Philips Gemini TF Research Articles

Comparing cardiac MRI and lung PET/CT after completion of tuberculosis treatment – preliminary findings of participants co-enrolled into the StatinTB trial and Cardiac Imaging After TB (CIA-TB) study

Abstract Introduction Mortality rates after successful tuberculosis (TB) treatment remain elevated and healthy survival is impaired by chronic pulmonary disease and persistent inflammation. Inflammation leads to a complex cascade of events and has been associated with cardiovascular disease (CVD). Recent epidemiological work described an increased risk of CVD in persons with history of TB. Aims and objectives In our ongoing StatinTB trial (NCT04147286), we evaluate safety and efficacy of 40 mg atorvastatin to reduce persistent lung inflammation (PLI) after TB treatment in HIV−/HIV+ adults measured by 18F-FDG-PET/CT. Participants are co-enrolled into the Cardiac Imaging After Tuberculosis (CIA-TB) study. Here we report preliminary findings of participants enrolled into the StatinTB trial and CIA-TB study with the aim to correlate PLI with cardiac magnetic resonance imaging (CMR) parameters. Methods Participants with clinical response to TB treatment and a negative sputum culture at 16 weeks were screened at the end of a 24-week course of treatment for drug-sensitive TB. PET/CT was performed on a Philips Gemini TF scanner and analysed using MIM®Software. PLI was defined as total lung glycolysis (TLG) ≥50 SUV·mL (Figure 1). CMR was performed using a 3T Siemens Magnetom Skyra scanner and parameters were derived from the following acquisitions: SSFP cine imaging of the short and long axes, T1 mapping (MOLLI, 5 (3) 3), T2 mapping and late gadolinium enhancement (LGE) imaging. StatinTB trial is conducted according to ICH-GCP. Results 26 participants (mean age 37.8±13.8 years, 38.5% women) underwent PET/CT and CMR. 23.1% were infected with HIV, 34.6% smokers, and 26.9% had a previous history of TB. Median BMI was 21.3 (IQR 19.7; 24.3). One participant reported hypertension. Common symptoms at enrollment were cough (26.9%), shortness of breath (11.5%), and chest pain (3.9%). On CT scan, 73.1% had pulmonary nodules, 57.7% scarring, 50.0% cavities, 42.3% bronchiectasis, 19.2% centrilobular disease, 19.2% distorted lung architecture, and 11.5% consolidations. On PET/CT, PLI was detected among 11 (42.3%) participants with a mean TLG of 216.0 (IQR 78.2; 384.0). In a multivariable regression analysis, left ventricular end systolic volume (LVESV) (−1.46; 95% CI: −2.89, −0.23; p=0.047) and distorted lung architecture (2.84; 95% CI: 1.17, 4.52; p=0.003) were significantly associated with TLG. Native T1 relaxation time showed a strong, but not significant association with TLG (15.01; 95% CI: −8.08, 38.10; p=184). Conclusion We present data of a young cohort at the end of TB treatment without significant cardiac risk factors. Structural lung disease is affecting at least half of participants, and many remain symptomatic. PLI is present in more than one-third of participants and our preliminary findings suggest an association with cardiac impairment. Further enrolment and ongoing laboratory studies will shed more light into post-TB cardio-pulmonary sequelae. Funding Acknowledgement Type of funding sources: Public grant(s) – EU funding. Main funding source(s): EDCTP

Investigating the impact of scatter correction on Centiloid

AbstractBackgroundCentiloid (CL) is a semiquantitative measure of amyloid‐β burden based on the ratio between neo‐cortical target region and the whole cerebellum. Photon scatter is one of the major sources of noise in PET data. Scatter correction methods are very different across PET cameras. In this study, we investigate how scatter correction affects the CL for inter/intra‐cameras by comparing PET images with/without scatter correction.Method203 subjects (Siemens mCT Biography (N=109), Philips Gemini TF 64 (N=94)) from the AIBL study were scanned with 18F‐NAV6240. All PET images were reconstructed off‐line with scatter correction enabled and then disabled. All images were processed by CapAIBL to generate CL values. The ΔCL, which was calculated as the difference between the scatter corrected CL and non‐scatter corrected CL, was investigated to quantify the impact of scatter correction. For the intra‐camera comparison, all subjects were categorised in three groups based on the head tilt along anteroposterior axis compared to the MNI‐152 template. For inter‐camera comparison, the head tilt angle was included as covariate into an ANOVA. Hierarchical linear regressions that included head tilt and camera models were used as covariates to investigate their effect on CL.ResultFigure 1 shows that scatter correction has a larger impact on the cerebellum cortex than in neocortex. Figure 2 shows that scatter correction depends on the head position, and it has more impact on subjects with higher CL for both cameras. Figure 3 presents that scatter corrections from two different cameras have different impact on CL values. Both head tilt and camera model had an impact on the scatter correction with (F=19.5, p=1.67e‐5) and (F=109.4, p=1.1e‐20), respectively. The b coefficients were 0.38CL/degree for head tilt and 8.2CL for the camera model, where more scatter is corrected on the Siemens CT than on the Philips Gemini TF.ConclusionOur results show that difference in scatter correction from different cameras have a significant impact on CL, it is yet to be known whether it is caused by hardware difference or software difference. Additional investigations using test‐retest data is needed to further characterize the real impact of scatter correction on Centiloid.

