 × 30 mm coupled to 18 SiPM devices with a 6 × 6 mm area, resulting in an axial FOV of 32 cm. The spatial resolution, sensitivity, count rate performance, and image quality delivered by PET were evaluated using the NEMA NU 2-2018 standard. PET images of 2 patients were evaluated to get a visual first impression of the Omni Legend 32 PET/CT system together with Precision DL. Results: The average spatial resolution at 1, 10, and 20 cm from the central axis was 4.3, 5.3, and 6.2 mm, respectively, for filtered backprojection and 3.7, 4.3, and 5.1 mm, respectively, for ordered-subset expectation maximization. The NEMA sensitivity was 47.30 and 47.05 cps/kBq at the axial center of the FOV and at a 10-cm radial offset, respectively. The scatter fraction, count rate accuracy, and peak noise-equivalent count rates were 35.4%, 1.7%, and 501.7 kcps, respectively, at 15.7 kBq/mL. Contrast recovery for the NEMA body phantom from the smallest to the largest sphere ranged from 61.3% to 93.0%, with a background variability of 5.4%–11.7% and a lung error of 5.1% for Q.Clear (β-value, 50). Good patient image quality was obtained with the Omni Legend 32. Conclusion: The Omni Legend 32 has class-leading sensitivity and count rates within the category of whole-body PET systems while maintaining spatial resolution broadly comparable to that of other current SiPM-based PET/CT systems. This combination of properties results in a very good image quality.

PET/CT is essential for the detection, localization, characterization, staging, and response evaluation to therapy of cancer; it also provides critical insights in neurology for neurodegenerative diseases and in cardiology for the assessment of myocardial function (1). Recent technologic advances in PET/CT imaging hardware and reconstruction methods have improved the performance of PET/CT systems (2–4).

Most commercial PET systems have recently incorporated good time-resolution detectors based on lutetium oxyorthosilicate (LSO) and lutetium-yttrium oxyorthosilicate scintillator crystals that have improved time-of-flight (TOF) localization (5–7). However, GE Healthcare maintains a series of PET/CT systems based on bismuth germanium oxide (BGO) detectors such as the Discovery IQ (D-IQ) (8). The high stopping power and detection efficiency of BGO crystals, as well as their relatively low production costs (9), are desirable properties for a PET scanner, although the timing performance of BGO crystals is unsuitable for use in PET systems that incorporate TOF information. The high sensitivity that can be achieved with BGO crystals can be attributed to their high stopping power—a result of the presence of bismuth, which, with an atomic number of 83, provides a high photofraction (38%) at 511 keV (10,11). Expanding the axial field of view (FOV) is another way to improve sensitivity (9). Thus, their sensitivity gains might be used to reduce the total acquisition duration or doses of administered radioactive tracers while maintaining image quality that might be comparable to or better than TOF PET scanners with lower sensitivity (9).

Silicon photomultipliers (SiPMs) are replacing photomultiplier tubes in PET systems. They are compact and rugged; have high gain, a good timing response, and high photon detection efficiency; and are insensitive to magnetic fields (12–14). Some studies have confirmed that SiPM-based PET/CT systems with better TOF capability, higher spatial resolution due to a smaller crystal size, and improved software allow for a decreased acquisition time or a reduced administered activity dose of radiopharmaceutical tracer compared with similar photomultiplier tube systems (15–17). To date, SiPMs have been included in the following clinical PET/CT systems: Discovery MI (GE Healthcare) (14,18,19), Vision (Siemens) (5), Vereos (Philips) (6), uMI (United Imaging) (20), and the Cartesion Prime (Canon).

The recently developed Omni Legend 32 (Omni Legend) PET/CT system (GE Healthcare) includes a new digital BGO detection technology that combines BGO crystals with SiPM technology and extends the size of the axial FOV to 32 cm. This extended FOV, combined with the stopping power of 30-mm-thick BGO and the photon detection efficiency of the SiPM, has resulted in high gains in sensitivity compared with other systems in this class of whole-body PET. In addition, a new image-processing algorithm using deep learning called Precision DL (GE Healthcare) can mimic subjective image quality obtained with a TOF-based system (21). Studies are evaluating whether the highly sensitive BGO-based digital detector inside the Omni Legend together with Precision DL can provide better image quality and improved small-lesion detectability on PET images. The present study aimed to define the performance characteristics of the Omni Legend according to the National Electrical Manufacturers Association (NEMA) standard. To our knowledge, this is the first publication to investigate the fundamental characteristics of the Omni Legend.

