EJNMMI Physics最新文献

Background: The application of semi-conductor detectors such as cadmium-zinc-telluride (CZT) in nuclear medicine improves extrinsic energy resolution and count sensitivity due to the direct conversion of gamma photons into electric signals. A 3D-ring pixelated CZT system named StarGuide was recently developed and implemented by GE HealthCare for SPECT acquisition. The system consists of 12 detector columns with seven modules of 16 × 16 CZT pixelated crystals, each with an integrated parallel-hole tungsten collimator. The axial coverage is 27.5 cm. The detector thickness is 7.25 mm, which allows acquisitions in the energy range [40-279] keV. Since there is currently no performance characterization specific to 3D-ring CZT SPECT systems, the National Electrical Manufacturers Association (NEMA) NU 1-2018 clinical standard can be tailored to these cameras. The aim of this study was to evaluate the performance of the SPECT/CT StarGuide system according to the NEMA NU 1-2018 clinical standard specifically adapted to characterize the new 3D-ring CZT.

Results: Due to the integrated collimator, the system geometry and the pixelated nature of the detector, some NEMA tests have been adapted to the features of the system. The extrinsic measured energy resolution was about 5-6% for the tested isotopes (^99mTc, ¹²³I and ⁵⁷Co); the maximum count rate was 760 kcps and the observed count rate at 20% loss was 917 kcps. The system spatial resolution in air extrapolated at 10 cm with ^99mTc was 7.2 mm, while the SPECT spatial resolutions with scatter were 4.2, 3.7 and 3.6 mm in a central, radial and tangential direction respectively. Single head sensitivity value for ^99mTc was 97 cps/MBq; with 12 detector columns, the system volumetric sensitivity reached 520 kcps MBq^-1 cc^-1.

Conclusions: The performance tests of the StarGuide can be performed according to the NEMA NU 1-2018 standard with some adaptations. The system has shown promising results, particularly in terms of energy resolution, spatial resolution and volumetric sensitivity, potentially leading to higher quality clinical images.

Purpose: Effective radiation therapy requires accurate segmentation of head and neck cancer, one of the most common types of cancer. With the advancement of deep learning, people have come up with various methods that use positron emission tomography-computed tomography to get complementary information. However, these approaches are computationally expensive because of the separation of feature extraction and fusion functions and do not make use of the high sensitivity of PET. We propose a new deep learning-based approach to alleviate these challenges.

Methods: We proposed a tumor region attention module that fully exploits the high sensitivity of PET and designed a network that learns the correlation between the PET and CT features using squeeze-and-excitation normalization (SE Norm) without separating the feature extraction and fusion functions. In addition, we introduce multi-scale context fusion, which exploits contextual information from different scales.

Results: The HECKTOR challenge 2021 dataset was used for training and testing. The proposed model outperformed the state-of-the-art models for medical image segmentation; in particular, the dice similarity coefficient increased by 8.78% compared to U-net.

Conclusion: The proposed network segmented the complex shape of the tumor better than the state-of-the-art medical image segmentation methods, accurately distinguishing between tumor and non-tumor regions.

Background: Low-dose ungated CT is commonly used for total-body PET attenuation and scatter correction (ASC). However, CT-based ASC (CT-ASC) is limited by radiation dose risks of CT examinations, propagation of CT-based artifacts and potential mismatches between PET and CT. We demonstrate the feasibility of direct ASC for multi-tracer total-body PET in the image domain.

Methods: Clinical uEXPLORER total-body PET/CT datasets of [¹⁸F]FDG (N = 52), [¹⁸F]FAPI (N = 46) and [⁶⁸Ga]FAPI (N = 60) were retrospectively enrolled in this study. We developed an improved 3D conditional generative adversarial network (cGAN) to directly estimate attenuation and scatter-corrected PET images from non-attenuation and scatter-corrected (NASC) PET images. The feasibility of the proposed 3D cGAN-based ASC was validated using four training strategies: (1) Paired 3D NASC and CT-ASC PET images from three tracers were pooled into one centralized server (CZ-ASC). (2) Paired 3D NASC and CT-ASC PET images from each tracer were individually used (DL-ASC). (3) Paired NASC and CT-ASC PET images from one tracer ([¹⁸F]FDG) were used to train the networks, while the other two tracers were used for testing without fine-tuning (NFT-ASC). (4) The pre-trained networks of (3) were fine-tuned with two other tracers individually (FT-ASC). We trained all networks in fivefold cross-validation. The performance of all ASC methods was evaluated by qualitative and quantitative metrics using CT-ASC as the reference.