An [18F]FDG-PET/CT deep learning method for fully automated detection of pathological mediastinal lymph nodes in lung cancer patients

PurposeThe identification of pathological mediastinal lymph nodes is an important step in the staging of lung cancer, with the presence of metastases significantly affecting survival rates. Nodes are currently identified by a physician, but this process is time-consuming and prone to errors. In this paper, we investigate the use of artificial intelligence–based methods to increase the accuracy and consistency of this process.MethodsWhole-body 18F-labelled fluoro-2-deoxyglucose ([18F]FDG) positron emission tomography/computed tomography ([18F]FDG-PET/CT) scans (Philips Gemini TF) from 134 patients were retrospectively analysed. The thorax was automatically located, and then slices were fed into a U-Net to identify candidate regions. These regions were split into overlapping 3D cubes, which were individually predicted as positive or negative using a 3D CNN. From these predictions, pathological mediastinal nodes could be identified. A second cohort of 71 patients was then acquired from a different, newer scanner (GE Discovery MI), and the performance of the model on this dataset was tested with and without transfer learning.ResultsOn the test set from the first scanner, our model achieved a sensitivity of 0.87 (95% confidence intervals [0.74, 0.94]) with 0.41 [0.22, 0.71] false positives/patient. This was comparable to the performance of an expert. Without transfer learning, on the test set from the second scanner, the corresponding results were 0.53 [0.35, 0.70] and 0.24 [0.10, 0.49], respectively. With transfer learning, these metrics were 0.88 [0.73, 0.97] and 0.69 [0.43, 1.04], respectively.ConclusionModel performance was comparable to that of an expert on data from the same scanner. With transfer learning, the model can be applied to data from a different scanner. To our knowledge it is the first study of its kind to go directly from whole-body [18F]FDG-PET/CT scans to pathological mediastinal lymph node localisation.

Pilot performance of a dedicated prostate PET suitable for diagnosis and biopsy guidance

BackgroundProstate cancer (PCa) represents one of the most common types of cancers facing the male population. Nowadays, to confirm PCa, systematic or multiparametric MRI-targeted transrectal or transperineal biopsies of the prostate are required. However, due to the lack of an accurate imaging technique capable to precisely locate cancerous cells in the prostate, ultrasound biopsies sample random parts of the prostate and, therefore, it is possible to miss regions where those cancerous cells are present. In spite of the improvement with multiparametric MRI, the low reproducibility of its reading undermines the specificity of the method. Recent development of prostate-specific radiotracers has grown the interest on using positron emission tomography (PET) scanners for this purpose, but technological improvements are still required (current scanners have resolutions in the range of 4–5 mm).ResultsThe main goal of this work is to improve state-of-the-art PCa imaging and diagnosis. We have focused our efforts on the design of a novel prostate-dedicated PET scanner, named ProsPET. This system has small scanner dimensions defined by a ring of just 41 cm inner diameter. In this work, we report the design, implementation, and evaluation (both through simulations and real data) of the ProsPET scanner. We have been able to achieve < 2 mm resolution in reconstructed images and high sensitivity. In addition, we have included a comparison with the Philips Gemini-TF scanner, which is used for routine imaging of PCa patients. The ProsPET exhibits better contrast, especially for rod sizes as small as 4.5 mm in diameter. Finally, we also show the first reconstructed image of a PCa patient acquired with the ProsPET.ConclusionsWe have designed and built a prostate specific PET system, with a small footprint and improved spatial resolution when compared to conventional whole-body PET scanners. The gamma ray impact within each detector block includes accurate DOI determination, correcting for the parallax error. The potential role of combined organ-dedicated prostate-specific membrane antigen (PSMA) PET and ultrasound devices, as a prebiopsy diagnostic tool, could be used to guide sampling of the most aggressive sites in the prostate.