MATERIALS AND METHODS

PET/CT Scanner Characteristics

We acquired PET data using an Omni Legend with an SiPM-based detector coupled with BGO scintillator crystals (digital BGO) and a large 32-cm axial FOV. The PET scanner has a detector consisting of 22 modules with 24 blocks each, resulting in a total of 528 BGO blocks; each block comprises an array of 6 × 12 crystals (4.1 × 4.1 × 30 mm each) coupled to 3 × 6 SiPMs (6 × 6 mm each). In total, this gives 38,016 BGO crystals and 9,504 SiPM channels covering axial and transaxial FOVs of 32 and 70 cm, respectively. The face-to-face distance of the detectors is 725.4 mm. The coincidence window is 6.93 ns, and the energy window is 435–580 keV. To improve count rate performance, each block has independent electronic processing, including digital pulse clipping and dual integration, with the short integration time as low as 320 ns. The CT component of the PET/CT system consists of a 64-slice helical multi–detector row scanner. The detector collimation is 64 × 0.625 mm. The maximum rotation speed of the CT component is 0.35 s per rotation.

Performance Evaluation According to NEMA NU 2-2018 Standard

Spatial Resolution

We measured spatial resolution using 3 18F point sources with activity concentrations greater than 500 MBq/mL. Droplets containing 18F (<1 mm axially) were suspended on flat trays, drawn into capillary tubes, and sealed. The tubes were placed in phantom holders that locked into accessory slots at the front of the cradle at 1, 10, and 20 cm from the isocenter. After an initial scan to ensure accurate positioning to within ±1 mm, the point source was imaged first at the center and then at one eighth of the axial FOV for 1 min each. The acquired data were reconstructed using 2 different methods: filtered backprojection and ordered-subset expectation maximization (OSEM) (VUE Point HD, with 6 iterations, 22 subsets, and 2-mm gaussian filtering) on a 384 × 384 matrix with a voxel size of 0.65 × 0.65 × 2.07 mm.

Sensitivity

Sensitivity was measured in counts per second detected by the scanner per unit of activity. A 70-cm-long line containing 2.3 mL (8 MBq of 18F) was threaded inside an aluminum sleeve and positioned at the center of the transverse FOV, and data were acquired for 1 min. This process was repeated 4 times while successively adding aluminum sleeves. The results were extrapolated to determine scanner sensitivity without attenuating material. The entire process was repeated with the line source placed 10 cm off isocenter.

Count Rate Performance

The count rate of the Omni Legend over a range of activity levels was measured via a long acquisition run starting at the highest activity level. A 70-cm-long 5.15-mL line containing 740 MBq of 18F was passed through a 70-cm-long polyethylene NEMA scatter phantom at the start of the scan. The patient table was moved to its lowest position, and this NEMA scatter phantom was placed in the center of the FOV on stacks of low-density material placed outside the FOV. Data were acquired for 14 h and 15 min and were reconstructed using OSEM, with 3 iterations, 22 subsets, a 5-mm gaussian filter, a 128 × 128 matrix, and an 18-cm FOV. The prompt, random, true, and scatter count rates associated with activity in the FOV were recorded. Peak true and noise-equivalent count rates (NECRs) were calculated at the corresponding activity concentrations, and scatter rates at the peak NECR were also calculated. Scatter rates at the same activity level were also quantified.

Count Loss Correction and Accuracy

In accordance with the NEMA standard, to evaluate count rate accuracy, the low count rate data (where randoms and dead time approach zero) were extrapolated to earlier time points (where the impact of count loss and randoms might be significant because of higher activity levels). Using the extrapolated activity as ground truth allows determination of the accuracy of count loss and randoms corrections. Relative count errors and the maximum absolute relative error below the peak NECR value were also calculated in the count rate performance section.