Results: CZ-ASC, DL-ASC and FT-ASC showed comparable visual quality with CT-ASC for all tracers. CZ-ASC and DL-ASC resulted in a normalized mean absolute error (NMAE) of 8.51 ± 7.32% versus 7.36 ± 6.77% (p < 0.05), outperforming NASC (p < 0.0001) in [¹⁸F]FDG dataset. CZ-ASC, FT-ASC and DL-ASC led to NMAE of 6.44 ± 7.02%, 6.55 ± 5.89%, and 7.25 ± 6.33% in [¹⁸F]FAPI dataset, and NMAE of 5.53 ± 3.99%, 5.60 ± 4.02%, and 5.68 ± 4.12% in [⁶⁸Ga]FAPI dataset, respectively. CZ-ASC, FT-ASC and DL-ASC were superior to NASC (p < 0.0001) and NFT-ASC (p < 0.0001) in terms of NMAE results.

Conclusions: CZ-ASC, DL-ASC and FT-ASC demonstrated the feasibility of providing accurate and robust ASC for multi-tracer total-body PET, thereby reducing the radiation hazards to patients from redundant CT examinations. CZ-ASC and FT-ASC could outperform DL-ASC for cross-tracer total-body PET AC.

${}^{177}\text{Lu}$ radiopharmaceutical therapy is a standardized systemic treatment, with a typical dose of 7.4 GBq per injection, but its response varies from patient to patient. Dosimetry provides the opportunity to personalize treatment, but it requires multiple post-injection images to monitor the radiopharmaceutical's biodistribution over time. This imposes an additional imaging burden on centers with limited resources. This review explores methods to lessen this burden by optimizing acquisition types and minimizing the number and duration of imaging sessions. After summarizing the different steps of dosimetry and providing examples of dosimetric workflows for ${}^{177}\text{Lu}$ -DOTATATE and ${}^{177}\text{Lu}$ -PSMA, we examine dosimetric workflows based on a reduced number of acquisitions, or even just one. We provide a non-exhaustive description of simplified methods and their assumptions, as well as their limitations. Next, we detail the specificities of each normal tissue and tumors, before reviewing dose-response relationships in the literature. In conclusion, we will discuss the current limitations of dosimetric workflows and propose avenues for improvement.

Purpose: To develop a deep learning (DL) model for generating automated regions of interest (ROIs) on ^99mTc-diethylenetriamine pentaacetic acid (DTPA) renal scans for glomerular filtration rate (GFR) measurement.

Methods: Manually-drawn ROIs retrieved from a Picture Archiving and Communications System were used as ground-truth (GT) labels. A two-dimensional U-Net convolutional neural network architecture with multichannel input was trained to generate DL ROIs. The agreement between GFR values from GT and DL ROIs was evaluated using Lin's concordance correlation coefficient (CCC) and slope coefficients for linear regression analyses. Bias and 95% limits of agreement (LOA) were assessed using Bland-Altman plots.

Results: A total of 24,364 scans (12,822 patients) were included. Excellent concordance between GT and DL GFR was found for left (CCC 0.982, 95% confidence interval [CI] 0.981-0.982; slope 1.004, 95% CI 1.003-1.004), right (CCC 0.969, 95% CI 0.968-0.969; slope 0.954, 95% CI 0.953-0.955) and both kidneys (CCC 0.978, 95% CI 0.978-0.979; slope 0.979, 95% CI 0.978-0.979). Bland-Altman analysis revealed minimal bias between GT and DL GFR, with mean differences of - 0.2 (95% LOA - 4.4-4.0), 1.4 (95% LOA - 3.5-6.3) and 1.2 (95% LOA - 6.5-8.8) mL/min/1.73 m² for left, right and both kidneys, respectively. Notably, 19,960 scans (81.9%) showed an absolute difference in GFR of less than 5 mL/min/1.73 m².

Conclusion: Our DL model exhibited excellent performance in the generation of ROIs on ^99mTc-DTPA renal scans. This automated approach could potentially reduce manual effort and enhance the precision of GFR measurement in clinical practice.