Optimization of injected 68Ga-PSMA activity based on list-mode phantom data and clinical validation

Optimization of injected gallium-68 (68Ga) activity for 68Ga-prostate-specific membrane antigen positron emission tomography/computed tomography (68Ga-PSMA PET/CT) studies is relevant for image quality, radiation protection, and from an economic point of view. However, no clear guidelines are available for 68Ga-PSMA studies. Therefore, a phantom study is performed to determine the highest coefficient of variation (COV) acceptable for reliable image interpretation and quantification.To evaluate image interpretation, the relationship of COV and contrast-to-noise ratio (CNR) was studied. The CNR should remain larger than five, according to the Rose criterion. To evaluate image quantification, the effect of COV on the percentage difference (PD) between quantification results of two studies was analyzed. Comparison was done by calculating the PD of the SUVmax. The maximum allowable PDSUVmax was set at 20%. The highest COV at which both criteria are still met is defined as COVmax. Of the NEMA Image Quality phantom, a 20 min/bed (2 bed positions) scan was acquired in list-mode PET (Philips Gemini TF PET/CT). The spheres to background activity ratio was approximately 9:1. To obtain images with different COV, lower activity was mimicked by reconstructions with acquisition times of 10 min/bed to 5 s/bed. Pairs of images were obtained by reconstruction of two non-overlapping parts of list-mode data.For the 10-mm diameter sphere, a COV of 25% still meets the criteria of CNRSUVmean ≥ 5 and PDSUVmax ≤ 20%. This phantom scan was acquired with an acquisition time of 116 s and a background activity concentration of 0.71 MBq/kg. Translation to a clinical protocol results in a clinical activity regimen of 3.5 MBq/kg min at injection. To verify this activity regimen, 15 patients (6 MBq/kg min) with a total of 22 lesions are included. Additional reconstructions were made to mimic the proposed activity regimen. Based on the CNRSUVmax, no lesions were missed with this proposed activity regimen.For our institution, a clinical activity regimen of 3.5 MBq/kg min at injection is acceptable, which indicates that activity can be reduced by almost 50% compared with the current code of practice. Our proposed method could be used to obtain an objective activity regimen for other PET/CT systems and tracers.

Impact of acquisition time on image quality and visibility of Yttrium-90 Positron Emission Tomography: A phantom study

90Y has internal pair production which enables the imaging of this isotope with positron emission tomography (PET). The acquisition time is a crucial factor which strongly influence the image quality due to very small decay branch of this isotope. Hence, the aim of this study was to evaluate the impact of acquisition time on image quality and visibility of 90Y PET. A NEMA IEC body phantom with set of fillable spheres were filled with 90Y to generate a hot lesion to background ratio of 8:1. PET image was acquired by Philips Gemini TF PET/CT scanner with acquisition time per bed varied from 15, 30, 60 and 120 minutes. Image quality was assessed by percentage hot contrast (HC), image noise (CV), signal-to-noise (SNR) ratio and visibility (VS). From a result, hot spheres and their edges were more defined as acquisition time increased. Additionally, SNR and visibility were increased with acquisition time and sphere size as expected but percentage hot contrast was decreased. Thus, this can be explained by the growing of noise at shorter acquisition times and smaller spheres. In conclusion, 90Y PET imaging was feasible but the imaging protocol should be carefully considered particularly in small object.

Assessment of the Targeting Quality of a Stereotactic Instrument to an Aiming Point in the Brain Using a Frameless Navigation System

A method using modeling on a human head phantom to evaluate the total targeting error of a stereotactic instrument to deep targets in patient’s brain is presented. The means of targeting the instrument to intracerebral aiming points was a Medtronic StealthStation S7 frameless navigation system combined with a Philips Achieva 3 T MRI scanner and a Philips Gemini TF PET/CT scanner.

O33. The value of different reconstruction algorithms for quantification of FDG PET brain imaging