Image Quality

We used a NEMA body phantom containing spheres measuring 10–37 mm (NEMA Body Phantom Set; Data Spectrum Corp.). The target-to-background ratio in the phantom was 4:1 at a background activity concentration of 5.3 kBq/mL. The lung insert contained Styrofoam (DuPont) beads surrounded by water to simulate the density (0.3 g/mL) of lung tissue. In accordance with Figure 7.3 of NEMA NU 2-2018, a scatter phantom was placed behind the body phantom and the line source in the phantom was filled with 105 MBq to simulate activity outside the FOV. Three consecutive scans were performed, and the results were then averaged. The acquisition mimicked a 100-cm, 30-min whole-body scan. The acquisition durations were 7:05, 7:24, and 7:45 min for the 32-cm axial FOV. Acquisition length was progressively prolonged to account for activity decay among 3 consecutive scans and to maintain similar counting statistics. Images were reconstructed using OSEM, with 6 iterations, 22 subsets, a 2-mm gaussian filter, and Q.Clear (GE Healthcare) using β-values of 50 and 300 and a 384 × 384 image matrix with a voxel size of 1.04 × 1.04 × 2.07 mm (22–24). Regions of interest were automatically drawn on the background and spheres, and contrast recovery, background variability, and mean lung error were measured according to NEMA specifications.

PET/CT Coregistration

Data were acquired using PET and CT fiducial markers at 6 locations within the PET and CT FOVs to determine the PET/CT coregistration accuracy. We used 3 NEMA phantom hollow spheres (17, 22, and 28 mm) and a coregistration phantom holder supplied by the manufacturer at 2 locations within the PET and CT fields to measure coregistration error between the 2 subsystems. A simulated patient weight of 115 kg was evenly distributed over cradles at 2 regions: the mass of 57.5 kg was evenly distributed over 65 cm, between 20 and 85 cm from the tip of the table, and the remaining mass was evenly distributed over 65 cm, between 115 and 180 cm from the tip of the table. The phantom holders were positioned axially 5 and 100 cm from the top of the table. A mixture was prepared by combining 44 MBq of 18F solution in 14 mL and CT contrast agent (iopamidol, 300 mg iodine/mL; Fuji Pharma Co.) in 6 mL. This solution was then dispensed into each of the 3 spheres. Thereafter, 17-, 22-, and 28-mm spheres were fixed at positions (0,1), (0,20), and (20,0) of the phantom holder in the transverse FOV. Phantom data were acquired for 3 min in each of the 2 positions, and the data were reconstructed using OSEM (VUE Point HD) (6 iterations; 22 subsets). The centroids of the spheres were calculated from the PET and CT data, and copositioning error was determined by calculating distances between the centroids.

Phantom Imaging

We acquired PET data for 60 min using the deluxe and mini deluxe Derenzo phantoms with hot rod portions of 4.8, 6.4, 7.9, 9.5, 11.1, and 12.7 mm and 1.2, 1.6, 2.4, 3.2, 4.0, and 4.8 mm, respectively. All PET images were reconstructed using Q.Clear (25,26) with β-values of 20, 50, 100, 200, 300, and 400. The phantom images were evaluated for conspicuity of the rods to assess the ability to detect small structures.

Patient Imaging

Two patient studies were included to provide a first impression of the clinical images of Omni Legend. The institutional review board approved this study, and both subjects gave written informed consent. First, a 67-y-old woman (height, 167 cm; weight, 62.2 kg) with confirmed postoperative pancreatic tail cancer was injected with 231 MBq of 18F-FDG. Second, a 62-y-old man (height, 160 cm; weight, 59.5 kg) with confirmed lung cancer was injected with 214 MBq of 18F-FDG. Datasets were acquired for 120 s/bed position from the 2 patients, starting 60 min after injection. The data generated by the Omni Legend were reconstructed under Q.Clear (25,26) with a β-value of 300; low, medium, and high Precision DL (21); MotionFree (27) on; a 384 × 384 matrix; a 60-cm FOV for the reconstruction; and a voxel size of 1.04 × 1.04 × 2.07 mm. The quality of the clinical PET images was analyzed using the noise level of the liver (26). The noise level (%) was defined as the SD normalized to the SUVmean of a large spheric reference volume of interest placed in normal liver.

RESULTS

Performance Evaluation

Table 1 summarizes the results for the Omni Legend. Tangential, radial, and axial spatial resolution was 4.22, 4.34, and 4.42 mm, respectively, at 1 cm from the center axis and 5.03, 7.83, and 5.59 mm, respectively, at 20 cm for filtered backprojection and 3.73, 3.89, and 3.61 mm, respectively, at 1 cm and 4.26, 7.59, and 3.58 mm, respectively, at 20 cm for OSEM.

View this table:
TABLE 1.