Background: Internal dosimetry in individual patients is essential for safe and effective radioligand therapy. Multiple time point imaging for accurate dosimetry is time consuming and hence can be demanding for nuclear medicine departments as well as patients. The objectives of this study were (1) to assess absorbed doses to organs at risk and tumor lesions for [¹⁷⁷Lu]Lu-PSMA-I&T using whole body SPECT imaging and (2) to investigate possible simplified dosimetry protocols.

Methods: This study included 16 patients each treated with 4 cycles of [¹⁷⁷Lu]Lu-PSMA-I&T. They underwent quantitative whole body SPECT/CT imaging (3 bed positions) at four time points (TP) comprising 2 h, 24 h, 48 h and 72-168 h post-injection (p.i.). Full 3D dosimetry (reference method) was performed for all patients and dose cycles for organs at risk (kidneys, parotid glands and submandibular glands) and up to ten tumor lesions per patient (resulting in 90 lesions overall). The simplified dosimetry methods (SM) included (1) generating time activity curves for subsequent cycles using a single TP of imaging applying the kinetics of dose cycle 1, and for organs at risk also (2) simple extrapolation from dose cycle 1 and (3) from both, dose cycle 1 and 2.

Results: Normalized absorbed doses were 0.71 ± 0.32 mGy/MBq, 0.28 ± 0.12 mGy/MBq and 0.22 ± 0.08 mGy/MBq for kidneys, parotid glands and submandibular glands, respectively. Tumor doses decreased from 3.86 ± 3.38 mGy/MBq in dose cycle 1 to 2.01 ± 2.65 mGy/MBq in dose cycle 4. Compared to the full dosimetry approach the SM 1 using single TP imaging at 48 h p.i. resulted in the most accurate and precise results for the organs at risk in terms of absorbed doses per cycle and total cumulated dose. For tumor lesions better results were achieved using the fourth TP (≥ 72 h p.i.).

Conclusion: Simplification of safety dosimetry protocols is possible for [¹⁷⁷Lu]Lu-PSMA-I&T therapy. If tumor dosimetry is of interest a later imaging TP (≥ 72 h p.i.) should be used/added to account for the slower kinetics of tumors compared to organs at risk.

Background: Several research groups have explored the potential of scandium radionuclides for theragnostic applications due to their longer half-lives and equal or similar coordination chemistry between their diagnostic and therapeutic counterparts, as well as lutetium-177 and terbium-161, respectively. Unlike the gallium-68/lutetium-177 pair, which may show different in-vivo uptake patterns, the use of scandium radioisotopes promises consistent behaviour between diagnostic and therapeutic radiopeptides. An advantage of scandium's longer half-life over gallium-68 is the ability to study radiopeptide uptake over extended periods and its suitability for centralized production and distribution. However, concerns arise from scandium-44's decay characteristics and scandium-43's high production costs. This study aimed to evaluate the dosimetric implications of using scandium radioisotopes with somatostatin analogues against gallium-68 for PET imaging of neuroendocrine tumours.

Methods: Absorbed dose per injected activity (AD/IA) from the generated time-integrated activity curve (TIAC) were estimated using the radiopeptides [^43/44/44mSc]Sc- and [⁶⁸Ga]Ga-DOTATATE. The kidneys, liver, spleen, and red bone marrow (RBM) were selected for dose estimation studies. The EGSnrc and MCNP6.1 Monte Carlo (MC) codes were used with female (AF) and male (AM) ICRP phantoms. The results were compared to Olinda/EXM software, and the effective dose concentrations assessed, varying composition between the scandium radioisotopes.

Results: Our findings showed good agreement between the MC codes, with - 3 ± 8% mean difference. Kidneys, liver, and spleen showed differences between the MC codes (min and max) in a range of - 4% to 8%. This was observed for both phantoms for all radiopeptides used in the study. Compared to Olinda/EXM the largest observed difference was for the RBM, of 21% for the AF and 16% for the AM for scandium- and gallium-based radiopeptides. Despite the differences, our findings showed a higher absorbed dose on [^43/44Sc]Sc-DOTATATE compared to its ⁶⁸Ga-based counterpart.

Conclusion: This study found that [^43/44Sc]Sc-DOTATATE delivers a higher absorbed dose to organs at risk compared to [⁶⁸Ga]Ga-DOTATATE, assuming equal distribution. This is due to the longer half-life of scandium radioisotopes compared to gallium-68. However, calculated doses are within acceptable ranges, making scandium radioisotopes a feasible replacement for gallium-68 in PET imaging, potentially offering enhanced diagnostic potential with later timepoint imaging.