Introduction Reconstruction influences the quantitative results in PET imaging. The aim of this study was the evaluation of different image reconstruction parameters and their impact on quantification for 18 F-FDG PET of the brain. The reconstruction parameters studied were the number of iterations, smoothing levels (relaxation parameter lambda), and the use of time of flight (TOF) information. In addition, we investigated the effect of signal-to-noise ratio on each of these reconstructed brain images, as well as the effect of scan duration on image quality. Materials and Methods A Philips® Gemini-TF Big bore PET/CT was used for acquiring data of a 3D Hoffman Brain phantom. Data was acquired for 25 minutes in list mode format after injection of 40 MBq FDG, and reconstructed with a voxel size of 2 × 2 × 2 mm 3 using two different iterative reconstruction algorithms: LOR-RAMLA and BLOB-OS. The number of iterations and subsets was varied successive from 3/33 (vendor default) to 30/33, acquisition scan duration from 1 to 25 minutes, lambda was selected as smooth (0.7) and normal (1.0), and TOF was switched on and off for BLOB-OS. The impact on image quality was analyzed in 15 cortical and subcortical brain regions (volumes of interest, VOIs) and for grey and white matter. Results Contrast increased for all regions of the brain and for grey matter/white matter (GM/WM) ratio if the number of iterations increased. Image convergence was reached after fifteen iterations for all different algorithms. When varying the smoothing filter it was found that lambda 1.0 resulted in a faster convergence than 0.7. The coefficient of variation (COV) for all VOIs showed BLOB-OS with TOF to be superior to the other algorithms. The COV results for different scan durations showed only a minimal improvement after 5 minutes in high-activity regions (GM), and after 10 minutes in low-activity regions (WM). Conclusion Based on phantom data 18 F-FDG brain imaging for 10 minutes and reconstructed with the BLOB-OS algorithm including TOF information with 15 iterations is optimal on the Philips Gemini TF Big bore PET/CT. Further analyses is planned using patient data to verify if these findings remain valid in a clinical setting.

JQC-PET, AN ImageJ macro for the standardization of PET/CT quality control

Introduction Several publications provide guidelines for the implementation of quality assurance and control programs concerning PET/CT systems. Sometimes the tests described involve complex calculus and stages that are difficult to reproduce or could be quite user dependent. In some cases the more cumbersome tests can be performed using proprietary software provided by the vendor of the equipment, but this “black box” solution makes the process rather opaque. Purpose The aim is to provide a standardization tool for the procedures of the PET/CT quality assurance for automating the generation of the regions of interest and the analysis. An ImageJ macro has been developed to facilitate the analysis of three PET/CT quality control procedures included in IAEA-Pub 1393. Materials and methods We have developed an open macro for the public domain Java image processing and analysis program ImageJ . The phantoms described in the IAEA-Pub 1393 have been used to run the quality assurance procedures in a Philips Gemini TF PET/CT (Philips Medical Systems). Three tests have been implemented: spatial resolution, image quality with accuracy of attenuation and scatter corrections and Uniformity of the reconstructed image. Results The results obtained with the software have been compared with those of the commercial software and the literature with good agreement in all cases. The software is available to the community of medical physicists at www.sefm.es . Conclusion The use of jQC-PET allows a standard analysis and the independence of the commercial software. Disclosure The authors are the developers of the software presented here.

Diagnostic importance of contrast enhanced (18)F-fluorodeoxyglucose positron emission computed tomography in patients with tumor induced osteomalacia: Our experience.

Aims and Objectives:To assess the diagnostic utility of contrast-enhanced 18F-fluorodeoxyglucose positron emission tomography-computed tomography (FDG PET-ceCT) in localization of tumors in patients with clinical diagnosis of tumor-induced osteomalacia (TIO), in correlation with histopathological results.Materials and Methods:Eight patients (five male and three female) aged 24–60 (mean 42) years with a clinical diagnosis of TIO were included in this prospective study. They underwent whole body (head to toe) FDG PET-ceCT following a standard protocol on Philips GEMINI TF PET-CT scanner. The FDG PET-ceCT results were correlated with postoperative histology findings and clinical follow-up.Results:All the patients had an abnormal PET-ceCT study. The sensitivity of PET-ceCT was 87.5%, and positive predictive value was 100%. The tumor was located in the craniofacial region in 6/8 patients and in bone in 2/8 patients. Hemangiopericytoma was the most common reported histology. All patients underwent surgery, following which they demonstrated clinical improvement. However, one patient with atypical findings on histology did not show any clinical improvement, hence, underwent 68Gallium-DOTANOC PET-ceCT scan for relocalization of the site of the tumor.Conclusion:The tumors causing TIO are small in size and usually located in obscure sites in the body. Hence, head to toe protocol should be followed for FDG PET-ceCT scans with the inclusion of upper limbs. Once the tumor is localized, regional magnetic resonance imaging can be performed for better characterization of soft tissue lesion. Imaging with FDG PET-ceCT plays an important role in detecting the site of the tumor and thereby facilitating timely management.