All NEMA NU 2-2018 Results for Omni Legend

Figure 1 shows sensitivity measured at the center and 10-cm radial offset. The NEMA sensitivity was 47.30 and 47.05 cps/kBq at the axial center of the FOV and at a 10-cm radial offset, respectively. In terms of count rate performance, the peak NECR was 501.7 kcps at 15.7 kBq/mL (Fig. 2). The scatter fraction at peak NECR was 35.4% (Fig. 3). The peak true counting rate was 2,089.1 kcps at 33.6 kBq/mL. Figure 4 shows that the maximum absolute error below the peak NECR and the mean error were 1.7% and 0.98%, respectively. The detector maintained a good linearity over a wide range of activity concentrations. Image quality was assessed using a NEMA body phantom across 6 spheres, and the contrast recovery, background variability, and lung error were 53.5%–88.9%, 8.0%–2.6%, and 12%, respectively, for OSEM and 61.3%–93.0%, 5.4%–1.7%, and 5.1%, respectively, for Q.Clear (β-value, 50). The PET/CT coregistration accuracy of spheres at positions (0,1), (0,20), and (20,0) was 0.85, 1.54, and 1.29 mm, respectively, at tables positioned at 5 cm and 2.47, 3.30, and 2.73 mm, respectively, at 100 cm.

FIGURE 1.
FIGURE 1.

Sensitivity values at 0- and 10-cm offset. (A) Sensitivity as function of numbers of attenuating aluminum sleeves. (B) Axial sensitivity profile.

FIGURE 2.
FIGURE 2.

Count rates for prompt, random, true, and scatter events and noise-equivalent count (NEC) as function of activity concentration.

FIGURE 3.
FIGURE 3.

Scatter fraction for Omni Legend as function of activity concentration.

FIGURE 4.
FIGURE 4.

Accuracy of extrapolated true rate (A) and maximum, mean, and minimum error (B) as function of activity concentration for Omni Legend.

Phantom Imaging

Figure 5 shows representative images of the deluxe and the mini deluxe Derenzo phantoms (β-value in Q.Clear, 50). The 2.4-mm hot rods are clearly visible. Supplemental Figure 1 provides the differences in the imaging performance of the mini deluxe Derenzo phantom for different β-values (supplemental materials are available at http://jnm.snmjournals.org).

FIGURE 5.
FIGURE 5.

Images of deluxe (A) and mini deluxe (B) Derenzo phantoms using Q.Clear (β-value, 50): original distribution of CT images (left) and actual acquired PET images (right).

Patient Imaging

Figure 6 shows example patient images acquired with the Omni Legend system. The results showed that this patient had gastric invasion and peritoneal pancreatic cancer dissemination. The maximum-intensity-projection and axial PET images acquired with the Omni Legend visually appeared to be especially clear across the liver, where the activity concentration was uniformly distributed. Moreover, the application of Precision DL reduced the noise level of the liver.

FIGURE 6.
FIGURE 6.

Pancreatic cancer patient images acquired using Omni Legend at 120 s/bed position (8 min total). (A) Maximum-intensity-projection PET images. Numbers indicate lesion SUVmax. (B) Transaxial PET images. HPDL = high-precision DL; LPDL = low precision DL; MPDL = medium-precision DL; noise = noise level of liver (%); non-PDL = nonprecision DL; PSF = point-spread function.

Figure 7 shows that the other patient had metastases in the right hilar mediastinum, supraclavicular fossa, and upper abdominal lymph nodes. Inflammatory uptake in the left axillary lymph nodes due to vaccination was also observed. The supraclavicular fossa and mediastinal lymph nodes in maximum-intensity-projection PET images were prominent. The coronal image showed uniform accumulation in the liver region and less accumulation in the lung cold background. The Omni Legend with Precision DL improved the feature sharpness of PET images. In particular, small foci, such as the mediastinal lymph nodes, were clearly visualized with high contrast. As the strength of the Precision DL was increased, the SUVmax of lesions increased whereas the SUVmean of liver remained stable.

FIGURE 7.
FIGURE 7.

Lung cancer patient images acquired using Omni Legend at 120 s/bed position (8 min total). (A) Maximum-intensity-projection PET images. Numbers indicate lesion SUVmax. (B) Coronal PET images. HPDL = high-precision DL; LPDL = low precision DL; MPDL = medium-precision DL; noise = noise level of liver (%); non-PDL = nonprecision DL; PSF = point-spread function.