Background: The aim was to investigate the noise and bias properties of quantitative ¹⁷⁷Lu-SPECT with respect to the number of projection angles, and the number of subsets and iterations in the OS-EM reconstruction, for different total acquisition times.

Methods: Experimental SPECT acquisition of six spheres in a NEMA body phantom filled with ¹⁷⁷Lu was performed, using medium-energy collimators and 120 projections with 180 s per projection. Bootstrapping was applied to generate data sets representing acquisitions with 20 to 120 projections for 10 min, 20 min, and 40 min, with 32 noise realizations per setting. Monte Carlo simulations were performed of ¹⁷⁷Lu-DOTA-TATE in an anthropomorphic computer phantom with three tumours (2.8 mL to 40.0 mL). Projections representing 24 h and 168 h post administration were simulated, each with 32 noise realizations. Images were reconstructed using OS-EM with compensation for attenuation, scatter, and distance-dependent resolution. The number of subsets and iterations were varied within a constrained range of the product number of iterations $\times$ number of projections $\le 2400$ . Volumes-of-interest were defined following the physical size of the spheres and tumours, the mean activity-concentrations estimated, and the absolute mean relative error and coefficient of variation (CV) over noise realizations calculated. Pareto fronts were established by analysis of CV versus mean relative error.

Results: Points at the Pareto fronts with low CV and high mean error resulted from using a low number of subsets, whilst points at the Pareto fronts associated with high CV but low mean error resulted from reconstructions with a high number of subsets. The number of projection angles had limited impact.

Conclusions: For accurate estimation of the ¹⁷⁷Lu activity-concentration from SPECT images, the number of projection angles has limited importance, whilst the total acquisition time and the number of subsets and iterations are parameters of importance.

Purpose: ¹²³I-Ioflupane SPECT is an effective tool for the diagnosis and progression assessment of Parkinson's disease (PD). Radiomics and deep learning (DL) can be used to track and analyze the underlying image texture and features to predict the Hoehn-Yahr stages (HYS) of PD. In this study, we aim to predict HYS at year 0 and year 4 after the first diagnosis with combined imaging, radiomics and DL-based features using ¹²³I-Ioflupane SPECT images at year 0.

Methods: In this study, 161 subjects from the Parkinson's Progressive Marker Initiative database underwent baseline 3T MRI and ¹²³I-Ioflupane SPECT, with HYS assessment at years 0 and 4 after first diagnosis. Conventional imaging features (IF) and radiomic features (RaF) for striatum uptakes were extracted from SPECT images using MRI- and SPECT-based (SPECT-V and SPECT-T) segmentations respectively. A 2D DenseNet was used to predict HYS of PD, and simultaneously generate deep features (DF). The random forest algorithm was applied to develop models based on DF, RaF, IF and combined features to predict HYS (stage 0, 1 and 2) at year 0 and (stage 0, 1 and ≥ 2) at year 4, respectively. Model predictive accuracy and receiver operating characteristic (ROC) analysis were assessed for various prediction models.

Results: For the diagnostic accuracy at year 0, DL (0.696) outperformed most models, except DF + IF in SPECT-V (0.704), significantly superior based on paired t-test. For year 4, accuracy of DF + RaF model in MRI-based method is the highest (0.835), significantly better than DF + IF, IF + RaF, RaF and IF models. And DL (0.820) surpassed models in both SPECT-based methods. The area under the ROC curve (AUC) highlighted DF + RaF model (0.854) in MRI-based method at year 0 and DF + RaF model (0.869) in SPECT-T method at year 4, outperforming DL models, respectively. And then, there was no significant differences between SPECT-based and MRI-based segmentation methods except for the imaging feature models.

Conclusion: The combination of radiomic and deep features enhances the prediction accuracy of PD HYS compared to only radiomics or DL. This suggests the potential for further advancements in predictive model performance for PD HYS at year 0 and year 4 after first diagnosis using ¹²³I-Ioflupane SPECT images at year 0, thereby facilitating early diagnosis and treatment for PD patients. No significant difference was observed in radiomics results obtained between MRI- and SPECT-based striatum segmentations for radiomic and deep features.