Evaluation of resistive-plate-chamber-based TOF-PET applied to in-beam particle therapy monitoring

Particle therapy is a highly conformal radiotherapy technique which reduces the dose deposited to the surrounding normal tissues. In order to fully exploit its advantages, treatment monitoring is necessary to minimize uncertainties related to the dose delivery. Up to now, the only clinically feasible technique for the monitoring of therapeutic irradiation with particle beams is Positron Emission Tomography (PET). In this work we have compared a Resistive Plate Chamber (RPC)-based PET scanner with a scintillation-crystal-based PET scanner for this application. In general, the main advantages of the RPC-PET system are its excellent timing resolution, low cost, and the possibility of building large area systems. We simulated a partial-ring scanner based on an RPC prototype under construction within the Fondazione per Adroterapia Oncologica (TERA). For comparison with the crystal-based PET scanner we have chosen the geometry of a commercially available PET scanner, the Philips Gemini TF. The coincidence time resolution used in the simulations takes into account the current achievable values as well as expected improvements of both technologies. Several scenarios (including patient data) have been simulated to evaluate the performance of different scanners. Initial results have shown that the low sensitivity of the RPC hampers its application to hadron-beam monitoring, which has an intrinsically low positron yield compared to diagnostic PET. In addition, for in-beam PET there is a further data loss due to the partial ring configuration. In order to improve the performance of the RPC-based scanner, an improved version of the RPC detector (modifying the thickness of the gas and glass layers), providing a larger sensitivity, has been simulated and compared with an axially extended version of the crystal-based device. The improved version of the RPC shows better performance than the prototype, but the extended version of the crystal-based PET outperforms all other options.

Patient doses from PET-CT procedures.

Positron emission tomography (PET) was installed for the first time in Bulgaria in 2009, and nowadays two hybrid PET-computed tomography (CT) systems are in operation. The aim of this work is to estimate patient doses from PET-CT procedures and to explore potential for optimisation. Data were retrospectively collected for 50 patients examined with the system Philips Gemini TF and for 58 patients examined with the system GE Discovery 600. Whole-body examinations with radiopharmaceutical (18)F-2-fluoro-2-deoxy-D-glucose (FDG) were performed on all patients. Patient effective doses from the CT component of the examination were calculated with CT Expo software and compared with doses estimated applying the National Radiological Protection Board (NRPB) conversion coefficients. Effective doses from the PET component were calculated applying the ICRP 80 conversion coefficients. For the first system, average effective doses from CT component were 8.0 and 8.9 mSv, applying CT Expo and NRPB coefficients, respectively, and 4.9 mSv from PET component. For the second system, the corresponding values were 7.8, 8.7 and 5.9 mSv. These results for patient effective doses are relatively lower or comparable to other similar surveys. Reasons for the observed differences are analysed and presented.

Monitoring of the Formation and Development Process for Infection and Inflammation Using F-18 FDG, PET/CT

Objective: Many radiopharmaceuticals have been evaluated extensively in both preclinical and clinical studies as potential diagnostic agents to identify the sites of infection. There is a definite role of FDG-PET in diagnosis, extent of assessing the disease, evaluation of treatment response and disease activity in patients with infections and inflammation. The aim of the study, the process of formation and development of infection and inflammation is monitored using (18 F) 2’-deoxy-2-fluoroD-glucose (F-18 FDG) by Positron Emission Computed Tomography (PET-CT). Methods: In this study, sterile abscess was induced by using turpentine and infected abscess was induced by using Staphylococcus aureus atcc 25923 strain on rats. In the abscess formation on rats, three grups rats were used as sterile, infected and control grups. There were examined male White Wistar Rats, the clinical healthy animals were 150-220 gr body weight. Bacterial strain and rat model for abscess formation for infected abscess formation on rats (n=7), S. aureus 0.5 ml 107 CFU/ml was inoculated in the right arm of the rats as subcutaneous. For sterile abscess formation on rats (n=7) 0.2-0.4 ml turpentine (sigma-aldrich) was injected into the right arm of the rats as subcutaneous. In control group (n=6), 0.5 ml 0.9% NaCl was injected into the right arm of the rats as subcutaneous. First day imsaging was acquired 24 hours after inoculation of S.aureus and turpentine. 1 mCi 18F-FDG was injected intravenously via the tail vein. Prior to 18F-FDG injection, rats fasted at least 4 hours and well hydrated. Imaging was done using PET-CT (PHILIPS Gemini TF), beginning 1 hour following injection of 18F-FDG IV in the first day and at intervals of 24 hours for five days. First day imaging was performed 1. hour after IV injection of 18F-FDG to obtain optimum imaging time. PET/CT images were visually and semiquantitatively assessed. For semiquantitative analysis of the PET images, a region of interest (ROI) was drawn around the abscess area in the right arm and the noninfected contralateral for control as background. Results: SUVmaxs were obtained from the images for evaluation glucose meatabolism of infection and inflammation detected by 18F-FDG, PET/CT. A soft tissue infection developed in the right arm site of the rats within 24 hours after bacterial inoculation. Swelling and redness of the abscess area was apparently seen in all rats. Abscess in rats were visualized by 18F-FDG, PET/CT. The higher abscess/ background rate was seen 1. hour than 2. hour after injection 18F-FDG. Imaging time was chosen as 1. hour post injection 18F-FDG for the following days. In the first day of the formation of S.aureus abscess SUVmax was about 3.9±0.9 while SUVmax was 1.5±0.2 in the control site. In sterile abscess, SUVmax of the first day was 2.2±0.8, while the SUVmax was 1.3±0.5 for control site. First day 2. hours following the injection of 18F-FDG SUVmax values were 2.8±0.6 for infected abscess, 1.2±0.09 for infected abscess control site and 1.9±0.9 for sterile abscess, 1.3±0.3 for sterile abscess control site. Conclusion: Steril and infected abscess differantiation can be evaulated by imaging with 18F-FDG PET. The rate of SUV(max) explores the correlation between sterile abscess and infected abscess. 18F-FDG PET is also useful technique to understand the extent of the infection and inflammation process. In addition 18F- FDG PET imaging method has rapid diagnosis and local availability of equipment and labeled agent.