DISCUSSION

We evaluated the performance characteristics and imaging capabilities of a solid-state 32-cm axial FOV digital BGO PET/CT system according to NEMA NU 2-2018. The Omni Legend is the most recent GE Healthcare BGO PET/CT system. The results showed that the Omni Legend has good spatial resolution close to the center of the FOV, high count rate characteristics, class-leading system sensitivity, and good image quality.

The spatial resolution (full width at half maximum) of OSEM was less than 4.0 mm at 1.0 cm off center with the Omni Legend, which was better than that of the D-IQ. The transaxial spatial resolution was better for the Omni Legend than for the D-IQ, by 0.64, 0.85, and 1.08 mm at a radius of 1, 10, and 20 cm, respectively. This improvement can be explained by the smaller BGO crystals (Omni Legend vs. D-IQ: 4.1 vs. 6.3 mm) (8). The contribution of crystal size to spatial resolution is approximately half the crystal size (28), which is 2.05 mm for the Omni Legend. In contrast, the resolution at 20 cm from the central axis was worse than for other modern scanners (5,6,20), as might be due to the depth-of-interaction effects expected with the long crystals of 30 mm.

The NEMA sensitivity of the Omni Legend is the highest of all current PET/CT systems except total-body PET/CT (9). A major contributor to the high sensitivity gain of the Omni Legend is its long axial FOV of 32 cm (7,29,30). We compared sensitivity between the Omni Legend and the GE Healthcare D-MI Generation 2 (D-MI) 6-ring system (30 cm) (29) with a comparable axial FOV. The sensitivity per unit length in the axial direction of the Omni Legend, at 1,478 cps/MBq/cm, was much higher than the value for the D-MI (1,092 cps/MBq/cm) (29). This difference can be attributed to the increased stopping power of the thicker (30 mm) BGO crystals compared with the thinner (25 mm) LSO crystals. The high density and stopping power of BGO (31,32), along with a crystal depth of 30 mm, provided an increase in sensitivity of up to 47 cps/kBq across a 32-cm axial FOV. The high sensitivity of the Omni Legend has the potential to improve the quality and quantitation of clinical images and reduce total scan duration.

The Omni Legend, like the D-IQ, features a dual-channel acquisition system that improves the scatter and the count rate (8). The scatter fraction tends to increase as an axial FOV is extended, specifically up to the axial FOV of about 30 cm for a detector ring diameter of 70 cm (33). However, the scatter fraction of the Omni Legend (35.4%) was equal to or lower than those of the D-IQ and the D-MI (36.2% (8) and 40.2%–41.0% (29,30), respectively). This finding can be explained by the good energy resolution (9.8%) of the Omni Legend and the improved energy window (435–580 keV) in the Omni Legend compared with 425–650 keV in the D-IQ and the D-MI. In addition, the Omni Legend, like the D-IQ, has additional end and front shielding that was specifically designed to reduce scatter from outside the FOV (8). The scatter fraction reached a peak and then decreased (Fig. 3). At low to intermediate count rates, pile-up can cause additional scatter events to appear above the lower-level discriminator (435 keV), increasing the scatter fraction. Because of the tight upper-level discriminator setting (580 keV), at high count rates piled-up events will more likely be rejected, which lowers the scatter fraction again.

The true count rate and NECR peaks for the Omni Legend were superior to those for the D-IQ (2,089.1 and 501.7 kcps at 33.6 and 15.7 kBq/mL, respectively, vs. 409.1 and 123.6 kcps at 25.8 and 9.1 kBq/mL, respectively) (8). We assumed that the increased true count rate was the result of the higher sensitivity, smaller block size, and better electronics, including digital pulse clipping and dual integration, of the Omni Legend than of the D-IQ. The maximum absolute error at activity below the NECR peak was slightly better for the Omni Legend than for the D-IQ (1.7% vs. 3.9%). The use of SiPM photodetectors adds flexibility in the design of the block; the use of smaller blocks (25 × 50 mm) is a key factor in reducing the system dead time and improving the high count rate performance. The combination of small block size and advanced pulse integration contributes to the 501.7-kcps NECR peak. This system is very good in correcting dead time effects far beyond the activity concentration where the noise-equivalent count peaks. Along with an especially high NECR curve, such correction becomes critical when considering the high count rate and short-lived 15O, 13N, and 82Rb tracers applied in neurology and cardiology, where the amounts of injected activity are much higher than in 18F-FDG oncologic imaging.

The contrast recovery and background variability of the Omni Legend are comparable to those published for the most recent GE Healthcare SiPM-based PET/CT system, D-MI (29). As a function of increasing β-value in Q.Clear, the contrast recovery and background variability decreased, and lung error increased, which are consistent with previous studies of LSO-based PET systems (22,25). Although the advances in LSO-based PET systems have led to superior TOF performance, currently around 200 ps, the choice between LSO and BGO should be carefully considered on the basis of the specific requirements of the application, including factors such as spatial resolution, sensitivity, and cost. The overall results of the Omni Legend are superior to those of the most recent GE Healthcare BGO-based PET/CT system, D-IQ (8), especially for the 4 smallest spheres. These findings can be explained by the high sensitivity and good spatial resolution near the center of the FOV of the Omni Legend.

The use of the Omni Legend together with Precision DL could potentially reduce noise and background variability. In addition, the SUVmax of small lesions increased as the strength of the Precision DL was increased (Figs. 6 and 7). These findings might be particularly beneficial when abnormalities are surrounded by a degree of background tracer uptake, such as liver and mediastinum. It might increase the detection and diagnostic confidence for abnormalities in these areas. However, the clinical images from this study should not be considered a valid objective comparison of clinical system performance and are presented only to provide readers a preliminary impression of the clinical images acquired by the Omni Legend. Further studies involving more patients with various types of tumors are required to assess the combined effect of the Omni Legend and Precision DL.

CONCLUSION

We evaluated the performance of the new-generation Omni Legend PET/CT system in accordance with the NEMA NU 2-2018 standard. The Omni Legend has a class-leading sensitivity and count rate while maintaining spatial resolution broadly comparable to that of other current SiPM-based PET systems.

DISCLOSURE

This work was supported in part by a KAKENHI Grant-in-Aid for Young Scientists (21K18097) and from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of the Japanese government. No other potential conflict of interest relevant to this article was reported.

KEY POINTS

QUESTION: What are the performance characteristics of the Omni Legend PET/CT system according to the NEMA NU 2-2018 standard?

PERTINENT FINDINGS: The Omni Legend has class-leading sensitivity and count rates while maintaining high spatial resolution similar to that of other current SiPM-based PET systems, substantially improving image quality.

IMPLICATIONS FOR PATIENT CARE: This highly sensitive PET/CT system with the large axial coverage of SiPM technology coupled with BGO crystals has already been implemented in routine clinical practice. It allows shorter acquisition protocols or reduced radiopharmaceutical doses.

ACKNOWLEDGMENTS

We thank Takayuki Miyachi, Yasuharu Sekiguchi, Naoki Suzuki. and Floris Jansen from GE Healthcare for valuable contributions to the data collection process for this publication.

Footnotes

  • ↵* Contributed equally to this work.

  • Published online Oct. 19, 2023.