Evaluation of PET quantification accuracy in vivo

SummaryQuantitative positron emission tomography (PET) requires accurate scanner calibration, which is commonly performed using phantoms. It is not clear to what extent this procedure ensures quantitatively correct results in vivo, since certain conditions differ between phantom and patient scans. Aim: We, therefore, have evaluated the actual quantification accuracy in vivo of PET under clinical routine conditions. Patients, methods: We determined the activity concentration in the bladder in patients undergoing routine [18F]FDG whole body investigations with three different PET scanners (Siemens ECAT EXACT HR+ PET: n = 21; Siemens Biograph 16 PET/CT: n = 16; Philips Gemini-TF PET/CT: n = 19). Urine samples were collected immediately after scan. Activity concentration in the samples was determined in well counters cross-calibrated against the respective scanner. The PET (bladder) to well counter (urine sample) activity concentration ratio was determined. Results: Activity concentration in the bladder (PET) was systematically lower than in the urine samples (well The patient-averaged PET to well counter ratios for the investigated scanners are (mean ± SEM): 0.881 ± 0.015 (ECAT HR+), 0.898 ± 0.024 (Biograph 16), 0.932 ± 0.024 (Gemini-TF). These values correspond to underestimates by PET of 11.9%, 10.2%, and 6.8%, respectively. Conclusions: The investigated PET systems consistently underestimate activity concentration in the bladder. The comparison of urine samples with PET scans of the bladder is a straightforward means for in vivo evaluation of the expectable quantification accuracy. The method might be interesting for multi-center trials, for additional quality assurance in PET and for investigation of PET/MR systems for which clear proof of sufficient quantitative accuracy in vivo is still missing.

Computed tomography automatic exposure control techniques in 18F-FDG oncology PET-CT scanning

Computed tomography (CT) automatic exposure control (AEC) systems are now used in all modern PET-CT scanners. A collaborative study was undertaken to compare AEC techniques of the three major PET-CT manufacturers for fluorine-18 fluorodeoxyglucose half-body oncology imaging. An audit of 70 patients was performed for half-body CT scans taken on a GE Discovery 690, Philips Gemini TF and Siemens Biograph mCT (all 64-slice CT). Patient demographic and dose information was recorded and image noise was calculated as the SD of Hounsfield units in the liver. A direct comparison of the AEC systems was made by scanning a Rando phantom on all three systems for a range of AEC settings. The variation in dose and image quality with patient weight was significantly different for all three systems, with the GE system showing the largest variation in dose with weight and Philips the least. Image noise varied with patient weight in Philips and Siemens systems but was constant for all weights in GE. The z-axis mA profiles from the Rando phantom demonstrate that these differences are caused by the nature of the tube current modulation techniques applied. The mA profiles varied considerably according to the AEC settings used. CT AEC techniques from the three manufacturers yield significantly different tube current modulation patterns and hence deliver different doses and levels of image quality across a range of patient weights. Users should be aware of how their system works and of steps that could be taken to optimize imaging protocols.