REFERENCES

  1. 1.↵
    1. Nishiyama Y,
    2. Kinuya S,
    3. Kato T,
    4. et al
    . Nuclear medicine practice in Japan: a report of the eighth nationwide survey in 2017. Ann Nucl Med. 2019;33:725–732.
    OpenUrl
  2. 2.↵
    1. van der Vos CS,
    2. Koopman D,
    3. Rijnsdorp S,
    4. et al
    . Quantification, improvement, and harmonization of small lesion detection with state-of-the-art PET. Eur J Nucl Med Mol Imaging. 2017;44:4–16.
    OpenUrlCrossRef
  3. 3.
    1. Aide N,
    2. Lasnon C,
    3. Kesner A,
    4. et al
    . New PET technologies: embracing progress and pushing the limits. Eur J Nucl Med Mol Imaging. 2021;48:2711–2726.
    OpenUrl
  4. 4.↵
    1. Miwa K,
    2. Wagatsuma K,
    3. Iimori T,
    4. et al
    . Multicenter study of quantitative PET system harmonization using NIST-traceable 68Ge/68Ga cross-calibration kit. Phys Med. 2018;52:98–103.
    OpenUrl
  5. 5.↵
    1. van Sluis J,
    2. de Jong J,
    3. Schaar J,
    4. et al
    . Performance characteristics of the digital Biograph Vision PET/CT system. J Nucl Med. 2019;60:1031–1036.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Rausch I,
    2. Ruiz A,
    3. Valverde-Pascual I,
    4. Cal-Gonzalez J,
    5. Beyer T,
    6. Carrio I
    . Performance evaluation of the Vereos PET/CT system according to the NEMA NU2-2012 standard. J Nucl Med. 2019;60:561–567.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    1. Spencer BA,
    2. Berg E,
    3. Schmall JP,
    4. et al
    . 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. J Nucl Med. 2021;62:861–870.
    OpenUrlAbstract/FREE Full Text
  8. 8.↵
    1. Reynés-Llompart G,
    2. Gamez-Cenzano C,
    3. Romero-Zayas I,
    4. Rodriguez-Bel L,
    5. Vercher-Conejero JL,
    6. Marti-Climent JM
    . Performance characteristics of the whole-body Discovery IQ PET/CT system. J Nucl Med. 2017;58:1155–1161.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Nadig V,
    2. Herrmann K,
    3. Mottaghy FM,
    4. Schulz V
    . Hybrid total-body pet scanners-current status and future perspectives. Eur J Nucl Med Mol Imaging. 2022;49:445–459.
    OpenUrl
  10. 10.↵
    1. Phunpueok CA,
    2. Chewpraditkul W,
    3. Thongpool V,
    4. Aphairaj D
    . Comparison of photofraction for LuYAP:Ce, LYSO:Ce and BGO crystals in gamma ray detection. Presented at: 15th International Conference of International Academy of Physical Sciences; December 9–13, 2012; Pathumthani, Thailand.
  11. 11.↵
    1. Yu X,
    2. Zhang X,
    3. Zhang H,
    4. et al
    . Requirements of scintillation crystals with the development of PET scanners. Crystals. 2022;12:1302.
  12. 12.↵
    1. Lecomte R
    . Novel detector technology for clinical PET. Eur J Nucl Med Mol Imaging. 2009;36(suppl 1):S69–S85.
    OpenUrlCrossRefPubMed
  13. 13.
    1. Wagatsuma K,
    2. Sakata M,
    3. Ishibashi K,
    4. et al
    . Direct comparison of brain [18F]FDG images acquired by SiPM-based and PMT-based PET/CT: phantom and clinical studies. EJNMMI Phys. 2020;7:70.
  14. 14.↵
    1. Wagatsuma K,
    2. Miwa K,
    3. Sakata M,
    4. et al
    . Comparison between new-generation SiPM-based and conventional PMT-based TOF-PET/CT. Phys Med. 2017;42:203–210.
    OpenUrl
  15. 15.↵
    1. Gnesin S,
    2. Kieffer C,
    3. Zeimpekis K,
    4. et al
    . Phantom-based image quality assessment of clinical 18F-FDG protocols in digital PET/CT and comparison to conventional PMT-based PET/CT. EJNMMI Phys. 2020;7:1.
    OpenUrl
  16. 16.
    1. Alberts I,
    2. Sachpekidis C,
    3. Prenosil G,
    4. et al
    . Digital PET/CT allows for shorter acquisition protocols or reduced radiopharmaceutical dose in [18F]-FDG PET/CT. Ann Nucl Med. 2021;35:485–492.
    OpenUrl
  17. 17.↵
    1. Schillaci O,
    2. Urbano N
    . Digital PET/CT: a new intriguing chance for clinical nuclear medicine and personalized molecular imaging. Eur J Nucl Med Mol Imaging. 2019;46:1222–1225.
    OpenUrl
  18. 18.↵
    1. Miwa K,
    2. Wagatsuma K,
    3. Nemoto R,
    4. et al
    . Detection of sub-centimeter lesions using digital TOF-PET/CT system combined with Bayesian penalized likelihood reconstruction algorithm. Ann Nucl Med. 2020;34:762–771.