Effect of Local TOF Kernel Miscalibrations on Contrast-Noise in TOF PET

TOF PET imaging requires specific calibrations: accurate characterization of the system timing resolution and timing offset is required to achieve the full potential image quality. Current system models used in image reconstruction assume a spatially uniform timing resolution kernel. Furthermore, although the timing offset errors are often pre-corrected, this correction becomes less accurate with the time since, especially in older scanners, the timing offsets are often calibrated only during the installation, as the procedure is time-consuming. In this study, we investigate and compare the effects of local mismatch of timing resolution when a uniform kernel is applied to systems with local variations in timing resolution and the effects of uncorrected time offset errors on image quality. A ring-like phantom was acquired on a Philips Gemini TF scanner and timing histograms were obtained from coincidence events to measure timing resolution along all sets of LORs crossing the scanner center. In addition, multiple acquisitions of a cylindrical phantom, 20 cm in diameter with spherical inserts, and a point source were simulated. A location-dependent timing resolution was simulated, with a median value of 500 ps and increasingly large local variations, and timing offset errors ranging from 0 to 350 ps were also simulated. Images were reconstructed with TOF MLEM with a uniform kernel corresponding to the effective timing resolution of the data, as well as with purposefully mismatched kernels. To CRC vs noise curves were measured over the simulated cylinder realizations, while the simulated point source was processed to generate timing histograms of the data. Results show that timing resolution is not uniform over the FOV of the considered scanner. The simulated phantom data indicate that CRC is moderately reduced in data sets with locally varying timing resolution reconstructed with a uniform kernel, while still performing better than non-TOF reconstruction. On the other hand, uncorrected offset errors in our setup have a larger potential for decreasing image quality and can lead to a reduction of CRC of up to 15% and an increase in the measured timing resolution kernel up to 40%. However, in realistic conditions in frequently calibrated systems, using a larger effective timing kernel in image reconstruction can compensate uncorrected offset errors.

Study-Parameter Impact in Quantitative 90-Yttrium PET Imaging for Radioembolization Treatment Monitoring and Dosimetry

A small positron-generating branch in 90-Yttrium ((90)Y) decay enables post-therapy dose assessment in liver cancer radioembolization treatment. The aim of this study was to validate clinical (90)Y positron emission tomography (PET) quantification, focusing on scanner linearity as well as acquisition and reconstruction parameter impact on scanner calibration. Data from three dedicated phantom studies (activity range: 55.2 MBq-2.1 GBq) carried out on a Philips Gemini TF 16 PET/CT scanner were analyzed after reconstruction with up to 361 parameter configurations. For activities above 200 MBq, scanner linearity could be confirmed with relative error margins 4%. An acquisition-time-normalized calibration factor of 1.04 MBq·s/CNTS was determined for the employed scanner. Stable activity convergence was found in hot phantom regions with relative differences in summed image intensities between -3.6% and +2.4%. Absolute differences in background noise artifacts between - 79.9% and + 350% were observed. Quantitative accuracy was dominated by subset size selection in the reconstruction. Using adequate segmentation and optimized acquisition parameters, the average activity recovery error induced by the axial scanner sensitivity profile was reduced to +2.4%±3.4% (mean ± standard deviation). We conclude that post-therapy dose assessment in (90)Y PET can be improved using adapted parameter setups.

Does 18F-Fluorodeoxyglucose Outperform 18F-Fluorothymidine When Using Positron Emission Tomography in Predicting Transformation of Indolent Non-Hodgkin's Lymphoma,

Abstract 3658Introduction Patients (pts) with transformed NHL have an extremely poor prognosis. However when diagnosed at an early stage outcome may be better, making early diagnosis crucial. Nowadays, transformation is diagnosed using biopsy of a lymph node when it starts growing rapidly or when the patient develops B symptoms or hypercalcemia. These pts often have advanced disease and diagnosis may be delayed because the lymph node biopsied is not representative. Positron emission tomography (PET) might be of value here: the commonly used tracer 18F-fluorodeoxyclucose (FDG)-PET can distinguish between indolent and aggressive lymphoma, but with considerable overlap in Standard Uptake Values (SUV). However, 18F-fluorothymidine (FLT) is thought to more directly reflect proliferation. We performed a head-to-head comparison of FDG and FLT PET in pts with indolent and transformed lymphoma to explore which tracer qualifies best for future research in timely diagnosis of transformation.Materials and methods Pts were selected based on histology: either follicular lymphoma or transformed lymphoma (defined as diffuse large B cell lymphoma diagnosed in a patient with former or simultaneous diagnosis of follicular lymphoma in a lymph node).In each patient two PET scans were made, one with FDG and one with FLT, maximum one week apart, before any treatment.Scans were made on the Philips Gemini TF PET-CT camera, 1 hour after injection of 185 MBq of FDG or FLT.Uptake (we present data as SUVmax normalized to body weight) was measured in all lymph nodes ≥ 2 cm (minimizing partial volume effects).Results 17 pts with indolent lymphoma and 10 with transformed lymphoma were included. Median age was 59 years (range 35–81) and a median of 9 lymph nodes were measured per patient (range 2–23). Between patients, SUVmax in the lymph node with the highest uptake for both FDG (p=0.01) as well as FLT (p=0.04) uptake was significantly higher in transformed lymphoma. For FDG values ranged from 4,9 to 19,6 in indolent and from 9,6 to 29,9 in transformed lymphoma, for FLT values were respectively 3,6 to 16,6 and 5,45 to 16,3. Accordingly, between patients, we found considerable overlap between highest values for indolent and transformed lymphoma with either tracer, making the determination of a cut off level for transformation difficult. The data suggest that, at least for FDG, the range between the lymph node with the highest and the lowest uptake within one patient was the best criterium to identify the transformed lymphomas, with significantly higher ranges for transformed lymphoma (p<0,001; ranges from 0,34 to 11,3 for indolent and from 5,1 to 20,2 for transformed lymphoma). A cut off value of 7 (highest SUVmax minus lowest SUVmax) has a positive predictive value of 90% for transformation.Conclusion Uptake of both FDG and FLT are significantly higher in transformed lymphoma than in indolent lymphoma. However, FDG-PET may better distinguish between indolent and transformed lymphoma. For further research, intrapatient heterogeneity of FDG uptake qualifies as a potential marker of transformation, ultimately to facilitate timely diagnosis of transformation and improved patient outcome. Disclosures:No relevant conflicts of interest to declare.