    OpenUrl
  19. 19.↵
    1. Hsu DFC,
    2. Ilan E,
    3. Peterson WT,
    4. Uribe J,
    5. Lubberink M,
    6. Levin CS
    . Studies of a next-generation silicon-photomultiplier-based time-of-flight PET/CT system. J Nucl Med. 2017;58:1511–1518.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Chen S,
    2. Hu P,
    3. Gu Y,
    4. Yu H,
    5. Shi H
    . Performance characteristics of the digital uMI550 PET/CT system according to the NEMA NU2-2018 standard. EJNMMI Phys. 2020;7:43.
    OpenUrl
  21. 21.↵
    1. Mehranian A,
    2. Wollenweber SD,
    3. Walker MD,
    4. et al
    . Deep learning-based time-of-flight (ToF) image enhancement of non-ToF PET scans. Eur J Nucl Med Mol Imaging. 2022;49:3740–3749.
    OpenUrl
  22. 22.↵
    1. Miwa K,
    2. Yoshii T,
    3. Wagatsuma K,
    4. et al
    . Impact of gamma factor in the penalty function of Bayesian penalized likelihood reconstruction (Q.Clear) to achieve high-resolution PET images. EJNMMI Phys. 2023;10:4.
    OpenUrl
  23. 23.
    1. Yoshii T,
    2. Miwa K,
    3. Yamaguchi M,
    4. et al
    . Optimization of a Bayesian penalized likelihood algorithm (Q.Clear) for 18F-NaF bone PET/CT images acquired over shorter durations using a custom-designed phantom. EJNMMI Phys. 2020;7:56.
    OpenUrl
  24. 24.↵
    1. Wagatsuma K,
    2. Miwa K,
    3. Kamitaka Y,
    4. et al
    . Determination of optimal regularization factor in Bayesian penalized likelihood reconstruction of brain PET images using [18F]FDG and [11C]PiB. Med Phys. 2022;49:2995–3005.
    OpenUrl
  25. 25.↵
    1. Teoh EJ,
    2. McGowan DR,
    3. Macpherson RE,
    4. Bradley KM,
    5. Gleeson FV
    . Phantom and clinical evaluation of the Bayesian penalized likelihood reconstruction algorithm Q.Clear on an LYSO PET/CT system. J Nucl Med. 2015;56:1447–1452.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    1. Lindström E,
    2. Sundin A,
    3. Trampal C,
    4. et al
    . Evaluation of penalized-likelihood estimation reconstruction on a digital time-of-flight PET/CT scanner for 18F-FDG whole-body examinations. J Nucl Med. 2018;59:1152–1158.
    OpenUrlAbstract/FREE Full Text
  27. 27.↵
    1. Kvernby S,
    2. Hult NK,
    3. Lindstrom E,
    4. et al
    . Quantitative comparison of data-driven gating and external hardware gating for 18F-FDG PET-MRI in patients with esophageal tumors. Eur J Hybrid Imaging. 2021;5:5.
    OpenUrl
  28. 28.↵
    1. Moses WW
    . Fundamental limits of spatial resolution in PET. Nucl Instrum Methods Phys Res A. 2011;648(suppl 1):S236–S240.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Zeimpekis KG,
    2. Kotasidis FA,
    3. Huellner M,
    4. Nemirovsky A,
    5. Kaufmann PA,
    6. Treyer V
    . NEMA NU 2-2018 performance evaluation of a new generation 30-cm axial field-of-view Discovery MI PET/CT. Eur J Nucl Med Mol Imaging. 2022;49:3023–3032.
    OpenUrl
  30. 30.↵
    1. Pan T,
    2. Einstein SA,
    3. Kappadath SC,
    4. et al
    . Performance evaluation of the 5-ring GE Discovery MI PET/CT system using the National Electrical Manufacturers Association NU 2-2012 standard. Med Phys. 2019;46:3025–3033.
    OpenUrl
  31. 31.↵
    1. Bolard G,
    2. Prior JO,
    3. Modolo L,
    4. et al
    . Performance comparison of two commercial BGO-based PET/CT scanners using NEMA NU 2-2001. Med Phys. 2007;34:2708–2717.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. De Ponti E,
    2. Morzenti S,
    3. Guerra L,
    4. et al
    . Performance measurements for the PET/CT Discovery-600 using NEMA NU 2-2007 standards. Med Phys. 2011;38:968–974.
    OpenUrlPubMed
  33. 33.↵
    1. Badawi RD,
    2. Kohlmyer SG,
    3. Harrison RL,
    4. Vannoy SD,
    5. Lewellen TK
    . The effect of camera geometry on singles flux, scatter fraction and trues and randoms sensitivity for cylindrical 3D PET: a simulation study. IEEE Trans Nucl Sci. 2000;47:1228–1232.
    OpenUrl
  • Received for publication June 7, 2023.
  • Revision received September 27, 2023.
View Abstract
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Bookmark this article

Jump to section

Keywords