The effect of errors in segmented attenuation maps on PET quantification

Accurate attenuation correction is important for PET quantification. Often, a segmented attenuation map is used, especially in MRI-based attenuation correction. As deriving the attenuation map from MRI images is difficult, different errors can be present in the segmented attenuation map. The goal of this paper is to determine the effect of these errors on quantification. The authors simulated the digital XCAT phantom using the GATE Monte Carlo simulation framework and a model of the Philips Gemini TF. A whole body scan was simulated, spanning an axial field of view of 70 cm. A total of fifteen lesions were placed in the lung, liver, spine, colon, prostate, and femur. The acquired data were reconstructed with a reference attenuation map and with different attenuation maps that were modified to reflect common segmentation errors. The quantitative difference between reconstructed images was evaluated. Segmentation into five tissue classes, namely cortical bone, spongeous bone, soft tissue, lung, and air yielded errors below 5%. Large errors were caused by ignoring lung tissue (up to 45%) or cortical bone (up to 17%). The interpatient variability of lung attenuation coefficients can lead to errors of 10% and more. Up to 20% tissue misclassification from bone to soft tissue yielded errors below 5%. The same applies for up to 10% misclassification from lung to air. When using a segmented attenuation map, at least five different tissue types should be considered: cortical bone, spongeous bone, soft tissue, lung, and air. Furthermore, the interpatient variability of lung attenuation coefficients should be taken into account. Limited misclassification from bone to soft tissue and from lung to air is acceptable, as these do not lead to relevant errors.

SU-E-T-146: Uncertainty of Stopping Powers of Tissue Equivalent Materials in Uniformly Scanned Proton Beam

Purpose: To quantify uncertainties associated with the extraction of CT HU to stopping power conversion curve for tissue equivalent materials in uniformly scanned (US)proton beams. Methods and Materials: The Philips Gemini TF Big Bore PET/CT scanner at the Hampton University Proton Therapy Institute (HUPTI), was used to extract the HU numbers for several tissue equivalent materials in the CIRS M062 phantom. Protonstopping power for each material was extracted by measuring the range change with and without the material, in a USproton beam pristine Bragg peak, and comparing to its physical thickness. The in‐beam measurements were performed using a Multilayer Ionization Chamber device (IBA Dosimetry, Zebra). Data were analyzed using both IBA OmniPro Incline and an in‐house software analysis package. Several proton beam energies were used: 143 MeV, 175 MeV, and 210 MeV. For each material and energy the difference between measured and predicted value of the stopping power was quantified. Results: The difference between measured and calculated stopping power with proton energy was found to be within 1–2.5 percent. The upper value of this variation was added to the machine specific systematic error of the CT HU number and quantified as the uncertainty of the CT HU to protonstopping power conversion curve. It was added to the total uncertainties (proton range and lateral profile), when PTV margins were implemented in the TPS. Conclusions: The relative linear stopping powers of 9 different tissue equivalent materials were measured using a range of uniformly scanned proton beam energies. These values were compared with predicted stopping powers calculated using current prescriptions and quantified as uncertainties of the CT HU to protonstopping power calibration curve. This work supports the accuracy of the protonstopping power implementation in the TPS and implicit the accuracy of patient treatment at HUPTI.